1DBI(3)                User Contributed Perl Documentation               DBI(3)
2
3
4

NAME

6       DBI - Database independent interface for Perl
7

SYNOPSIS

9         use DBI;
10
11         @driver_names = DBI->available_drivers;
12         %drivers      = DBI->installed_drivers;
13         @data_sources = DBI->data_sources($driver_name, \%attr);
14
15         $dbh = DBI->connect($data_source, $username, $auth, \%attr);
16
17         $rv  = $dbh->do($statement);
18         $rv  = $dbh->do($statement, \%attr);
19         $rv  = $dbh->do($statement, \%attr, @bind_values);
20
21         $ary_ref  = $dbh->selectall_arrayref($statement);
22         $hash_ref = $dbh->selectall_hashref($statement, $key_field);
23
24         $ary_ref  = $dbh->selectcol_arrayref($statement);
25         $ary_ref  = $dbh->selectcol_arrayref($statement, \%attr);
26
27         @row_ary  = $dbh->selectrow_array($statement);
28         $ary_ref  = $dbh->selectrow_arrayref($statement);
29         $hash_ref = $dbh->selectrow_hashref($statement);
30
31         $sth = $dbh->prepare($statement);
32         $sth = $dbh->prepare_cached($statement);
33
34         $rc = $sth->bind_param($p_num, $bind_value);
35         $rc = $sth->bind_param($p_num, $bind_value, $bind_type);
36         $rc = $sth->bind_param($p_num, $bind_value, \%attr);
37
38         $rv = $sth->execute;
39         $rv = $sth->execute(@bind_values);
40         $rv = $sth->execute_array(\%attr, ...);
41
42         $rc = $sth->bind_col($col_num, \$col_variable);
43         $rc = $sth->bind_columns(@list_of_refs_to_vars_to_bind);
44
45         @row_ary  = $sth->fetchrow_array;
46         $ary_ref  = $sth->fetchrow_arrayref;
47         $hash_ref = $sth->fetchrow_hashref;
48
49         $ary_ref  = $sth->fetchall_arrayref;
50         $ary_ref  = $sth->fetchall_arrayref( $slice, $max_rows );
51
52         $hash_ref = $sth->fetchall_hashref( $key_field );
53
54         $rv  = $sth->rows;
55
56         $rc  = $dbh->begin_work;
57         $rc  = $dbh->commit;
58         $rc  = $dbh->rollback;
59
60         $quoted_string = $dbh->quote($string);
61
62         $rc  = $h->err;
63         $str = $h->errstr;
64         $rv  = $h->state;
65
66         $rc  = $dbh->disconnect;
67
68       The synopsis above only lists the major methods and parameters.
69
70   GETTING HELP
71       General
72
73       Before asking any questions, reread this document, consult the archives
74       and read the DBI FAQ. The archives are listed at the end of this
75       document and on the DBI home page <http://dbi.perl.org/support/>
76
77       You might also like to read the Advanced DBI Tutorial at
78       <http://www.slideshare.net/Tim.Bunce/dbi-advanced-tutorial-2007>
79
80       To help you make the best use of the dbi-users mailing list, and any
81       other lists or forums you may use, I recommend that you read "Getting
82       Answers" by Mike Ash: <http://mikeash.com/getting_answers.html>.
83
84       Mailing Lists
85
86       If you have questions about DBI, or DBD driver modules, you can get
87       help from the dbi-users@perl.org mailing list. This is the best way to
88       get help. You don't have to subscribe to the list in order to post,
89       though I'd recommend it. You can get help on subscribing and using the
90       list by emailing dbi-users-help@perl.org.
91
92       Please note that Tim Bunce does not maintain the mailing lists or the
93       web pages (generous volunteers do that).  So please don't send mail
94       directly to him; he just doesn't have the time to answer questions
95       personally. The dbi-users mailing list has lots of experienced people
96       who should be able to help you if you need it. If you do email Tim he
97       is very likely to just forward it to the mailing list.
98
99       IRC
100
101       DBI IRC Channel: #dbi on irc.perl.org (<irc://irc.perl.org/#dbi>)
102
103       Online
104
105       StackOverflow has a DBI tag
106       <http://stackoverflow.com/questions/tagged/dbi> with over 800
107       questions.
108
109       The DBI home page at <http://dbi.perl.org/> and the DBI FAQ at
110       <http://faq.dbi-support.com/> may be worth a visit.  They include links
111       to other resources, but are rather out-dated.
112
113       Reporting a Bug
114
115       If you think you've found a bug then please read "How to Report Bugs
116       Effectively" by Simon Tatham:
117       <http://www.chiark.greenend.org.uk/~sgtatham/bugs.html>.
118
119       If you think you've found a memory leak then read "Memory Leaks".
120
121       Your problem is most likely related to the specific DBD driver module
122       you're using. If that's the case then click on the 'Bugs' link on the
123       <http://metacpan.org> page for your driver. Only submit a bug report
124       against the DBI itself if you're sure that your issue isn't related to
125       the driver you're using.
126
127   NOTES
128       This is the DBI specification that corresponds to DBI version 1.642
129       (see DBI::Changes for details).
130
131       The DBI is evolving at a steady pace, so it's good to check that you
132       have the latest copy.
133
134       The significant user-visible changes in each release are documented in
135       the DBI::Changes module so you can read them by executing "perldoc
136       DBI::Changes".
137
138       Some DBI changes require changes in the drivers, but the drivers can
139       take some time to catch up. Newer versions of the DBI have added
140       features that may not yet be supported by the drivers you use.  Talk to
141       the authors of your drivers if you need a new feature that is not yet
142       supported.
143
144       Features added after DBI 1.21 (February 2002) are marked in the text
145       with the version number of the DBI release they first appeared in.
146
147       Extensions to the DBI API often use the "DBIx::*" namespace.  See
148       "Naming Conventions and Name Space". DBI extension modules can be found
149       at <https://metacpan.org/search?q=DBIx>.  And all modules related to
150       the DBI can be found at <https://metacpan.org/search?q=DBI>.
151

DESCRIPTION

153       The DBI is a database access module for the Perl programming language.
154       It defines a set of methods, variables, and conventions that provide a
155       consistent database interface, independent of the actual database being
156       used.
157
158       It is important to remember that the DBI is just an interface.  The DBI
159       is a layer of "glue" between an application and one or more database
160       driver modules.  It is the driver modules which do most of the real
161       work. The DBI provides a standard interface and framework for the
162       drivers to operate within.
163
164       This document often uses terms like references, objects, methods.  If
165       you're not familiar with those terms then it would be a good idea to
166       read at least the following perl manuals first: perlreftut, perldsc,
167       perllol, and perlboot.
168
169   Architecture of a DBI Application
170                    |<- Scope of DBI ->|
171                         .-.   .--------------.   .-------------.
172         .-------.       | |---| XYZ Driver   |---| XYZ Engine  |
173         | Perl  |       | |   `--------------'   `-------------'
174         | script|  |A|  |D|   .--------------.   .-------------.
175         | using |--|P|--|B|---|Oracle Driver |---|Oracle Engine|
176         | DBI   |  |I|  |I|   `--------------'   `-------------'
177         | API   |       | |...
178         |methods|       | |... Other drivers
179         `-------'       | |...
180                         `-'
181
182       The API, or Application Programming Interface, defines the call
183       interface and variables for Perl scripts to use. The API is implemented
184       by the Perl DBI extension.
185
186       The DBI "dispatches" the method calls to the appropriate driver for
187       actual execution.  The DBI is also responsible for the dynamic loading
188       of drivers, error checking and handling, providing default
189       implementations for methods, and many other non-database specific
190       duties.
191
192       Each driver contains implementations of the DBI methods using the
193       private interface functions of the corresponding database engine.  Only
194       authors of sophisticated/multi-database applications or generic library
195       functions need be concerned with drivers.
196
197   Notation and Conventions
198       The following conventions are used in this document:
199
200         $dbh    Database handle object
201         $sth    Statement handle object
202         $drh    Driver handle object (rarely seen or used in applications)
203         $h      Any of the handle types above ($dbh, $sth, or $drh)
204         $rc     General Return Code  (boolean: true=ok, false=error)
205         $rv     General Return Value (typically an integer)
206         @ary    List of values returned from the database, typically a row of data
207         $rows   Number of rows processed (if available, else -1)
208         $fh     A filehandle
209         undef   NULL values are represented by undefined values in Perl
210         \%attr  Reference to a hash of attribute values passed to methods
211
212       Note that Perl will automatically destroy database and statement handle
213       objects if all references to them are deleted.
214
215   Outline Usage
216       To use DBI, first you need to load the DBI module:
217
218         use DBI;
219         use strict;
220
221       (The "use strict;" isn't required but is strongly recommended.)
222
223       Then you need to "connect" to your data source and get a handle for
224       that connection:
225
226         $dbh = DBI->connect($dsn, $user, $password,
227                             { RaiseError => 1, AutoCommit => 0 });
228
229       Since connecting can be expensive, you generally just connect at the
230       start of your program and disconnect at the end.
231
232       Explicitly defining the required "AutoCommit" behaviour is strongly
233       recommended and may become mandatory in a later version.  This
234       determines whether changes are automatically committed to the database
235       when executed, or need to be explicitly committed later.
236
237       The DBI allows an application to "prepare" statements for later
238       execution.  A prepared statement is identified by a statement handle
239       held in a Perl variable.  We'll call the Perl variable $sth in our
240       examples.
241
242       The typical method call sequence for a "SELECT" statement is:
243
244         prepare,
245           execute, fetch, fetch, ...
246           execute, fetch, fetch, ...
247           execute, fetch, fetch, ...
248
249       for example:
250
251         $sth = $dbh->prepare("SELECT foo, bar FROM table WHERE baz=?");
252
253         $sth->execute( $baz );
254
255         while ( @row = $sth->fetchrow_array ) {
256           print "@row\n";
257         }
258
259       For queries that are not executed many times at once, it is often
260       cleaner to use the higher level select wrappers:
261
262         $row_hashref = $dbh->selectrow_hashref("SELECT foo, bar FROM table WHERE baz=?", undef, $baz);
263
264         $arrayref_of_row_hashrefs = $dbh->selectall_arrayref(
265           "SELECT foo, bar FROM table WHERE baz BETWEEN ? AND ?",
266           { Slice => {} }, $baz_min, $baz_max);
267
268       The typical method call sequence for a non-"SELECT" statement is:
269
270         prepare,
271           execute,
272           execute,
273           execute.
274
275       for example:
276
277         $sth = $dbh->prepare("INSERT INTO table(foo,bar,baz) VALUES (?,?,?)");
278
279         while(<CSV>) {
280           chomp;
281           my ($foo,$bar,$baz) = split /,/;
282               $sth->execute( $foo, $bar, $baz );
283         }
284
285       The "do()" method is a wrapper of prepare and execute that can be
286       simpler for non repeated non-"SELECT" statements (or with drivers that
287       don't support placeholders):
288
289         $rows_affected = $dbh->do("UPDATE your_table SET foo = foo + 1");
290
291         $rows_affected = $dbh->do("DELETE FROM table WHERE foo = ?", undef, $foo);
292
293       To commit your changes to the database (when "AutoCommit" is off):
294
295         $dbh->commit;  # or call $dbh->rollback; to undo changes
296
297       Finally, when you have finished working with the data source, you
298       should "disconnect" from it:
299
300         $dbh->disconnect;
301
302   General Interface Rules & Caveats
303       The DBI does not have a concept of a "current session". Every session
304       has a handle object (i.e., a $dbh) returned from the "connect" method.
305       That handle object is used to invoke database related methods.
306
307       Most data is returned to the Perl script as strings. (Null values are
308       returned as "undef".)  This allows arbitrary precision numeric data to
309       be handled without loss of accuracy.  Beware that Perl may not preserve
310       the same accuracy when the string is used as a number.
311
312       Dates and times are returned as character strings in the current
313       default format of the corresponding database engine.  Time zone effects
314       are database/driver dependent.
315
316       Perl supports binary data in Perl strings, and the DBI will pass binary
317       data to and from the driver without change. It is up to the driver
318       implementors to decide how they wish to handle such binary data.
319
320       Perl supports two kinds of strings: Unicode (utf8 internally) and non-
321       Unicode (defaults to iso-8859-1 if forced to assume an encoding).
322       Drivers should accept both kinds of strings and, if required, convert
323       them to the character set of the database being used. Similarly, when
324       fetching from the database character data that isn't iso-8859-1 the
325       driver should convert it into utf8.
326
327       Multiple SQL statements may not be combined in a single statement
328       handle ($sth), although some databases and drivers do support this
329       (notably Sybase and SQL Server).
330
331       Non-sequential record reads are not supported in this version of the
332       DBI.  In other words, records can only be fetched in the order that the
333       database returned them, and once fetched they are forgotten.
334
335       Positioned updates and deletes are not directly supported by the DBI.
336       See the description of the "CursorName" attribute for an alternative.
337
338       Individual driver implementors are free to provide any private
339       functions and/or handle attributes that they feel are useful.  Private
340       driver functions can be invoked using the DBI "func()" method.  Private
341       driver attributes are accessed just like standard attributes.
342
343       Many methods have an optional "\%attr" parameter which can be used to
344       pass information to the driver implementing the method. Except where
345       specifically documented, the "\%attr" parameter can only be used to
346       pass driver specific hints. In general, you can ignore "\%attr"
347       parameters or pass it as "undef".
348
349   Naming Conventions and Name Space
350       The DBI package and all packages below it ("DBI::*") are reserved for
351       use by the DBI. Extensions and related modules use the "DBIx::"
352       namespace (see <http://www.perl.com/CPAN/modules/by-module/DBIx/>).
353       Package names beginning with "DBD::" are reserved for use by DBI
354       database drivers.  All environment variables used by the DBI or by
355       individual DBDs begin with ""DBI_"" or ""DBD_"".
356
357       The letter case used for attribute names is significant and plays an
358       important part in the portability of DBI scripts.  The case of the
359       attribute name is used to signify who defined the meaning of that name
360       and its values.
361
362         Case of name  Has a meaning defined by
363         ------------  ------------------------
364         UPPER_CASE    Standards, e.g.,  X/Open, ISO SQL92 etc (portable)
365         MixedCase     DBI API (portable), underscores are not used.
366         lower_case    Driver or database engine specific (non-portable)
367
368       It is of the utmost importance that Driver developers only use
369       lowercase attribute names when defining private attributes. Private
370       attribute names must be prefixed with the driver name or suitable
371       abbreviation (e.g., ""ora_"" for Oracle, ""ing_"" for Ingres, etc).
372
373   SQL - A Query Language
374       Most DBI drivers require applications to use a dialect of SQL
375       (Structured Query Language) to interact with the database engine.  The
376       "Standards Reference Information" section provides links to useful
377       information about SQL.
378
379       The DBI itself does not mandate or require any particular language to
380       be used; it is language independent. In ODBC terms, the DBI is in
381       "pass-thru" mode, although individual drivers might not be. The only
382       requirement is that queries and other statements must be expressed as a
383       single string of characters passed as the first argument to the
384       "prepare" or "do" methods.
385
386       For an interesting diversion on the real history of RDBMS and SQL, from
387       the people who made it happen, see:
388
389         http://www.mcjones.org/System_R/SQL_Reunion_95/sqlr95.html
390
391       Follow the "Full Contents" then "Intergalactic dataspeak" links for the
392       SQL history.
393
394   Placeholders and Bind Values
395       Some drivers support placeholders and bind values.  Placeholders, also
396       called parameter markers, are used to indicate values in a database
397       statement that will be supplied later, before the prepared statement is
398       executed.  For example, an application might use the following to
399       insert a row of data into the SALES table:
400
401         INSERT INTO sales (product_code, qty, price) VALUES (?, ?, ?)
402
403       or the following, to select the description for a product:
404
405         SELECT description FROM products WHERE product_code = ?
406
407       The "?" characters are the placeholders.  The association of actual
408       values with placeholders is known as binding, and the values are
409       referred to as bind values.  Note that the "?" is not enclosed in
410       quotation marks, even when the placeholder represents a string.
411
412       Some drivers also allow placeholders like ":"name and ":"N (e.g., ":1",
413       ":2", and so on) in addition to "?", but their use is not portable.
414
415       If the ":"N form of placeholder is supported by the driver you're
416       using, then you should be able to use either "bind_param" or "execute"
417       to bind values. Check your driver documentation.
418
419       Some drivers allow you to prevent the recognition of a placeholder by
420       placing a single backslash character ("\") immediately before it. The
421       driver will remove the backslash character and ignore the placeholder,
422       passing it unchanged to the backend. If the driver supports this then
423       "get_info"(9000) will return true.
424
425       With most drivers, placeholders can't be used for any element of a
426       statement that would prevent the database server from validating the
427       statement and creating a query execution plan for it. For example:
428
429         "SELECT name, age FROM ?"         # wrong (will probably fail)
430         "SELECT name, ?   FROM people"    # wrong (but may not 'fail')
431
432       Also, placeholders can only represent single scalar values.  For
433       example, the following statement won't work as expected for more than
434       one value:
435
436         "SELECT name, age FROM people WHERE name IN (?)"    # wrong
437         "SELECT name, age FROM people WHERE name IN (?,?)"  # two names
438
439       When using placeholders with the SQL "LIKE" qualifier, you must
440       remember that the placeholder substitutes for the whole string.  So you
441       should use ""... LIKE ? ..."" and include any wildcard characters in
442       the value that you bind to the placeholder.
443
444       NULL Values
445
446       Undefined values, or "undef", are used to indicate NULL values.  You
447       can insert and update columns with a NULL value as you would a non-NULL
448       value.  These examples insert and update the column "age" with a NULL
449       value:
450
451         $sth = $dbh->prepare(qq{
452           INSERT INTO people (fullname, age) VALUES (?, ?)
453         });
454         $sth->execute("Joe Bloggs", undef);
455
456         $sth = $dbh->prepare(qq{
457           UPDATE people SET age = ? WHERE fullname = ?
458         });
459         $sth->execute(undef, "Joe Bloggs");
460
461       However, care must be taken when trying to use NULL values in a "WHERE"
462       clause.  Consider:
463
464         SELECT fullname FROM people WHERE age = ?
465
466       Binding an "undef" (NULL) to the placeholder will not select rows which
467       have a NULL "age"!  At least for database engines that conform to the
468       SQL standard.  Refer to the SQL manual for your database engine or any
469       SQL book for the reasons for this.  To explicitly select NULLs you have
470       to say ""WHERE age IS NULL"".
471
472       A common issue is to have a code fragment handle a value that could be
473       either "defined" or "undef" (non-NULL or NULL) at runtime.  A simple
474       technique is to prepare the appropriate statement as needed, and
475       substitute the placeholder for non-NULL cases:
476
477         $sql_clause = defined $age? "age = ?" : "age IS NULL";
478         $sth = $dbh->prepare(qq{
479           SELECT fullname FROM people WHERE $sql_clause
480         });
481         $sth->execute(defined $age ? $age : ());
482
483       The following technique illustrates qualifying a "WHERE" clause with
484       several columns, whose associated values ("defined" or "undef") are in
485       a hash %h:
486
487         for my $col ("age", "phone", "email") {
488           if (defined $h{$col}) {
489             push @sql_qual, "$col = ?";
490             push @sql_bind, $h{$col};
491           }
492           else {
493             push @sql_qual, "$col IS NULL";
494           }
495         }
496         $sql_clause = join(" AND ", @sql_qual);
497         $sth = $dbh->prepare(qq{
498             SELECT fullname FROM people WHERE $sql_clause
499         });
500         $sth->execute(@sql_bind);
501
502       The techniques above call prepare for the SQL statement with each call
503       to execute.  Because calls to prepare() can be expensive, performance
504       can suffer when an application iterates many times over statements like
505       the above.
506
507       A better solution is a single "WHERE" clause that supports both NULL
508       and non-NULL comparisons.  Its SQL statement would need to be prepared
509       only once for all cases, thus improving performance.  Several examples
510       of "WHERE" clauses that support this are presented below.  But each
511       example lacks portability, robustness, or simplicity.  Whether an
512       example is supported on your database engine depends on what SQL
513       extensions it provides, and where it supports the "?"  placeholder in a
514       statement.
515
516         0)  age = ?
517         1)  NVL(age, xx) = NVL(?, xx)
518         2)  ISNULL(age, xx) = ISNULL(?, xx)
519         3)  DECODE(age, ?, 1, 0) = 1
520         4)  age = ? OR (age IS NULL AND ? IS NULL)
521         5)  age = ? OR (age IS NULL AND SP_ISNULL(?) = 1)
522         6)  age = ? OR (age IS NULL AND ? = 1)
523
524       Statements formed with the above "WHERE" clauses require execute
525       statements as follows.  The arguments are required, whether their
526       values are "defined" or "undef".
527
528         0,1,2,3)  $sth->execute($age);
529         4,5)      $sth->execute($age, $age);
530         6)        $sth->execute($age, defined($age) ? 0 : 1);
531
532       Example 0 should not work (as mentioned earlier), but may work on a few
533       database engines anyway (e.g. Sybase).  Example 0 is part of examples
534       4, 5, and 6, so if example 0 works, these other examples may work, even
535       if the engine does not properly support the right hand side of the "OR"
536       expression.
537
538       Examples 1 and 2 are not robust: they require that you provide a valid
539       column value xx (e.g. '~') which is not present in any row.  That means
540       you must have some notion of what data won't be stored in the column,
541       and expect clients to adhere to that.
542
543       Example 5 requires that you provide a stored procedure (SP_ISNULL in
544       this example) that acts as a function: it checks whether a value is
545       null, and returns 1 if it is, or 0 if not.
546
547       Example 6, the least simple, is probably the most portable, i.e., it
548       should work with most, if not all, database engines.
549
550       Here is a table that indicates which examples above are known to work
551       on various database engines:
552
553                          -----Examples------
554                          0  1  2  3  4  5  6
555                          -  -  -  -  -  -  -
556         Oracle 9         N  Y  N  Y  Y  ?  Y
557         Informix IDS 9   N  N  N  Y  N  Y  Y
558         MS SQL           N  N  Y  N  Y  ?  Y
559         Sybase           Y  N  N  N  N  N  Y
560         AnyData,DBM,CSV  Y  N  N  N  Y  Y* Y
561         SQLite 3.3       N  N  N  N  Y  N  N
562         MSAccess         N  N  N  N  Y  N  Y
563
564       * Works only because Example 0 works.
565
566       DBI provides a sample perl script that will test the examples above on
567       your database engine and tell you which ones work.  It is located in
568       the ex/ subdirectory of the DBI source distribution, or here:
569       <https://github.com/perl5-dbi/dbi/blob/master/ex/perl_dbi_nulls_test.pl>
570       Please use the script to help us fill-in and maintain this table.
571
572       Performance
573
574       Without using placeholders, the insert statement shown previously would
575       have to contain the literal values to be inserted and would have to be
576       re-prepared and re-executed for each row. With placeholders, the insert
577       statement only needs to be prepared once. The bind values for each row
578       can be given to the "execute" method each time it's called. By avoiding
579       the need to re-prepare the statement for each row, the application
580       typically runs many times faster. Here's an example:
581
582         my $sth = $dbh->prepare(q{
583           INSERT INTO sales (product_code, qty, price) VALUES (?, ?, ?)
584         }) or die $dbh->errstr;
585         while (<>) {
586             chomp;
587             my ($product_code, $qty, $price) = split /,/;
588             $sth->execute($product_code, $qty, $price) or die $dbh->errstr;
589         }
590         $dbh->commit or die $dbh->errstr;
591
592       See "execute" and "bind_param" for more details.
593
594       The "q{...}" style quoting used in this example avoids clashing with
595       quotes that may be used in the SQL statement. Use the double-quote like
596       "qq{...}" operator if you want to interpolate variables into the
597       string.  See "Quote and Quote-like Operators" in perlop for more
598       details.
599
600       See also the "bind_columns" method, which is used to associate Perl
601       variables with the output columns of a "SELECT" statement.
602

THE DBI PACKAGE AND CLASS

604       In this section, we cover the DBI class methods, utility functions, and
605       the dynamic attributes associated with generic DBI handles.
606
607   DBI Constants
608       Constants representing the values of the SQL standard types can be
609       imported individually by name, or all together by importing the special
610       ":sql_types" tag.
611
612       The names and values of all the defined SQL standard types can be
613       produced like this:
614
615         foreach (@{ $DBI::EXPORT_TAGS{sql_types} }) {
616           printf "%s=%d\n", $_, &{"DBI::$_"};
617         }
618
619       These constants are defined by SQL/CLI, ODBC or both.  "SQL_BIGINT" has
620       conflicting codes in SQL/CLI and ODBC, DBI uses the ODBC one.
621
622       See the "type_info", "type_info_all", and "bind_param" methods for
623       possible uses.
624
625       Note that just because the DBI defines a named constant for a given
626       data type doesn't mean that drivers will support that data type.
627
628   DBI Class Methods
629       The following methods are provided by the DBI class:
630
631       "parse_dsn"
632
633         ($scheme, $driver, $attr_string, $attr_hash, $driver_dsn) = DBI->parse_dsn($dsn)
634             or die "Can't parse DBI DSN '$dsn'";
635
636       Breaks apart a DBI Data Source Name (DSN) and returns the individual
637       parts. If $dsn doesn't contain a valid DSN then parse_dsn() returns an
638       empty list.
639
640       $scheme is the first part of the DSN and is currently always 'dbi'.
641       $driver is the driver name, possibly defaulted to $ENV{DBI_DRIVER}, and
642       may be undefined.  $attr_string is the contents of the optional
643       attribute string, which may be undefined.  If $attr_string is not empty
644       then $attr_hash is a reference to a hash containing the parsed
645       attribute names and values.  $driver_dsn is the last part of the DBI
646       DSN string. For example:
647
648         ($scheme, $driver, $attr_string, $attr_hash, $driver_dsn)
649             = DBI->parse_dsn("dbi:MyDriver(RaiseError=>1):db=test;port=42");
650         $scheme      = 'dbi';
651         $driver      = 'MyDriver';
652         $attr_string = 'RaiseError=>1';
653         $attr_hash   = { 'RaiseError' => '1' };
654         $driver_dsn  = 'db=test;port=42';
655
656       The parse_dsn() method was added in DBI 1.43.
657
658       "connect"
659
660         $dbh = DBI->connect($data_source, $username, $password)
661                   or die $DBI::errstr;
662         $dbh = DBI->connect($data_source, $username, $password, \%attr)
663                   or die $DBI::errstr;
664
665       Establishes a database connection, or session, to the requested
666       $data_source.  Returns a database handle object if the connection
667       succeeds. Use "$dbh->disconnect" to terminate the connection.
668
669       If the connect fails (see below), it returns "undef" and sets both
670       $DBI::err and $DBI::errstr. (It does not explicitly set $!.) You should
671       generally test the return status of "connect" and "print $DBI::errstr"
672       if it has failed.
673
674       Multiple simultaneous connections to multiple databases through
675       multiple drivers can be made via the DBI. Simply make one "connect"
676       call for each database and keep a copy of each returned database
677       handle.
678
679       The $data_source value must begin with ""dbi:"driver_name":"".  The
680       driver_name specifies the driver that will be used to make the
681       connection. (Letter case is significant.)
682
683       As a convenience, if the $data_source parameter is undefined or empty,
684       the DBI will substitute the value of the environment variable
685       "DBI_DSN".  If just the driver_name part is empty (i.e., the
686       $data_source prefix is ""dbi::""), the environment variable
687       "DBI_DRIVER" is used. If neither variable is set, then "connect" dies.
688
689       Examples of $data_source values are:
690
691         dbi:DriverName:database_name
692         dbi:DriverName:database_name@hostname:port
693         dbi:DriverName:database=database_name;host=hostname;port=port
694
695       There is no standard for the text following the driver name. Each
696       driver is free to use whatever syntax it wants. The only requirement
697       the DBI makes is that all the information is supplied in a single
698       string.  You must consult the documentation for the drivers you are
699       using for a description of the syntax they require.
700
701       It is recommended that drivers support the ODBC style, shown in the
702       last example above. It is also recommended that they support the three
703       common names '"host"', '"port"', and '"database"' (plus '"db"' as an
704       alias for "database"). This simplifies automatic construction of basic
705       DSNs: "dbi:$driver:database=$db;host=$host;port=$port".  Drivers should
706       aim to 'do something reasonable' when given a DSN in this form, but if
707       any part is meaningless for that driver (such as 'port' for Informix)
708       it should generate an error if that part is not empty.
709
710       If the environment variable "DBI_AUTOPROXY" is defined (and the driver
711       in $data_source is not ""Proxy"") then the connect request will
712       automatically be changed to:
713
714         $ENV{DBI_AUTOPROXY};dsn=$data_source
715
716       "DBI_AUTOPROXY" is typically set as
717       ""dbi:Proxy:hostname=...;port=..."".  If $ENV{DBI_AUTOPROXY} doesn't
718       begin with '"dbi:"' then "dbi:Proxy:" will be prepended to it first.
719       See the DBD::Proxy documentation for more details.
720
721       If $username or $password are undefined (rather than just empty), then
722       the DBI will substitute the values of the "DBI_USER" and "DBI_PASS"
723       environment variables, respectively.  The DBI will warn if the
724       environment variables are not defined.  However, the everyday use of
725       these environment variables is not recommended for security reasons.
726       The mechanism is primarily intended to simplify testing.  See below for
727       alternative way to specify the username and password.
728
729       "DBI->connect" automatically installs the driver if it has not been
730       installed yet. Driver installation either returns a valid driver
731       handle, or it dies with an error message that includes the string
732       ""install_driver"" and the underlying problem. So "DBI->connect" will
733       die on a driver installation failure and will only return "undef" on a
734       connect failure, in which case $DBI::errstr will hold the error
735       message.  Use "eval" if you need to catch the ""install_driver"" error.
736
737       The $data_source argument (with the ""dbi:...:"" prefix removed) and
738       the $username and $password arguments are then passed to the driver for
739       processing. The DBI does not define any interpretation for the contents
740       of these fields.  The driver is free to interpret the $data_source,
741       $username, and $password fields in any way, and supply whatever
742       defaults are appropriate for the engine being accessed.  (Oracle, for
743       example, uses the ORACLE_SID and TWO_TASK environment variables if no
744       $data_source is specified.)
745
746       The "AutoCommit" and "PrintError" attributes for each connection
747       default to "on". (See "AutoCommit" and "PrintError" for more
748       information.)  However, it is strongly recommended that you explicitly
749       define "AutoCommit" rather than rely on the default. The "PrintWarn"
750       attribute defaults to true.  The "RaiseWarn" attribute defaults to
751       false.
752
753       The "\%attr" parameter can be used to alter the default settings of
754       "PrintError", "RaiseError", "AutoCommit", and other attributes. For
755       example:
756
757         $dbh = DBI->connect($data_source, $user, $pass, {
758               PrintError => 0,
759               AutoCommit => 0
760         });
761
762       The username and password can also be specified using the attributes
763       "Username" and "Password", in which case they take precedence over the
764       $username and $password parameters.
765
766       You can also define connection attribute values within the $data_source
767       parameter. For example:
768
769         dbi:DriverName(PrintWarn=>0,PrintError=>0,Taint=>1):...
770
771       Individual attributes values specified in this way take precedence over
772       any conflicting values specified via the "\%attr" parameter to
773       "connect".
774
775       The "dbi_connect_method" attribute can be used to specify which driver
776       method should be called to establish the connection. The only useful
777       values are 'connect', 'connect_cached', or some specialized case like
778       'Apache::DBI::connect' (which is automatically the default when running
779       within Apache).
780
781       Where possible, each session ($dbh) is independent from the
782       transactions in other sessions. This is useful when you need to hold
783       cursors open across transactions--for example, if you use one session
784       for your long lifespan cursors (typically read-only) and another for
785       your short update transactions.
786
787       For compatibility with old DBI scripts, the driver can be specified by
788       passing its name as the fourth argument to "connect" (instead of
789       "\%attr"):
790
791         $dbh = DBI->connect($data_source, $user, $pass, $driver);
792
793       In this "old-style" form of "connect", the $data_source should not
794       start with ""dbi:driver_name:"". (If it does, the embedded driver_name
795       will be ignored). Also note that in this older form of "connect", the
796       "$dbh->{AutoCommit}" attribute is undefined, the "$dbh->{PrintError}"
797       attribute is off, and the old "DBI_DBNAME" environment variable is
798       checked if "DBI_DSN" is not defined. Beware that this "old-style"
799       "connect" will soon be withdrawn in a future version of DBI.
800
801       "connect_cached"
802
803         $dbh = DBI->connect_cached($data_source, $username, $password)
804                   or die $DBI::errstr;
805         $dbh = DBI->connect_cached($data_source, $username, $password, \%attr)
806                   or die $DBI::errstr;
807
808       "connect_cached" is like "connect", except that the database handle
809       returned is also stored in a hash associated with the given parameters.
810       If another call is made to "connect_cached" with the same parameter
811       values, then the corresponding cached $dbh will be returned if it is
812       still valid.  The cached database handle is replaced with a new
813       connection if it has been disconnected or if the "ping" method fails.
814
815       Note that the behaviour of this method differs in several respects from
816       the behaviour of persistent connections implemented by Apache::DBI.
817       However, if Apache::DBI is loaded then "connect_cached" will use it.
818
819       Caching connections can be useful in some applications, but it can also
820       cause problems, such as too many connections, and so should be used
821       with care. In particular, avoid changing the attributes of a database
822       handle created via connect_cached() because it will affect other code
823       that may be using the same handle. When connect_cached() returns a
824       handle the attributes will be reset to their initial values.  This can
825       cause problems, especially with the "AutoCommit" attribute.
826
827       Also, to ensure that the attributes passed are always the same, avoid
828       passing references inline. For example, the "Callbacks" attribute is
829       specified as a hash reference. Be sure to declare it external to the
830       call to connect_cached(), such that the hash reference is not re-
831       created on every call. A package-level lexical works well:
832
833         package MyDBH;
834         my $cb = {
835             'connect_cached.reused' => sub { delete $_[4]->{AutoCommit} },
836         };
837
838         sub dbh {
839             DBI->connect_cached( $dsn, $username, $auth, { Callbacks => $cb });
840         }
841
842       Where multiple separate parts of a program are using connect_cached()
843       to connect to the same database with the same (initial) attributes it
844       is a good idea to add a private attribute to the connect_cached() call
845       to effectively limit the scope of the caching. For example:
846
847         DBI->connect_cached(..., { private_foo_cachekey => "Bar", ... });
848
849       Handles returned from that connect_cached() call will only be returned
850       by other connect_cached() call elsewhere in the code if those other
851       calls also pass in the same attribute values, including the private
852       one.  (I've used "private_foo_cachekey" here as an example, you can use
853       any attribute name with a "private_" prefix.)
854
855       Taking that one step further, you can limit a particular
856       connect_cached() call to return handles unique to that one place in the
857       code by setting the private attribute to a unique value for that place:
858
859         DBI->connect_cached(..., { private_foo_cachekey => __FILE__.__LINE__, ... });
860
861       By using a private attribute you still get connection caching for the
862       individual calls to connect_cached() but, by making separate database
863       connections for separate parts of the code, the database handles are
864       isolated from any attribute changes made to other handles.
865
866       The cache can be accessed (and cleared) via the "CachedKids" attribute:
867
868         my $CachedKids_hashref = $dbh->{Driver}->{CachedKids};
869         %$CachedKids_hashref = () if $CachedKids_hashref;
870
871       "available_drivers"
872
873         @ary = DBI->available_drivers;
874         @ary = DBI->available_drivers($quiet);
875
876       Returns a list of all available drivers by searching for "DBD::*"
877       modules through the directories in @INC. By default, a warning is given
878       if some drivers are hidden by others of the same name in earlier
879       directories. Passing a true value for $quiet will inhibit the warning.
880
881       "installed_drivers"
882
883         %drivers = DBI->installed_drivers();
884
885       Returns a list of driver name and driver handle pairs for all drivers
886       'installed' (loaded) into the current process.  The driver name does
887       not include the 'DBD::' prefix.
888
889       To get a list of all drivers available in your perl installation you
890       can use "available_drivers".
891
892       Added in DBI 1.49.
893
894       "installed_versions"
895
896         DBI->installed_versions;
897         @ary  = DBI->installed_versions;
898         $hash = DBI->installed_versions;
899
900       Calls available_drivers() and attempts to load each of them in turn
901       using install_driver().  For each load that succeeds the driver name
902       and version number are added to a hash. When running under
903       DBI::PurePerl drivers which appear not be pure-perl are ignored.
904
905       When called in array context the list of successfully loaded drivers is
906       returned (without the 'DBD::' prefix).
907
908       When called in scalar context an extra entry for the "DBI" is added
909       (and "DBI::PurePerl" if appropriate) and a reference to the hash is
910       returned.
911
912       When called in a void context the installed_versions() method will
913       print out a formatted list of the hash contents, one per line, along
914       with some other information about the DBI version and OS.
915
916       Due to the potentially high memory cost and unknown risks of loading in
917       an unknown number of drivers that just happen to be installed on the
918       system, this method is not recommended for general use.  Use
919       available_drivers() instead.
920
921       The installed_versions() method is primarily intended as a quick way to
922       see from the command line what's installed. For example:
923
924         perl -MDBI -e 'DBI->installed_versions'
925
926       The installed_versions() method was added in DBI 1.38.
927
928       "data_sources"
929
930         @ary = DBI->data_sources($driver);
931         @ary = DBI->data_sources($driver, \%attr);
932
933       Returns a list of data sources (databases) available via the named
934       driver.  If $driver is empty or "undef", then the value of the
935       "DBI_DRIVER" environment variable is used.
936
937       The driver will be loaded if it hasn't been already. Note that if the
938       driver loading fails then data_sources() dies with an error message
939       that includes the string ""install_driver"" and the underlying problem.
940
941       Data sources are returned in a form suitable for passing to the
942       "connect" method (that is, they will include the ""dbi:$driver:""
943       prefix).
944
945       Note that many drivers have no way of knowing what data sources might
946       be available for it. These drivers return an empty or incomplete list
947       or may require driver-specific attributes.
948
949       There is also a data_sources() method defined for database handles.
950
951       "trace"
952
953         DBI->trace($trace_setting)
954         DBI->trace($trace_setting, $trace_filename)
955         DBI->trace($trace_setting, $trace_filehandle)
956         $trace_setting = DBI->trace;
957
958       The "DBI->trace" method sets the global default trace settings and
959       returns the previous trace settings. It can also be used to change
960       where the trace output is sent.
961
962       There's a similar method, "$h->trace", which sets the trace settings
963       for the specific handle it's called on.
964
965       See the "TRACING" section for full details about the DBI's powerful
966       tracing facilities.
967
968       "visit_handles"
969
970         DBI->visit_handles( $coderef );
971         DBI->visit_handles( $coderef, $info );
972
973       Where $coderef is a reference to a subroutine and $info is an arbitrary
974       value which, if undefined, defaults to a reference to an empty hash.
975       Returns $info.
976
977       For each installed driver handle, if any, $coderef is invoked as:
978
979         $coderef->($driver_handle, $info);
980
981       If the execution of $coderef returns a true value then
982       "visit_child_handles" is called on that child handle and passed the
983       returned value as $info.
984
985       For example:
986
987         my $info = $dbh->{Driver}->visit_child_handles(sub {
988             my ($h, $info) = @_;
989             ++$info->{ $h->{Type} }; # count types of handles (dr/db/st)
990             return $info; # visit kids
991         });
992
993       See also "visit_child_handles".
994
995   DBI Utility Functions
996       In addition to the DBI methods listed in the previous section, the DBI
997       package also provides several utility functions.
998
999       These can be imported into your code by listing them in the "use"
1000       statement. For example:
1001
1002         use DBI qw(neat data_diff);
1003
1004       Alternatively, all these utility functions (except hash) can be
1005       imported using the ":utils" import tag. For example:
1006
1007         use DBI qw(:utils);
1008
1009       "data_string_desc"
1010
1011         $description = data_string_desc($string);
1012
1013       Returns an informal description of the string. For example:
1014
1015         UTF8 off, ASCII, 42 characters 42 bytes
1016         UTF8 off, non-ASCII, 42 characters 42 bytes
1017         UTF8 on, non-ASCII, 4 characters 6 bytes
1018         UTF8 on but INVALID encoding, non-ASCII, 4 characters 6 bytes
1019         UTF8 off, undef
1020
1021       The initial "UTF8" on/off refers to Perl's internal SvUTF8 flag.  If
1022       $string has the SvUTF8 flag set but the sequence of bytes it contains
1023       are not a valid UTF-8 encoding then data_string_desc() will report
1024       "UTF8 on but INVALID encoding".
1025
1026       The "ASCII" vs "non-ASCII" portion shows "ASCII" if all the characters
1027       in the string are ASCII (have code points <= 127).
1028
1029       The data_string_desc() function was added in DBI 1.46.
1030
1031       "data_string_diff"
1032
1033         $diff = data_string_diff($a, $b);
1034
1035       Returns an informal description of the first character difference
1036       between the strings. If both $a and $b contain the same sequence of
1037       characters then data_string_diff() returns an empty string.  For
1038       example:
1039
1040        Params a & b     Result
1041        ------------     ------
1042        'aaa', 'aaa'     ''
1043        'aaa', 'abc'     'Strings differ at index 2: a[2]=a, b[2]=b'
1044        'aaa', undef     'String b is undef, string a has 3 characters'
1045        'aaa', 'aa'      'String b truncated after 2 characters'
1046
1047       Unicode characters are reported in "\x{XXXX}" format. Unicode code
1048       points in the range U+0800 to U+08FF are unassigned and most likely to
1049       occur due to double-encoding. Characters in this range are reported as
1050       "\x{08XX}='C'" where "C" is the corresponding latin-1 character.
1051
1052       The data_string_diff() function only considers logical characters and
1053       not the underlying encoding. See "data_diff" for an alternative.
1054
1055       The data_string_diff() function was added in DBI 1.46.
1056
1057       "data_diff"
1058
1059         $diff = data_diff($a, $b);
1060         $diff = data_diff($a, $b, $logical);
1061
1062       Returns an informal description of the difference between two strings.
1063       It calls "data_string_desc" and "data_string_diff" and returns the
1064       combined results as a multi-line string.
1065
1066       For example, "data_diff("abc", "ab\x{263a}")" will return:
1067
1068         a: UTF8 off, ASCII, 3 characters 3 bytes
1069         b: UTF8 on, non-ASCII, 3 characters 5 bytes
1070         Strings differ at index 2: a[2]=c, b[2]=\x{263A}
1071
1072       If $a and $b are identical in both the characters they contain and
1073       their physical encoding then data_diff() returns an empty string.  If
1074       $logical is true then physical encoding differences are ignored (but
1075       are still reported if there is a difference in the characters).
1076
1077       The data_diff() function was added in DBI 1.46.
1078
1079       "neat"
1080
1081         $str = neat($value);
1082         $str = neat($value, $maxlen);
1083
1084       Return a string containing a neat (and tidy) representation of the
1085       supplied value.
1086
1087       Strings will be quoted, although internal quotes will not be escaped.
1088       Values known to be numeric will be unquoted. Undefined (NULL) values
1089       will be shown as "undef" (without quotes).
1090
1091       If the string is flagged internally as utf8 then double quotes will be
1092       used, otherwise single quotes are used and unprintable characters will
1093       be replaced by dot (.).
1094
1095       For result strings longer than $maxlen the result string will be
1096       truncated to "$maxlen-4" and ""...'"" will be appended.  If $maxlen is
1097       0 or "undef", it defaults to $DBI::neat_maxlen which, in turn, defaults
1098       to 400.
1099
1100       This function is designed to format values for human consumption.  It
1101       is used internally by the DBI for "trace" output. It should typically
1102       not be used for formatting values for database use.  (See also
1103       "quote".)
1104
1105       "neat_list"
1106
1107         $str = neat_list(\@listref, $maxlen, $field_sep);
1108
1109       Calls "neat" on each element of the list and returns a string
1110       containing the results joined with $field_sep. $field_sep defaults to
1111       ", ".
1112
1113       "looks_like_number"
1114
1115         @bool = looks_like_number(@array);
1116
1117       Returns true for each element that looks like a number.  Returns false
1118       for each element that does not look like a number.  Returns "undef" for
1119       each element that is undefined or empty.
1120
1121       "hash"
1122
1123         $hash_value = DBI::hash($buffer, $type);
1124
1125       Return a 32-bit integer 'hash' value corresponding to the contents of
1126       $buffer.  The $type parameter selects which kind of hash algorithm
1127       should be used.
1128
1129       For the technically curious, type 0 (which is the default if $type
1130       isn't specified) is based on the Perl 5.1 hash except that the value is
1131       forced to be negative (for obscure historical reasons).  Type 1 is the
1132       better "Fowler / Noll / Vo" (FNV) hash. See
1133       <http://www.isthe.com/chongo/tech/comp/fnv/> for more information.
1134       Both types are implemented in C and are very fast.
1135
1136       This function doesn't have much to do with databases, except that it
1137       can sometimes be handy to store such values in a database.  It also
1138       doesn't have much to do with perl hashes, like %foo.
1139
1140       "sql_type_cast"
1141
1142         $sts = DBI::sql_type_cast($sv, $sql_type, $flags);
1143
1144       sql_type_cast attempts to cast $sv to the SQL type (see DBI Constants)
1145       specified in $sql_type. At present only the SQL types "SQL_INTEGER",
1146       "SQL_DOUBLE" and "SQL_NUMERIC" are supported.
1147
1148       For "SQL_INTEGER" the effect is similar to using the value in an
1149       expression that requires an integer. It gives the perl scalar an
1150       'integer aspect'.  (Technically the value gains an IV, or possibly a UV
1151       or NV if the value is too large for an IV.)
1152
1153       For "SQL_DOUBLE" the effect is similar to using the value in an
1154       expression that requires a general numeric value. It gives the perl
1155       scalar a 'numeric aspect'.  (Technically the value gains an NV.)
1156
1157       "SQL_NUMERIC" is similar to "SQL_INTEGER" or "SQL_DOUBLE" but more
1158       general and more cautious.  It will look at the string first and if it
1159       looks like an integer (that will fit in an IV or UV) it will act like
1160       "SQL_INTEGER", if it looks like a floating point value it will act like
1161       "SQL_DOUBLE", if it looks like neither then it will do nothing - and
1162       thereby avoid the warnings that would be generated by "SQL_INTEGER" and
1163       "SQL_DOUBLE" when given non-numeric data.
1164
1165       $flags may be:
1166
1167       "DBIstcf_DISCARD_STRING"
1168           If this flag is specified then when the driver successfully casts
1169           the bound perl scalar to a non-string type then the string portion
1170           of the scalar will be discarded.
1171
1172       "DBIstcf_STRICT"
1173           If $sv cannot be cast to the requested $sql_type then by default it
1174           is left untouched and no error is generated. If you specify
1175           "DBIstcf_STRICT" and the cast fails, this will generate an error.
1176
1177       The returned $sts value is:
1178
1179         -2 sql_type is not handled
1180         -1 sv is undef so unchanged
1181          0 sv could not be cast cleanly and DBIstcf_STRICT was used
1182          1 sv could not be cast and DBIstcf_STRICT was not used
1183          2 sv was cast successfully
1184
1185       This method is exported by the :utils tag and was introduced in DBI
1186       1.611.
1187
1188   DBI Dynamic Attributes
1189       Dynamic attributes are always associated with the last handle used
1190       (that handle is represented by $h in the descriptions below).
1191
1192       Where an attribute is equivalent to a method call, then refer to the
1193       method call for all related documentation.
1194
1195       Warning: these attributes are provided as a convenience but they do
1196       have limitations. Specifically, they have a short lifespan: because
1197       they are associated with the last handle used, they should only be used
1198       immediately after calling the method that "sets" them.  If in any
1199       doubt, use the corresponding method call.
1200
1201       $DBI::err
1202
1203       Equivalent to "$h->err".
1204
1205       $DBI::errstr
1206
1207       Equivalent to "$h->errstr".
1208
1209       $DBI::state
1210
1211       Equivalent to "$h->state".
1212
1213       $DBI::rows
1214
1215       Equivalent to "$h->rows". Please refer to the documentation for the
1216       "rows" method.
1217
1218       $DBI::lasth
1219
1220       Returns the DBI object handle used for the most recent DBI method call.
1221       If the last DBI method call was a DESTROY then $DBI::lasth will return
1222       the handle of the parent of the destroyed handle, if there is one.
1223

METHODS COMMON TO ALL HANDLES

1225       The following methods can be used by all types of DBI handles.
1226
1227       "err"
1228
1229         $rv = $h->err;
1230
1231       Returns the native database engine error code from the last driver
1232       method called. The code is typically an integer but you should not
1233       assume that.
1234
1235       The DBI resets $h->err to undef before almost all DBI method calls, so
1236       the value only has a short lifespan. Also, for most drivers, the
1237       statement handles share the same error variable as the parent database
1238       handle, so calling a method on one handle may reset the error on the
1239       related handles.
1240
1241       (Methods which don't reset err before being called include err() and
1242       errstr(), obviously, state(), rows(), func(), trace(), trace_msg(),
1243       ping(), and the tied hash attribute FETCH() and STORE() methods.)
1244
1245       If you need to test for specific error conditions and have your program
1246       be portable to different database engines, then you'll need to
1247       determine what the corresponding error codes are for all those engines
1248       and test for all of them.
1249
1250       The DBI uses the value of $DBI::stderr as the "err" value for internal
1251       errors.  Drivers should also do likewise.  The default value for
1252       $DBI::stderr is 2000000000.
1253
1254       A driver may return 0 from err() to indicate a warning condition after
1255       a method call. Similarly, a driver may return an empty string to
1256       indicate a 'success with information' condition. In both these cases
1257       the value is false but not undef. The errstr() and state() methods may
1258       be used to retrieve extra information in these cases.
1259
1260       See "set_err" for more information.
1261
1262       "errstr"
1263
1264         $str = $h->errstr;
1265
1266       Returns the native database engine error message from the last DBI
1267       method called. This has the same lifespan issues as the "err" method
1268       described above.
1269
1270       The returned string may contain multiple messages separated by newline
1271       characters.
1272
1273       The errstr() method should not be used to test for errors, use err()
1274       for that, because drivers may return 'success with information' or
1275       warning messages via errstr() for methods that have not 'failed'.
1276
1277       See "set_err" for more information.
1278
1279       "state"
1280
1281         $str = $h->state;
1282
1283       Returns a state code in the standard SQLSTATE five character format.
1284       Note that the specific success code 00000 is translated to any empty
1285       string (false). If the driver does not support SQLSTATE (and most
1286       don't), then state() will return "S1000" (General Error) for all
1287       errors.
1288
1289       The driver is free to return any value via "state", e.g., warning
1290       codes, even if it has not declared an error by returning a true value
1291       via the "err" method described above.
1292
1293       The state() method should not be used to test for errors, use err() for
1294       that, because drivers may return a 'success with information' or
1295       warning state code via state() for methods that have not 'failed'.
1296
1297       "set_err"
1298
1299         $rv = $h->set_err($err, $errstr);
1300         $rv = $h->set_err($err, $errstr, $state);
1301         $rv = $h->set_err($err, $errstr, $state, $method);
1302         $rv = $h->set_err($err, $errstr, $state, $method, $rv);
1303
1304       Set the "err", "errstr", and "state" values for the handle.  This
1305       method is typically only used by DBI drivers and DBI subclasses.
1306
1307       If the "HandleSetErr" attribute holds a reference to a subroutine it is
1308       called first. The subroutine can alter the $err, $errstr, $state, and
1309       $method values. See "HandleSetErr" for full details.  If the subroutine
1310       returns a true value then the handle "err", "errstr", and "state"
1311       values are not altered and set_err() returns an empty list (it normally
1312       returns $rv which defaults to undef, see below).
1313
1314       Setting "err" to a true value indicates an error and will trigger the
1315       normal DBI error handling mechanisms, such as "RaiseError" and
1316       "HandleError", if they are enabled, when execution returns from the DBI
1317       back to the application.
1318
1319       Setting "err" to "" indicates an 'information' state, and setting it to
1320       "0" indicates a 'warning' state. Setting "err" to "undef" also sets
1321       "errstr" to undef, and "state" to "", irrespective of the values of the
1322       $errstr and $state parameters.
1323
1324       The $method parameter provides an alternate method name for the
1325       "RaiseError"/"PrintError"/"RaiseWarn"/"PrintWarn" error string instead
1326       of the fairly unhelpful '"set_err"'.
1327
1328       The "set_err" method normally returns undef.  The $rv parameter
1329       provides an alternate return value.
1330
1331       Some special rules apply if the "err" or "errstr" values for the handle
1332       are already set...
1333
1334       If "errstr" is true then: "" [err was %s now %s]"" is appended if $err
1335       is true and "err" is already true and the new err value differs from
1336       the original one. Similarly "" [state was %s now %s]"" is appended if
1337       $state is true and "state" is already true and the new state value
1338       differs from the original one. Finally ""\n"" and the new $errstr are
1339       appended if $errstr differs from the existing errstr value. Obviously
1340       the %s's above are replaced by the corresponding values.
1341
1342       The handle "err" value is set to $err if: $err is true; or handle "err"
1343       value is undef; or $err is defined and the length is greater than the
1344       handle "err" length. The effect is that an 'information' state only
1345       overrides undef; a 'warning' overrides undef or 'information', and an
1346       'error' state overrides anything.
1347
1348       The handle "state" value is set to $state if $state is true and the
1349       handle "err" value was set (by the rules above).
1350
1351       Support for warning and information states was added in DBI 1.41.
1352
1353       "trace"
1354
1355         $h->trace($trace_settings);
1356         $h->trace($trace_settings, $trace_filename);
1357         $trace_settings = $h->trace;
1358
1359       The trace() method is used to alter the trace settings for a handle
1360       (and any future children of that handle).  It can also be used to
1361       change where the trace output is sent.
1362
1363       There's a similar method, "DBI->trace", which sets the global default
1364       trace settings.
1365
1366       See the "TRACING" section for full details about the DBI's powerful
1367       tracing facilities.
1368
1369       "trace_msg"
1370
1371         $h->trace_msg($message_text);
1372         $h->trace_msg($message_text, $min_level);
1373
1374       Writes $message_text to the trace file if the trace level is greater
1375       than or equal to $min_level (which defaults to 1).  Can also be called
1376       as "DBI->trace_msg($msg)".
1377
1378       See "TRACING" for more details.
1379
1380       "func"
1381
1382         $h->func(@func_arguments, $func_name) or die ...;
1383
1384       The "func" method can be used to call private non-standard and non-
1385       portable methods implemented by the driver. Note that the function name
1386       is given as the last argument.
1387
1388       It's also important to note that the func() method does not clear a
1389       previous error ($DBI::err etc.) and it does not trigger automatic error
1390       detection (RaiseError etc.) so you must check the return status and/or
1391       $h->err to detect errors.
1392
1393       (This method is not directly related to calling stored procedures.
1394       Calling stored procedures is currently not defined by the DBI.  Some
1395       drivers, such as DBD::Oracle, support it in non-portable ways.  See
1396       driver documentation for more details.)
1397
1398       See also install_method() in DBI::DBD for how you can avoid needing to
1399       use func() and gain direct access to driver-private methods.
1400
1401       "can"
1402
1403         $is_implemented = $h->can($method_name);
1404
1405       Returns true if $method_name is implemented by the driver or a default
1406       method is provided by the DBI's driver base class.  It returns false
1407       where a driver hasn't implemented a method and the default method is
1408       provided by the DBI's driver base class is just an empty stub.
1409
1410       "parse_trace_flags"
1411
1412         $trace_settings_integer = $h->parse_trace_flags($trace_settings);
1413
1414       Parses a string containing trace settings and returns the corresponding
1415       integer value used internally by the DBI and drivers.
1416
1417       The $trace_settings argument is a string containing a trace level
1418       between 0 and 15 and/or trace flag names separated by vertical bar
1419       (""|"") or comma ("","") characters. For example: "SQL|3|foo".
1420
1421       It uses the parse_trace_flag() method, described below, to process the
1422       individual trace flag names.
1423
1424       The parse_trace_flags() method was added in DBI 1.42.
1425
1426       "parse_trace_flag"
1427
1428         $bit_flag = $h->parse_trace_flag($trace_flag_name);
1429
1430       Returns the bit flag corresponding to the trace flag name in
1431       $trace_flag_name.  Drivers are expected to override this method and
1432       check if $trace_flag_name is a driver specific trace flags and, if not,
1433       then call the DBI's default parse_trace_flag().
1434
1435       The parse_trace_flag() method was added in DBI 1.42.
1436
1437       "private_attribute_info"
1438
1439         $hash_ref = $h->private_attribute_info();
1440
1441       Returns a reference to a hash whose keys are the names of driver-
1442       private handle attributes available for the kind of handle (driver,
1443       database, statement) that the method was called on.
1444
1445       For example, the return value when called with a DBD::Sybase $dbh could
1446       look like this:
1447
1448         {
1449             syb_dynamic_supported => undef,
1450             syb_oc_version => undef,
1451             syb_server_version => undef,
1452             syb_server_version_string => undef,
1453         }
1454
1455       and when called with a DBD::Sybase $sth they could look like this:
1456
1457         {
1458             syb_types => undef,
1459             syb_proc_status => undef,
1460             syb_result_type => undef,
1461         }
1462
1463       The values should be undef. Meanings may be assigned to particular
1464       values in future.
1465
1466       "swap_inner_handle"
1467
1468         $rc = $h1->swap_inner_handle( $h2 );
1469         $rc = $h1->swap_inner_handle( $h2, $allow_reparent );
1470
1471       Brain transplants for handles. You don't need to know about this unless
1472       you want to become a handle surgeon.
1473
1474       A DBI handle is a reference to a tied hash. A tied hash has an inner
1475       hash that actually holds the contents.  The swap_inner_handle() method
1476       swaps the inner hashes between two handles. The $h1 and $h2 handles
1477       still point to the same tied hashes, but what those hashes are tied to
1478       has been swapped.  In effect $h1 becomes $h2 and vice-versa. This is
1479       powerful stuff, expect problems. Use with care.
1480
1481       As a small safety measure, the two handles, $h1 and $h2, have to share
1482       the same parent unless $allow_reparent is true.
1483
1484       The swap_inner_handle() method was added in DBI 1.44.
1485
1486       Here's a quick kind of 'diagram' as a worked example to help think
1487       about what's happening:
1488
1489           Original state:
1490                   dbh1o -> dbh1i
1491                   sthAo -> sthAi(dbh1i)
1492                   dbh2o -> dbh2i
1493
1494           swap_inner_handle dbh1o with dbh2o:
1495                   dbh2o -> dbh1i
1496                   sthAo -> sthAi(dbh1i)
1497                   dbh1o -> dbh2i
1498
1499           create new sth from dbh1o:
1500                   dbh2o -> dbh1i
1501                   sthAo -> sthAi(dbh1i)
1502                   dbh1o -> dbh2i
1503                   sthBo -> sthBi(dbh2i)
1504
1505           swap_inner_handle sthAo with sthBo:
1506                   dbh2o -> dbh1i
1507                   sthBo -> sthAi(dbh1i)
1508                   dbh1o -> dbh2i
1509                   sthAo -> sthBi(dbh2i)
1510
1511       "visit_child_handles"
1512
1513         $h->visit_child_handles( $coderef );
1514         $h->visit_child_handles( $coderef, $info );
1515
1516       Where $coderef is a reference to a subroutine and $info is an arbitrary
1517       value which, if undefined, defaults to a reference to an empty hash.
1518       Returns $info.
1519
1520       For each child handle of $h, if any, $coderef is invoked as:
1521
1522         $coderef->($child_handle, $info);
1523
1524       If the execution of $coderef returns a true value then
1525       "visit_child_handles" is called on that child handle and passed the
1526       returned value as $info.
1527
1528       For example:
1529
1530         # count database connections with names (DSN) matching a pattern
1531         my $connections = 0;
1532         $dbh->{Driver}->visit_child_handles(sub {
1533             my ($h, $info) = @_;
1534             ++$connections if $h->{Name} =~ /foo/;
1535             return 0; # don't visit kids
1536         })
1537
1538       See also "visit_handles".
1539

ATTRIBUTES COMMON TO ALL HANDLES

1541       These attributes are common to all types of DBI handles.
1542
1543       Some attributes are inherited by child handles. That is, the value of
1544       an inherited attribute in a newly created statement handle is the same
1545       as the value in the parent database handle. Changes to attributes in
1546       the new statement handle do not affect the parent database handle and
1547       changes to the database handle do not affect existing statement
1548       handles, only future ones.
1549
1550       Attempting to set or get the value of an unknown attribute generates a
1551       warning, except for private driver specific attributes (which all have
1552       names starting with a lowercase letter).
1553
1554       Example:
1555
1556         $h->{AttributeName} = ...;    # set/write
1557         ... = $h->{AttributeName};    # get/read
1558
1559       "Warn"
1560
1561       Type: boolean, inherited
1562
1563       The "Warn" attribute enables useful warnings for certain bad practices.
1564       It is enabled by default and should only be disabled in rare
1565       circumstances.  Since warnings are generated using the Perl "warn"
1566       function, they can be intercepted using the Perl $SIG{__WARN__} hook.
1567
1568       The "Warn" attribute is not related to the "PrintWarn" attribute.
1569
1570       "Active"
1571
1572       Type: boolean, read-only
1573
1574       The "Active" attribute is true if the handle object is "active". This
1575       is rarely used in applications. The exact meaning of active is somewhat
1576       vague at the moment. For a database handle it typically means that the
1577       handle is connected to a database ("$dbh->disconnect" sets "Active"
1578       off).  For a statement handle it typically means that the handle is a
1579       "SELECT" that may have more data to fetch. (Fetching all the data or
1580       calling "$sth->finish" sets "Active" off.)
1581
1582       "Executed"
1583
1584       Type: boolean
1585
1586       The "Executed" attribute is true if the handle object has been
1587       "executed".  Currently only the $dbh do() method and the $sth
1588       execute(), execute_array(), and execute_for_fetch() methods set the
1589       "Executed" attribute.
1590
1591       When it's set on a handle it is also set on the parent handle at the
1592       same time. So calling execute() on a $sth also sets the "Executed"
1593       attribute on the parent $dbh.
1594
1595       The "Executed" attribute for a database handle is cleared by the
1596       commit() and rollback() methods (even if they fail). The "Executed"
1597       attribute of a statement handle is not cleared by the DBI under any
1598       circumstances and so acts as a permanent record of whether the
1599       statement handle was ever used.
1600
1601       The "Executed" attribute was added in DBI 1.41.
1602
1603       "Kids"
1604
1605       Type: integer, read-only
1606
1607       For a driver handle, "Kids" is the number of currently existing
1608       database handles that were created from that driver handle.  For a
1609       database handle, "Kids" is the number of currently existing statement
1610       handles that were created from that database handle.  For a statement
1611       handle, the value is zero.
1612
1613       "ActiveKids"
1614
1615       Type: integer, read-only
1616
1617       Like "Kids", but only counting those that are "Active" (as above).
1618
1619       "CachedKids"
1620
1621       Type: hash ref
1622
1623       For a database handle, "CachedKids" returns a reference to the cache
1624       (hash) of statement handles created by the "prepare_cached" method.
1625       For a driver handle, returns a reference to the cache (hash) of
1626       database handles created by the "connect_cached" method.
1627
1628       "Type"
1629
1630       Type: scalar, read-only
1631
1632       The "Type" attribute identifies the type of a DBI handle.  Returns "dr"
1633       for driver handles, "db" for database handles and "st" for statement
1634       handles.
1635
1636       "ChildHandles"
1637
1638       Type: array ref
1639
1640       The ChildHandles attribute contains a reference to an array of all the
1641       handles created by this handle which are still accessible.  The
1642       contents of the array are weak-refs and will become undef when the
1643       handle goes out of scope. (They're cleared out occasionally.)
1644
1645       "ChildHandles" returns undef if your perl version does not support weak
1646       references (check the Scalar::Util module).  The referenced array
1647       returned should be treated as read-only.
1648
1649       For example, to enumerate all driver handles, database handles and
1650       statement handles:
1651
1652           sub show_child_handles {
1653               my ($h, $level) = @_;
1654               printf "%sh %s %s\n", $h->{Type}, "\t" x $level, $h;
1655               show_child_handles($_, $level + 1)
1656                   for (grep { defined } @{$h->{ChildHandles}});
1657           }
1658
1659           my %drivers = DBI->installed_drivers();
1660           show_child_handles($_, 0) for (values %drivers);
1661
1662       "CompatMode"
1663
1664       Type: boolean, inherited
1665
1666       The "CompatMode" attribute is used by emulation layers (such as
1667       Oraperl) to enable compatible behaviour in the underlying driver (e.g.,
1668       DBD::Oracle) for this handle. Not normally set by application code.
1669
1670       It also has the effect of disabling the 'quick FETCH' of attribute
1671       values from the handles attribute cache. So all attribute values are
1672       handled by the drivers own FETCH method. This makes them slightly
1673       slower but is useful for special-purpose drivers like DBD::Multiplex.
1674
1675       "InactiveDestroy"
1676
1677       Type: boolean
1678
1679       The default value, false, means a handle will be fully destroyed as
1680       normal when the last reference to it is removed, just as you'd expect.
1681
1682       If set true then the handle will be treated by the DESTROY as if it was
1683       no longer Active, and so the database engine related effects of
1684       DESTROYing a handle will be skipped.  Think of the name as meaning
1685       'treat the handle as not-Active in the DESTROY method'.
1686
1687       For a database handle, this attribute does not disable an explicit call
1688       to the disconnect method, only the implicit call from DESTROY that
1689       happens if the handle is still marked as "Active".
1690
1691       This attribute is specifically designed for use in Unix applications
1692       that "fork" child processes.  For some drivers, when the child process
1693       exits the destruction of inherited handles cause the corresponding
1694       handles in the parent process to cease working.
1695
1696       Either the parent or the child process, but not both, should set
1697       "InactiveDestroy" true on all their shared handles. Alternatively, and
1698       preferably, the "AutoInactiveDestroy" can be set in the parent on
1699       connect.
1700
1701       To help tracing applications using fork the process id is shown in the
1702       trace log whenever a DBI or handle trace() method is called.  The
1703       process id also shown for every method call if the DBI trace level (not
1704       handle trace level) is set high enough to show the trace from the DBI's
1705       method dispatcher, e.g. >= 9.
1706
1707       "AutoInactiveDestroy"
1708
1709       Type: boolean, inherited
1710
1711       The "InactiveDestroy" attribute, described above, needs to be
1712       explicitly set in the child process after a fork(), on every active
1713       database and statement handle.  This is a problem if the code that
1714       performs the fork() is not under your control, perhaps in a third-party
1715       module.  Use "AutoInactiveDestroy" to get around this situation.
1716
1717       If set true, the DESTROY method will check the process id of the handle
1718       and, if different from the current process id, it will set the
1719       InactiveDestroy attribute.  It is strongly recommended that
1720       "AutoInactiveDestroy" is enabled on all new code (it's only not enabled
1721       by default to avoid backwards compatibility problems).
1722
1723       This is the example it's designed to deal with:
1724
1725           my $dbh = DBI->connect(...);
1726           some_code_that_forks(); # Perhaps without your knowledge
1727           # Child process dies, destroying the inherited dbh
1728           $dbh->do(...); # Breaks because parent $dbh is now broken
1729
1730       The "AutoInactiveDestroy" attribute was added in DBI 1.614.
1731
1732       "PrintWarn"
1733
1734       Type: boolean, inherited
1735
1736       The "PrintWarn" attribute controls the printing of warnings recorded by
1737       the driver.  When set to a true value (the default) the DBI will check
1738       method calls to see if a warning condition has been set. If so, the DBI
1739       will effectively do a "warn("$class $method warning: $DBI::errstr")"
1740       where $class is the driver class and $method is the name of the method
1741       which failed. E.g.,
1742
1743         DBD::Oracle::db execute warning: ... warning text here ...
1744
1745       If desired, the warnings can be caught and processed using a
1746       $SIG{__WARN__} handler or modules like CGI::Carp and CGI::ErrorWrap.
1747
1748       See also "set_err" for how warnings are recorded and "HandleSetErr" for
1749       how to influence it.
1750
1751       Fetching the full details of warnings can require an extra round-trip
1752       to the database server for some drivers. In which case the driver may
1753       opt to only fetch the full details of warnings if the "PrintWarn"
1754       attribute is true. If "PrintWarn" is false then these drivers should
1755       still indicate the fact that there were warnings by setting the warning
1756       string to, for example: "3 warnings".
1757
1758       "PrintError"
1759
1760       Type: boolean, inherited
1761
1762       The "PrintError" attribute can be used to force errors to generate
1763       warnings (using "warn") in addition to returning error codes in the
1764       normal way.  When set "on", any method which results in an error
1765       occurring will cause the DBI to effectively do a "warn("$class $method
1766       failed: $DBI::errstr")" where $class is the driver class and $method is
1767       the name of the method which failed. E.g.,
1768
1769         DBD::Oracle::db prepare failed: ... error text here ...
1770
1771       By default, "DBI->connect" sets "PrintError" "on".
1772
1773       If desired, the warnings can be caught and processed using a
1774       $SIG{__WARN__} handler or modules like CGI::Carp and CGI::ErrorWrap.
1775
1776       "RaiseWarn"
1777
1778       Type: boolean, inherited
1779
1780       The "RaiseWarn" attribute can be used to force warnings to raise
1781       exceptions rather then simply printing them. It is "off" by default.
1782       When set "on", any method which sets warning condition will cause the
1783       DBI to effectively do a "die("$class $method warning: $DBI::errstr")",
1784       where $class is the driver class and $method is the name of the method
1785       that sets warning condition. E.g.,
1786
1787         DBD::Oracle::db execute warning: ... warning text here ...
1788
1789       If you turn "RaiseWarn" on then you'd normally turn "PrintWarn" off.
1790       If "PrintWarn" is also on, then the "PrintWarn" is done first
1791       (naturally).
1792
1793       This attribute was added in DBI 1.643.
1794
1795       "RaiseError"
1796
1797       Type: boolean, inherited
1798
1799       The "RaiseError" attribute can be used to force errors to raise
1800       exceptions rather than simply return error codes in the normal way. It
1801       is "off" by default.  When set "on", any method which results in an
1802       error will cause the DBI to effectively do a "die("$class $method
1803       failed: $DBI::errstr")", where $class is the driver class and $method
1804       is the name of the method that failed. E.g.,
1805
1806         DBD::Oracle::db prepare failed: ... error text here ...
1807
1808       If you turn "RaiseError" on then you'd normally turn "PrintError" off.
1809       If "PrintError" is also on, then the "PrintError" is done first
1810       (naturally).
1811
1812       Typically "RaiseError" is used in conjunction with "eval", or a module
1813       like Try::Tiny or TryCatch, to catch the exception that's been thrown
1814       and handle it.  For example:
1815
1816         use Try::Tiny;
1817
1818         try {
1819           ...
1820           $sth->execute();
1821           ...
1822         } catch {
1823           # $sth->err and $DBI::err will be true if error was from DBI
1824           warn $_; # print the error (which Try::Tiny puts into $_)
1825           ... # do whatever you need to deal with the error
1826         };
1827
1828       In the catch block the $DBI::lasth variable can be useful for diagnosis
1829       and reporting if you can't be sure which handle triggered the error.
1830       For example, $DBI::lasth->{Type} and $DBI::lasth->{Statement}.
1831
1832       See also "Transactions".
1833
1834       If you want to temporarily turn "RaiseError" off (inside a library
1835       function that is likely to fail, for example), the recommended way is
1836       like this:
1837
1838         {
1839           local $h->{RaiseError};  # localize and turn off for this block
1840           ...
1841         }
1842
1843       The original value will automatically and reliably be restored by Perl,
1844       regardless of how the block is exited.  The same logic applies to other
1845       attributes, including "PrintError".
1846
1847       "HandleError"
1848
1849       Type: code ref, inherited
1850
1851       The "HandleError" attribute can be used to provide your own alternative
1852       behaviour in case of errors. If set to a reference to a subroutine then
1853       that subroutine is called when an error is detected (at the same point
1854       that "RaiseError" and "PrintError" are handled). It is called also when
1855       "RaiseWarn" is enabled and a warning is detected.
1856
1857       The subroutine is called with three parameters: the error message
1858       string that "RaiseError", "RaiseWarn" or "PrintError" would use, the
1859       DBI handle being used, and the first value being returned by the method
1860       that failed (typically undef).
1861
1862       If the subroutine returns a false value then the "RaiseError",
1863       "RaiseWarn" and/or "PrintError" attributes are checked and acted upon
1864       as normal.
1865
1866       For example, to "die" with a full stack trace for any error:
1867
1868         use Carp;
1869         $h->{HandleError} = sub { confess(shift) };
1870
1871       Or to turn errors into exceptions:
1872
1873         use Exception; # or your own favourite exception module
1874         $h->{HandleError} = sub { Exception->new('DBI')->raise($_[0]) };
1875
1876       It is possible to 'stack' multiple HandleError handlers by using
1877       closures:
1878
1879         sub your_subroutine {
1880           my $previous_handler = $h->{HandleError};
1881           $h->{HandleError} = sub {
1882             return 1 if $previous_handler and &$previous_handler(@_);
1883             ... your code here ...
1884           };
1885         }
1886
1887       Using a "my" inside a subroutine to store the previous "HandleError"
1888       value is important.  See perlsub and perlref for more information about
1889       closures.
1890
1891       It is possible for "HandleError" to alter the error message that will
1892       be used by "RaiseError", "RaiseWarn" and "PrintError" if it returns
1893       false.  It can do that by altering the value of $_[0]. This example
1894       appends a stack trace to all errors and, unlike the previous example
1895       using Carp::confess, this will work "PrintError" as well as
1896       "RaiseError":
1897
1898         $h->{HandleError} = sub { $_[0]=Carp::longmess($_[0]); 0; };
1899
1900       It is also possible for "HandleError" to hide an error, to a limited
1901       degree, by using "set_err" to reset $DBI::err and $DBI::errstr, and
1902       altering the return value of the failed method. For example:
1903
1904         $h->{HandleError} = sub {
1905           return 0 unless $_[0] =~ /^\S+ fetchrow_arrayref failed:/;
1906           return 0 unless $_[1]->err == 1234; # the error to 'hide'
1907           $h->set_err(undef,undef);   # turn off the error
1908           $_[2] = [ ... ];    # supply alternative return value
1909           return 1;
1910         };
1911
1912       This only works for methods which return a single value and is hard to
1913       make reliable (avoiding infinite loops, for example) and so isn't
1914       recommended for general use!  If you find a good use for it then please
1915       let me know.
1916
1917       "HandleSetErr"
1918
1919       Type: code ref, inherited
1920
1921       The "HandleSetErr" attribute can be used to intercept the setting of
1922       handle "err", "errstr", and "state" values.  If set to a reference to a
1923       subroutine then that subroutine is called whenever set_err() is called,
1924       typically by the driver or a subclass.
1925
1926       The subroutine is called with five arguments, the first five that were
1927       passed to set_err(): the handle, the "err", "errstr", and "state"
1928       values being set, and the method name. These can be altered by changing
1929       the values in the @_ array. The return value affects set_err()
1930       behaviour, see "set_err" for details.
1931
1932       It is possible to 'stack' multiple HandleSetErr handlers by using
1933       closures. See "HandleError" for an example.
1934
1935       The "HandleSetErr" and "HandleError" subroutines differ in subtle but
1936       significant ways. HandleError is only invoked at the point where the
1937       DBI is about to return to the application with "err" set true.  It's
1938       not invoked by the failure of a method that's been called by another
1939       DBI method.  HandleSetErr, on the other hand, is called whenever
1940       set_err() is called with a defined "err" value, even if false.  So it's
1941       not just for errors, despite the name, but also warn and info states.
1942       The set_err() method, and thus HandleSetErr, may be called multiple
1943       times within a method and is usually invoked from deep within driver
1944       code.
1945
1946       In theory a driver can use the return value from HandleSetErr via
1947       set_err() to decide whether to continue or not. If set_err() returns an
1948       empty list, indicating that the HandleSetErr code has 'handled' the
1949       'error', the driver could then continue instead of failing (if that's a
1950       reasonable thing to do).  This isn't excepted to be common and any such
1951       cases should be clearly marked in the driver documentation and
1952       discussed on the dbi-dev mailing list.
1953
1954       The "HandleSetErr" attribute was added in DBI 1.41.
1955
1956       "ErrCount"
1957
1958       Type: unsigned integer
1959
1960       The "ErrCount" attribute is incremented whenever the set_err() method
1961       records an error. It isn't incremented by warnings or information
1962       states. It is not reset by the DBI at any time.
1963
1964       The "ErrCount" attribute was added in DBI 1.41. Older drivers may not
1965       have been updated to use set_err() to record errors and so this
1966       attribute may not be incremented when using them.
1967
1968       "ShowErrorStatement"
1969
1970       Type: boolean, inherited
1971
1972       The "ShowErrorStatement" attribute can be used to cause the relevant
1973       Statement text to be appended to the error messages generated by the
1974       "RaiseError", "PrintError", "RaiseWarn" and "PrintWarn" attributes.
1975       Only applies to errors on statement handles plus the prepare(), do(),
1976       and the various "select*()" database handle methods.  (The exact format
1977       of the appended text is subject to change.)
1978
1979       If "$h->{ParamValues}" returns a hash reference of parameter
1980       (placeholder) values then those are formatted and appended to the end
1981       of the Statement text in the error message.
1982
1983       "TraceLevel"
1984
1985       Type: integer, inherited
1986
1987       The "TraceLevel" attribute can be used as an alternative to the "trace"
1988       method to set the DBI trace level and trace flags for a specific
1989       handle.  See "TRACING" for more details.
1990
1991       The "TraceLevel" attribute is especially useful combined with "local"
1992       to alter the trace settings for just a single block of code.
1993
1994       "FetchHashKeyName"
1995
1996       Type: string, inherited
1997
1998       The "FetchHashKeyName" attribute is used to specify whether the
1999       fetchrow_hashref() method should perform case conversion on the field
2000       names used for the hash keys. For historical reasons it defaults to
2001       '"NAME"' but it is recommended to set it to '"NAME_lc"' (convert to
2002       lower case) or '"NAME_uc"' (convert to upper case) according to your
2003       preference.  It can only be set for driver and database handles.  For
2004       statement handles the value is frozen when prepare() is called.
2005
2006       "ChopBlanks"
2007
2008       Type: boolean, inherited
2009
2010       The "ChopBlanks" attribute can be used to control the trimming of
2011       trailing space characters from fixed width character (CHAR) fields. No
2012       other field types are affected, even where field values have trailing
2013       spaces.
2014
2015       The default is false (although it is possible that the default may
2016       change).  Applications that need specific behaviour should set the
2017       attribute as needed.
2018
2019       Drivers are not required to support this attribute, but any driver
2020       which does not support it must arrange to return "undef" as the
2021       attribute value.
2022
2023       "LongReadLen"
2024
2025       Type: unsigned integer, inherited
2026
2027       The "LongReadLen" attribute may be used to control the maximum length
2028       of 'long' type fields (LONG, BLOB, CLOB, MEMO, etc.) which the driver
2029       will read from the database automatically when it fetches each row of
2030       data.
2031
2032       The "LongReadLen" attribute only relates to fetching and reading long
2033       values; it is not involved in inserting or updating them.
2034
2035       A value of 0 means not to automatically fetch any long data.  Drivers
2036       may return undef or an empty string for long fields when "LongReadLen"
2037       is 0.
2038
2039       The default is typically 0 (zero) or 80 bytes but may vary between
2040       drivers.  Applications fetching long fields should set this value to
2041       slightly larger than the longest long field value to be fetched.
2042
2043       Some databases return some long types encoded as pairs of hex digits.
2044       For these types, "LongReadLen" relates to the underlying data length
2045       and not the doubled-up length of the encoded string.
2046
2047       Changing the value of "LongReadLen" for a statement handle after it has
2048       been "prepare"'d will typically have no effect, so it's common to set
2049       "LongReadLen" on the $dbh before calling "prepare".
2050
2051       For most drivers the value used here has a direct effect on the memory
2052       used by the statement handle while it's active, so don't be too
2053       generous. If you can't be sure what value to use you could execute an
2054       extra select statement to determine the longest value.  For example:
2055
2056         $dbh->{LongReadLen} = $dbh->selectrow_array(qq{
2057             SELECT MAX(OCTET_LENGTH(long_column_name))
2058             FROM table WHERE ...
2059         });
2060         $sth = $dbh->prepare(qq{
2061             SELECT long_column_name, ... FROM table WHERE ...
2062         });
2063
2064       You may need to take extra care if the table can be modified between
2065       the first select and the second being executed. You may also need to
2066       use a different function if OCTET_LENGTH() does not work for long types
2067       in your database. For example, for Sybase use DATALENGTH() and for
2068       Oracle use LENGTHB().
2069
2070       See also "LongTruncOk" for information on truncation of long types.
2071
2072       "LongTruncOk"
2073
2074       Type: boolean, inherited
2075
2076       The "LongTruncOk" attribute may be used to control the effect of
2077       fetching a long field value which has been truncated (typically because
2078       it's longer than the value of the "LongReadLen" attribute).
2079
2080       By default, "LongTruncOk" is false and so fetching a long value that
2081       needs to be truncated will cause the fetch to fail.  (Applications
2082       should always be sure to check for errors after a fetch loop in case an
2083       error, such as a divide by zero or long field truncation, caused the
2084       fetch to terminate prematurely.)
2085
2086       If a fetch fails due to a long field truncation when "LongTruncOk" is
2087       false, many drivers will allow you to continue fetching further rows.
2088
2089       See also "LongReadLen".
2090
2091       "TaintIn"
2092
2093       Type: boolean, inherited
2094
2095       If the "TaintIn" attribute is set to a true value and Perl is running
2096       in taint mode (e.g., started with the "-T" option), then all the
2097       arguments to most DBI method calls are checked for being tainted. This
2098       may change.
2099
2100       The attribute defaults to off, even if Perl is in taint mode.  See
2101       perlsec for more about taint mode.  If Perl is not running in taint
2102       mode, this attribute has no effect.
2103
2104       When fetching data that you trust you can turn off the TaintIn
2105       attribute, for that statement handle, for the duration of the fetch
2106       loop.
2107
2108       The "TaintIn" attribute was added in DBI 1.31.
2109
2110       "TaintOut"
2111
2112       Type: boolean, inherited
2113
2114       If the "TaintOut" attribute is set to a true value and Perl is running
2115       in taint mode (e.g., started with the "-T" option), then most data
2116       fetched from the database is considered tainted. This may change.
2117
2118       The attribute defaults to off, even if Perl is in taint mode.  See
2119       perlsec for more about taint mode.  If Perl is not running in taint
2120       mode, this attribute has no effect.
2121
2122       When fetching data that you trust you can turn off the TaintOut
2123       attribute, for that statement handle, for the duration of the fetch
2124       loop.
2125
2126       Currently only fetched data is tainted. It is possible that the results
2127       of other DBI method calls, and the value of fetched attributes, may
2128       also be tainted in future versions. That change may well break your
2129       applications unless you take great care now. If you use DBI Taint mode,
2130       please report your experience and any suggestions for changes.
2131
2132       The "TaintOut" attribute was added in DBI 1.31.
2133
2134       "Taint"
2135
2136       Type: boolean, inherited
2137
2138       The "Taint" attribute is a shortcut for "TaintIn" and "TaintOut" (it is
2139       also present for backwards compatibility).
2140
2141       Setting this attribute sets both "TaintIn" and "TaintOut", and
2142       retrieving it returns a true value if and only if "TaintIn" and
2143       "TaintOut" are both set to true values.
2144
2145       "Profile"
2146
2147       Type: inherited
2148
2149       The "Profile" attribute enables the collection and reporting of method
2150       call timing statistics.  See the DBI::Profile module documentation for
2151       much more detail.
2152
2153       The "Profile" attribute was added in DBI 1.24.
2154
2155       "ReadOnly"
2156
2157       Type: boolean, inherited
2158
2159       An application can set the "ReadOnly" attribute of a handle to a true
2160       value to indicate that it will not be attempting to make any changes
2161       using that handle or any children of it.
2162
2163       Note that the exact definition of 'read only' is rather fuzzy.  For
2164       more details see the documentation for the driver you're using.
2165
2166       If the driver can make the handle truly read-only then it should
2167       (unless doing so would have unpleasant side effect, like changing the
2168       consistency level from per-statement to per-session).  Otherwise the
2169       attribute is simply advisory.
2170
2171       A driver can set the "ReadOnly" attribute itself to indicate that the
2172       data it is connected to cannot be changed for some reason.
2173
2174       If the driver cannot ensure the "ReadOnly" attribute is adhered to it
2175       will record a warning.  In this case reading the "ReadOnly" attribute
2176       back after it is set true will return true even if the underlying
2177       driver cannot ensure this (so any application knows the application
2178       declared itself ReadOnly).
2179
2180       Library modules and proxy drivers can use the attribute to influence
2181       their behavior.  For example, the DBD::Gofer driver considers the
2182       "ReadOnly" attribute when making a decision about whether to retry an
2183       operation that failed.
2184
2185       The attribute should be set to 1 or 0 (or undef). Other values are
2186       reserved.
2187
2188       "Callbacks"
2189
2190       Type: hash ref
2191
2192       The DBI callback mechanism lets you intercept, and optionally replace,
2193       any method call on a DBI handle. At the extreme, it lets you become a
2194       puppet master, deceiving the application in any way you want.
2195
2196       The "Callbacks" attribute is a hash reference where the keys are DBI
2197       method names and the values are code references. For each key naming a
2198       method, the DBI will execute the associated code reference before
2199       executing the method.
2200
2201       The arguments to the code reference will be the same as to the method,
2202       including the invocant (a database handle or statement handle). For
2203       example, say that to callback to some code on a call to "prepare()":
2204
2205         $dbh->{Callbacks} = {
2206             prepare => sub {
2207                 my ($dbh, $query, $attrs) = @_;
2208                 print "Preparing q{$query}\n"
2209             },
2210         };
2211
2212       The callback would then be executed when you called the "prepare()"
2213       method:
2214
2215         $dbh->prepare('SELECT 1');
2216
2217       And the output of course would be:
2218
2219         Preparing q{SELECT 1}
2220
2221       Because callbacks are executed before the methods they're associated
2222       with, you can modify the arguments before they're passed on to the
2223       method call. For example, to make sure that all calls to "prepare()"
2224       are immediately prepared by DBD::Pg, add a callback that makes sure
2225       that the "pg_prepare_now" attribute is always set:
2226
2227         my $dbh = DBI->connect($dsn, $username, $auth, {
2228             Callbacks => {
2229                 prepare => sub {
2230                     $_[2] ||= {};
2231                     $_[2]->{pg_prepare_now} = 1;
2232                     return; # must return nothing
2233                 },
2234             }
2235         });
2236
2237       Note that we are editing the contents of @_ directly. In this case
2238       we've created the attributes hash if it's not passed to the "prepare"
2239       call.
2240
2241       You can also prevent the associated method from ever executing. While a
2242       callback executes, $_ holds the method name. (This allows multiple
2243       callbacks to share the same code reference and still know what method
2244       was called.)  To prevent the method from executing, simply "undef $_".
2245       For example, if you wanted to disable calls to "ping()", you could do
2246       this:
2247
2248         $dbh->{Callbacks} = {
2249             ping => sub {
2250                 # tell dispatch to not call the method:
2251                 undef $_;
2252                 # return this value instead:
2253                 return "42 bells";
2254             }
2255         };
2256
2257       As with other attributes, Callbacks can be specified on a handle or via
2258       the attributes to "connect()". Callbacks can also be applied to a
2259       statement methods on a statement handle. For example:
2260
2261         $sth->{Callbacks} = {
2262             execute => sub {
2263                 print "Executing ", shift->{Statement}, "\n";
2264             }
2265         };
2266
2267       The "Callbacks" attribute of a database handle isn't copied to any
2268       statement handles it creates. So setting callbacks for a statement
2269       handle requires you to set the "Callbacks" attribute on the statement
2270       handle yourself, as in the example above, or use the special
2271       "ChildCallbacks" key described below.
2272
2273       Special Keys in Callbacks Attribute
2274
2275       In addition to DBI handle method names, the "Callbacks" hash reference
2276       supports four additional keys.
2277
2278       The first is the "ChildCallbacks" key. When a statement handle is
2279       created from a database handle the "ChildCallbacks" key of the database
2280       handle's "Callbacks" attribute, if any, becomes the new "Callbacks"
2281       attribute of the statement handle.  This allows you to define callbacks
2282       for all statement handles created from a database handle. For example,
2283       if you wanted to count how many times "execute" was called in your
2284       application, you could write:
2285
2286         my $exec_count = 0;
2287         my $dbh = DBI->connect( $dsn, $username, $auth, {
2288             Callbacks => {
2289                 ChildCallbacks => {
2290                     execute => sub { $exec_count++; return; }
2291                 }
2292             }
2293         });
2294
2295         END {
2296             print "The execute method was called $exec_count times\n";
2297         }
2298
2299       The other three special keys are "connect_cached.new",
2300       "connect_cached.connected", and "connect_cached.reused". These keys
2301       define callbacks that are called when "connect_cached()" is called, but
2302       allow different behaviors depending on whether a new handle is created
2303       or a handle is returned. The callback is invoked with these arguments:
2304       "$dbh, $dsn, $user, $auth, $attr".
2305
2306       For example, some applications uses "connect_cached()" to connect with
2307       "AutoCommit" enabled and then disable "AutoCommit" temporarily for
2308       transactions. If "connect_cached()" is called during a transaction,
2309       perhaps in a utility method, then it might select the same cached
2310       handle and then force "AutoCommit" on, forcing a commit of the
2311       transaction. See the "connect_cached" documentation for one way to deal
2312       with that. Here we'll describe an alternative approach using a
2313       callback.
2314
2315       Because the "connect_cached.new" and "connect_cached.reused" callbacks
2316       are invoked before "connect_cached()" has applied the connect
2317       attributes, you can use them to edit the attributes that will be
2318       applied. To prevent a cached handle from having its transactions
2319       committed before it's returned, you can eliminate the "AutoCommit"
2320       attribute in a "connect_cached.reused" callback, like so:
2321
2322         my $cb = {
2323             'connect_cached.reused' => sub { delete $_[4]->{AutoCommit} },
2324         };
2325
2326         sub dbh {
2327             my $self = shift;
2328             DBI->connect_cached( $dsn, $username, $auth, {
2329                 PrintError => 0,
2330                 RaiseError => 1,
2331                 AutoCommit => 1,
2332                 Callbacks  => $cb,
2333             });
2334         }
2335
2336       The upshot is that new database handles are created with "AutoCommit"
2337       enabled, while cached database handles are left in whatever transaction
2338       state they happened to be in when retrieved from the cache.
2339
2340       Note that we've also used a lexical for the callbacks hash reference.
2341       This is because "connect_cached()" returns a new database handle if any
2342       of the attributes passed to is have changed. If we used an inline hash
2343       reference, "connect_cached()" would return a new database handle every
2344       time. Which would rather defeat the purpose.
2345
2346       A more common application for callbacks is setting connection state
2347       only when a new connection is made (by connect() or connect_cached()).
2348       Adding a callback to the connected method (when using "connect") or via
2349       "connect_cached.connected" (when useing connect_cached()>) makes this
2350       easy.  The connected() method is a no-op by default (unless you
2351       subclass the DBI and change it). The DBI calls it to indicate that a
2352       new connection has been made and the connection attributes have all
2353       been set. You can give it a bit of added functionality by applying a
2354       callback to it. For example, to make sure that MySQL understands your
2355       application's ANSI-compliant SQL, set it up like so:
2356
2357         my $dbh = DBI->connect($dsn, $username, $auth, {
2358             Callbacks => {
2359                 connected => sub {
2360                     shift->do(q{
2361                         SET SESSION sql_mode='ansi,strict_trans_tables,no_auto_value_on_zero';
2362                     });
2363                     return;
2364                 },
2365             }
2366         });
2367
2368       If you're using "connect_cached()", use the "connect_cached.connected"
2369       callback, instead. This is because "connected()" is called for both new
2370       and reused database handles, but you want to execute a callback only
2371       the when a new database handle is returned. For example, to set the
2372       time zone on connection to a PostgreSQL database, try this:
2373
2374         my $cb = {
2375             'connect_cached.connected' => sub {
2376                 shift->do('SET timezone = UTC');
2377             }
2378         };
2379
2380         sub dbh {
2381             my $self = shift;
2382             DBI->connect_cached( $dsn, $username, $auth, { Callbacks => $cb });
2383         }
2384
2385       One significant limitation with callbacks is that there can only be one
2386       per method per handle. This means it's easy for one use of callbacks to
2387       interfere with, or typically simply overwrite, another use of
2388       callbacks. For this reason modules using callbacks should document the
2389       fact clearly so application authors can tell if use of callbacks by the
2390       module will clash with use of callbacks by the application.
2391
2392       You might be able to work around this issue by taking a copy of the
2393       original callback and calling it within your own. For example:
2394
2395         my $prev_cb = $h->{Callbacks}{method_name};
2396         $h->{Callbacks}{method_name} = sub {
2397           if ($prev_cb) {
2398               my @result = $prev_cb->(@_);
2399               return @result if not $_; # $prev_cb vetoed call
2400           }
2401           ... your callback logic here ...
2402         };
2403
2404       "private_your_module_name_*"
2405
2406       The DBI provides a way to store extra information in a DBI handle as
2407       "private" attributes. The DBI will allow you to store and retrieve any
2408       attribute which has a name starting with ""private_"".
2409
2410       It is strongly recommended that you use just one private attribute
2411       (e.g., use a hash ref) and give it a long and unambiguous name that
2412       includes the module or application name that the attribute relates to
2413       (e.g., ""private_YourFullModuleName_thingy"").
2414
2415       Because of the way the Perl tie mechanism works you cannot reliably use
2416       the "||=" operator directly to initialise the attribute, like this:
2417
2418         my $foo = $dbh->{private_yourmodname_foo} ||= { ... }; # WRONG
2419
2420       you should use a two step approach like this:
2421
2422         my $foo = $dbh->{private_yourmodname_foo};
2423         $foo ||= $dbh->{private_yourmodname_foo} = { ... };
2424
2425       This attribute is primarily of interest to people sub-classing DBI, or
2426       for applications to piggy-back extra information onto DBI handles.
2427

DBI DATABASE HANDLE OBJECTS

2429       This section covers the methods and attributes associated with database
2430       handles.
2431
2432   Database Handle Methods
2433       The following methods are specified for DBI database handles:
2434
2435       "clone"
2436
2437         $new_dbh = $dbh->clone(\%attr);
2438
2439       The "clone" method duplicates the $dbh connection by connecting with
2440       the same parameters ($dsn, $user, $password) as originally used.
2441
2442       The attributes for the cloned connect are the same as those used for
2443       the original connect, with any other attributes in "\%attr" merged over
2444       them.  Effectively the same as doing:
2445
2446         %attributes_used = ( %original_attributes, %attr );
2447
2448       If \%attr is not given then it defaults to a hash containing all the
2449       attributes in the attribute cache of $dbh excluding any non-code
2450       references, plus the main boolean attributes (RaiseError, PrintError,
2451       AutoCommit, etc.). This behaviour is unreliable and so use of clone
2452       without an argument is deprecated and may cause a warning in a future
2453       release.
2454
2455       The clone method can be used even if the database handle is
2456       disconnected.
2457
2458       The "clone" method was added in DBI 1.33.
2459
2460       "data_sources"
2461
2462         @ary = $dbh->data_sources();
2463         @ary = $dbh->data_sources(\%attr);
2464
2465       Returns a list of data sources (databases) available via the $dbh
2466       driver's data_sources() method, plus any extra data sources that the
2467       driver can discover via the connected $dbh. Typically the extra data
2468       sources are other databases managed by the same server process that the
2469       $dbh is connected to.
2470
2471       Data sources are returned in a form suitable for passing to the
2472       "connect" method (that is, they will include the ""dbi:$driver:""
2473       prefix).
2474
2475       The data_sources() method, for a $dbh, was added in DBI 1.38.
2476
2477       "do"
2478
2479         $rows = $dbh->do($statement)           or die $dbh->errstr;
2480         $rows = $dbh->do($statement, \%attr)   or die $dbh->errstr;
2481         $rows = $dbh->do($statement, \%attr, @bind_values) or die ...
2482
2483       Prepare and execute a single statement. Returns the number of rows
2484       affected or "undef" on error. A return value of "-1" means the number
2485       of rows is not known, not applicable, or not available.
2486
2487       This method is typically most useful for non-"SELECT" statements that
2488       either cannot be prepared in advance (due to a limitation of the
2489       driver) or do not need to be executed repeatedly. It should not be used
2490       for "SELECT" statements because it does not return a statement handle
2491       (so you can't fetch any data).
2492
2493       The default "do" method is logically similar to:
2494
2495         sub do {
2496             my($dbh, $statement, $attr, @bind_values) = @_;
2497             my $sth = $dbh->prepare($statement, $attr) or return undef;
2498             $sth->execute(@bind_values) or return undef;
2499             my $rows = $sth->rows;
2500             ($rows == 0) ? "0E0" : $rows; # always return true if no error
2501         }
2502
2503       For example:
2504
2505         my $rows_deleted = $dbh->do(q{
2506             DELETE FROM table
2507             WHERE status = ?
2508         }, undef, 'DONE') or die $dbh->errstr;
2509
2510       Using placeholders and @bind_values with the "do" method can be useful
2511       because it avoids the need to correctly quote any variables in the
2512       $statement. But if you'll be executing the statement many times then
2513       it's more efficient to "prepare" it once and call "execute" many times
2514       instead.
2515
2516       The "q{...}" style quoting used in this example avoids clashing with
2517       quotes that may be used in the SQL statement. Use the double-quote-like
2518       "qq{...}" operator if you want to interpolate variables into the
2519       string.  See "Quote and Quote-like Operators" in perlop for more
2520       details.
2521
2522       Note drivers are free to avoid the overhead of creating an DBI
2523       statement handle for do(), especially if there are no parameters. In
2524       this case error handlers, if invoked during do(), will be passed the
2525       database handle.
2526
2527       "last_insert_id"
2528
2529         $rv = $dbh->last_insert_id();
2530         $rv = $dbh->last_insert_id($catalog, $schema, $table, $field);
2531         $rv = $dbh->last_insert_id($catalog, $schema, $table, $field, \%attr);
2532
2533       Returns a value 'identifying' the row just inserted, if possible.
2534       Typically this would be a value assigned by the database server to a
2535       column with an auto_increment or serial type.  Returns undef if the
2536       driver does not support the method or can't determine the value.
2537
2538       The $catalog, $schema, $table, and $field parameters may be required
2539       for some drivers (see below).  If you don't know the parameter values
2540       and your driver does not need them, then use "undef" for each.
2541
2542       There are several caveats to be aware of with this method if you want
2543       to use it for portable applications:
2544
2545       * For some drivers the value may only available immediately after the
2546       insert statement has executed (e.g., mysql, Informix).
2547
2548       * For some drivers the $catalog, $schema, $table, and $field parameters
2549       are required, for others they are ignored (e.g., mysql).
2550
2551       * Drivers may return an indeterminate value if no insert has been
2552       performed yet.
2553
2554       * For some drivers the value may only be available if placeholders have
2555       not been used (e.g., Sybase, MS SQL). In this case the value returned
2556       would be from the last non-placeholder insert statement.
2557
2558       * Some drivers may need driver-specific hints about how to get the
2559       value. For example, being told the name of the database 'sequence'
2560       object that holds the value. Any such hints are passed as driver-
2561       specific attributes in the \%attr parameter.
2562
2563       * If the underlying database offers nothing better, then some drivers
2564       may attempt to implement this method by executing ""select max($field)
2565       from $table"". Drivers using any approach like this should issue a
2566       warning if "AutoCommit" is true because it is generally unsafe -
2567       another process may have modified the table between your insert and the
2568       select. For situations where you know it is safe, such as when you have
2569       locked the table, you can silence the warning by passing "Warn" => 0 in
2570       \%attr.
2571
2572       * If no insert has been performed yet, or the last insert failed, then
2573       the value is implementation defined.
2574
2575       Given all the caveats above, it's clear that this method must be used
2576       with care.
2577
2578       The "last_insert_id" method was added in DBI 1.38.
2579
2580       "selectrow_array"
2581
2582         @row_ary = $dbh->selectrow_array($statement);
2583         @row_ary = $dbh->selectrow_array($statement, \%attr);
2584         @row_ary = $dbh->selectrow_array($statement, \%attr, @bind_values);
2585
2586       This utility method combines "prepare", "execute" and "fetchrow_array"
2587       into a single call. If called in a list context, it returns the first
2588       row of data from the statement.  The $statement parameter can be a
2589       previously prepared statement handle, in which case the "prepare" is
2590       skipped.
2591
2592       If any method fails, and "RaiseError" is not set, "selectrow_array"
2593       will return an empty list.
2594
2595       If called in a scalar context for a statement handle that has more than
2596       one column, it is undefined whether the driver will return the value of
2597       the first column or the last. So don't do that.  Also, in a scalar
2598       context, an "undef" is returned if there are no more rows or if an
2599       error occurred. That "undef" can't be distinguished from an "undef"
2600       returned because the first field value was NULL.  For these reasons you
2601       should exercise some caution if you use "selectrow_array" in a scalar
2602       context, or just don't do that.
2603
2604       "selectrow_arrayref"
2605
2606         $ary_ref = $dbh->selectrow_arrayref($statement);
2607         $ary_ref = $dbh->selectrow_arrayref($statement, \%attr);
2608         $ary_ref = $dbh->selectrow_arrayref($statement, \%attr, @bind_values);
2609
2610       This utility method combines "prepare", "execute" and
2611       "fetchrow_arrayref" into a single call. It returns the first row of
2612       data from the statement.  The $statement parameter can be a previously
2613       prepared statement handle, in which case the "prepare" is skipped.
2614
2615       If any method fails, and "RaiseError" is not set, "selectrow_arrayref"
2616       will return undef.
2617
2618       "selectrow_hashref"
2619
2620         $hash_ref = $dbh->selectrow_hashref($statement);
2621         $hash_ref = $dbh->selectrow_hashref($statement, \%attr);
2622         $hash_ref = $dbh->selectrow_hashref($statement, \%attr, @bind_values);
2623
2624       This utility method combines "prepare", "execute" and
2625       "fetchrow_hashref" into a single call. It returns the first row of data
2626       from the statement.  The $statement parameter can be a previously
2627       prepared statement handle, in which case the "prepare" is skipped.
2628
2629       If any method fails, and "RaiseError" is not set, "selectrow_hashref"
2630       will return undef.
2631
2632       "selectall_arrayref"
2633
2634         $ary_ref = $dbh->selectall_arrayref($statement);
2635         $ary_ref = $dbh->selectall_arrayref($statement, \%attr);
2636         $ary_ref = $dbh->selectall_arrayref($statement, \%attr, @bind_values);
2637
2638       This utility method combines "prepare", "execute" and
2639       "fetchall_arrayref" into a single call. It returns a reference to an
2640       array containing a reference to an array (or hash, see below) for each
2641       row of data fetched.
2642
2643       The $statement parameter can be a previously prepared statement handle,
2644       in which case the "prepare" is skipped. This is recommended if the
2645       statement is going to be executed many times.
2646
2647       If "RaiseError" is not set and any method except "fetchall_arrayref"
2648       fails then "selectall_arrayref" will return "undef"; if
2649       "fetchall_arrayref" fails then it will return with whatever data has
2650       been fetched thus far. You should check "$dbh->err" afterwards (or use
2651       the "RaiseError" attribute) to discover if the data is complete or was
2652       truncated due to an error.
2653
2654       The "fetchall_arrayref" method called by "selectall_arrayref" supports
2655       a $max_rows parameter. You can specify a value for $max_rows by
2656       including a '"MaxRows"' attribute in \%attr. In which case finish() is
2657       called for you after fetchall_arrayref() returns.
2658
2659       The "fetchall_arrayref" method called by "selectall_arrayref" also
2660       supports a $slice parameter. You can specify a value for $slice by
2661       including a '"Slice"' or '"Columns"' attribute in \%attr. The only
2662       difference between the two is that if "Slice" is not defined and
2663       "Columns" is an array ref, then the array is assumed to contain column
2664       index values (which count from 1), rather than perl array index values.
2665       In which case the array is copied and each value decremented before
2666       passing to "/fetchall_arrayref".
2667
2668       You may often want to fetch an array of rows where each row is stored
2669       as a hash. That can be done simply using:
2670
2671         my $emps = $dbh->selectall_arrayref(
2672             "SELECT ename FROM emp ORDER BY ename",
2673             { Slice => {} }
2674         );
2675         foreach my $emp ( @$emps ) {
2676             print "Employee: $emp->{ename}\n";
2677         }
2678
2679       Or, to fetch into an array instead of an array ref:
2680
2681         @result = @{ $dbh->selectall_arrayref($sql, { Slice => {} }) };
2682
2683       See "fetchall_arrayref" method for more details.
2684
2685       "selectall_array"
2686
2687         @ary = $dbh->selectall_array($statement);
2688         @ary = $dbh->selectall_array($statement, \%attr);
2689         @ary = $dbh->selectall_array($statement, \%attr, @bind_values);
2690
2691       This is a convenience wrapper around selectall_arrayref that returns
2692       the rows directly as a list, rather than a reference to an array of
2693       rows.
2694
2695       Note that if "RaiseError" is not set then you can't tell the difference
2696       between returning no rows and an error. Using RaiseError is best
2697       practice.
2698
2699       The "selectall_array" method was added in DBI 1.635.
2700
2701       "selectall_hashref"
2702
2703         $hash_ref = $dbh->selectall_hashref($statement, $key_field);
2704         $hash_ref = $dbh->selectall_hashref($statement, $key_field, \%attr);
2705         $hash_ref = $dbh->selectall_hashref($statement, $key_field, \%attr, @bind_values);
2706
2707       This utility method combines "prepare", "execute" and
2708       "fetchall_hashref" into a single call. It returns a reference to a hash
2709       containing one entry, at most, for each row, as returned by
2710       fetchall_hashref().
2711
2712       The $statement parameter can be a previously prepared statement handle,
2713       in which case the "prepare" is skipped.  This is recommended if the
2714       statement is going to be executed many times.
2715
2716       The $key_field parameter defines which column, or columns, are used as
2717       keys in the returned hash. It can either be the name of a single field,
2718       or a reference to an array containing multiple field names. Using
2719       multiple names yields a tree of nested hashes.
2720
2721       If a row has the same key as an earlier row then it replaces the
2722       earlier row.
2723
2724       If any method except "fetchall_hashref" fails, and "RaiseError" is not
2725       set, "selectall_hashref" will return "undef".  If "fetchall_hashref"
2726       fails and "RaiseError" is not set, then it will return with whatever
2727       data it has fetched thus far. $DBI::err should be checked to catch
2728       that.
2729
2730       See fetchall_hashref() for more details.
2731
2732       "selectcol_arrayref"
2733
2734         $ary_ref = $dbh->selectcol_arrayref($statement);
2735         $ary_ref = $dbh->selectcol_arrayref($statement, \%attr);
2736         $ary_ref = $dbh->selectcol_arrayref($statement, \%attr, @bind_values);
2737
2738       This utility method combines "prepare", "execute", and fetching one
2739       column from all the rows, into a single call. It returns a reference to
2740       an array containing the values of the first column from each row.
2741
2742       The $statement parameter can be a previously prepared statement handle,
2743       in which case the "prepare" is skipped. This is recommended if the
2744       statement is going to be executed many times.
2745
2746       If any method except "fetch" fails, and "RaiseError" is not set,
2747       "selectcol_arrayref" will return "undef".  If "fetch" fails and
2748       "RaiseError" is not set, then it will return with whatever data it has
2749       fetched thus far. $DBI::err should be checked to catch that.
2750
2751       The "selectcol_arrayref" method defaults to pushing a single column
2752       value (the first) from each row into the result array. However, it can
2753       also push another column, or even multiple columns per row, into the
2754       result array. This behaviour can be specified via a '"Columns"'
2755       attribute which must be a ref to an array containing the column number
2756       or numbers to use. For example:
2757
2758         # get array of id and name pairs:
2759         my $ary_ref = $dbh->selectcol_arrayref("select id, name from table", { Columns=>[1,2] });
2760         my %hash = @$ary_ref; # build hash from key-value pairs so $hash{$id} => name
2761
2762       You can specify a maximum number of rows to fetch by including a
2763       '"MaxRows"' attribute in \%attr.
2764
2765       "prepare"
2766
2767         $sth = $dbh->prepare($statement)          or die $dbh->errstr;
2768         $sth = $dbh->prepare($statement, \%attr)  or die $dbh->errstr;
2769
2770       Prepares a statement for later execution by the database engine and
2771       returns a reference to a statement handle object.
2772
2773       The returned statement handle can be used to get attributes of the
2774       statement and invoke the "execute" method. See "Statement Handle
2775       Methods".
2776
2777       Drivers for engines without the concept of preparing a statement will
2778       typically just store the statement in the returned handle and process
2779       it when "$sth->execute" is called. Such drivers are unlikely to give
2780       much useful information about the statement, such as
2781       "$sth->{NUM_OF_FIELDS}", until after "$sth->execute" has been called.
2782       Portable applications should take this into account.
2783
2784       In general, DBI drivers do not parse the contents of the statement
2785       (other than simply counting any Placeholders).  The statement is passed
2786       directly to the database engine, sometimes known as pass-thru mode.
2787       This has advantages and disadvantages. On the plus side, you can access
2788       all the functionality of the engine being used. On the downside, you're
2789       limited if you're using a simple engine, and you need to take extra
2790       care if writing applications intended to be portable between engines.
2791
2792       Portable applications should not assume that a new statement can be
2793       prepared and/or executed while still fetching results from a previous
2794       statement.
2795
2796       Some command-line SQL tools use statement terminators, like a
2797       semicolon, to indicate the end of a statement. Such terminators should
2798       not normally be used with the DBI.
2799
2800       "prepare_cached"
2801
2802         $sth = $dbh->prepare_cached($statement)
2803         $sth = $dbh->prepare_cached($statement, \%attr)
2804         $sth = $dbh->prepare_cached($statement, \%attr, $if_active)
2805
2806       Like "prepare" except that the statement handle returned will be stored
2807       in a hash associated with the $dbh. If another call is made to
2808       "prepare_cached" with the same $statement and %attr parameter values,
2809       then the corresponding cached $sth will be returned without contacting
2810       the database server. Be sure to understand the cautions and caveats
2811       noted below.
2812
2813       The $if_active parameter lets you adjust the behaviour if an already
2814       cached statement handle is still Active.  There are several
2815       alternatives:
2816
2817       0: A warning will be generated, and finish() will be called on the
2818       statement handle before it is returned.  This is the default behaviour
2819       if $if_active is not passed.
2820       1: finish() will be called on the statement handle, but the warning is
2821       suppressed.
2822       2: Disables any checking.
2823       3: The existing active statement handle will be removed from the cache
2824       and a new statement handle prepared and cached in its place. This is
2825       the safest option because it doesn't affect the state of the old
2826       handle, it just removes it from the cache. [Added in DBI 1.40]
2827
2828       Here are some examples of "prepare_cached":
2829
2830         sub insert_hash {
2831           my ($table, $field_values) = @_;
2832           # sort to keep field order, and thus sql, stable for prepare_cached
2833           my @fields = sort keys %$field_values;
2834           my @values = @{$field_values}{@fields};
2835           my $sql = sprintf "insert into %s (%s) values (%s)",
2836               $table, join(",", @fields), join(",", ("?")x@fields);
2837           my $sth = $dbh->prepare_cached($sql);
2838           return $sth->execute(@values);
2839         }
2840
2841         sub search_hash {
2842           my ($table, $field_values) = @_;
2843           # sort to keep field order, and thus sql, stable for prepare_cached
2844           my @fields = sort keys %$field_values;
2845           my @values = @{$field_values}{@fields};
2846           my $qualifier = "";
2847           $qualifier = "where ".join(" and ", map { "$_=?" } @fields) if @fields;
2848           $sth = $dbh->prepare_cached("SELECT * FROM $table $qualifier");
2849           return $dbh->selectall_arrayref($sth, {}, @values);
2850         }
2851
2852       Caveat emptor: This caching can be useful in some applications, but it
2853       can also cause problems and should be used with care. Here is a
2854       contrived case where caching would cause a significant problem:
2855
2856         my $sth = $dbh->prepare_cached('SELECT * FROM foo WHERE bar=?');
2857         $sth->execute(...);
2858         while (my $data = $sth->fetchrow_hashref) {
2859
2860           # later, in some other code called within the loop...
2861           my $sth2 = $dbh->prepare_cached('SELECT * FROM foo WHERE bar=?');
2862           $sth2->execute(...);
2863           while (my $data2 = $sth2->fetchrow_arrayref) {
2864             do_stuff(...);
2865           }
2866         }
2867
2868       In this example, since both handles are preparing the exact same
2869       statement, $sth2 will not be its own statement handle, but a duplicate
2870       of $sth returned from the cache. The results will certainly not be what
2871       you expect.  Typically the inner fetch loop will work normally,
2872       fetching all the records and terminating when there are no more, but
2873       now that $sth is the same as $sth2 the outer fetch loop will also
2874       terminate.
2875
2876       You'll know if you run into this problem because prepare_cached() will
2877       generate a warning by default (when $if_active is false).
2878
2879       The cache used by prepare_cached() is keyed by both the statement and
2880       any attributes so you can also avoid this issue by doing something
2881       like:
2882
2883         $sth = $dbh->prepare_cached("...", { dbi_dummy => __FILE__.__LINE__ });
2884
2885       which will ensure that prepare_cached only returns statements cached by
2886       that line of code in that source file.
2887
2888       Also, to ensure the attributes passed are always the same, avoid
2889       passing references inline. For example, the Slice attribute is
2890       specified as a reference. Be sure to declare it external to the call to
2891       prepare_cached(), such that a new hash reference is not created on
2892       every call. See "connect_cached" for more details and examples.
2893
2894       If you'd like the cache to managed intelligently, you can tie the
2895       hashref returned by "CachedKids" to an appropriate caching module, such
2896       as Tie::Cache::LRU:
2897
2898         my $cache;
2899         tie %$cache, 'Tie::Cache::LRU', 500;
2900         $dbh->{CachedKids} = $cache;
2901
2902       "commit"
2903
2904         $rc  = $dbh->commit     or die $dbh->errstr;
2905
2906       Commit (make permanent) the most recent series of database changes if
2907       the database supports transactions and AutoCommit is off.
2908
2909       If "AutoCommit" is on, then calling "commit" will issue a "commit
2910       ineffective with AutoCommit" warning.
2911
2912       See also "Transactions" in the "FURTHER INFORMATION" section below.
2913
2914       "rollback"
2915
2916         $rc  = $dbh->rollback   or die $dbh->errstr;
2917
2918       Rollback (undo) the most recent series of uncommitted database changes
2919       if the database supports transactions and AutoCommit is off.
2920
2921       If "AutoCommit" is on, then calling "rollback" will issue a "rollback
2922       ineffective with AutoCommit" warning.
2923
2924       See also "Transactions" in the "FURTHER INFORMATION" section below.
2925
2926       "begin_work"
2927
2928         $rc  = $dbh->begin_work   or die $dbh->errstr;
2929
2930       Enable transactions (by turning "AutoCommit" off) until the next call
2931       to "commit" or "rollback". After the next "commit" or "rollback",
2932       "AutoCommit" will automatically be turned on again.
2933
2934       If "AutoCommit" is already off when "begin_work" is called then it does
2935       nothing except return an error. If the driver does not support
2936       transactions then when "begin_work" attempts to set "AutoCommit" off
2937       the driver will trigger a fatal error.
2938
2939       See also "Transactions" in the "FURTHER INFORMATION" section below.
2940
2941       "disconnect"
2942
2943         $rc = $dbh->disconnect  or warn $dbh->errstr;
2944
2945       Disconnects the database from the database handle. "disconnect" is
2946       typically only used before exiting the program. The handle is of little
2947       use after disconnecting.
2948
2949       The transaction behaviour of the "disconnect" method is, sadly,
2950       undefined.  Some database systems (such as Oracle and Ingres) will
2951       automatically commit any outstanding changes, but others (such as
2952       Informix) will rollback any outstanding changes.  Applications not
2953       using "AutoCommit" should explicitly call "commit" or "rollback" before
2954       calling "disconnect".
2955
2956       The database is automatically disconnected by the "DESTROY" method if
2957       still connected when there are no longer any references to the handle.
2958       The "DESTROY" method for each driver should implicitly call "rollback"
2959       to undo any uncommitted changes. This is vital behaviour to ensure that
2960       incomplete transactions don't get committed simply because Perl calls
2961       "DESTROY" on every object before exiting. Also, do not rely on the
2962       order of object destruction during "global destruction", as it is
2963       undefined.
2964
2965       Generally, if you want your changes to be committed or rolled back when
2966       you disconnect, then you should explicitly call "commit" or "rollback"
2967       before disconnecting.
2968
2969       If you disconnect from a database while you still have active statement
2970       handles (e.g., SELECT statement handles that may have more data to
2971       fetch), you will get a warning. The warning may indicate that a fetch
2972       loop terminated early, perhaps due to an uncaught error.  To avoid the
2973       warning call the "finish" method on the active handles.
2974
2975       "ping"
2976
2977         $rc = $dbh->ping;
2978
2979       Attempts to determine, in a reasonably efficient way, if the database
2980       server is still running and the connection to it is still working.
2981       Individual drivers should implement this function in the most suitable
2982       manner for their database engine.
2983
2984       The current default implementation always returns true without actually
2985       doing anything. Actually, it returns ""0 but true"" which is true but
2986       zero. That way you can tell if the return value is genuine or just the
2987       default. Drivers should override this method with one that does the
2988       right thing for their type of database.
2989
2990       Few applications would have direct use for this method. See the
2991       specialized Apache::DBI module for one example usage.
2992
2993       "get_info"
2994
2995         $value = $dbh->get_info( $info_type );
2996
2997       Returns information about the implementation, i.e. driver and data
2998       source capabilities, restrictions etc. It returns "undef" for unknown
2999       or unimplemented information types. For example:
3000
3001         $database_version  = $dbh->get_info(  18 ); # SQL_DBMS_VER
3002         $max_select_tables = $dbh->get_info( 106 ); # SQL_MAXIMUM_TABLES_IN_SELECT
3003
3004       See "Standards Reference Information" for more detailed information
3005       about the information types and their meanings and possible return
3006       values.
3007
3008       The DBI::Const::GetInfoType module exports a %GetInfoType hash that can
3009       be used to map info type names to numbers. For example:
3010
3011         $database_version = $dbh->get_info( $GetInfoType{SQL_DBMS_VER} );
3012
3013       The names are a merging of the ANSI and ODBC standards (which differ in
3014       some cases). See DBI::Const::GetInfoType for more details.
3015
3016       Because some DBI methods make use of get_info(), drivers are strongly
3017       encouraged to support at least the following very minimal set of
3018       information types to ensure the DBI itself works properly:
3019
3020        Type  Name                        Example A     Example B
3021        ----  --------------------------  ------------  ----------------
3022          17  SQL_DBMS_NAME               'ACCESS'      'Oracle'
3023          18  SQL_DBMS_VER                '03.50.0000'  '08.01.0721 ...'
3024          29  SQL_IDENTIFIER_QUOTE_CHAR   '`'           '"'
3025          41  SQL_CATALOG_NAME_SEPARATOR  '.'           '@'
3026         114  SQL_CATALOG_LOCATION        1             2
3027
3028       Values from 9000 to 9999 for get_info are officially reserved for use
3029       by Perl DBI.  Values in that range which have been assigned a meaning
3030       are defined here:
3031
3032       9000: true if a backslash character ("\") before placeholder-like text
3033       (e.g. "?", ":foo") will prevent it being treated as a placeholder by
3034       the driver.  The backslash will be removed before the text is passed to
3035       the backend.
3036
3037       "table_info"
3038
3039         $sth = $dbh->table_info( $catalog, $schema, $table, $type );
3040         $sth = $dbh->table_info( $catalog, $schema, $table, $type, \%attr );
3041
3042         # then $sth->fetchall_arrayref or $sth->fetchall_hashref etc
3043
3044       Returns an active statement handle that can be used to fetch
3045       information about tables and views that exist in the database.
3046
3047       The arguments $catalog, $schema and $table may accept search patterns
3048       according to the database/driver, for example: $table = '%FOO%';
3049       Remember that the underscore character ('"_"') is a search pattern that
3050       means match any character, so 'FOO_%' is the same as 'FOO%' and
3051       'FOO_BAR%' will match names like 'FOO1BAR'.
3052
3053       The value of $type is a comma-separated list of one or more types of
3054       tables to be returned in the result set. Each value may optionally be
3055       quoted, e.g.:
3056
3057         $type = "TABLE";
3058         $type = "'TABLE','VIEW'";
3059
3060       In addition the following special cases may also be supported by some
3061       drivers:
3062
3063       •   If the value of $catalog is '%' and $schema and $table name are
3064           empty strings, the result set contains a list of catalog names.
3065           For example:
3066
3067             $sth = $dbh->table_info('%', '', '');
3068
3069       •   If the value of $schema is '%' and $catalog and $table are empty
3070           strings, the result set contains a list of schema names.
3071
3072       •   If the value of $type is '%' and $catalog, $schema, and $table are
3073           all empty strings, the result set contains a list of table types.
3074
3075       If your driver doesn't support one or more of the selection filter
3076       parameters then you may get back more than you asked for and can do the
3077       filtering yourself.
3078
3079       This method can be expensive, and can return a large amount of data.
3080       (For example, small Oracle installation returns over 2000 rows.)  So
3081       it's a good idea to use the filters to limit the data as much as
3082       possible.
3083
3084       The statement handle returned has at least the following fields in the
3085       order show below. Other fields, after these, may also be present.
3086
3087       TABLE_CAT: Table catalog identifier. This field is NULL ("undef") if
3088       not applicable to the data source, which is usually the case. This
3089       field is empty if not applicable to the table.
3090
3091       TABLE_SCHEM: The name of the schema containing the TABLE_NAME value.
3092       This field is NULL ("undef") if not applicable to data source, and
3093       empty if not applicable to the table.
3094
3095       TABLE_NAME: Name of the table (or view, synonym, etc).
3096
3097       TABLE_TYPE: One of the following: "TABLE", "VIEW", "SYSTEM TABLE",
3098       "GLOBAL TEMPORARY", "LOCAL TEMPORARY", "ALIAS", "SYNONYM" or a type
3099       identifier that is specific to the data source.
3100
3101       REMARKS: A description of the table. May be NULL ("undef").
3102
3103       Note that "table_info" might not return records for all tables.
3104       Applications can use any valid table regardless of whether it's
3105       returned by "table_info".
3106
3107       See also "tables", "Catalog Methods" and "Standards Reference
3108       Information".
3109
3110       "column_info"
3111
3112         $sth = $dbh->column_info( $catalog, $schema, $table, $column );
3113
3114         # then $sth->fetchall_arrayref or $sth->fetchall_hashref etc
3115
3116       Returns an active statement handle that can be used to fetch
3117       information about columns in specified tables.
3118
3119       The arguments $schema, $table and $column may accept search patterns
3120       according to the database/driver, for example: $table = '%FOO%';
3121
3122       Note: The support for the selection criteria is driver specific. If the
3123       driver doesn't support one or more of them then you may get back more
3124       than you asked for and can do the filtering yourself.
3125
3126       Note: If your driver does not support column_info an undef is returned.
3127       This is distinct from asking for something which does not exist in a
3128       driver which supports column_info as a valid statement handle to an
3129       empty result-set will be returned in this case.
3130
3131       If the arguments don't match any tables then you'll still get a
3132       statement handle, it'll just return no rows.
3133
3134       The statement handle returned has at least the following fields in the
3135       order shown below. Other fields, after these, may also be present.
3136
3137       TABLE_CAT: The catalog identifier.  This field is NULL ("undef") if not
3138       applicable to the data source, which is often the case.  This field is
3139       empty if not applicable to the table.
3140
3141       TABLE_SCHEM: The schema identifier.  This field is NULL ("undef") if
3142       not applicable to the data source, and empty if not applicable to the
3143       table.
3144
3145       TABLE_NAME: The table identifier.  Note: A driver may provide column
3146       metadata not only for base tables, but also for derived objects like
3147       SYNONYMS etc.
3148
3149       COLUMN_NAME: The column identifier.
3150
3151       DATA_TYPE: The concise data type code.
3152
3153       TYPE_NAME: A data source dependent data type name.
3154
3155       COLUMN_SIZE: The column size.  This is the maximum length in characters
3156       for character data types, the number of digits or bits for numeric data
3157       types or the length in the representation of temporal types.  See the
3158       relevant specifications for detailed information.
3159
3160       BUFFER_LENGTH: The length in bytes of transferred data.
3161
3162       DECIMAL_DIGITS: The total number of significant digits to the right of
3163       the decimal point.
3164
3165       NUM_PREC_RADIX: The radix for numeric precision.  The value is 10 or 2
3166       for numeric data types and NULL ("undef") if not applicable.
3167
3168       NULLABLE: Indicates if a column can accept NULLs.  The following values
3169       are defined:
3170
3171         SQL_NO_NULLS          0
3172         SQL_NULLABLE          1
3173         SQL_NULLABLE_UNKNOWN  2
3174
3175       REMARKS: A description of the column.
3176
3177       COLUMN_DEF: The default value of the column, in a format that can be
3178       used directly in an SQL statement.
3179
3180       Note that this may be an expression and not simply the text used for
3181       the default value in the original CREATE TABLE statement. For example,
3182       given:
3183
3184           col1 char(30) default current_user    -- a 'function'
3185           col2 char(30) default 'string'        -- a string literal
3186
3187       where "current_user" is the name of a function, the corresponding
3188       "COLUMN_DEF" values would be:
3189
3190           Database        col1                     col2
3191           --------        ----                     ----
3192           Oracle:         current_user             'string'
3193           Postgres:       "current_user"()         'string'::text
3194           MS SQL:         (user_name())            ('string')
3195
3196       SQL_DATA_TYPE: The SQL data type.
3197
3198       SQL_DATETIME_SUB: The subtype code for datetime and interval data
3199       types.
3200
3201       CHAR_OCTET_LENGTH: The maximum length in bytes of a character or binary
3202       data type column.
3203
3204       ORDINAL_POSITION: The column sequence number (starting with 1).
3205
3206       IS_NULLABLE: Indicates if the column can accept NULLs.  Possible values
3207       are: 'NO', 'YES' and ''.
3208
3209       SQL/CLI defines the following additional columns:
3210
3211         CHAR_SET_CAT
3212         CHAR_SET_SCHEM
3213         CHAR_SET_NAME
3214         COLLATION_CAT
3215         COLLATION_SCHEM
3216         COLLATION_NAME
3217         UDT_CAT
3218         UDT_SCHEM
3219         UDT_NAME
3220         DOMAIN_CAT
3221         DOMAIN_SCHEM
3222         DOMAIN_NAME
3223         SCOPE_CAT
3224         SCOPE_SCHEM
3225         SCOPE_NAME
3226         MAX_CARDINALITY
3227         DTD_IDENTIFIER
3228         IS_SELF_REF
3229
3230       Drivers capable of supplying any of those values should do so in the
3231       corresponding column and supply undef values for the others.
3232
3233       Drivers wishing to provide extra database/driver specific information
3234       should do so in extra columns beyond all those listed above, and use
3235       lowercase field names with the driver-specific prefix (i.e.,
3236       'ora_...'). Applications accessing such fields should do so by name and
3237       not by column number.
3238
3239       The result set is ordered by TABLE_CAT, TABLE_SCHEM, TABLE_NAME and
3240       ORDINAL_POSITION.
3241
3242       Note: There is some overlap with statement handle attributes (in perl)
3243       and SQLDescribeCol (in ODBC). However, SQLColumns provides more
3244       metadata.
3245
3246       See also "Catalog Methods" and "Standards Reference Information".
3247
3248       "primary_key_info"
3249
3250         $sth = $dbh->primary_key_info( $catalog, $schema, $table );
3251
3252         # then $sth->fetchall_arrayref or $sth->fetchall_hashref etc
3253
3254       Returns an active statement handle that can be used to fetch
3255       information about columns that make up the primary key for a table.
3256       The arguments don't accept search patterns (unlike table_info()).
3257
3258       The statement handle will return one row per column, ordered by
3259       TABLE_CAT, TABLE_SCHEM, TABLE_NAME, and KEY_SEQ.  If there is no
3260       primary key then the statement handle will fetch no rows.
3261
3262       Note: The support for the selection criteria, such as $catalog, is
3263       driver specific.  If the driver doesn't support catalogs and/or
3264       schemas, it may ignore these criteria.
3265
3266       The statement handle returned has at least the following fields in the
3267       order shown below. Other fields, after these, may also be present.
3268
3269       TABLE_CAT: The catalog identifier.  This field is NULL ("undef") if not
3270       applicable to the data source, which is often the case.  This field is
3271       empty if not applicable to the table.
3272
3273       TABLE_SCHEM: The schema identifier.  This field is NULL ("undef") if
3274       not applicable to the data source, and empty if not applicable to the
3275       table.
3276
3277       TABLE_NAME: The table identifier.
3278
3279       COLUMN_NAME: The column identifier.
3280
3281       KEY_SEQ: The column sequence number (starting with 1).  Note: This
3282       field is named ORDINAL_POSITION in SQL/CLI.
3283
3284       PK_NAME: The primary key constraint identifier.  This field is NULL
3285       ("undef") if not applicable to the data source.
3286
3287       See also "Catalog Methods" and "Standards Reference Information".
3288
3289       "primary_key"
3290
3291         @key_column_names = $dbh->primary_key( $catalog, $schema, $table );
3292
3293       Simple interface to the primary_key_info() method. Returns a list of
3294       the column names that comprise the primary key of the specified table.
3295       The list is in primary key column sequence order.  If there is no
3296       primary key then an empty list is returned.
3297
3298       "foreign_key_info"
3299
3300         $sth = $dbh->foreign_key_info( $pk_catalog, $pk_schema, $pk_table
3301                                      , $fk_catalog, $fk_schema, $fk_table );
3302
3303         $sth = $dbh->foreign_key_info( $pk_catalog, $pk_schema, $pk_table
3304                                      , $fk_catalog, $fk_schema, $fk_table
3305                                      , \%attr );
3306
3307         # then $sth->fetchall_arrayref or $sth->fetchall_hashref etc
3308
3309       Returns an active statement handle that can be used to fetch
3310       information about foreign keys in and/or referencing the specified
3311       table(s).  The arguments don't accept search patterns (unlike
3312       table_info()).
3313
3314       $pk_catalog, $pk_schema, $pk_table identify the primary (unique) key
3315       table (PKT).
3316
3317       $fk_catalog, $fk_schema, $fk_table identify the foreign key table
3318       (FKT).
3319
3320       If both PKT and FKT are given, the function returns the foreign key, if
3321       any, in table FKT that refers to the primary (unique) key of table PKT.
3322       (Note: In SQL/CLI, the result is implementation-defined.)
3323
3324       If only PKT is given, then the result set contains the primary key of
3325       that table and all foreign keys that refer to it.
3326
3327       If only FKT is given, then the result set contains all foreign keys in
3328       that table and the primary keys to which they refer.  (Note: In
3329       SQL/CLI, the result includes unique keys too.)
3330
3331       For example:
3332
3333         $sth = $dbh->foreign_key_info( undef, $user, 'master');
3334         $sth = $dbh->foreign_key_info( undef, undef,   undef , undef, $user, 'detail');
3335         $sth = $dbh->foreign_key_info( undef, $user, 'master', undef, $user, 'detail');
3336
3337         # then $sth->fetchall_arrayref or $sth->fetchall_hashref etc
3338
3339       Note: The support for the selection criteria, such as $catalog, is
3340       driver specific.  If the driver doesn't support catalogs and/or
3341       schemas, it may ignore these criteria.
3342
3343       The statement handle returned has the following fields in the order
3344       shown below.  Because ODBC never includes unique keys, they define
3345       different columns in the result set than SQL/CLI. SQL/CLI column names
3346       are shown in parentheses.
3347
3348       PKTABLE_CAT    ( UK_TABLE_CAT      ): The primary (unique) key table
3349       catalog identifier.  This field is NULL ("undef") if not applicable to
3350       the data source, which is often the case.  This field is empty if not
3351       applicable to the table.
3352
3353       PKTABLE_SCHEM  ( UK_TABLE_SCHEM    ): The primary (unique) key table
3354       schema identifier.  This field is NULL ("undef") if not applicable to
3355       the data source, and empty if not applicable to the table.
3356
3357       PKTABLE_NAME   ( UK_TABLE_NAME     ): The primary (unique) key table
3358       identifier.
3359
3360       PKCOLUMN_NAME  (UK_COLUMN_NAME    ): The primary (unique) key column
3361       identifier.
3362
3363       FKTABLE_CAT    ( FK_TABLE_CAT      ): The foreign key table catalog
3364       identifier.  This field is NULL ("undef") if not applicable to the data
3365       source, which is often the case.  This field is empty if not applicable
3366       to the table.
3367
3368       FKTABLE_SCHEM  ( FK_TABLE_SCHEM    ): The foreign key table schema
3369       identifier.  This field is NULL ("undef") if not applicable to the data
3370       source, and empty if not applicable to the table.
3371
3372       FKTABLE_NAME   ( FK_TABLE_NAME     ): The foreign key table identifier.
3373
3374       FKCOLUMN_NAME  ( FK_COLUMN_NAME    ): The foreign key column
3375       identifier.
3376
3377       KEY_SEQ        ( ORDINAL_POSITION  ): The column sequence number
3378       (starting with 1).
3379
3380       UPDATE_RULE    ( UPDATE_RULE       ): The referential action for the
3381       UPDATE rule.  The following codes are defined:
3382
3383         CASCADE              0
3384         RESTRICT             1
3385         SET NULL             2
3386         NO ACTION            3
3387         SET DEFAULT          4
3388
3389       DELETE_RULE    ( DELETE_RULE       ): The referential action for the
3390       DELETE rule.  The codes are the same as for UPDATE_RULE.
3391
3392       FK_NAME        ( FK_NAME           ): The foreign key name.
3393
3394       PK_NAME        ( UK_NAME           ): The primary (unique) key name.
3395
3396       DEFERRABILITY  ( DEFERABILITY      ): The deferrability of the foreign
3397       key constraint.  The following codes are defined:
3398
3399         INITIALLY DEFERRED   5
3400         INITIALLY IMMEDIATE  6
3401         NOT DEFERRABLE       7
3402
3403                      ( UNIQUE_OR_PRIMARY ): This column is necessary if a
3404       driver includes all candidate (i.e. primary and alternate) keys in the
3405       result set (as specified by SQL/CLI).  The value of this column is
3406       UNIQUE if the foreign key references an alternate key and PRIMARY if
3407       the foreign key references a primary key, or it may be undefined if the
3408       driver doesn't have access to the information.
3409
3410       See also "Catalog Methods" and "Standards Reference Information".
3411
3412       "statistics_info"
3413
3414       Warning: This method is experimental and may change.
3415
3416         $sth = $dbh->statistics_info( $catalog, $schema, $table, $unique_only, $quick );
3417
3418         # then $sth->fetchall_arrayref or $sth->fetchall_hashref etc
3419
3420       Returns an active statement handle that can be used to fetch
3421       statistical information about a table and its indexes.
3422
3423       The arguments don't accept search patterns (unlike "table_info").
3424
3425       If the boolean argument $unique_only is true, only UNIQUE indexes will
3426       be returned in the result set, otherwise all indexes will be returned.
3427
3428       If the boolean argument $quick is set, the actual statistical
3429       information columns (CARDINALITY and PAGES) will only be returned if
3430       they are readily available from the server, and might not be current.
3431       Some databases may return stale statistics or no statistics at all with
3432       this flag set.
3433
3434       The statement handle will return at most one row per column name per
3435       index, plus at most one row for the entire table itself, ordered by
3436       NON_UNIQUE, TYPE, INDEX_QUALIFIER, INDEX_NAME, and ORDINAL_POSITION.
3437
3438       Note: The support for the selection criteria, such as $catalog, is
3439       driver specific.  If the driver doesn't support catalogs and/or
3440       schemas, it may ignore these criteria.
3441
3442       The statement handle returned has at least the following fields in the
3443       order shown below. Other fields, after these, may also be present.
3444
3445       TABLE_CAT: The catalog identifier.  This field is NULL ("undef") if not
3446       applicable to the data source, which is often the case.  This field is
3447       empty if not applicable to the table.
3448
3449       TABLE_SCHEM: The schema identifier.  This field is NULL ("undef") if
3450       not applicable to the data source, and empty if not applicable to the
3451       table.
3452
3453       TABLE_NAME: The table identifier.
3454
3455       NON_UNIQUE: Unique index indicator.  Returns 0 for unique indexes, 1
3456       for non-unique indexes
3457
3458       INDEX_QUALIFIER: Index qualifier identifier.  The identifier that is
3459       used to qualify the index name when doing a "DROP INDEX"; NULL
3460       ("undef") is returned if an index qualifier is not supported by the
3461       data source.  If a non-NULL (defined) value is returned in this column,
3462       it must be used to qualify the index name on a "DROP INDEX" statement;
3463       otherwise, the TABLE_SCHEM should be used to qualify the index name.
3464
3465       INDEX_NAME: The index identifier.
3466
3467       TYPE: The type of information being returned.  Can be any of the
3468       following values: 'table', 'btree', 'clustered', 'content', 'hashed',
3469       or 'other'.
3470
3471       In the case that this field is 'table', all fields other than
3472       TABLE_CAT, TABLE_SCHEM, TABLE_NAME, TYPE, CARDINALITY, and PAGES will
3473       be NULL ("undef").
3474
3475       ORDINAL_POSITION: Column sequence number (starting with 1).
3476
3477       COLUMN_NAME: The column identifier.
3478
3479       ASC_OR_DESC: Column sort sequence.  "A" for Ascending, "D" for
3480       Descending, or NULL ("undef") if not supported for this index.
3481
3482       CARDINALITY: Cardinality of the table or index.  For indexes, this is
3483       the number of unique values in the index.  For tables, this is the
3484       number of rows in the table.  If not supported, the value will be NULL
3485       ("undef").
3486
3487       PAGES: Number of storage pages used by this table or index.  If not
3488       supported, the value will be NULL ("undef").
3489
3490       FILTER_CONDITION: The index filter condition as a string.  If the index
3491       is not a filtered index, or it cannot be determined whether the index
3492       is a filtered index, this value is NULL ("undef").  If the index is a
3493       filtered index, but the filter condition cannot be determined, this
3494       value is the empty string ''.  Otherwise it will be the literal filter
3495       condition as a string, such as "SALARY <= 4500".
3496
3497       See also "Catalog Methods" and "Standards Reference Information".
3498
3499       "tables"
3500
3501         @names = $dbh->tables( $catalog, $schema, $table, $type );
3502         @names = $dbh->tables;        # deprecated
3503
3504       Simple interface to table_info(). Returns a list of matching table
3505       names, possibly including a catalog/schema prefix.
3506
3507       See "table_info" for a description of the parameters.
3508
3509       If "$dbh->get_info(29)" returns true (29 is SQL_IDENTIFIER_QUOTE_CHAR)
3510       then the table names are constructed and quoted by "quote_identifier"
3511       to ensure they are usable even if they contain whitespace or reserved
3512       words etc. This means that the table names returned will include quote
3513       characters.
3514
3515       "type_info_all"
3516
3517         $type_info_all = $dbh->type_info_all;
3518
3519       Returns a reference to an array which holds information about each data
3520       type variant supported by the database and driver. The array and its
3521       contents should be treated as read-only.
3522
3523       The first item is a reference to an 'index' hash of "Name ="> "Index"
3524       pairs.  The items following that are references to arrays, one per
3525       supported data type variant. The leading index hash defines the names
3526       and order of the fields within the arrays that follow it.  For example:
3527
3528         $type_info_all = [
3529           {   TYPE_NAME         => 0,
3530               DATA_TYPE         => 1,
3531               COLUMN_SIZE       => 2,     # was PRECISION originally
3532               LITERAL_PREFIX    => 3,
3533               LITERAL_SUFFIX    => 4,
3534               CREATE_PARAMS     => 5,
3535               NULLABLE          => 6,
3536               CASE_SENSITIVE    => 7,
3537               SEARCHABLE        => 8,
3538               UNSIGNED_ATTRIBUTE=> 9,
3539               FIXED_PREC_SCALE  => 10,    # was MONEY originally
3540               AUTO_UNIQUE_VALUE => 11,    # was AUTO_INCREMENT originally
3541               LOCAL_TYPE_NAME   => 12,
3542               MINIMUM_SCALE     => 13,
3543               MAXIMUM_SCALE     => 14,
3544               SQL_DATA_TYPE     => 15,
3545               SQL_DATETIME_SUB  => 16,
3546               NUM_PREC_RADIX    => 17,
3547               INTERVAL_PRECISION=> 18,
3548           },
3549           [ 'VARCHAR', SQL_VARCHAR,
3550               undef, "'","'", undef,0, 1,1,0,0,0,undef,1,255, undef
3551           ],
3552           [ 'INTEGER', SQL_INTEGER,
3553               undef,  "", "", undef,0, 0,1,0,0,0,undef,0,  0, 10
3554           ],
3555         ];
3556
3557       More than one row may have the same value in the "DATA_TYPE" field if
3558       there are different ways to spell the type name and/or there are
3559       variants of the type with different attributes (e.g., with and without
3560       "AUTO_UNIQUE_VALUE" set, with and without "UNSIGNED_ATTRIBUTE", etc).
3561
3562       The rows are ordered by "DATA_TYPE" first and then by how closely each
3563       type maps to the corresponding ODBC SQL data type, closest first.
3564
3565       The meaning of the fields is described in the documentation for the
3566       "type_info" method.
3567
3568       An 'index' hash is provided so you don't need to rely on index values
3569       defined above.  However, using DBD::ODBC with some old ODBC drivers may
3570       return older names, shown as comments in the example above.  Another
3571       issue with the index hash is that the lettercase of the keys is not
3572       defined. It is usually uppercase, as show here, but drivers may return
3573       names with any lettercase.
3574
3575       Drivers are also free to return extra driver-specific columns of
3576       information - though it's recommended that they start at column index
3577       50 to leave room for expansion of the DBI/ODBC specification.
3578
3579       The type_info_all() method is not normally used directly.  The
3580       "type_info" method provides a more usable and useful interface to the
3581       data.
3582
3583       "type_info"
3584
3585         @type_info = $dbh->type_info($data_type);
3586
3587       Returns a list of hash references holding information about one or more
3588       variants of $data_type. The list is ordered by "DATA_TYPE" first and
3589       then by how closely each type maps to the corresponding ODBC SQL data
3590       type, closest first.  If called in a scalar context then only the first
3591       (best) element is returned.
3592
3593       If $data_type is undefined or "SQL_ALL_TYPES", then the list will
3594       contain hashes for all data type variants supported by the database and
3595       driver.
3596
3597       If $data_type is an array reference then "type_info" returns the
3598       information for the first type in the array that has any matches.
3599
3600       The keys of the hash follow the same letter case conventions as the
3601       rest of the DBI (see "Naming Conventions and Name Space"). The
3602       following uppercase items should always exist, though may be undef:
3603
3604       TYPE_NAME (string)
3605           Data type name for use in CREATE TABLE statements etc.
3606
3607       DATA_TYPE (integer)
3608           SQL data type number.
3609
3610       COLUMN_SIZE (integer)
3611           For numeric types, this is either the total number of digits (if
3612           the NUM_PREC_RADIX value is 10) or the total number of bits allowed
3613           in the column (if NUM_PREC_RADIX is 2).
3614
3615           For string types, this is the maximum size of the string in
3616           characters.
3617
3618           For date and interval types, this is the maximum number of
3619           characters needed to display the value.
3620
3621       LITERAL_PREFIX (string)
3622           Characters used to prefix a literal. A typical prefix is ""'"" for
3623           characters, or possibly ""0x"" for binary values passed as
3624           hexadecimal.  NULL ("undef") is returned for data types for which
3625           this is not applicable.
3626
3627       LITERAL_SUFFIX (string)
3628           Characters used to suffix a literal. Typically ""'"" for
3629           characters.  NULL ("undef") is returned for data types where this
3630           is not applicable.
3631
3632       CREATE_PARAMS (string)
3633           Parameter names for data type definition. For example,
3634           "CREATE_PARAMS" for a "DECIMAL" would be ""precision,scale"" if the
3635           DECIMAL type should be declared as "DECIMAL("precision,scale")"
3636           where precision and scale are integer values.  For a "VARCHAR" it
3637           would be ""max length"".  NULL ("undef") is returned for data types
3638           for which this is not applicable.
3639
3640       NULLABLE (integer)
3641           Indicates whether the data type accepts a NULL value: 0 or an empty
3642           string = no, 1 = yes, 2 = unknown.
3643
3644       CASE_SENSITIVE (boolean)
3645           Indicates whether the data type is case sensitive in collations and
3646           comparisons.
3647
3648       SEARCHABLE (integer)
3649           Indicates how the data type can be used in a WHERE clause, as
3650           follows:
3651
3652             0 - Cannot be used in a WHERE clause
3653             1 - Only with a LIKE predicate
3654             2 - All comparison operators except LIKE
3655             3 - Can be used in a WHERE clause with any comparison operator
3656
3657       UNSIGNED_ATTRIBUTE (boolean)
3658           Indicates whether the data type is unsigned.  NULL ("undef") is
3659           returned for data types for which this is not applicable.
3660
3661       FIXED_PREC_SCALE (boolean)
3662           Indicates whether the data type always has the same precision and
3663           scale (such as a money type).  NULL ("undef") is returned for data
3664           types for which this is not applicable.
3665
3666       AUTO_UNIQUE_VALUE (boolean)
3667           Indicates whether a column of this data type is automatically set
3668           to a unique value whenever a new row is inserted.  NULL ("undef")
3669           is returned for data types for which this is not applicable.
3670
3671       LOCAL_TYPE_NAME (string)
3672           Localized version of the "TYPE_NAME" for use in dialog with users.
3673           NULL ("undef") is returned if a localized name is not available (in
3674           which case "TYPE_NAME" should be used).
3675
3676       MINIMUM_SCALE (integer)
3677           The minimum scale of the data type. If a data type has a fixed
3678           scale, then "MAXIMUM_SCALE" holds the same value.  NULL ("undef")
3679           is returned for data types for which this is not applicable.
3680
3681       MAXIMUM_SCALE (integer)
3682           The maximum scale of the data type. If a data type has a fixed
3683           scale, then "MINIMUM_SCALE" holds the same value.  NULL ("undef")
3684           is returned for data types for which this is not applicable.
3685
3686       SQL_DATA_TYPE (integer)
3687           This column is the same as the "DATA_TYPE" column, except for
3688           interval and datetime data types.  For interval and datetime data
3689           types, the "SQL_DATA_TYPE" field will return "SQL_INTERVAL" or
3690           "SQL_DATETIME", and the "SQL_DATETIME_SUB" field below will return
3691           the subcode for the specific interval or datetime data type. If
3692           this field is NULL, then the driver does not support or report on
3693           interval or datetime subtypes.
3694
3695       SQL_DATETIME_SUB (integer)
3696           For interval or datetime data types, where the "SQL_DATA_TYPE"
3697           field above is "SQL_INTERVAL" or "SQL_DATETIME", this field will
3698           hold the subcode for the specific interval or datetime data type.
3699           Otherwise it will be NULL ("undef").
3700
3701           Although not mentioned explicitly in the standards, it seems there
3702           is a simple relationship between these values:
3703
3704             DATA_TYPE == (10 * SQL_DATA_TYPE) + SQL_DATETIME_SUB
3705
3706       NUM_PREC_RADIX (integer)
3707           The radix value of the data type. For approximate numeric types,
3708           "NUM_PREC_RADIX" contains the value 2 and "COLUMN_SIZE" holds the
3709           number of bits. For exact numeric types, "NUM_PREC_RADIX" contains
3710           the value 10 and "COLUMN_SIZE" holds the number of decimal digits.
3711           NULL ("undef") is returned either for data types for which this is
3712           not applicable or if the driver cannot report this information.
3713
3714       INTERVAL_PRECISION (integer)
3715           The interval leading precision for interval types. NULL is returned
3716           either for data types for which this is not applicable or if the
3717           driver cannot report this information.
3718
3719       For example, to find the type name for the fields in a select statement
3720       you can do:
3721
3722         @names = map { scalar $dbh->type_info($_)->{TYPE_NAME} } @{ $sth->{TYPE} }
3723
3724       Since DBI and ODBC drivers vary in how they map their types into the
3725       ISO standard types you may need to search for more than one type.
3726       Here's an example looking for a usable type to store a date:
3727
3728         $my_date_type = $dbh->type_info( [ SQL_DATE, SQL_TIMESTAMP ] );
3729
3730       Similarly, to more reliably find a type to store small integers, you
3731       could use a list starting with "SQL_SMALLINT", "SQL_INTEGER",
3732       "SQL_DECIMAL", etc.
3733
3734       See also "Standards Reference Information".
3735
3736       "quote"
3737
3738         $sql = $dbh->quote($value);
3739         $sql = $dbh->quote($value, $data_type);
3740
3741       Quote a string literal for use as a literal value in an SQL statement,
3742       by escaping any special characters (such as quotation marks) contained
3743       within the string and adding the required type of outer quotation
3744       marks.
3745
3746         $sql = sprintf "SELECT foo FROM bar WHERE baz = %s",
3747                       $dbh->quote("Don't");
3748
3749       For most database types, at least those that conform to SQL standards,
3750       quote would return 'Don''t' (including the outer quotation marks). For
3751       others it may return something like 'Don\'t'
3752
3753       An undefined $value value will be returned as the string "NULL"
3754       (without single quotation marks) to match how NULLs are represented in
3755       SQL.
3756
3757       If $data_type is supplied, it is used to try to determine the required
3758       quoting behaviour by using the information returned by "type_info".  As
3759       a special case, the standard numeric types are optimized to return
3760       $value without calling "type_info".
3761
3762       Quote will probably not be able to deal with all possible input (such
3763       as binary data or data containing newlines), and is not related in any
3764       way with escaping or quoting shell meta-characters.
3765
3766       It is valid for the quote() method to return an SQL expression that
3767       evaluates to the desired string. For example:
3768
3769         $quoted = $dbh->quote("one\ntwo\0three")
3770
3771       may return something like:
3772
3773         CONCAT('one', CHAR(12), 'two', CHAR(0), 'three')
3774
3775       The quote() method should not be used with "Placeholders and Bind
3776       Values".
3777
3778       "quote_identifier"
3779
3780         $sql = $dbh->quote_identifier( $name );
3781         $sql = $dbh->quote_identifier( $catalog, $schema, $table, \%attr );
3782
3783       Quote an identifier (table name etc.) for use in an SQL statement, by
3784       escaping any special characters (such as double quotation marks) it
3785       contains and adding the required type of outer quotation marks.
3786
3787       Undefined names are ignored and the remainder are quoted and then
3788       joined together, typically with a dot (".") character. For example:
3789
3790         $id = $dbh->quote_identifier( undef, 'Her schema', 'My table' );
3791
3792       would, for most database types, return "Her schema"."My table"
3793       (including all the double quotation marks).
3794
3795       If three names are supplied then the first is assumed to be a catalog
3796       name and special rules may be applied based on what "get_info" returns
3797       for SQL_CATALOG_NAME_SEPARATOR (41) and SQL_CATALOG_LOCATION (114).
3798       For example, for Oracle:
3799
3800         $id = $dbh->quote_identifier( 'link', 'schema', 'table' );
3801
3802       would return "schema"."table"@"link".
3803
3804       "take_imp_data"
3805
3806         $imp_data = $dbh->take_imp_data;
3807
3808       Leaves the $dbh in an almost dead, zombie-like, state and returns a
3809       binary string of raw implementation data from the driver which
3810       describes the current database connection. Effectively it detaches the
3811       underlying database API connection data from the DBI handle.  After
3812       calling take_imp_data(), all other methods except "DESTROY" will
3813       generate a warning and return undef.
3814
3815       Why would you want to do this? You don't, forget I even mentioned it.
3816       Unless, that is, you're implementing something advanced like a multi-
3817       threaded connection pool. See DBI::Pool.
3818
3819       The returned $imp_data can be passed as a "dbi_imp_data" attribute to a
3820       later connect() call, even in a separate thread in the same process,
3821       where the driver can use it to 'adopt' the existing connection that the
3822       implementation data was taken from.
3823
3824       Some things to keep in mind...
3825
3826       * the $imp_data holds the only reference to the underlying database API
3827       connection data. That connection is still 'live' and won't be cleaned
3828       up properly unless the $imp_data is used to create a new $dbh which is
3829       then allowed to disconnect() normally.
3830
3831       * using the same $imp_data to create more than one other new $dbh at a
3832       time may well lead to unpleasant problems. Don't do that.
3833
3834       Any child statement handles are effectively destroyed when
3835       take_imp_data() is called.
3836
3837       The "take_imp_data" method was added in DBI 1.36 but wasn't useful till
3838       1.49.
3839
3840   Database Handle Attributes
3841       This section describes attributes specific to database handles.
3842
3843       Changes to these database handle attributes do not affect any other
3844       existing or future database handles.
3845
3846       Attempting to set or get the value of an unknown attribute generates a
3847       warning, except for private driver-specific attributes (which all have
3848       names starting with a lowercase letter).
3849
3850       Example:
3851
3852         $h->{AutoCommit} = ...;       # set/write
3853         ... = $h->{AutoCommit};       # get/read
3854
3855       "AutoCommit"
3856
3857       Type: boolean
3858
3859       If true, then database changes cannot be rolled-back (undone).  If
3860       false, then database changes automatically occur within a
3861       "transaction", which must either be committed or rolled back using the
3862       "commit" or "rollback" methods.
3863
3864       Drivers should always default to "AutoCommit" mode (an unfortunate
3865       choice largely forced on the DBI by ODBC and JDBC conventions.)
3866
3867       Attempting to set "AutoCommit" to an unsupported value is a fatal
3868       error.  This is an important feature of the DBI. Applications that need
3869       full transaction behaviour can set "$dbh->{AutoCommit} = 0" (or set
3870       "AutoCommit" to 0 via "connect") without having to check that the value
3871       was assigned successfully.
3872
3873       For the purposes of this description, we can divide databases into
3874       three categories:
3875
3876         Databases which don't support transactions at all.
3877         Databases in which a transaction is always active.
3878         Databases in which a transaction must be explicitly started (C<'BEGIN WORK'>).
3879
3880       * Databases which don't support transactions at all
3881
3882       For these databases, attempting to turn "AutoCommit" off is a fatal
3883       error.  "commit" and "rollback" both issue warnings about being
3884       ineffective while "AutoCommit" is in effect.
3885
3886       * Databases in which a transaction is always active
3887
3888       These are typically mainstream commercial relational databases with
3889       "ANSI standard" transaction behaviour.  If "AutoCommit" is off, then
3890       changes to the database won't have any lasting effect unless "commit"
3891       is called (but see also "disconnect"). If "rollback" is called then any
3892       changes since the last commit are undone.
3893
3894       If "AutoCommit" is on, then the effect is the same as if the DBI called
3895       "commit" automatically after every successful database operation. So
3896       calling "commit" or "rollback" explicitly while "AutoCommit" is on
3897       would be ineffective because the changes would have already been
3898       committed.
3899
3900       Changing "AutoCommit" from off to on will trigger a "commit".
3901
3902       For databases which don't support a specific auto-commit mode, the
3903       driver has to commit each statement automatically using an explicit
3904       "COMMIT" after it completes successfully (and roll it back using an
3905       explicit "ROLLBACK" if it fails).  The error information reported to
3906       the application will correspond to the statement which was executed,
3907       unless it succeeded and the commit or rollback failed.
3908
3909       * Databases in which a transaction must be explicitly started
3910
3911       For these databases, the intention is to have them act like databases
3912       in which a transaction is always active (as described above).
3913
3914       To do this, the driver will automatically begin an explicit transaction
3915       when "AutoCommit" is turned off, or after a "commit" or "rollback" (or
3916       when the application issues the next database operation after one of
3917       those events).
3918
3919       In this way, the application does not have to treat these databases as
3920       a special case.
3921
3922       See "commit", "disconnect" and "Transactions" for other important notes
3923       about transactions.
3924
3925       "Driver"
3926
3927       Type: handle
3928
3929       Holds the handle of the parent driver. The only recommended use for
3930       this is to find the name of the driver using:
3931
3932         $dbh->{Driver}->{Name}
3933
3934       "Name"
3935
3936       Type: string
3937
3938       Holds the "name" of the database. Usually (and recommended to be) the
3939       same as the ""dbi:DriverName:..."" string used to connect to the
3940       database, but with the leading ""dbi:DriverName:"" removed.
3941
3942       "Statement"
3943
3944       Type: string, read-only
3945
3946       Returns the statement string passed to the most recent "prepare" or
3947       "do" method called in this database handle, even if that method failed.
3948       This is especially useful where "RaiseError" is enabled and the
3949       exception handler checks $@ and sees that a 'prepare' method call
3950       failed.
3951
3952       "RowCacheSize"
3953
3954       Type: integer
3955
3956       A hint to the driver indicating the size of the local row cache that
3957       the application would like the driver to use for future "SELECT"
3958       statements.  If a row cache is not implemented, then setting
3959       "RowCacheSize" is ignored and getting the value returns "undef".
3960
3961       Some "RowCacheSize" values have special meaning, as follows:
3962
3963         0 - Automatically determine a reasonable cache size for each C<SELECT>
3964         1 - Disable the local row cache
3965        >1 - Cache this many rows
3966        <0 - Cache as many rows that will fit into this much memory for each C<SELECT>.
3967
3968       Note that large cache sizes may require a very large amount of memory
3969       (cached rows * maximum size of row). Also, a large cache will cause a
3970       longer delay not only for the first fetch, but also whenever the cache
3971       needs refilling.
3972
3973       See also the "RowsInCache" statement handle attribute.
3974
3975       "Username"
3976
3977       Type: string
3978
3979       Returns the username used to connect to the database.
3980

DBI STATEMENT HANDLE OBJECTS

3982       This section lists the methods and attributes associated with DBI
3983       statement handles.
3984
3985   Statement Handle Methods
3986       The DBI defines the following methods for use on DBI statement handles:
3987
3988       "bind_param"
3989
3990         $sth->bind_param($p_num, $bind_value)
3991         $sth->bind_param($p_num, $bind_value, \%attr)
3992         $sth->bind_param($p_num, $bind_value, $bind_type)
3993
3994       The "bind_param" method takes a copy of $bind_value and associates it
3995       (binds it) with a placeholder, identified by $p_num, embedded in the
3996       prepared statement. Placeholders are indicated with question mark
3997       character ("?"). For example:
3998
3999         $dbh->{RaiseError} = 1;        # save having to check each method call
4000         $sth = $dbh->prepare("SELECT name, age FROM people WHERE name LIKE ?");
4001         $sth->bind_param(1, "John%");  # placeholders are numbered from 1
4002         $sth->execute;
4003         DBI::dump_results($sth);
4004
4005       See "Placeholders and Bind Values" for more information.
4006
4007       Data Types for Placeholders
4008
4009       The "\%attr" parameter can be used to hint at the data type the
4010       placeholder should have. This is rarely needed. Typically, the driver
4011       is only interested in knowing if the placeholder should be bound as a
4012       number or a string.
4013
4014         $sth->bind_param(1, $value, { TYPE => SQL_INTEGER });
4015
4016       As a short-cut for the common case, the data type can be passed
4017       directly, in place of the "\%attr" hash reference. This example is
4018       equivalent to the one above:
4019
4020         $sth->bind_param(1, $value, SQL_INTEGER);
4021
4022       The "TYPE" value indicates the standard (non-driver-specific) type for
4023       this parameter. To specify the driver-specific type, the driver may
4024       support a driver-specific attribute, such as "{ ora_type => 97 }".
4025
4026       The SQL_INTEGER and other related constants can be imported using
4027
4028         use DBI qw(:sql_types);
4029
4030       See "DBI Constants" for more information.
4031
4032       The data type is 'sticky' in that bind values passed to execute() are
4033       bound with the data type specified by earlier bind_param() calls, if
4034       any.  Portable applications should not rely on being able to change the
4035       data type after the first "bind_param" call.
4036
4037       Perl only has string and number scalar data types. All database types
4038       that aren't numbers are bound as strings and must be in a format the
4039       database will understand except where the bind_param() TYPE attribute
4040       specifies a type that implies a particular format. For example, given:
4041
4042         $sth->bind_param(1, $value, SQL_DATETIME);
4043
4044       the driver should expect $value to be in the ODBC standard SQL_DATETIME
4045       format, which is 'YYYY-MM-DD HH:MM:SS'. Similarly for SQL_DATE,
4046       SQL_TIME etc.
4047
4048       As an alternative to specifying the data type in the "bind_param" call,
4049       you can let the driver pass the value as the default type ("VARCHAR").
4050       You can then use an SQL function to convert the type within the
4051       statement.  For example:
4052
4053         INSERT INTO price(code, price) VALUES (?, CONVERT(MONEY,?))
4054
4055       The "CONVERT" function used here is just an example. The actual
4056       function and syntax will vary between different databases and is non-
4057       portable.
4058
4059       See also "Placeholders and Bind Values" for more information.
4060
4061       "bind_param_inout"
4062
4063         $rc = $sth->bind_param_inout($p_num, \$bind_value, $max_len)  or die $sth->errstr;
4064         $rv = $sth->bind_param_inout($p_num, \$bind_value, $max_len, \%attr)     or ...
4065         $rv = $sth->bind_param_inout($p_num, \$bind_value, $max_len, $bind_type) or ...
4066
4067       This method acts like "bind_param", but also enables values to be
4068       updated by the statement. The statement is typically a call to a stored
4069       procedure. The $bind_value must be passed as a reference to the actual
4070       value to be used.
4071
4072       Note that unlike "bind_param", the $bind_value variable is not copied
4073       when "bind_param_inout" is called. Instead, the value in the variable
4074       is read at the time "execute" is called.
4075
4076       The additional $max_len parameter specifies the minimum amount of
4077       memory to allocate to $bind_value for the new value. If the value
4078       returned from the database is too big to fit, then the execution should
4079       fail. If unsure what value to use, pick a generous length, i.e., a
4080       length larger than the longest value that would ever be returned.  The
4081       only cost of using a larger value than needed is wasted memory.
4082
4083       Undefined values or "undef" are used to indicate null values.  See also
4084       "Placeholders and Bind Values" for more information.
4085
4086       "bind_param_array"
4087
4088         $rc = $sth->bind_param_array($p_num, $array_ref_or_value)
4089         $rc = $sth->bind_param_array($p_num, $array_ref_or_value, \%attr)
4090         $rc = $sth->bind_param_array($p_num, $array_ref_or_value, $bind_type)
4091
4092       The "bind_param_array" method is used to bind an array of values to a
4093       placeholder embedded in the prepared statement which is to be executed
4094       with "execute_array". For example:
4095
4096         $dbh->{RaiseError} = 1;        # save having to check each method call
4097         $sth = $dbh->prepare("INSERT INTO staff (first_name, last_name, dept) VALUES(?, ?, ?)");
4098         $sth->bind_param_array(1, [ 'John', 'Mary', 'Tim' ]);
4099         $sth->bind_param_array(2, [ 'Booth', 'Todd', 'Robinson' ]);
4100         $sth->bind_param_array(3, "SALES"); # scalar will be reused for each row
4101         $sth->execute_array( { ArrayTupleStatus => \my @tuple_status } );
4102
4103       The %attr ($bind_type) argument is the same as defined for
4104       "bind_param".  Refer to "bind_param" for general details on using
4105       placeholders.
4106
4107       (Note that bind_param_array() can not be used to expand a placeholder
4108       into a list of values for a statement like "SELECT foo WHERE bar IN
4109       (?)".  A placeholder can only ever represent one value per execution.)
4110
4111       Scalar values, including "undef", may also be bound by
4112       "bind_param_array". In which case the same value will be used for each
4113       "execute" call. Driver-specific implementations may behave differently,
4114       e.g., when binding to a stored procedure call, some databases may
4115       permit mixing scalars and arrays as arguments.
4116
4117       The default implementation provided by DBI (for drivers that have not
4118       implemented array binding) is to iteratively call "execute" for each
4119       parameter tuple provided in the bound arrays.  Drivers may provide more
4120       optimized implementations using whatever bulk operation support the
4121       database API provides. The default driver behaviour should match the
4122       default DBI behaviour, but always consult your driver documentation as
4123       there may be driver specific issues to consider.
4124
4125       Note that the default implementation currently only supports non-data
4126       returning statements (INSERT, UPDATE, but not SELECT). Also,
4127       "bind_param_array" and "bind_param" cannot be mixed in the same
4128       statement execution, and "bind_param_array" must be used with
4129       "execute_array"; using "bind_param_array" will have no effect for
4130       "execute".
4131
4132       The "bind_param_array" method was added in DBI 1.22.
4133
4134       "execute"
4135
4136         $rv = $sth->execute                or die $sth->errstr;
4137         $rv = $sth->execute(@bind_values)  or die $sth->errstr;
4138
4139       Perform whatever processing is necessary to execute the prepared
4140       statement.  An "undef" is returned if an error occurs.  A successful
4141       "execute" always returns true regardless of the number of rows
4142       affected, even if it's zero (see below). It is always important to
4143       check the return status of "execute" (and most other DBI methods) for
4144       errors if you're not using "RaiseError".
4145
4146       For a non-"SELECT" statement, "execute" returns the number of rows
4147       affected, if known. If no rows were affected, then "execute" returns
4148       "0E0", which Perl will treat as 0 but will regard as true. Note that it
4149       is not an error for no rows to be affected by a statement. If the
4150       number of rows affected is not known, then "execute" returns -1.
4151
4152       For "SELECT" statements, execute simply "starts" the query within the
4153       database engine. Use one of the fetch methods to retrieve the data
4154       after calling "execute".  The "execute" method does not return the
4155       number of rows that will be returned by the query (because most
4156       databases can't tell in advance), it simply returns a true value.
4157
4158       You can tell if the statement was a "SELECT" statement by checking if
4159       "$sth->{NUM_OF_FIELDS}" is greater than zero after calling "execute".
4160
4161       If any arguments are given, then "execute" will effectively call
4162       "bind_param" for each value before executing the statement.  Values
4163       bound in this way are usually treated as "SQL_VARCHAR" types unless the
4164       driver can determine the correct type (which is rare), or unless
4165       "bind_param" (or "bind_param_inout") has already been used to specify
4166       the type.
4167
4168       Note that passing "execute" an empty array is the same as passing no
4169       arguments at all, which will execute the statement with previously
4170       bound values.  That's probably not what you want.
4171
4172       If execute() is called on a statement handle that's still active
4173       ($sth->{Active} is true) then it should effectively call finish() to
4174       tidy up the previous execution results before starting this new
4175       execution.
4176
4177       "execute_array"
4178
4179         $tuples = $sth->execute_array(\%attr) or die $sth->errstr;
4180         $tuples = $sth->execute_array(\%attr, @bind_values) or die $sth->errstr;
4181
4182         ($tuples, $rows) = $sth->execute_array(\%attr) or die $sth->errstr;
4183         ($tuples, $rows) = $sth->execute_array(\%attr, @bind_values) or die $sth->errstr;
4184
4185       Execute the prepared statement once for each parameter tuple (group of
4186       values) provided either in the @bind_values, or by prior calls to
4187       "bind_param_array", or via a reference passed in \%attr.
4188
4189       When called in scalar context the execute_array() method returns the
4190       number of tuples executed, or "undef" if an error occurred.  Like
4191       execute(), a successful execute_array() always returns true regardless
4192       of the number of tuples executed, even if it's zero. If there were any
4193       errors the ArrayTupleStatus array can be used to discover which tuples
4194       failed and with what errors.
4195
4196       When called in list context the execute_array() method returns two
4197       scalars; $tuples is the same as calling execute_array() in scalar
4198       context and $rows is the number of rows affected for each tuple, if
4199       available or -1 if the driver cannot determine this. NOTE, some drivers
4200       cannot determine the number of rows affected per tuple but can provide
4201       the number of rows affected for the batch.  If you are doing an update
4202       operation the returned rows affected may not be what you expect if, for
4203       instance, one or more of the tuples affected the same row multiple
4204       times.  Some drivers may not yet support list context, in which case
4205       $rows will be undef, or may not be able to provide the number of rows
4206       affected when performing this batch operation, in which case $rows will
4207       be -1.
4208
4209       Bind values for the tuples to be executed may be supplied row-wise by
4210       an "ArrayTupleFetch" attribute, or else column-wise in the @bind_values
4211       argument, or else column-wise by prior calls to "bind_param_array".
4212
4213       Where column-wise binding is used (via the @bind_values argument or
4214       calls to bind_param_array()) the maximum number of elements in any one
4215       of the bound value arrays determines the number of tuples executed.
4216       Placeholders with fewer values in their parameter arrays are treated as
4217       if padded with undef (NULL) values.
4218
4219       If a scalar value is bound, instead of an array reference, it is
4220       treated as a variable length array with all elements having the same
4221       value. It does not influence the number of tuples executed, so if all
4222       bound arrays have zero elements then zero tuples will be executed. If
4223       all bound values are scalars then one tuple will be executed, making
4224       execute_array() act just like execute().
4225
4226       The "ArrayTupleFetch" attribute can be used to specify a reference to a
4227       subroutine that will be called to provide the bind values for each
4228       tuple execution. The subroutine should return an reference to an array
4229       which contains the appropriate number of bind values, or return an
4230       undef if there is no more data to execute.
4231
4232       As a convenience, the "ArrayTupleFetch" attribute can also be used to
4233       specify a statement handle. In which case the fetchrow_arrayref()
4234       method will be called on the given statement handle in order to provide
4235       the bind values for each tuple execution.
4236
4237       The values specified via bind_param_array() or the @bind_values
4238       parameter may be either scalars, or arrayrefs.  If any @bind_values are
4239       given, then "execute_array" will effectively call "bind_param_array"
4240       for each value before executing the statement.  Values bound in this
4241       way are usually treated as "SQL_VARCHAR" types unless the driver can
4242       determine the correct type (which is rare), or unless "bind_param",
4243       "bind_param_inout", "bind_param_array", or "bind_param_inout_array" has
4244       already been used to specify the type.  See "bind_param_array" for
4245       details.
4246
4247       The "ArrayTupleStatus" attribute can be used to specify a reference to
4248       an array which will receive the execute status of each executed
4249       parameter tuple. Note the "ArrayTupleStatus" attribute was mandatory
4250       until DBI 1.38.
4251
4252       For tuples which are successfully executed, the element at the same
4253       ordinal position in the status array is the resulting rowcount (or -1
4254       if unknown).  If the execution of a tuple causes an error, then the
4255       corresponding status array element will be set to a reference to an
4256       array containing "err", "errstr" and "state" set by the failed
4257       execution.
4258
4259       If any tuple execution returns an error, "execute_array" will return
4260       "undef". In that case, the application should inspect the status array
4261       to determine which parameter tuples failed.  Some databases may not
4262       continue executing tuples beyond the first failure. In this case the
4263       status array will either hold fewer elements, or the elements beyond
4264       the failure will be undef.
4265
4266       If all parameter tuples are successfully executed, "execute_array"
4267       returns the number tuples executed.  If no tuples were executed, then
4268       execute_array() returns "0E0", just like execute() does, which Perl
4269       will treat as 0 but will regard as true.
4270
4271       For example:
4272
4273         $sth = $dbh->prepare("INSERT INTO staff (first_name, last_name) VALUES (?, ?)");
4274         my $tuples = $sth->execute_array(
4275             { ArrayTupleStatus => \my @tuple_status },
4276             \@first_names,
4277             \@last_names,
4278         );
4279         if ($tuples) {
4280             print "Successfully inserted $tuples records\n";
4281         }
4282         else {
4283             for my $tuple (0..@last_names-1) {
4284                 my $status = $tuple_status[$tuple];
4285                 $status = [0, "Skipped"] unless defined $status;
4286                 next unless ref $status;
4287                 printf "Failed to insert (%s, %s): %s\n",
4288                     $first_names[$tuple], $last_names[$tuple], $status->[1];
4289             }
4290         }
4291
4292       Support for data returning statements such as SELECT is driver-specific
4293       and subject to change. At present, the default implementation provided
4294       by DBI only supports non-data returning statements.
4295
4296       Transaction semantics when using array binding are driver and database
4297       specific.  If "AutoCommit" is on, the default DBI implementation will
4298       cause each parameter tuple to be individually committed (or rolled back
4299       in the event of an error). If "AutoCommit" is off, the application is
4300       responsible for explicitly committing the entire set of bound parameter
4301       tuples.  Note that different drivers and databases may have different
4302       behaviours when some parameter tuples cause failures. In some cases,
4303       the driver or database may automatically rollback the effect of all
4304       prior parameter tuples that succeeded in the transaction; other drivers
4305       or databases may retain the effect of prior successfully executed
4306       parameter tuples. Be sure to check your driver and database for its
4307       specific behaviour.
4308
4309       Note that, in general, performance will usually be better with
4310       "AutoCommit" turned off, and using explicit "commit" after each
4311       "execute_array" call.
4312
4313       The "execute_array" method was added in DBI 1.22, and ArrayTupleFetch
4314       was added in 1.36.
4315
4316       "execute_for_fetch"
4317
4318         $tuples = $sth->execute_for_fetch($fetch_tuple_sub);
4319         $tuples = $sth->execute_for_fetch($fetch_tuple_sub, \@tuple_status);
4320
4321         ($tuples, $rows) = $sth->execute_for_fetch($fetch_tuple_sub);
4322         ($tuples, $rows) = $sth->execute_for_fetch($fetch_tuple_sub, \@tuple_status);
4323
4324       The execute_for_fetch() method is used to perform bulk operations and
4325       although it is most often used via the execute_array() method you can
4326       use it directly. The main difference between execute_array and
4327       execute_for_fetch is the former does column or row-wise binding and the
4328       latter uses row-wise binding.
4329
4330       The fetch subroutine, referenced by $fetch_tuple_sub, is expected to
4331       return a reference to an array (known as a 'tuple') or undef.
4332
4333       The execute_for_fetch() method calls $fetch_tuple_sub, without any
4334       parameters, until it returns a false value. Each tuple returned is used
4335       to provide bind values for an $sth->execute(@$tuple) call.
4336
4337       In scalar context execute_for_fetch() returns "undef" if there were any
4338       errors and the number of tuples executed otherwise. Like execute() and
4339       execute_array() a zero is returned as "0E0" so execute_for_fetch() is
4340       only false on error.  If there were any errors the @tuple_status array
4341       can be used to discover which tuples failed and with what errors.
4342
4343       When called in list context execute_for_fetch() returns two scalars;
4344       $tuples is the same as calling execute_for_fetch() in scalar context
4345       and $rows is the sum of the number of rows affected for each tuple, if
4346       available or -1 if the driver cannot determine this.  If you are doing
4347       an update operation the returned rows affected may not be what you
4348       expect if, for instance, one or more of the tuples affected the same
4349       row multiple times.  Some drivers may not yet support list context, in
4350       which case $rows will be undef, or may not be able to provide the
4351       number of rows affected when performing this batch operation, in which
4352       case $rows will be -1.
4353
4354       If \@tuple_status is passed then the execute_for_fetch method uses it
4355       to return status information. The tuple_status array holds one element
4356       per tuple. If the corresponding execute() did not fail then the element
4357       holds the return value from execute(), which is typically a row count.
4358       If the execute() did fail then the element holds a reference to an
4359       array containing ($sth->err, $sth->errstr, $sth->state).
4360
4361       If the driver detects an error that it knows means no further tuples
4362       can be executed then it may return, with an error status, even though
4363       $fetch_tuple_sub may still have more tuples to be executed.
4364
4365       Although each tuple returned by $fetch_tuple_sub is effectively used to
4366       call $sth->execute(@$tuple_array_ref) the exact timing may vary.
4367       Drivers are free to accumulate sets of tuples to pass to the database
4368       server in bulk group operations for more efficient execution.  However,
4369       the $fetch_tuple_sub is specifically allowed to return the same array
4370       reference each time (which is what fetchrow_arrayref() usually does).
4371
4372       For example:
4373
4374         my $sel = $dbh1->prepare("select foo, bar from table1");
4375         $sel->execute;
4376
4377         my $ins = $dbh2->prepare("insert into table2 (foo, bar) values (?,?)");
4378         my $fetch_tuple_sub = sub { $sel->fetchrow_arrayref };
4379
4380         my @tuple_status;
4381         $rc = $ins->execute_for_fetch($fetch_tuple_sub, \@tuple_status);
4382         my @errors = grep { ref $_ } @tuple_status;
4383
4384       Similarly, if you already have an array containing the data rows to be
4385       processed you'd use a subroutine to shift off and return each array ref
4386       in turn:
4387
4388         $ins->execute_for_fetch( sub { shift @array_of_arrays }, \@tuple_status);
4389
4390       The "execute_for_fetch" method was added in DBI 1.38.
4391
4392       "last_insert_id"
4393
4394         $rv = $sth->last_insert_id();
4395         $rv = $sth->last_insert_id($catalog, $schema, $table, $field);
4396         $rv = $sth->last_insert_id($catalog, $schema, $table, $field, \%attr);
4397
4398       Returns a value 'identifying' the row inserted by last execution of the
4399       statement $sth, if possible.
4400
4401       For some drivers the value may be 'identifying' the row inserted by the
4402       last executed statement, not by $sth.
4403
4404       See database handle method last_insert_id for all details.
4405
4406       The "last_insert_id" statement method was added in DBI 1.642.
4407
4408       "fetchrow_arrayref"
4409
4410         $ary_ref = $sth->fetchrow_arrayref;
4411         $ary_ref = $sth->fetch;    # alias
4412
4413       Fetches the next row of data and returns a reference to an array
4414       holding the field values.  Null fields are returned as "undef" values
4415       in the array.  This is the fastest way to fetch data, particularly if
4416       used with "$sth->bind_columns".
4417
4418       If there are no more rows or if an error occurs, then
4419       "fetchrow_arrayref" returns an "undef". You should check "$sth->err"
4420       afterwards (or use the "RaiseError" attribute) to discover if the
4421       "undef" returned was due to an error.
4422
4423       Note that the same array reference is returned for each fetch, so don't
4424       store the reference and then use it after a later fetch.  Also, the
4425       elements of the array are also reused for each row, so take care if you
4426       want to take a reference to an element. See also "bind_columns".
4427
4428       "fetchrow_array"
4429
4430        @ary = $sth->fetchrow_array;
4431
4432       An alternative to "fetchrow_arrayref". Fetches the next row of data and
4433       returns it as a list containing the field values.  Null fields are
4434       returned as "undef" values in the list.
4435
4436       If there are no more rows or if an error occurs, then "fetchrow_array"
4437       returns an empty list. You should check "$sth->err" afterwards (or use
4438       the "RaiseError" attribute) to discover if the empty list returned was
4439       due to an error.
4440
4441       If called in a scalar context for a statement handle that has more than
4442       one column, it is undefined whether the driver will return the value of
4443       the first column or the last. So don't do that.  Also, in a scalar
4444       context, an "undef" is returned if there are no more rows or if an
4445       error occurred. That "undef" can't be distinguished from an "undef"
4446       returned because the first field value was NULL.  For these reasons you
4447       should exercise some caution if you use "fetchrow_array" in a scalar
4448       context.
4449
4450       "fetchrow_hashref"
4451
4452        $hash_ref = $sth->fetchrow_hashref;
4453        $hash_ref = $sth->fetchrow_hashref($name);
4454
4455       An alternative to "fetchrow_arrayref". Fetches the next row of data and
4456       returns it as a reference to a hash containing field name and field
4457       value pairs.  Null fields are returned as "undef" values in the hash.
4458
4459       If there are no more rows or if an error occurs, then
4460       "fetchrow_hashref" returns an "undef". You should check "$sth->err"
4461       afterwards (or use the "RaiseError" attribute) to discover if the
4462       "undef" returned was due to an error.
4463
4464       The optional $name parameter specifies the name of the statement handle
4465       attribute. For historical reasons it defaults to ""NAME"", however
4466       using either ""NAME_lc"" or ""NAME_uc"" is recommended for portability.
4467
4468       The keys of the hash are the same names returned by "$sth->{$name}". If
4469       more than one field has the same name, there will only be one entry in
4470       the returned hash for those fields, so statements like ""select foo,
4471       foo from bar"" will return only a single key from "fetchrow_hashref".
4472       In these cases use column aliases or "fetchrow_arrayref".  Note that it
4473       is the database server (and not the DBD implementation) which provides
4474       the name for fields containing functions like "count(*)" or
4475       ""max(c_foo)"" and they may clash with existing column names (most
4476       databases don't care about duplicate column names in a result-set). If
4477       you want these to return as unique names that are the same across
4478       databases, use aliases, as in ""select count(*) as cnt"" or ""select
4479       max(c_foo) mx_foo, ..."" depending on the syntax your database
4480       supports.
4481
4482       Because of the extra work "fetchrow_hashref" and Perl have to perform,
4483       it is not as efficient as "fetchrow_arrayref" or "fetchrow_array".
4484
4485       By default a reference to a new hash is returned for each row.  It is
4486       likely that a future version of the DBI will support an attribute which
4487       will enable the same hash to be reused for each row. This will give a
4488       significant performance boost, but it won't be enabled by default
4489       because of the risk of breaking old code.
4490
4491       "fetchall_arrayref"
4492
4493         $tbl_ary_ref = $sth->fetchall_arrayref;
4494         $tbl_ary_ref = $sth->fetchall_arrayref( $slice );
4495         $tbl_ary_ref = $sth->fetchall_arrayref( $slice, $max_rows  );
4496
4497       The "fetchall_arrayref" method can be used to fetch all the data to be
4498       returned from a prepared and executed statement handle. It returns a
4499       reference to an array that contains one reference per row.
4500
4501       If called on an inactive statement handle, "fetchall_arrayref" returns
4502       undef.
4503
4504       If there are no rows left to return from an active statement handle,
4505       "fetchall_arrayref" returns a reference to an empty array. If an error
4506       occurs, "fetchall_arrayref" returns the data fetched thus far, which
4507       may be none.  You should check "$sth->err" afterwards (or use the
4508       "RaiseError" attribute) to discover if the data is complete or was
4509       truncated due to an error.
4510
4511       If $slice is an array reference, "fetchall_arrayref" uses
4512       "fetchrow_arrayref" to fetch each row as an array ref. If the $slice
4513       array is not empty then it is used as a slice to select individual
4514       columns by perl array index number (starting at 0, unlike column and
4515       parameter numbers which start at 1).
4516
4517       With no parameters, or if $slice is undefined, "fetchall_arrayref" acts
4518       as if passed an empty array ref.
4519
4520       For example, to fetch just the first column of every row:
4521
4522         $tbl_ary_ref = $sth->fetchall_arrayref([0]);
4523
4524       To fetch the second to last and last column of every row:
4525
4526         $tbl_ary_ref = $sth->fetchall_arrayref([-2,-1]);
4527
4528       Those two examples both return a reference to an array of array refs.
4529
4530       If $slice is a hash reference, "fetchall_arrayref" fetches each row as
4531       a hash reference. If the $slice hash is empty then the keys in the
4532       hashes have whatever name lettercase is returned by default. (See
4533       "FetchHashKeyName" attribute.) If the $slice hash is not empty, then it
4534       is used as a slice to select individual columns by name. The values of
4535       the hash should be set to 1.  The key names of the returned hashes
4536       match the letter case of the names in the parameter hash, regardless of
4537       the "FetchHashKeyName" attribute.
4538
4539       For example, to fetch all fields of every row as a hash ref:
4540
4541         $tbl_ary_ref = $sth->fetchall_arrayref({});
4542
4543       To fetch only the fields called "foo" and "bar" of every row as a hash
4544       ref (with keys named "foo" and "BAR", regardless of the original
4545       capitalization):
4546
4547         $tbl_ary_ref = $sth->fetchall_arrayref({ foo=>1, BAR=>1 });
4548
4549       Those two examples both return a reference to an array of hash refs.
4550
4551       If $slice is a reference to a hash reference, that hash is used to
4552       select and rename columns. The keys are 0-based column index numbers
4553       and the values are the corresponding keys for the returned row hashes.
4554
4555       For example, to fetch only the first and second columns of every row as
4556       a hash ref (with keys named "k" and "v" regardless of their original
4557       names):
4558
4559         $tbl_ary_ref = $sth->fetchall_arrayref( \{ 0 => 'k', 1 => 'v' } );
4560
4561       If $max_rows is defined and greater than or equal to zero then it is
4562       used to limit the number of rows fetched before returning.
4563       fetchall_arrayref() can then be called again to fetch more rows.  This
4564       is especially useful when you need the better performance of
4565       fetchall_arrayref() but don't have enough memory to fetch and return
4566       all the rows in one go.
4567
4568       Here's an example (assumes RaiseError is enabled):
4569
4570         my $rows = []; # cache for batches of rows
4571         while( my $row = ( shift(@$rows) || # get row from cache, or reload cache:
4572                            shift(@{$rows=$sth->fetchall_arrayref(undef,10_000)||[]}) )
4573         ) {
4574           ...
4575         }
4576
4577       That might be the fastest way to fetch and process lots of rows using
4578       the DBI, but it depends on the relative cost of method calls vs memory
4579       allocation.
4580
4581       A standard "while" loop with column binding is often faster because the
4582       cost of allocating memory for the batch of rows is greater than the
4583       saving by reducing method calls. It's possible that the DBI may provide
4584       a way to reuse the memory of a previous batch in future, which would
4585       then shift the balance back towards fetchall_arrayref().
4586
4587       "fetchall_hashref"
4588
4589         $hash_ref = $sth->fetchall_hashref($key_field);
4590
4591       The "fetchall_hashref" method can be used to fetch all the data to be
4592       returned from a prepared and executed statement handle. It returns a
4593       reference to a hash containing a key for each distinct value of the
4594       $key_field column that was fetched. For each key the corresponding
4595       value is a reference to a hash containing all the selected columns and
4596       their values, as returned by "fetchrow_hashref()".
4597
4598       If there are no rows to return, "fetchall_hashref" returns a reference
4599       to an empty hash. If an error occurs, "fetchall_hashref" returns the
4600       data fetched thus far, which may be none.  You should check "$sth->err"
4601       afterwards (or use the "RaiseError" attribute) to discover if the data
4602       is complete or was truncated due to an error.
4603
4604       The $key_field parameter provides the name of the field that holds the
4605       value to be used for the key for the returned hash.  For example:
4606
4607         $dbh->{FetchHashKeyName} = 'NAME_lc';
4608         $sth = $dbh->prepare("SELECT FOO, BAR, ID, NAME, BAZ FROM TABLE");
4609         $sth->execute;
4610         $hash_ref = $sth->fetchall_hashref('id');
4611         print "Name for id 42 is $hash_ref->{42}->{name}\n";
4612
4613       The $key_field parameter can also be specified as an integer column
4614       number (counting from 1).  If $key_field doesn't match any column in
4615       the statement, as a name first then as a number, then an error is
4616       returned.
4617
4618       For queries returning more than one 'key' column, you can specify
4619       multiple column names by passing $key_field as a reference to an array
4620       containing one or more key column names (or index numbers).  For
4621       example:
4622
4623         $sth = $dbh->prepare("SELECT foo, bar, baz FROM table");
4624         $sth->execute;
4625         $hash_ref = $sth->fetchall_hashref( [ qw(foo bar) ] );
4626         print "For foo 42 and bar 38, baz is $hash_ref->{42}->{38}->{baz}\n";
4627
4628       The fetchall_hashref() method is normally used only where the key
4629       fields values for each row are unique.  If multiple rows are returned
4630       with the same values for the key fields then later rows overwrite
4631       earlier ones.
4632
4633       "finish"
4634
4635         $rc  = $sth->finish;
4636
4637       Indicate that no more data will be fetched from this statement handle
4638       before it is either executed again or destroyed.  You almost certainly
4639       do not need to call this method.
4640
4641       Adding calls to "finish" after loop that fetches all rows is a common
4642       mistake, don't do it, it can mask genuine problems like uncaught fetch
4643       errors.
4644
4645       When all the data has been fetched from a "SELECT" statement, the
4646       driver will automatically call "finish" for you. So you should not call
4647       it explicitly except when you know that you've not fetched all the data
4648       from a statement handle and the handle won't be destroyed soon.
4649
4650       The most common example is when you only want to fetch just one row,
4651       but in that case the "selectrow_*" methods are usually better anyway.
4652
4653       Consider a query like:
4654
4655         SELECT foo FROM table WHERE bar=? ORDER BY baz
4656
4657       on a very large table. When executed, the database server will have to
4658       use temporary buffer space to store the sorted rows. If, after
4659       executing the handle and selecting just a few rows, the handle won't be
4660       re-executed for some time and won't be destroyed, the "finish" method
4661       can be used to tell the server that the buffer space can be freed.
4662
4663       Calling "finish" resets the "Active" attribute for the statement.  It
4664       may also make some statement handle attributes (such as "NAME" and
4665       "TYPE") unavailable if they have not already been accessed (and thus
4666       cached).
4667
4668       The "finish" method does not affect the transaction status of the
4669       database connection.  It has nothing to do with transactions. It's
4670       mostly an internal "housekeeping" method that is rarely needed.  See
4671       also "disconnect" and the "Active" attribute.
4672
4673       The "finish" method should have been called "discard_pending_rows".
4674
4675       "rows"
4676
4677         $rv = $sth->rows;
4678
4679       Returns the number of rows affected by the last row affecting command,
4680       or -1 if the number of rows is not known or not available.
4681
4682       Generally, you can only rely on a row count after a non-"SELECT"
4683       "execute" (for some specific operations like "UPDATE" and "DELETE"), or
4684       after fetching all the rows of a "SELECT" statement.
4685
4686       For "SELECT" statements, it is generally not possible to know how many
4687       rows will be returned except by fetching them all.  Some drivers will
4688       return the number of rows the application has fetched so far, but
4689       others may return -1 until all rows have been fetched.  So use of the
4690       "rows" method or $DBI::rows with "SELECT" statements is not
4691       recommended.
4692
4693       One alternative method to get a row count for a "SELECT" is to execute
4694       a "SELECT COUNT(*) FROM ..." SQL statement with the same "..." as your
4695       query and then fetch the row count from that.
4696
4697       "bind_col"
4698
4699         $rc = $sth->bind_col($column_number, \$var_to_bind);
4700         $rc = $sth->bind_col($column_number, \$var_to_bind, \%attr );
4701         $rc = $sth->bind_col($column_number, \$var_to_bind, $bind_type );
4702
4703       Binds a Perl variable and/or some attributes to an output column
4704       (field) of a "SELECT" statement.  Column numbers count up from 1.  You
4705       do not need to bind output columns in order to fetch data.  For maximum
4706       portability between drivers, bind_col() should be called after
4707       execute() and not before.  See also "bind_columns" for an example.
4708
4709       The binding is performed at a low level using Perl aliasing.  Whenever
4710       a row is fetched from the database $var_to_bind appears to be
4711       automatically updated simply because it now refers to the same memory
4712       location as the corresponding column value.  This makes using bound
4713       variables very efficient.  Binding a tied variable doesn't work,
4714       currently.
4715
4716       The "bind_param" method performs a similar, but opposite, function for
4717       input variables.
4718
4719       Data Types for Column Binding
4720
4721       The "\%attr" parameter can be used to hint at the data type formatting
4722       the column should have. For example, you can use:
4723
4724         $sth->bind_col(1, undef, { TYPE => SQL_DATETIME });
4725
4726       to specify that you'd like the column (which presumably is some kind of
4727       datetime type) to be returned in the standard format for SQL_DATETIME,
4728       which is 'YYYY-MM-DD HH:MM:SS', rather than the native formatting the
4729       database would normally use.
4730
4731       There's no $var_to_bind in that example to emphasize the point that
4732       bind_col() works on the underlying column and not just a particular
4733       bound variable.
4734
4735       As a short-cut for the common case, the data type can be passed
4736       directly, in place of the "\%attr" hash reference. This example is
4737       equivalent to the one above:
4738
4739         $sth->bind_col(1, undef, SQL_DATETIME);
4740
4741       The "TYPE" value indicates the standard (non-driver-specific) type for
4742       this parameter. To specify the driver-specific type, the driver may
4743       support a driver-specific attribute, such as "{ ora_type => 97 }".
4744
4745       The SQL_DATETIME and other related constants can be imported using
4746
4747         use DBI qw(:sql_types);
4748
4749       See "DBI Constants" for more information.
4750
4751       Few drivers support specifying a data type via a "bind_col" call (most
4752       will simply ignore the data type). Fewer still allow the data type to
4753       be altered once set. If you do set a column type the type should remain
4754       sticky through further calls to bind_col for the same column if the
4755       type is not overridden (this is important for instance when you are
4756       using a slice in fetchall_arrayref).
4757
4758       The TYPE attribute for bind_col() was first specified in DBI 1.41.
4759
4760       From DBI 1.611, drivers can use the "TYPE" attribute to attempt to cast
4761       the bound scalar to a perl type which more closely matches "TYPE". At
4762       present DBI supports "SQL_INTEGER", "SQL_DOUBLE" and "SQL_NUMERIC". See
4763       "sql_type_cast" for details of how types are cast.
4764
4765       Other attributes for Column Binding
4766
4767       The "\%attr" parameter may also contain the following attributes:
4768
4769       "StrictlyTyped"
4770           If a "TYPE" attribute is passed to bind_col, then the driver will
4771           attempt to change the bound perl scalar to match the type more
4772           closely. If the bound value cannot be cast to the requested "TYPE"
4773           then by default it is left untouched and no error is generated. If
4774           you specify "StrictlyTyped" as 1 and the cast fails, this will
4775           generate an error.
4776
4777           This attribute was first added in DBI 1.611. When 1.611 was
4778           released few drivers actually supported this attribute but
4779           DBD::Oracle and DBD::ODBC should from versions 1.24.
4780
4781       "DiscardString"
4782           When the "TYPE" attribute is passed to "bind_col" and the driver
4783           successfully casts the bound perl scalar to a non-string type then
4784           if "DiscardString" is set to 1, the string portion of the scalar
4785           will be discarded. By default, "DiscardString" is not set.
4786
4787           This attribute was first added in DBI 1.611. When 1.611 was
4788           released few drivers actually supported this attribute but
4789           DBD::Oracle and DBD::ODBC should from versions 1.24.
4790
4791       "bind_columns"
4792
4793         $rc = $sth->bind_columns(@list_of_refs_to_vars_to_bind);
4794
4795       Calls "bind_col" for each column of the "SELECT" statement.
4796
4797       The list of references should have the same number of elements as the
4798       number of columns in the "SELECT" statement. If it doesn't then
4799       "bind_columns" will bind the elements given, up to the number of
4800       columns, and then return an error.
4801
4802       For maximum portability between drivers, bind_columns() should be
4803       called after execute() and not before.
4804
4805       For example:
4806
4807         $dbh->{RaiseError} = 1; # do this, or check every call for errors
4808         $sth = $dbh->prepare(q{ SELECT region, sales FROM sales_by_region });
4809         $sth->execute;
4810         my ($region, $sales);
4811
4812         # Bind Perl variables to columns:
4813         $rv = $sth->bind_columns(\$region, \$sales);
4814
4815         # you can also use Perl's \(...) syntax (see perlref docs):
4816         #     $sth->bind_columns(\($region, $sales));
4817
4818         # Column binding is the most efficient way to fetch data
4819         while ($sth->fetch) {
4820             print "$region: $sales\n";
4821         }
4822
4823       For compatibility with old scripts, the first parameter will be ignored
4824       if it is "undef" or a hash reference.
4825
4826       Here's a more fancy example that binds columns to the values inside a
4827       hash (thanks to H.Merijn Brand):
4828
4829         $sth->execute;
4830         my %row;
4831         $sth->bind_columns( \( @row{ @{$sth->{NAME_lc} } } ));
4832         while ($sth->fetch) {
4833             print "$row{region}: $row{sales}\n";
4834         }
4835
4836       "dump_results"
4837
4838         $rows = $sth->dump_results($maxlen, $lsep, $fsep, $fh);
4839
4840       Fetches all the rows from $sth, calls "DBI::neat_list" for each row,
4841       and prints the results to $fh (defaults to "STDOUT") separated by $lsep
4842       (default "\n"). $fsep defaults to ", " and $maxlen defaults to 35.
4843
4844       This method is designed as a handy utility for prototyping and testing
4845       queries. Since it uses "neat_list" to format and edit the string for
4846       reading by humans, it is not recommended for data transfer
4847       applications.
4848
4849   Statement Handle Attributes
4850       This section describes attributes specific to statement handles. Most
4851       of these attributes are read-only.
4852
4853       Changes to these statement handle attributes do not affect any other
4854       existing or future statement handles.
4855
4856       Attempting to set or get the value of an unknown attribute generates a
4857       warning, except for private driver specific attributes (which all have
4858       names starting with a lowercase letter).
4859
4860       Example:
4861
4862         ... = $h->{NUM_OF_FIELDS};    # get/read
4863
4864       Some drivers cannot provide valid values for some or all of these
4865       attributes until after "$sth->execute" has been successfully called.
4866       Typically the attribute will be "undef" in these situations.
4867
4868       Some attributes, like NAME, are not appropriate to some types of
4869       statement, like SELECT. Typically the attribute will be "undef" in
4870       these situations.
4871
4872       For drivers which support stored procedures and multiple result sets
4873       (see "more_results") these attributes relate to the current result set.
4874
4875       See also "finish" to learn more about the effect it may have on some
4876       attributes.
4877
4878       "NUM_OF_FIELDS"
4879
4880       Type: integer, read-only
4881
4882       Number of fields (columns) in the data the prepared statement may
4883       return.  Statements that don't return rows of data, like "DELETE" and
4884       "CREATE" set "NUM_OF_FIELDS" to 0 (though it may be undef in some
4885       drivers).
4886
4887       "NUM_OF_PARAMS"
4888
4889       Type: integer, read-only
4890
4891       The number of parameters (placeholders) in the prepared statement.  See
4892       SUBSTITUTION VARIABLES below for more details.
4893
4894       "NAME"
4895
4896       Type: array-ref, read-only
4897
4898       Returns a reference to an array of field names for each column. The
4899       names may contain spaces but should not be truncated or have any
4900       trailing space. Note that the names have the letter case (upper, lower
4901       or mixed) as returned by the driver being used. Portable applications
4902       should use "NAME_lc" or "NAME_uc".
4903
4904         print "First column name: $sth->{NAME}->[0]\n";
4905
4906       Also note that the name returned for (aggregate) functions like
4907       count(*) or "max(c_foo)" is determined by the database server and not
4908       by "DBI" or the "DBD" backend.
4909
4910       "NAME_lc"
4911
4912       Type: array-ref, read-only
4913
4914       Like "/NAME" but always returns lowercase names.
4915
4916       "NAME_uc"
4917
4918       Type: array-ref, read-only
4919
4920       Like "/NAME" but always returns uppercase names.
4921
4922       "NAME_hash"
4923
4924       Type: hash-ref, read-only
4925
4926       "NAME_lc_hash"
4927
4928       Type: hash-ref, read-only
4929
4930       "NAME_uc_hash"
4931
4932       Type: hash-ref, read-only
4933
4934       The "NAME_hash", "NAME_lc_hash", and "NAME_uc_hash" attributes return
4935       column name information as a reference to a hash.
4936
4937       The keys of the hash are the names of the columns.  The letter case of
4938       the keys corresponds to the letter case returned by the "NAME",
4939       "NAME_lc", and "NAME_uc" attributes respectively (as described above).
4940
4941       The value of each hash entry is the perl index number of the
4942       corresponding column (counting from 0). For example:
4943
4944         $sth = $dbh->prepare("select Id, Name from table");
4945         $sth->execute;
4946         @row = $sth->fetchrow_array;
4947         print "Name $row[ $sth->{NAME_lc_hash}{name} ]\n";
4948
4949       "TYPE"
4950
4951       Type: array-ref, read-only
4952
4953       Returns a reference to an array of integer values for each column. The
4954       value indicates the data type of the corresponding column.
4955
4956       The values correspond to the international standards (ANSI X3.135 and
4957       ISO/IEC 9075) which, in general terms, means ODBC. Driver-specific
4958       types that don't exactly match standard types should generally return
4959       the same values as an ODBC driver supplied by the makers of the
4960       database. That might include private type numbers in ranges the vendor
4961       has officially registered with the ISO working group:
4962
4963         ftp://sqlstandards.org/SC32/SQL_Registry/
4964
4965       Where there's no vendor-supplied ODBC driver to be compatible with, the
4966       DBI driver can use type numbers in the range that is now officially
4967       reserved for use by the DBI: -9999 to -9000.
4968
4969       All possible values for "TYPE" should have at least one entry in the
4970       output of the "type_info_all" method (see "type_info_all").
4971
4972       "PRECISION"
4973
4974       Type: array-ref, read-only
4975
4976       Returns a reference to an array of integer values for each column.
4977
4978       For numeric columns, the value is the maximum number of digits (without
4979       considering a sign character or decimal point). Note that the "display
4980       size" for floating point types (REAL, FLOAT, DOUBLE) can be up to 7
4981       characters greater than the precision (for the sign + decimal point +
4982       the letter E + a sign + 2 or 3 digits).
4983
4984       For any character type column the value is the OCTET_LENGTH, in other
4985       words the number of bytes, not characters.
4986
4987       (More recent standards refer to this as COLUMN_SIZE but we stick with
4988       PRECISION for backwards compatibility.)
4989
4990       "SCALE"
4991
4992       Type: array-ref, read-only
4993
4994       Returns a reference to an array of integer values for each column.
4995       NULL ("undef") values indicate columns where scale is not applicable.
4996
4997       "NULLABLE"
4998
4999       Type: array-ref, read-only
5000
5001       Returns a reference to an array indicating the possibility of each
5002       column returning a null.  Possible values are 0 (or an empty string) =
5003       no, 1 = yes, 2 = unknown.
5004
5005         print "First column may return NULL\n" if $sth->{NULLABLE}->[0];
5006
5007       "CursorName"
5008
5009       Type: string, read-only
5010
5011       Returns the name of the cursor associated with the statement handle, if
5012       available. If not available or if the database driver does not support
5013       the "where current of ..." SQL syntax, then it returns "undef".
5014
5015       "Database"
5016
5017       Type: dbh, read-only
5018
5019       Returns the parent $dbh of the statement handle.
5020
5021       "Statement"
5022
5023       Type: string, read-only
5024
5025       Returns the statement string passed to the "prepare" method.
5026
5027       "ParamValues"
5028
5029       Type: hash ref, read-only
5030
5031       Returns a reference to a hash containing the values currently bound to
5032       placeholders.  The keys of the hash are the 'names' of the
5033       placeholders, typically integers starting at 1.  Returns undef if not
5034       supported by the driver.
5035
5036       See "ShowErrorStatement" for an example of how this is used.
5037
5038       * Keys:
5039
5040       If the driver supports "ParamValues" but no values have been bound yet
5041       then the driver should return a hash with placeholders names in the
5042       keys but all the values undef, but some drivers may return a ref to an
5043       empty hash because they can't pre-determine the names.
5044
5045       It is possible that the keys in the hash returned by "ParamValues" are
5046       not exactly the same as those implied by the prepared statement.  For
5047       example, DBD::Oracle translates '"?"' placeholders into '":pN"' where N
5048       is a sequence number starting at 1.
5049
5050       * Values:
5051
5052       It is possible that the values in the hash returned by "ParamValues"
5053       are not exactly the same as those passed to bind_param() or execute().
5054       The driver may have slightly modified values in some way based on the
5055       TYPE the value was bound with. For example a floating point value bound
5056       as an SQL_INTEGER type may be returned as an integer.  The values
5057       returned by "ParamValues" can be passed to another bind_param() method
5058       with the same TYPE and will be seen by the database as the same value.
5059       See also "ParamTypes" below.
5060
5061       The "ParamValues" attribute was added in DBI 1.28.
5062
5063       "ParamTypes"
5064
5065       Type: hash ref, read-only
5066
5067       Returns a reference to a hash containing the type information currently
5068       bound to placeholders.  Returns undef if not supported by the driver.
5069
5070       * Keys:
5071
5072       See "ParamValues" above.
5073
5074       * Values:
5075
5076       The hash values are hashrefs of type information in the same form as
5077       that passed to the various bind_param() methods (See "bind_param" for
5078       the format and values).
5079
5080       It is possible that the values in the hash returned by "ParamTypes" are
5081       not exactly the same as those passed to bind_param() or execute().
5082       Param attributes specified using the abbreviated form, like this:
5083
5084           $sth->bind_param(1, SQL_INTEGER);
5085
5086       are returned in the expanded form, as if called like this:
5087
5088           $sth->bind_param(1, { TYPE => SQL_INTEGER });
5089
5090       The driver may have modified the type information in some way based on
5091       the bound values, other hints provided by the prepare()'d SQL
5092       statement, or alternate type mappings required by the driver or target
5093       database system. The driver may also add private keys (with names
5094       beginning with the drivers reserved prefix, e.g., odbc_xxx).
5095
5096       * Example:
5097
5098       The keys and values in the returned hash can be passed to the various
5099       bind_param() methods to effectively reproduce a previous param binding.
5100       For example:
5101
5102         # assuming $sth1 is a previously prepared statement handle
5103         my $sth2 = $dbh->prepare( $sth1->{Statement} );
5104         my $ParamValues = $sth1->{ParamValues} || {};
5105         my $ParamTypes  = $sth1->{ParamTypes}  || {};
5106         $sth2->bind_param($_, $ParamValues->{$_}, $ParamTypes->{$_})
5107           for keys %{ {%$ParamValues, %$ParamTypes} };
5108         $sth2->execute();
5109
5110       The "ParamTypes" attribute was added in DBI 1.49. Implementation is the
5111       responsibility of individual drivers; the DBI layer default
5112       implementation simply returns undef.
5113
5114       "ParamArrays"
5115
5116       Type: hash ref, read-only
5117
5118       Returns a reference to a hash containing the values currently bound to
5119       placeholders with "execute_array" or "bind_param_array".  The keys of
5120       the hash are the 'names' of the placeholders, typically integers
5121       starting at 1.  Returns undef if not supported by the driver or no
5122       arrays of parameters are bound.
5123
5124       Each key value is an array reference containing a list of the bound
5125       parameters for that column.
5126
5127       For example:
5128
5129         $sth = $dbh->prepare("INSERT INTO staff (id, name) values (?,?)");
5130         $sth->execute_array({},[1,2], ['fred','dave']);
5131         if ($sth->{ParamArrays}) {
5132             foreach $param (keys %{$sth->{ParamArrays}}) {
5133                 printf "Parameters for %s : %s\n", $param,
5134                 join(",", @{$sth->{ParamArrays}->{$param}});
5135             }
5136         }
5137
5138       It is possible that the values in the hash returned by "ParamArrays"
5139       are not exactly the same as those passed to "bind_param_array" or
5140       "execute_array".  The driver may have slightly modified values in some
5141       way based on the TYPE the value was bound with. For example a floating
5142       point value bound as an SQL_INTEGER type may be returned as an integer.
5143
5144       It is also possible that the keys in the hash returned by "ParamArrays"
5145       are not exactly the same as those implied by the prepared statement.
5146       For example, DBD::Oracle translates '"?"'  placeholders into '":pN"'
5147       where N is a sequence number starting at 1.
5148
5149       "RowsInCache"
5150
5151       Type: integer, read-only
5152
5153       If the driver supports a local row cache for "SELECT" statements, then
5154       this attribute holds the number of un-fetched rows in the cache. If the
5155       driver doesn't, then it returns "undef". Note that some drivers pre-
5156       fetch rows on execute, whereas others wait till the first fetch.
5157
5158       See also the "RowCacheSize" database handle attribute.
5159

FURTHER INFORMATION

5161   Catalog Methods
5162       An application can retrieve metadata information from the DBMS by
5163       issuing appropriate queries on the views of the Information Schema.
5164       Unfortunately, "INFORMATION_SCHEMA" views are seldom supported by the
5165       DBMS.  Special methods (catalog methods) are available to return result
5166       sets for a small but important portion of that metadata:
5167
5168         column_info
5169         foreign_key_info
5170         primary_key_info
5171         table_info
5172         statistics_info
5173
5174       All catalog methods accept arguments in order to restrict the result
5175       sets.  Passing "undef" to an optional argument does not constrain the
5176       search for that argument.  However, an empty string ('') is treated as
5177       a regular search criteria and will only match an empty value.
5178
5179       Note: SQL/CLI and ODBC differ in the handling of empty strings. An
5180       empty string will not restrict the result set in SQL/CLI.
5181
5182       Most arguments in the catalog methods accept only ordinary values, e.g.
5183       the arguments of "primary_key_info()".  Such arguments are treated as a
5184       literal string, i.e. the case is significant and quote characters are
5185       taken literally.
5186
5187       Some arguments in the catalog methods accept search patterns (strings
5188       containing '_' and/or '%'), e.g. the $table argument of
5189       "column_info()".  Passing '%' is equivalent to leaving the argument
5190       "undef".
5191
5192       Caveat: The underscore ('_') is valid and often used in SQL
5193       identifiers.  Passing such a value to a search pattern argument may
5194       return more rows than expected!  To include pattern characters as
5195       literals, they must be preceded by an escape character which can be
5196       achieved with
5197
5198         $esc = $dbh->get_info( 14 );  # SQL_SEARCH_PATTERN_ESCAPE
5199         $search_pattern =~ s/([_%])/$esc$1/g;
5200
5201       The ODBC and SQL/CLI specifications define a way to change the default
5202       behaviour described above: All arguments (except list value arguments)
5203       are treated as identifier if the "SQL_ATTR_METADATA_ID" attribute is
5204       set to "SQL_TRUE".  Quoted identifiers are very similar to ordinary
5205       values, i.e. their body (the string within the quotes) is interpreted
5206       literally.  Unquoted identifiers are compared in UPPERCASE.
5207
5208       The DBI (currently) does not support the "SQL_ATTR_METADATA_ID"
5209       attribute, i.e. it behaves like an ODBC driver where
5210       "SQL_ATTR_METADATA_ID" is set to "SQL_FALSE".
5211
5212   Transactions
5213       Transactions are a fundamental part of any robust database system. They
5214       protect against errors and database corruption by ensuring that sets of
5215       related changes to the database take place in atomic (indivisible, all-
5216       or-nothing) units.
5217
5218       This section applies to databases that support transactions and where
5219       "AutoCommit" is off.  See "AutoCommit" for details of using
5220       "AutoCommit" with various types of databases.
5221
5222       The recommended way to implement robust transactions in Perl
5223       applications is to enable "RaiseError" and catch the error that's
5224       'thrown' as an exception.  For example, using Try::Tiny:
5225
5226         use Try::Tiny;
5227         $dbh->{AutoCommit} = 0;  # enable transactions, if possible
5228         $dbh->{RaiseError} = 1;
5229         try {
5230             foo(...)        # do lots of work here
5231             bar(...)        # including inserts
5232             baz(...)        # and updates
5233             $dbh->commit;   # commit the changes if we get this far
5234         } catch {
5235             warn "Transaction aborted because $_"; # Try::Tiny copies $@ into $_
5236             # now rollback to undo the incomplete changes
5237             # but do it in an eval{} as it may also fail
5238             eval { $dbh->rollback };
5239             # add other application on-error-clean-up code here
5240         };
5241
5242       If the "RaiseError" attribute is not set, then DBI calls would need to
5243       be manually checked for errors, typically like this:
5244
5245         $h->method(@args) or die $h->errstr;
5246
5247       With "RaiseError" set, the DBI will automatically "die" if any DBI
5248       method call on that handle (or a child handle) fails, so you don't have
5249       to test the return value of each method call. See "RaiseError" for more
5250       details.
5251
5252       A major advantage of the "eval" approach is that the transaction will
5253       be properly rolled back if any code (not just DBI calls) in the inner
5254       application dies for any reason. The major advantage of using the
5255       "$h->{RaiseError}" attribute is that all DBI calls will be checked
5256       automatically. Both techniques are strongly recommended.
5257
5258       After calling "commit" or "rollback" many drivers will not let you
5259       fetch from a previously active "SELECT" statement handle that's a child
5260       of the same database handle. A typical way round this is to connect the
5261       the database twice and use one connection for "SELECT" statements.
5262
5263       See "AutoCommit" and "disconnect" for other important information about
5264       transactions.
5265
5266   Handling BLOB / LONG / Memo Fields
5267       Many databases support "blob" (binary large objects), "long", or
5268       similar datatypes for holding very long strings or large amounts of
5269       binary data in a single field. Some databases support variable length
5270       long values over 2,000,000,000 bytes in length.
5271
5272       Since values of that size can't usually be held in memory, and because
5273       databases can't usually know in advance the length of the longest long
5274       that will be returned from a "SELECT" statement (unlike other data
5275       types), some special handling is required.
5276
5277       In this situation, the value of the "$h->{LongReadLen}" attribute is
5278       used to determine how much buffer space to allocate when fetching such
5279       fields.  The "$h->{LongTruncOk}" attribute is used to determine how to
5280       behave if a fetched value can't fit into the buffer.
5281
5282       See the description of "LongReadLen" for more information.
5283
5284       When trying to insert long or binary values, placeholders should be
5285       used since there are often limits on the maximum size of an "INSERT"
5286       statement and the "quote" method generally can't cope with binary data.
5287       See "Placeholders and Bind Values".
5288
5289   Simple Examples
5290       Here's a complete example program to select and fetch some data:
5291
5292         my $data_source = "dbi::DriverName:db_name";
5293         my $dbh = DBI->connect($data_source, $user, $password)
5294             or die "Can't connect to $data_source: $DBI::errstr";
5295
5296         my $sth = $dbh->prepare( q{
5297                 SELECT name, phone
5298                 FROM mytelbook
5299         }) or die "Can't prepare statement: $DBI::errstr";
5300
5301         my $rc = $sth->execute
5302             or die "Can't execute statement: $DBI::errstr";
5303
5304         print "Query will return $sth->{NUM_OF_FIELDS} fields.\n\n";
5305         print "Field names: @{ $sth->{NAME} }\n";
5306
5307         while (($name, $phone) = $sth->fetchrow_array) {
5308             print "$name: $phone\n";
5309         }
5310         # check for problems which may have terminated the fetch early
5311         die $sth->errstr if $sth->err;
5312
5313         $dbh->disconnect;
5314
5315       Here's a complete example program to insert some data from a file.
5316       (This example uses "RaiseError" to avoid needing to check each call).
5317
5318         my $dbh = DBI->connect("dbi:DriverName:db_name", $user, $password, {
5319             RaiseError => 1, AutoCommit => 0
5320         });
5321
5322         my $sth = $dbh->prepare( q{
5323             INSERT INTO table (name, phone) VALUES (?, ?)
5324         });
5325
5326         open FH, "<phone.csv" or die "Unable to open phone.csv: $!";
5327         while (<FH>) {
5328             chomp;
5329             my ($name, $phone) = split /,/;
5330             $sth->execute($name, $phone);
5331         }
5332         close FH;
5333
5334         $dbh->commit;
5335         $dbh->disconnect;
5336
5337       Here's how to convert fetched NULLs (undefined values) into empty
5338       strings:
5339
5340         while($row = $sth->fetchrow_arrayref) {
5341           # this is a fast and simple way to deal with nulls:
5342           foreach (@$row) { $_ = '' unless defined }
5343           print "@$row\n";
5344         }
5345
5346       The "q{...}" style quoting used in these examples avoids clashing with
5347       quotes that may be used in the SQL statement. Use the double-quote like
5348       "qq{...}" operator if you want to interpolate variables into the
5349       string.  See "Quote and Quote-like Operators" in perlop for more
5350       details.
5351
5352   Threads and Thread Safety
5353       Perl 5.7 and later support a new threading model called iThreads.  (The
5354       old "5.005 style" threads are not supported by the DBI.)
5355
5356       In the iThreads model each thread has its own copy of the perl
5357       interpreter.  When a new thread is created the original perl
5358       interpreter is 'cloned' to create a new copy for the new thread.
5359
5360       If the DBI and drivers are loaded and handles created before the thread
5361       is created then it will get a cloned copy of the DBI, the drivers and
5362       the handles.
5363
5364       However, the internal pointer data within the handles will refer to the
5365       DBI and drivers in the original interpreter. Using those handles in the
5366       new interpreter thread is not safe, so the DBI detects this and croaks
5367       on any method call using handles that don't belong to the current
5368       thread (except for DESTROY).
5369
5370       Because of this (possibly temporary) restriction, newly created threads
5371       must make their own connections to the database. Handles can't be
5372       shared across threads.
5373
5374       But BEWARE, some underlying database APIs (the code the DBD driver uses
5375       to talk to the database, often supplied by the database vendor) are not
5376       thread safe. If it's not thread safe, then allowing more than one
5377       thread to enter the code at the same time may cause subtle/serious
5378       problems. In some cases allowing more than one thread to enter the
5379       code, even if not at the same time, can cause problems. You have been
5380       warned.
5381
5382       Using DBI with perl threads is not yet recommended for production
5383       environments. For more information see
5384       <http://www.perlmonks.org/index.pl?node_id=288022>
5385
5386       Note: There is a bug in perl 5.8.2 when configured with threads and
5387       debugging enabled (bug #24463) which causes a DBI test to fail.
5388
5389   Signal Handling and Canceling Operations
5390       [The following only applies to systems with unix-like signal handling.
5391       I'd welcome additions for other systems, especially Windows.]
5392
5393       The first thing to say is that signal handling in Perl versions less
5394       than 5.8 is not safe. There is always a small risk of Perl crashing
5395       and/or core dumping when, or after, handling a signal because the
5396       signal could arrive and be handled while internal data structures are
5397       being changed. If the signal handling code used those same internal
5398       data structures it could cause all manner of subtle and not-so-subtle
5399       problems.  The risk was reduced with 5.4.4 but was still present in all
5400       perls up through 5.8.0.
5401
5402       Beginning in perl 5.8.0 perl implements 'safe' signal handling if your
5403       system has the POSIX sigaction() routine. Now when a signal is
5404       delivered perl just makes a note of it but does not run the %SIG
5405       handler. The handling is 'deferred' until a 'safe' moment.
5406
5407       Although this change made signal handling safe, it also lead to a
5408       problem with signals being deferred for longer than you'd like.  If a
5409       signal arrived while executing a system call, such as waiting for data
5410       on a network connection, the signal is noted and then the system call
5411       that was executing returns with an EINTR error code to indicate that it
5412       was interrupted. All fine so far.
5413
5414       The problem comes when the code that made the system call sees the
5415       EINTR code and decides it's going to call it again. Perl doesn't do
5416       that, but database code sometimes does. If that happens then the signal
5417       handler doesn't get called until later. Maybe much later.
5418
5419       Fortunately there are ways around this which we'll discuss below.
5420       Unfortunately they make signals unsafe again.
5421
5422       The two most common uses of signals in relation to the DBI are for
5423       canceling operations when the user types Ctrl-C (interrupt), and for
5424       implementing a timeout using "alarm()" and $SIG{ALRM}.
5425
5426       Cancel
5427           The DBI provides a "cancel" method for statement handles. The
5428           "cancel" method should abort the current operation and is designed
5429           to be called from a signal handler.  For example:
5430
5431             $SIG{INT} = sub { $sth->cancel };
5432
5433           However, few drivers implement this (the DBI provides a default
5434           method that just returns "undef") and, even if implemented, there
5435           is still a possibility that the statement handle, and even the
5436           parent database handle, will not be usable afterwards.
5437
5438           If "cancel" returns true, then it has successfully invoked the
5439           database engine's own cancel function.  If it returns false, then
5440           "cancel" failed. If it returns "undef", then the database driver
5441           does not have cancel implemented - very few do.
5442
5443       Timeout
5444           The traditional way to implement a timeout is to set $SIG{ALRM} to
5445           refer to some code that will be executed when an ALRM signal
5446           arrives and then to call alarm($seconds) to schedule an ALRM signal
5447           to be delivered $seconds in the future. For example:
5448
5449             my $failed;
5450             eval {
5451               local $SIG{ALRM} = sub { die "TIMEOUT\n" }; # N.B. \n required
5452               eval {
5453                 alarm($seconds);
5454                 ... code to execute with timeout here (which may die) ...
5455                 1;
5456               } or $failed = 1;
5457               # outer eval catches alarm that might fire JUST before this alarm(0)
5458               alarm(0);  # cancel alarm (if code ran fast)
5459               die "$@" if $failed;
5460               1;
5461             } or $failed = 1;
5462             if ( $failed ) {
5463               if ( defined $@ and $@ eq "TIMEOUT\n" ) { ... }
5464               else { ... } # some other error
5465             }
5466
5467           The first (outer) eval is used to avoid the unlikely but possible
5468           chance that the "code to execute" dies and the alarm fires before
5469           it is cancelled. Without the outer eval, if this happened your
5470           program will die if you have no ALRM handler or a non-local alarm
5471           handler will be called.
5472
5473           Unfortunately, as described above, this won't always work as
5474           expected, depending on your perl version and the underlying
5475           database code.
5476
5477           With Oracle for instance (DBD::Oracle), if the system which hosts
5478           the database is down the DBI->connect() call will hang for several
5479           minutes before returning an error.
5480
5481       The solution on these systems is to use the "POSIX::sigaction()"
5482       routine to gain low level access to how the signal handler is
5483       installed.
5484
5485       The code would look something like this (for the DBD-Oracle connect()):
5486
5487          use POSIX qw(:signal_h);
5488
5489          my $mask = POSIX::SigSet->new( SIGALRM ); # signals to mask in the handler
5490          my $action = POSIX::SigAction->new(
5491              sub { die "connect timeout\n" },        # the handler code ref
5492              $mask,
5493              # not using (perl 5.8.2 and later) 'safe' switch or sa_flags
5494          );
5495          my $oldaction = POSIX::SigAction->new();
5496          sigaction( SIGALRM, $action, $oldaction );
5497          my $dbh;
5498          my $failed;
5499          eval {
5500             eval {
5501               alarm(5); # seconds before time out
5502               $dbh = DBI->connect("dbi:Oracle:$dsn" ... );
5503               1;
5504             } or $failed = 1;
5505             alarm(0); # cancel alarm (if connect worked fast)
5506             die "$@\n" if $failed; # connect died
5507             1;
5508          } or $failed = 1;
5509          sigaction( SIGALRM, $oldaction );  # restore original signal handler
5510          if ( $failed ) {
5511            if ( defined $@ and $@ eq "connect timeout\n" ) {...}
5512            else { # connect died }
5513          }
5514
5515       See previous example for the reasoning around the double eval.
5516
5517       Similar techniques can be used for canceling statement execution.
5518
5519       Unfortunately, this solution is somewhat messy, and it does not work
5520       with perl versions less than perl 5.8 where "POSIX::sigaction()"
5521       appears to be broken.
5522
5523       For a cleaner implementation that works across perl versions, see
5524       Lincoln Baxter's Sys::SigAction module at Sys::SigAction.  The
5525       documentation for Sys::SigAction includes an longer discussion of this
5526       problem, and a DBD::Oracle test script.
5527
5528       Be sure to read all the signal handling sections of the perlipc manual.
5529
5530       And finally, two more points to keep firmly in mind. Firstly, remember
5531       that what we've done here is essentially revert to old style unsafe
5532       handling of these signals. So do as little as possible in the handler.
5533       Ideally just die(). Secondly, the handles in use at the time the signal
5534       is handled may not be safe to use afterwards.
5535
5536   Subclassing the DBI
5537       DBI can be subclassed and extended just like any other object oriented
5538       module.  Before we talk about how to do that, it's important to be
5539       clear about the various DBI classes and how they work together.
5540
5541       By default "$dbh = DBI->connect(...)" returns a $dbh blessed into the
5542       "DBI::db" class.  And the "$dbh->prepare" method returns an $sth
5543       blessed into the "DBI::st" class (actually it simply changes the last
5544       four characters of the calling handle class to be "::st").
5545
5546       The leading '"DBI"' is known as the 'root class' and the extra '"::db"'
5547       or '"::st"' are the 'handle type suffixes'. If you want to subclass the
5548       DBI you'll need to put your overriding methods into the appropriate
5549       classes.  For example, if you want to use a root class of "MySubDBI"
5550       and override the do(), prepare() and execute() methods, then your do()
5551       and prepare() methods should be in the "MySubDBI::db" class and the
5552       execute() method should be in the "MySubDBI::st" class.
5553
5554       To setup the inheritance hierarchy the @ISA variable in "MySubDBI::db"
5555       should include "DBI::db" and the @ISA variable in "MySubDBI::st" should
5556       include "DBI::st".  The "MySubDBI" root class itself isn't currently
5557       used for anything visible and so, apart from setting @ISA to include
5558       "DBI", it can be left empty.
5559
5560       So, having put your overriding methods into the right classes, and
5561       setup the inheritance hierarchy, how do you get the DBI to use them?
5562       You have two choices, either a static method call using the name of
5563       your subclass:
5564
5565         $dbh = MySubDBI->connect(...);
5566
5567       or specifying a "RootClass" attribute:
5568
5569         $dbh = DBI->connect(..., { RootClass => 'MySubDBI' });
5570
5571       If both forms are used then the attribute takes precedence.
5572
5573       The only differences between the two are that using an explicit
5574       RootClass attribute will a) make the DBI automatically attempt to load
5575       a module by that name if the class doesn't exist, and b) won't call
5576       your MySubDBI::connect() method, if you have one.
5577
5578       When subclassing is being used then, after a successful new connect,
5579       the DBI->connect method automatically calls:
5580
5581         $dbh->connected($dsn, $user, $pass, \%attr);
5582
5583       The default method does nothing. The call is made just to simplify any
5584       post-connection setup that your subclass may want to perform.  The
5585       parameters are the same as passed to DBI->connect.  If your subclass
5586       supplies a connected method, it should be part of the MySubDBI::db
5587       package.
5588
5589       One more thing to note: you must let the DBI do the handle creation.
5590       If you want to override the connect() method in your *::dr class then
5591       it must still call SUPER::connect to get a $dbh to work with.
5592       Similarly, an overridden prepare() method in *::db must still call
5593       SUPER::prepare to get a $sth.  If you try to create your own handles
5594       using bless() then you'll find the DBI will reject them with an "is not
5595       a DBI handle (has no magic)" error.
5596
5597       Here's a brief example of a DBI subclass.  A more thorough example can
5598       be found in t/subclass.t in the DBI distribution.
5599
5600         package MySubDBI;
5601
5602         use strict;
5603
5604         use DBI;
5605         use vars qw(@ISA);
5606         @ISA = qw(DBI);
5607
5608         package MySubDBI::db;
5609         use vars qw(@ISA);
5610         @ISA = qw(DBI::db);
5611
5612         sub prepare {
5613           my ($dbh, @args) = @_;
5614           my $sth = $dbh->SUPER::prepare(@args)
5615               or return;
5616           $sth->{private_mysubdbi_info} = { foo => 'bar' };
5617           return $sth;
5618         }
5619
5620         package MySubDBI::st;
5621         use vars qw(@ISA);
5622         @ISA = qw(DBI::st);
5623
5624         sub fetch {
5625           my ($sth, @args) = @_;
5626           my $row = $sth->SUPER::fetch(@args)
5627               or return;
5628           do_something_magical_with_row_data($row)
5629               or return $sth->set_err(1234, "The magic failed", undef, "fetch");
5630           return $row;
5631         }
5632
5633       When calling a SUPER::method that returns a handle, be careful to check
5634       the return value before trying to do other things with it in your
5635       overridden method. This is especially important if you want to set a
5636       hash attribute on the handle, as Perl's autovivification will bite you
5637       by (in)conveniently creating an unblessed hashref, which your method
5638       will then return with usually baffling results later on like the error
5639       "dbih_getcom handle HASH(0xa4451a8) is not a DBI handle (has no magic".
5640       It's best to check right after the call and return undef immediately on
5641       error, just like DBI would and just like the example above.
5642
5643       If your method needs to record an error it should call the set_err()
5644       method with the error code and error string, as shown in the example
5645       above. The error code and error string will be recorded in the handle
5646       and available via "$h->err" and $DBI::errstr etc.  The set_err() method
5647       always returns an undef or empty list as appropriate. Since your method
5648       should nearly always return an undef or empty list as soon as an error
5649       is detected it's handy to simply return what set_err() returns, as
5650       shown in the example above.
5651
5652       If the handle has "RaiseError", "PrintError", or "HandleError" etc. set
5653       then the set_err() method will honour them. This means that if
5654       "RaiseError" is set then set_err() won't return in the normal way but
5655       will 'throw an exception' that can be caught with an "eval" block.
5656
5657       You can stash private data into DBI handles via "$h->{private_..._*}".
5658       See the entry under "ATTRIBUTES COMMON TO ALL HANDLES" for info and
5659       important caveats.
5660
5661   Memory Leaks
5662       When tracking down memory leaks using tools like Devel::Leak you'll
5663       find that some DBI internals are reported as 'leaking' memory.  This is
5664       very unlikely to be a real leak.  The DBI has various caches to improve
5665       performance and the apparrent leaks are simply the normal operation of
5666       these caches.
5667
5668       The most frequent sources of the apparrent leaks are "ChildHandles",
5669       "prepare_cached" and "connect_cached".
5670
5671       For example
5672       http://stackoverflow.com/questions/13338308/perl-dbi-memory-leak
5673
5674       Given how widely the DBI is used, you can rest assured that if a new
5675       release of the DBI did have a real leak it would be discovered,
5676       reported, and fixed immediately. The leak you're looking for is
5677       probably elsewhere. Good luck!
5678

TRACING

5680       The DBI has a powerful tracing mechanism built in. It enables you to
5681       see what's going on 'behind the scenes', both within the DBI and the
5682       drivers you're using.
5683
5684   Trace Settings
5685       Which details are written to the trace output is controlled by a
5686       combination of a trace level, an integer from 0 to 15, and a set of
5687       trace flags that are either on or off. Together these are known as the
5688       trace settings and are stored together in a single integer.  For normal
5689       use you only need to set the trace level, and generally only to a value
5690       between 1 and 4.
5691
5692       Each handle has its own trace settings, and so does the DBI.  When you
5693       call a method the DBI merges the handles settings into its own for the
5694       duration of the call: the trace flags of the handle are OR'd into the
5695       trace flags of the DBI, and if the handle has a higher trace level then
5696       the DBI trace level is raised to match it.  The previous DBI trace
5697       settings are restored when the called method returns.
5698
5699   Trace Levels
5700       Trace levels are as follows:
5701
5702         0 - Trace disabled.
5703         1 - Trace top-level DBI method calls returning with results or errors.
5704         2 - As above, adding tracing of top-level method entry with parameters.
5705         3 - As above, adding some high-level information from the driver
5706             and some internal information from the DBI.
5707         4 - As above, adding more detailed information from the driver.
5708             This is the first level to trace all the rows being fetched.
5709         5 to 15 - As above but with more and more internal information.
5710
5711       Trace level 1 is best for a simple overview of what's happening.  Trace
5712       levels 2 thru 4 a good choice for general purpose tracing.  Levels 5
5713       and above are best reserved for investigating a specific problem, when
5714       you need to see "inside" the driver and DBI.
5715
5716       The trace output is detailed and typically very useful. Much of the
5717       trace output is formatted using the "neat" function, so strings in the
5718       trace output may be edited and truncated by that function.
5719
5720   Trace Flags
5721       Trace flags are used to enable tracing of specific activities within
5722       the DBI and drivers. The DBI defines some trace flags and drivers can
5723       define others. DBI trace flag names begin with a capital letter and
5724       driver specific names begin with a lowercase letter, as usual.
5725
5726       Currently the DBI defines these trace flags:
5727
5728         ALL - turn on all DBI and driver flags (not recommended)
5729         SQL - trace SQL statements executed
5730               (not yet implemented in DBI but implemented in some DBDs)
5731         CON - trace connection process
5732         ENC - trace encoding (unicode translations etc)
5733               (not yet implemented in DBI but implemented in some DBDs)
5734         DBD - trace only DBD messages
5735               (not implemented by all DBDs yet)
5736         TXN - trace transactions
5737               (not implemented in all DBDs yet)
5738
5739       The "parse_trace_flags" and "parse_trace_flag" methods are used to
5740       convert trace flag names into the corresponding integer bit flags.
5741
5742   Enabling Trace
5743       The "$h->trace" method sets the trace settings for a handle and
5744       "DBI->trace" does the same for the DBI.
5745
5746       In addition to the "trace" method, you can enable the same trace
5747       information, and direct the output to a file, by setting the
5748       "DBI_TRACE" environment variable before starting Perl.  See "DBI_TRACE"
5749       for more information.
5750
5751       Finally, you can set, or get, the trace settings for a handle using the
5752       "TraceLevel" attribute.
5753
5754       All of those methods use parse_trace_flags() and so allow you set both
5755       the trace level and multiple trace flags by using a string containing
5756       the trace level and/or flag names separated by vertical bar (""|"") or
5757       comma ("","") characters. For example:
5758
5759         local $h->{TraceLevel} = "3|SQL|foo";
5760
5761   Trace Output
5762       Initially trace output is written to "STDERR".  Both the "$h->trace"
5763       and "DBI->trace" methods take an optional $trace_file parameter, which
5764       may be either the name of a file to be opened by DBI in append mode, or
5765       a reference to an existing writable (possibly layered) filehandle. If
5766       $trace_file is a filename, and can be opened in append mode, or
5767       $trace_file is a writable filehandle, then all trace output (currently
5768       including that from other handles) is redirected to that file. A
5769       warning is generated if $trace_file can't be opened or is not writable.
5770
5771       Further calls to trace() without $trace_file do not alter where the
5772       trace output is sent. If $trace_file is undefined, then trace output is
5773       sent to "STDERR" and, if the prior trace was opened with $trace_file as
5774       a filename, the previous trace file is closed; if $trace_file was a
5775       filehandle, the filehandle is not closed.
5776
5777       NOTE: If $trace_file is specified as a filehandle, the filehandle
5778       should not be closed until all DBI operations are completed, or the
5779       application has reset the trace file via another call to "trace()" that
5780       changes the trace file.
5781
5782   Tracing to Layered Filehandles
5783       NOTE:
5784
5785       •   Tied filehandles are not currently supported, as tie operations are
5786           not available to the PerlIO methods used by the DBI.
5787
5788       •   PerlIO layer support requires Perl version 5.8 or higher.
5789
5790       As of version 5.8, Perl provides the ability to layer various
5791       "disciplines" on an open filehandle via the PerlIO module.
5792
5793       A simple example of using PerlIO layers is to use a scalar as the
5794       output:
5795
5796           my $scalar = '';
5797           open( my $fh, "+>:scalar", \$scalar );
5798           $dbh->trace( 2, $fh );
5799
5800       Now all trace output is simply appended to $scalar.
5801
5802       A more complex application of tracing to a layered filehandle is the
5803       use of a custom layer (Refer to Perlio::via for details on creating
5804       custom PerlIO layers.). Consider an application with the following
5805       logger module:
5806
5807           package MyFancyLogger;
5808
5809           sub new
5810           {
5811               my $self = {};
5812               my $fh;
5813               open $fh, '>', 'fancylog.log';
5814               $self->{_fh} = $fh;
5815               $self->{_buf} = '';
5816               return bless $self, shift;
5817           }
5818
5819           sub log
5820           {
5821               my $self = shift;
5822               return unless exists $self->{_fh};
5823               my $fh = $self->{_fh};
5824               $self->{_buf} .= shift;
5825           #
5826           # DBI feeds us pieces at a time, so accumulate a complete line
5827           # before outputing
5828           #
5829               print $fh "At ", scalar localtime(), ':', $self->{_buf}, "\n" and
5830               $self->{_buf} = ''
5831                   if $self->{_buf}=~tr/\n//;
5832           }
5833
5834           sub close {
5835               my $self = shift;
5836               return unless exists $self->{_fh};
5837               my $fh = $self->{_fh};
5838               print $fh "At ", scalar localtime(), ':', $self->{_buf}, "\n" and
5839               $self->{_buf} = ''
5840                   if $self->{_buf};
5841               close $fh;
5842               delete $self->{_fh};
5843           }
5844
5845           1;
5846
5847       To redirect DBI traces to this logger requires creating a package for
5848       the layer:
5849
5850           package PerlIO::via::MyFancyLogLayer;
5851
5852           sub PUSHED
5853           {
5854               my ($class,$mode,$fh) = @_;
5855               my $logger;
5856               return bless \$logger,$class;
5857           }
5858
5859           sub OPEN {
5860               my ($self, $path, $mode, $fh) = @_;
5861               #
5862               # $path is actually our logger object
5863               #
5864               $$self = $path;
5865               return 1;
5866           }
5867
5868           sub WRITE
5869           {
5870               my ($self, $buf, $fh) = @_;
5871               $$self->log($buf);
5872               return length($buf);
5873           }
5874
5875           sub CLOSE {
5876               my $self = shift;
5877               $$self->close();
5878               return 0;
5879           }
5880
5881           1;
5882
5883       The application can then cause DBI traces to be routed to the logger
5884       using
5885
5886           use PerlIO::via::MyFancyLogLayer;
5887
5888           open my $fh, '>:via(MyFancyLogLayer)', MyFancyLogger->new();
5889
5890           $dbh->trace('SQL', $fh);
5891
5892       Now all trace output will be processed by MyFancyLogger's log() method.
5893
5894   Trace Content
5895       Many of the values embedded in trace output are formatted using the
5896       neat() utility function. This means they may be quoted, sanitized, and
5897       possibly truncated if longer than $DBI::neat_maxlen. See "neat" for
5898       more details.
5899
5900   Tracing Tips
5901       You can add tracing to your own application code using the "trace_msg"
5902       method.
5903
5904       It can sometimes be handy to compare trace files from two different
5905       runs of the same script. However using a tool like "diff" on the
5906       original log output doesn't work well because the trace file is full of
5907       object addresses that may differ on each run.
5908
5909       The DBI includes a handy utility called dbilogstrip that can be used to
5910       'normalize' the log content. It can be used as a filter like this:
5911
5912           DBI_TRACE=2 perl yourscript.pl ...args1... 2>&1 | dbilogstrip > dbitrace1.log
5913           DBI_TRACE=2 perl yourscript.pl ...args2... 2>&1 | dbilogstrip > dbitrace2.log
5914           diff -u dbitrace1.log dbitrace2.log
5915
5916       See dbilogstrip for more information.
5917

DBI ENVIRONMENT VARIABLES

5919       The DBI module recognizes a number of environment variables, but most
5920       of them should not be used most of the time.  It is better to be
5921       explicit about what you are doing to avoid the need for environment
5922       variables, especially in a web serving system where web servers are
5923       stingy about which environment variables are available.
5924
5925   DBI_DSN
5926       The DBI_DSN environment variable is used by DBI->connect if you do not
5927       specify a data source when you issue the connect.  It should have a
5928       format such as "dbi:Driver:databasename".
5929
5930   DBI_DRIVER
5931       The DBI_DRIVER environment variable is used to fill in the database
5932       driver name in DBI->connect if the data source string starts "dbi::"
5933       (thereby omitting the driver).  If DBI_DSN omits the driver name,
5934       DBI_DRIVER can fill the gap.
5935
5936   DBI_AUTOPROXY
5937       The DBI_AUTOPROXY environment variable takes a string value that starts
5938       "dbi:Proxy:" and is typically followed by "hostname=...;port=...".  It
5939       is used to alter the behaviour of DBI->connect.  For full details, see
5940       DBI::Proxy documentation.
5941
5942   DBI_USER
5943       The DBI_USER environment variable takes a string value that is used as
5944       the user name if the DBI->connect call is given undef (as distinct from
5945       an empty string) as the username argument.  Be wary of the security
5946       implications of using this.
5947
5948   DBI_PASS
5949       The DBI_PASS environment variable takes a string value that is used as
5950       the password if the DBI->connect call is given undef (as distinct from
5951       an empty string) as the password argument.  Be extra wary of the
5952       security implications of using this.
5953
5954   DBI_DBNAME (obsolete)
5955       The DBI_DBNAME environment variable takes a string value that is used
5956       only when the obsolescent style of DBI->connect (with driver name as
5957       fourth parameter) is used, and when no value is provided for the first
5958       (database name) argument.
5959
5960   DBI_TRACE
5961       The DBI_TRACE environment variable specifies the global default trace
5962       settings for the DBI at startup. Can also be used to direct trace
5963       output to a file. When the DBI is loaded it does:
5964
5965         DBI->trace(split /=/, $ENV{DBI_TRACE}, 2) if $ENV{DBI_TRACE};
5966
5967       So if "DBI_TRACE" contains an ""="" character then what follows it is
5968       used as the name of the file to append the trace to.
5969
5970       output appended to that file. If the name begins with a number followed
5971       by an equal sign ("="), then the number and the equal sign are stripped
5972       off from the name, and the number is used to set the trace level. For
5973       example:
5974
5975         DBI_TRACE=1=dbitrace.log perl your_test_script.pl
5976
5977       On Unix-like systems using a Bourne-like shell, you can do this easily
5978       on the command line:
5979
5980         DBI_TRACE=2 perl your_test_script.pl
5981
5982       See "TRACING" for more information.
5983
5984   PERL_DBI_DEBUG (obsolete)
5985       An old variable that should no longer be used; equivalent to DBI_TRACE.
5986
5987   DBI_PROFILE
5988       The DBI_PROFILE environment variable can be used to enable profiling of
5989       DBI method calls. See DBI::Profile for more information.
5990
5991   DBI_PUREPERL
5992       The DBI_PUREPERL environment variable can be used to enable the use of
5993       DBI::PurePerl.  See DBI::PurePerl for more information.
5994

WARNING AND ERROR MESSAGES

5996   Fatal Errors
5997       Can't call method "prepare" without a package or object reference
5998           The $dbh handle you're using to call "prepare" is probably
5999           undefined because the preceding "connect" failed. You should always
6000           check the return status of DBI methods, or use the "RaiseError"
6001           attribute.
6002
6003       Can't call method "execute" without a package or object reference
6004           The $sth handle you're using to call "execute" is probably
6005           undefined because the preceding "prepare" failed. You should always
6006           check the return status of DBI methods, or use the "RaiseError"
6007           attribute.
6008
6009       DBI/DBD internal version mismatch
6010           The DBD driver module was built with a different version of DBI
6011           than the one currently being used.  You should rebuild the DBD
6012           module under the current version of DBI.
6013
6014           (Some rare platforms require "static linking". On those platforms,
6015           there may be an old DBI or DBD driver version actually embedded in
6016           the Perl executable being used.)
6017
6018       DBD driver has not implemented the AutoCommit attribute
6019           The DBD driver implementation is incomplete. Consult the author.
6020
6021       Can't [sg]et %s->{%s}: unrecognised attribute
6022           You attempted to set or get an unknown attribute of a handle.  Make
6023           sure you have spelled the attribute name correctly; case is
6024           significant (e.g., "Autocommit" is not the same as "AutoCommit").
6025

Pure-Perl DBI

6027       A pure-perl emulation of the DBI is included in the distribution for
6028       people using pure-perl drivers who, for whatever reason, can't install
6029       the compiled DBI. See DBI::PurePerl.
6030

SEE ALSO

6032   Driver and Database Documentation
6033       Refer to the documentation for the DBD driver that you are using.
6034
6035       Refer to the SQL Language Reference Manual for the database engine that
6036       you are using.
6037
6038   ODBC and SQL/CLI Standards Reference Information
6039       More detailed information about the semantics of certain DBI methods
6040       that are based on ODBC and SQL/CLI standards is available on-line via
6041       microsoft.com, for ODBC, and www.jtc1sc32.org for the SQL/CLI standard:
6042
6043        DBI method        ODBC function     SQL/CLI Working Draft
6044        ----------        -------------     ---------------------
6045        column_info       SQLColumns        Page 124
6046        foreign_key_info  SQLForeignKeys    Page 163
6047        get_info          SQLGetInfo        Page 214
6048        primary_key_info  SQLPrimaryKeys    Page 254
6049        table_info        SQLTables         Page 294
6050        type_info         SQLGetTypeInfo    Page 239
6051        statistics_info   SQLStatistics
6052
6053       To find documentation on the ODBC function you can use the MSDN search
6054       facility at:
6055
6056           http://msdn.microsoft.com/Search
6057
6058       and search for something like "SQLColumns returns".
6059
6060       And for SQL/CLI standard information on SQLColumns you'd read page 124
6061       of the (very large) SQL/CLI Working Draft available from:
6062
6063         http://jtc1sc32.org/doc/N0701-0750/32N0744T.pdf
6064
6065   Standards Reference Information
6066       A hyperlinked, browsable version of the BNF syntax for SQL92 (plus
6067       Oracle 7 SQL and PL/SQL) is available here:
6068
6069         http://cui.unige.ch/db-research/Enseignement/analyseinfo/SQL92/BNFindex.html
6070
6071       You can find more information about SQL standards online by searching
6072       for the appropriate standard names and numbers. For example, searching
6073       for "ANSI/ISO/IEC International Standard (IS) Database Language SQL -
6074       Part 1: SQL/Framework" you'll find a copy at:
6075
6076         ftp://ftp.iks-jena.de/mitarb/lutz/standards/sql/ansi-iso-9075-1-1999.pdf
6077
6078   Books and Articles
6079       Programming the Perl DBI, by Alligator Descartes and Tim Bunce.
6080       <http://books.perl.org/book/154>
6081
6082       Programming Perl 3rd Ed. by Larry Wall, Tom Christiansen & Jon Orwant.
6083       <http://books.perl.org/book/134>
6084
6085       Learning Perl by Randal Schwartz.  <http://books.perl.org/book/101>
6086
6087       Details of many other books related to perl can be found at
6088       <http://books.perl.org>
6089
6090   Perl Modules
6091       Index of DBI related modules available from CPAN:
6092
6093        L<https://metacpan.org/search?q=DBD%3A%3A>
6094        L<https://metacpan.org/search?q=DBIx%3A%3A>
6095        L<https://metacpan.org/search?q=DBI>
6096
6097       For a good comparison of RDBMS-OO mappers and some OO-RDBMS mappers
6098       (including Class::DBI, Alzabo, and DBIx::RecordSet in the former
6099       category and Tangram and SPOPS in the latter) see the Perl Object-
6100       Oriented Persistence project pages at:
6101
6102        http://poop.sourceforge.net
6103
6104       A similar page for Java toolkits can be found at:
6105
6106        http://c2.com/cgi-bin/wiki?ObjectRelationalToolComparison
6107
6108   Mailing List
6109       The dbi-users mailing list is the primary means of communication among
6110       users of the DBI and its related modules. For details send email to:
6111
6112        L<dbi-users-help@perl.org>
6113
6114       There are typically between 700 and 900 messages per month.  You have
6115       to subscribe in order to be able to post. However you can opt for a
6116       'post-only' subscription.
6117
6118       Mailing list archives (of variable quality) are held at:
6119
6120        http://groups.google.com/groups?group=perl.dbi.users
6121        http://www.xray.mpe.mpg.de/mailing-lists/dbi/
6122        http://www.mail-archive.com/dbi-users%40perl.org/
6123
6124   Assorted Related Links
6125       The DBI "Home Page":
6126
6127        http://dbi.perl.org/
6128
6129       Other DBI related links:
6130
6131        http://www.perlmonks.org/?node=DBI%20recipes
6132        http://www.perlmonks.org/?node=Speeding%20up%20the%20DBI
6133
6134       Other database related links:
6135
6136        http://www.connectionstrings.com/
6137
6138       Security, especially the "SQL Injection" attack:
6139
6140        http://bobby-tables.com/
6141        http://online.securityfocus.com/infocus/1644
6142
6143   FAQ
6144       See <http://faq.dbi-support.com/>
6145

AUTHORS

6147       DBI by Tim Bunce, <http://www.tim.bunce.name>
6148
6149       This pod text by Tim Bunce, J. Douglas Dunlop, Jonathan Leffler and
6150       others.  Perl by Larry Wall and the "perl5-porters".
6151
6153       The DBI module is Copyright (c) 1994-2012 Tim Bunce. Ireland.  All
6154       rights reserved.
6155
6156       You may distribute under the terms of either the GNU General Public
6157       License or the Artistic License, as specified in the Perl 5.10.0 README
6158       file.
6159

SUPPORT / WARRANTY

6161       The DBI is free Open Source software. IT COMES WITHOUT WARRANTY OF ANY
6162       KIND.
6163
6164   Support
6165       My consulting company, Data Plan Services, offers annual and multi-
6166       annual support contracts for the DBI. These provide sustained support
6167       for DBI development, and sustained value for you in return.  Contact me
6168       for details.
6169
6170   Sponsor Enhancements
6171       If your company would benefit from a specific new DBI feature, please
6172       consider sponsoring its development.  Work is performed rapidly, and
6173       usually on a fixed-price payment-on-delivery basis.  Contact me for
6174       details.
6175
6176       Using such targeted financing allows you to contribute to DBI
6177       development, and rapidly get something specific and valuable in return.
6178

ACKNOWLEDGEMENTS

6180       I would like to acknowledge the valuable contributions of the many
6181       people I have worked with on the DBI project, especially in the early
6182       years (1992-1994). In no particular order: Kevin Stock, Buzz Moschetti,
6183       Kurt Andersen, Ted Lemon, William Hails, Garth Kennedy, Michael
6184       Peppler, Neil S. Briscoe, Jeff Urlwin, David J. Hughes, Jeff Stander,
6185       Forrest D Whitcher, Larry Wall, Jeff Fried, Roy Johnson, Paul Hudson,
6186       Georg Rehfeld, Steve Sizemore, Ron Pool, Jon Meek, Tom Christiansen,
6187       Steve Baumgarten, Randal Schwartz, and a whole lot more.
6188
6189       Then, of course, there are the poor souls who have struggled through
6190       untold and undocumented obstacles to actually implement DBI drivers.
6191       Among their ranks are Jochen Wiedmann, Alligator Descartes, Jonathan
6192       Leffler, Jeff Urlwin, Michael Peppler, Henrik Tougaard, Edwin Pratomo,
6193       Davide Migliavacca, Jan Pazdziora, Peter Haworth, Edmund Mergl, Steve
6194       Williams, Thomas Lowery, and Phlip Plumlee. Without them, the DBI would
6195       not be the practical reality it is today.  I'm also especially grateful
6196       to Alligator Descartes for starting work on the first edition of the
6197       "Programming the Perl DBI" book and letting me jump on board.
6198
6199       The DBI and DBD::Oracle were originally developed while I was Technical
6200       Director (CTO) of the Paul Ingram Group in the UK.  So I'd especially
6201       like to thank Paul for his generosity and vision in supporting this
6202       work for many years.
6203
6204       A couple of specific DBI features have been sponsored by enlightened
6205       companies:
6206
6207       The development of the swap_inner_handle() method was sponsored by
6208       BizRate.com (<http://BizRate.com>)
6209
6210       The development of DBD::Gofer and related modules was sponsored by
6211       Shopzilla.com (<http://Shopzilla.com>), where I currently work.
6212

CONTRIBUTING

6214       As you can see above, many people have contributed to the DBI and
6215       drivers in many ways over many years.
6216
6217       If you'd like to help then see <http://dbi.perl.org/contributing>.
6218
6219       If you'd like the DBI to do something new or different then a good way
6220       to make that happen is to do it yourself and send me a patch to the
6221       source code that shows the changes. (But read "Speak before you patch"
6222       below.)
6223
6224   Browsing the source code repository
6225       Use https://github.com/perl5-dbi/dbi
6226
6227   How to create a patch using Git
6228       The DBI source code is maintained using Git.  To access the source
6229       you'll need to install a Git client. Then, to get the source code, do:
6230
6231         git clone https://github.com/perl5-dbi/dbi.git DBI-git
6232
6233       The source code will now be available in the new subdirectory
6234       "DBI-git".
6235
6236       When you want to synchronize later, issue the command
6237
6238         git pull --all
6239
6240       Make your changes, test them, test them again until everything passes.
6241       If there are no tests for the new feature you added or a behaviour
6242       change, the change should include a new test. Then commit the changes.
6243       Either use
6244
6245         git gui
6246
6247       or
6248
6249         git commit -a -m 'Message to my changes'
6250
6251       If you get any conflicts reported you'll need to fix them first.
6252
6253       Then generate the patch file to be mailed:
6254
6255         git format-patch -1 --attach
6256
6257       which will create a file 0001-*.patch (where * relates to the commit
6258       message).  Read the patch file, as a sanity check, and then email it to
6259       dbi-dev@perl.org.
6260
6261       If you have a github <https://github.com> account, you can also fork
6262       the repository, commit your changes to the forked repository and then
6263       do a pull request.
6264
6265   How to create a patch without Git
6266       Unpack a fresh copy of the distribution:
6267
6268         wget http://cpan.metacpan.org/authors/id/T/TI/TIMB/DBI-1.627.tar.gz
6269         tar xfz DBI-1.627.tar.gz
6270
6271       Rename the newly created top level directory:
6272
6273         mv DBI-1.627 DBI-1.627.your_foo
6274
6275       Edit the contents of DBI-1.627.your_foo/* till it does what you want.
6276
6277       Test your changes and then remove all temporary files:
6278
6279         make test && make distclean
6280
6281       Go back to the directory you originally unpacked the distribution:
6282
6283         cd ..
6284
6285       Unpack another copy of the original distribution you started with:
6286
6287         tar xfz DBI-1.627.tar.gz
6288
6289       Then create a patch file by performing a recursive "diff" on the two
6290       top level directories:
6291
6292         diff -purd DBI-1.627 DBI-1.627.your_foo > DBI-1.627.your_foo.patch
6293
6294   Speak before you patch
6295       For anything non-trivial or possibly controversial it's a good idea to
6296       discuss (on dbi-dev@perl.org) the changes you propose before actually
6297       spending time working on them. Otherwise you run the risk of them being
6298       rejected because they don't fit into some larger plans you may not be
6299       aware of.
6300
6301       You can also reach the developers on IRC (chat). If they are on-line,
6302       the most likely place to talk to them is the #dbi channel on
6303       irc.perl.org
6304

TRANSLATIONS

6306       A German translation of this manual (possibly slightly out of date) is
6307       available, thanks to O'Reilly, at:
6308
6309         http://www.oreilly.de/catalog/perldbiger/
6310
6312       Apache::DBI
6313           To be used with the Apache daemon together with an embedded Perl
6314           interpreter like "mod_perl". Establishes a database connection
6315           which remains open for the lifetime of the HTTP daemon. This way
6316           the CGI connect and disconnect for every database access becomes
6317           superfluous.
6318
6319       SQL Parser
6320           See also the SQL::Statement module, SQL parser and engine.
6321
6322
6323
6324perl v5.32.1                      2021-01-27                            DBI(3)
Impressum