1Template::Manual::DirecUtsievresC(o3n)tributed Perl DocuTmeemnptlaattieo:n:Manual::Directives(3)
2
3
4

NAME

6       Template::Manual::Directives - Template directives
7

Accessing and Updating Template Variables

9   GET
10       The "GET" directive retrieves and outputs the value of the named
11       variable.
12
13           [% GET foo %]
14
15       The "GET" keyword is optional.  A variable can be specified in a
16       directive tag by itself.
17
18           [% foo %]
19
20       The variable can have an unlimited number of elements, each separated
21       by a dot. Each element can have arguments specified within parentheses.
22
23           [% foo %]
24           [% bar.baz %]
25           [% biz.baz(10) %]
26           ...etc...
27
28       See Template::Manual::Variables for a full discussion on template
29       variables.
30
31       You can also specify expressions using the logical ("and", "or", "not",
32       "?", ":") and mathematic operators ("+", "-", "*", "/", "%", "mod",
33       "div").
34
35           [% template.title or default.title %]
36
37           [% score * 100 %]
38
39           [% order.nitems ? checkout(order.total) : 'no items' %]
40
41       The "div" operator returns the integer result of division.  Both "%"
42       and "mod" return the modulus (i.e. remainder) of division.
43
44           [% 15 / 6 %]            # 2.5
45           [% 15 div 6 %]          # 2
46           [% 15 mod 6 %]          # 3
47
48   CALL
49       The "CALL" directive is similar to "GET" in evaluating the variable
50       named, but doesn't print the result returned.  This can be useful when
51       a variable is bound to a sub-routine or object method which you want to
52       call but aren't interested in the value returned.
53
54           [% CALL dbi.disconnect %]
55
56           [% CALL inc_page_counter(page_count) %]
57
58   SET
59       The "SET" directive allows you to assign new values to existing
60       variables or create new temporary variables.
61
62           [% SET title = 'Hello World' %]
63
64       The "SET" keyword is also optional.
65
66           [% title = 'Hello World' %]
67
68       Variables may be assigned the values of other variables, unquoted
69       numbers (2.718), literal text ('single quotes') or quoted text ("double
70       quotes").  In the latter case, any variable references within the text
71       will be interpolated when the string is evaluated.  Variables should be
72       prefixed by "$", using curly braces to explicitly scope the variable
73       name where necessary.
74
75           [% foo  = 'Foo'  %]               # literal value 'Foo'
76           [% bar  =  foo   %]               # value of variable 'foo'
77           [% cost = '$100' %]               # literal value '$100'
78           [% item = "$bar: ${cost}.00" %]   # value "Foo: $100.00"
79
80       Multiple variables may be assigned in the same directive and are
81       evaluated in the order specified.  Thus, the above could have been
82       written:
83
84           [% foo  = 'Foo'
85              bar  = foo
86              cost = '$100'
87              item = "$bar: ${cost}.00"
88           %]
89
90       Simple expressions can also be used, as per "GET".
91
92           [% ten    = 10
93              twenty = 20
94              thirty = twenty + ten
95              forty  = 2 * twenty
96              fifty  = 100 div 2
97              six    = twenty mod 7
98           %]
99
100       You can concatenate strings together using the '_' underscore operator.
101       In Perl 5, the "." dot is used for string concatenation, but in Perl 6,
102       as in the Template Toolkit, the "." dot will be used as the method
103       calling operator and '_' underscore will be used for string
104       concatenation.  Note that the operator must be specified with
105       surrounding whitespace which, as Larry says, is construed as a feature:
106
107           [% copyright = '(C) Copyright' _ year _ ' ' _ author %]
108
109       You can, of course, achieve a similar effect with double quoted string
110       interpolation.
111
112           [% copyright = "(C) Copyright $year $author" %]
113
114   DEFAULT
115       The "DEFAULT" directive is similar to "SET" but only updates variables
116       that are currently undefined or have no "true" value (in the Perl
117       sense).
118
119           [% DEFAULT
120               name = 'John Doe'
121               id   = 'jdoe'
122           %]
123
124       This can be particularly useful in common template components to ensure
125       that some sensible default are provided for otherwise undefined
126       variables.
127
128           [% DEFAULT
129              title = 'Hello World'
130              bgcol = '#ffffff'
131           %]
132           <html>
133             <head>
134               <title>[% title %]</title>
135             </head>
136             <body bgcolor="[% bgcol %]">
137               ...etc...
138

Processing Template Files and Blocks

140   INSERT
141       The "INSERT" directive is used to insert the contents of an external
142       file at the current position.
143
144           [% INSERT myfile %]
145
146       No attempt to parse or process the file is made.  The contents,
147       possibly including any embedded template directives, are inserted
148       intact.
149
150       The filename specified should be relative to one of the "INCLUDE_PATH"
151       directories.  Absolute (i.e. starting with "/") and relative (i.e.
152       starting with ".") filenames may be used if the "ABSOLUTE" and
153       "RELATIVE" options are set, respectively.  Both these options are
154       disabled by default.
155
156           my $template = Template->new({
157               INCLUDE_PATH => '/here:/there',
158           });
159
160           $template->process('myfile');
161
162       myfile:
163
164           [% INSERT foo %]            # looks for /here/foo then /there/foo
165           [% INSERT /etc/passwd %]    # file error: ABSOLUTE not set
166           [% INSERT ../secret %]      # file error: RELATIVE not set
167
168       For convenience, the filename does not need to be quoted as long as it
169       contains only alphanumeric characters, underscores, dots or forward
170       slashes.  Names containing any other characters should be quoted.
171
172           [% INSERT misc/legalese.txt            %]
173           [% INSERT 'dos98/Program Files/stupid' %]
174
175       To evaluate a variable to specify a filename, you should explicitly
176       prefix it with a "$" or use double-quoted string interpolation.
177
178           [% language = 'en'
179              legalese = 'misc/legalese.txt'
180           %]
181
182           [% INSERT $legalese %]              # misc/legalese.txt
183           [% INSERT "$language/$legalese" %]  # en/misc/legalese.txt
184
185       Multiple files can be specified using "+" as a delimiter.  All files
186       should be unquoted names or quoted strings.  Any variables should be
187       interpolated into double-quoted strings.
188
189           [% INSERT legalese.txt + warning.txt %]
190           [% INSERT  "$legalese" + warning.txt %]  # requires quoting
191
192   INCLUDE
193       The "INCLUDE" directive is used to process and include the output of
194       another template file or block.
195
196           [% INCLUDE header %]
197
198       If a "BLOCK" of the specified name is defined in the same file, or in a
199       file from which the current template has been called (i.e. a parent
200       template) then it will be used in preference to any file of the same
201       name.
202
203           [% INCLUDE table %]     # uses BLOCK defined below
204
205           [% BLOCK table %]
206              <table>
207                ...
208              </table>
209           [% END %]
210
211       If a "BLOCK" definition is not currently visible then the template name
212       should be a file relative to one of the "INCLUDE_PATH" directories, or
213       an absolute or relative file name if the "ABSOLUTE"/"RELATIVE" options
214       are appropriately enabled.  The "INCLUDE" directive automatically
215       quotes the filename specified, as per "INSERT" described above.  When a
216       variable contains the name of the template for the "INCLUDE" directive,
217       it should be explicitly prefixed by "$" or double-quoted
218
219           [% myheader = 'my/misc/header' %]
220           [% INCLUDE   myheader  %]           # 'myheader'
221           [% INCLUDE  $myheader  %]           # 'my/misc/header'
222           [% INCLUDE "$myheader" %]           # 'my/misc/header'
223
224       Any template directives embedded within the file will be processed
225       accordingly.  All variables currently defined will be visible and
226       accessible from within the included template.
227
228           [% title = 'Hello World' %]
229           [% INCLUDE header %]
230           <body>
231           ...
232
233       header:
234
235           <html>
236           <title>[% title %]</title>
237
238       output:
239
240           <html>
241           <title>Hello World</title>
242           <body>
243           ...
244
245       Local variable definitions may be specified after the template name,
246       temporarily masking any existing variables.  Insignificant whitespace
247       is ignored within directives so you can add variable definitions on the
248       same line, the next line or split across several line with comments
249       interspersed, if you prefer.
250
251           [% INCLUDE table %]
252
253           [% INCLUDE table title="Active Projects" %]
254
255           [% INCLUDE table
256                title   = "Active Projects"
257                bgcolor = "#80ff00"    # chartreuse
258                border  = 2
259           %]
260
261       The "INCLUDE" directive localises (i.e. copies) all variables before
262       processing the template.  Any changes made within the included template
263       will not affect variables in the including template.
264
265           [% foo = 10 %]
266
267           foo is originally [% foo %]
268           [% INCLUDE bar %]
269           foo is still [% foo %]
270
271           [% BLOCK bar %]
272              foo was [% foo %]
273              [% foo = 20 %]
274              foo is now [% foo %]
275           [% END %]
276
277       output:
278
279           foo is originally 10
280              foo was 10
281              foo is now 20
282           foo is still 10
283
284       Technical Note: the localisation of the stash (that is, the process by
285       which variables are copied before an "INCLUDE" to prevent being
286       overwritten) is only skin deep.  The top-level variable namespace
287       (hash) is copied, but no attempt is made to perform a deep-copy of
288       other structures (hashes, arrays, objects, etc.)  Therefore, a "foo"
289       variable referencing a hash will be copied to create a new "foo"
290       variable but which points to the same hash array.  Thus, if you update
291       compound variables (e.g. "foo.bar") then you will change the original
292       copy, regardless of any stash localisation.  If you're not worried
293       about preserving variable values, or you trust the templates you're
294       including then you might prefer to use the "PROCESS" directive which is
295       faster by virtue of not performing any localisation.
296
297       You can specify dotted variables as "local" variables to an "INCLUDE"
298       directive.  However, be aware that because of the localisation issues
299       explained above (if you skipped the previous Technical Note above then
300       you might want to go back and read it or skip this section too), the
301       variables might not actually be "local". If the first element of the
302       variable name already references a hash array then the variable update
303       will affect the original variable.
304
305           [% foo = {
306                  bar = 'Baz'
307              }
308           %]
309
310           [% INCLUDE somefile foo.bar='Boz' %]
311
312           [% foo.bar %]           # Boz
313
314       This behaviour can be a little unpredictable (and may well be improved
315       upon in a future version).  If you know what you're doing with it and
316       you're sure that the variables in question are defined (nor not) as you
317       expect them to be, then you can rely on this feature to implement some
318       powerful "global" data sharing techniques.  Otherwise, you might prefer
319       to steer well clear and always pass simple (undotted) variables as
320       parameters to "INCLUDE" and other similar directives.
321
322       If you want to process several templates in one go then you can specify
323       each of their names (quoted or unquoted names only, no unquoted
324       $variables) joined together by "+".  The "INCLUDE" directive will then
325       process them in order.
326
327           [% INCLUDE html/header + "site/$header" + site/menu
328                title = "My Groovy Web Site"
329           %]
330
331       The variable stash is localised once and then the templates specified
332       are processed in order, all within that same variable context.  This
333       makes it slightly faster than specifying several separate "INCLUDE"
334       directives (because you only clone the variable stash once instead of n
335       times), but not quite as "safe" because any variable changes in the
336       first file will be visible in the second, third and so on.  This might
337       be what you want, of course, but then again, it might not.
338
339   PROCESS
340       The PROCESS directive is similar to "INCLUDE" but does not perform any
341       localisation of variables before processing the template.  Any changes
342       made to variables within the included template will be visible in the
343       including template.
344
345           [% foo = 10 %]
346
347           foo is [% foo %]
348           [% PROCESS bar %]
349           foo is [% foo %]
350
351           [% BLOCK bar %]
352              [% foo = 20 %]
353              changed foo to [% foo %]
354           [% END %]
355
356       output:
357
358           foo is 10
359              changed foo to 20
360           foo is 20
361
362       Parameters may be specified in the "PROCESS" directive, but these too
363       will become visible changes to current variable values.
364
365           [% foo = 10 %]
366           foo is [% foo %]
367           [% PROCESS bar
368              foo = 20
369           %]
370           foo is [% foo %]
371
372           [% BLOCK bar %]
373              this is bar, foo is [% foo %]
374           [% END %]
375
376       output:
377
378           foo is 10
379              this is bar, foo is 20
380           foo is 20
381
382       The "PROCESS" directive is slightly faster than "INCLUDE" because it
383       avoids the need to localise (i.e. copy) the variable stash before
384       processing the template.  As with "INSERT" and "INCLUDE", the first
385       parameter does not need to be quoted as long as it contains only
386       alphanumeric characters, underscores, periods or forward slashes.  A
387       "$" prefix can be used to explicitly indicate a variable which should
388       be interpolated to provide the template name:
389
390           [% myheader = 'my/misc/header' %]
391           [% PROCESS  myheader %]              # 'myheader'
392           [% PROCESS $myheader %]              # 'my/misc/header'
393
394       As with "INCLUDE", multiple templates can be specified, delimited by
395       "+", and are processed in order.
396
397           [% PROCESS html/header + my/header %]
398
399   WRAPPER
400       It's not unusual to find yourself adding common headers and footers to
401       pages or sub-sections within a page.  Something like this:
402
403           [% INCLUDE section/header
404              title = 'Quantum Mechanics'
405           %]
406              Quantum mechanics is a very interesting subject wish
407              should prove easy for the layman to fully comprehend.
408           [% INCLUDE section/footer %]
409
410           [% INCLUDE section/header
411              title = 'Desktop Nuclear Fusion for under $50'
412           %]
413              This describes a simple device which generates significant
414              sustainable electrical power from common tap water by process
415              of nuclear fusion.
416           [% INCLUDE section/footer %]
417
418       The individual template components being included might look like
419       these:
420
421       section/header:
422
423           <p>
424           <h2>[% title %]</h2>
425
426       section/footer:
427
428           </p>
429
430       The "WRAPPER" directive provides a way of simplifying this a little. It
431       encloses a block up to a matching "END" directive, which is first
432       processed to generate some output. This is then passed to the named
433       template file or "BLOCK" as the "content" variable.
434
435           [% WRAPPER section
436              title = 'Quantum Mechanics'
437           %]
438              Quantum mechanics is a very interesting subject wish
439              should prove easy for the layman to fully comprehend.
440           [% END %]
441
442           [% WRAPPER section
443              title = 'Desktop Nuclear Fusion for under $50'
444           %]
445              This describes a simple device which generates significant
446              sustainable electrical power from common tap water by process
447              of nuclear fusion.
448           [% END %]
449
450       The single 'section' template can then be defined as:
451
452           <h2>[% title %]</h2>
453           <p>
454             [% content %]
455           </p>
456
457       Like other block directives, it can be used in side-effect notation:
458
459           [% INSERT legalese.txt WRAPPER big_bold_table %]
460
461       It's also possible to specify multiple templates to a "WRAPPER"
462       directive.  The specification order indicates outermost to innermost
463       wrapper templates.  For example, given the following template block
464       definitions:
465
466           [% BLOCK bold   %]<b>[% content %]</b>[% END %]
467           [% BLOCK italic %]<i>[% content %]</i>[% END %]
468
469       the directive
470
471           [% WRAPPER bold+italic %]Hello World[% END %]
472
473       would generate the following output:
474
475           <b><i>Hello World</i></b>
476
477   BLOCK
478       The "BLOCK"..."END" construct can be used to define template component
479       blocks which can be processed with the "INCLUDE", "PROCESS" and
480       "WRAPPER" directives.
481
482           [% BLOCK tabrow %]
483           <tr>
484             <td>[% name %]<td>
485             <td>[% email %]</td>
486           </tr>
487           [% END %]
488
489           <table>
490             [% PROCESS tabrow  name='Fred'  email='fred@nowhere.com' %]
491             [% PROCESS tabrow  name='Alan'  email='alan@nowhere.com' %]
492           </table>
493
494       A "BLOCK" definition can be used before it is defined, as long as the
495       definition resides in the same file.  The block definition itself does
496       not generate any output.
497
498           [% PROCESS tmpblk %]
499
500           [% BLOCK tmpblk %] This is OK [% END %]
501
502       You can use an anonymous "BLOCK" to capture the output of a template
503       fragment.
504
505           [% julius = BLOCK %]
506              And Caesar's spirit, ranging for revenge,
507              With Ate by his side come hot from hell,
508              Shall in these confines with a monarch's voice
509              Cry  'Havoc', and let slip the dogs of war;
510              That this foul deed shall smell above the earth
511              With carrion men, groaning for burial.
512           [% END %]
513
514       Like a named block, it can contain any other template directives which
515       are processed when the block is defined.  The output generated by the
516       block is then assigned to the variable "julius".
517
518       Anonymous "BLOCK"s can also be used to define block macros.  The
519       enclosing block is processed each time the macro is called.
520
521           [% MACRO locate BLOCK %]
522              The [% animal %] sat on the [% place %].
523           [% END %]
524
525           [% locate(animal='cat', place='mat') %]    # The cat sat on the mat
526           [% locate(animal='dog', place='log') %]    # The dog sat on the log
527

Conditional Processing

529   IF / UNLESS / ELSIF / ELSE
530       The "IF" and "UNLESS" directives can be used to process or ignore a
531       block based on some run-time condition.
532
533           [% IF frames %]
534              [% INCLUDE frameset %]
535           [% END %]
536
537           [% UNLESS text_mode %]
538              [% INCLUDE biglogo %]
539           [% END %]
540
541       Multiple conditions may be joined with "ELSIF" and/or "ELSE" blocks.
542
543           [% IF age < 10 %]
544              Hello [% name %], does your mother know you're
545              using her AOL account?
546           [% ELSIF age < 18 %]
547              Sorry, you're not old enough to enter
548              (and too dumb to lie about your age)
549           [% ELSE %]
550              Welcome [% name %].
551           [% END %]
552
553       The following conditional and boolean operators may be used:
554
555           == != < <= > >= && || ! and or not
556
557       Conditions may be arbitrarily complex and are evaluated with the same
558       precedence as in Perl.  Parenthesis may be used to explicitly determine
559       evaluation order.
560
561           # ridiculously contrived complex example
562           [% IF (name == 'admin' || uid <= 0) && mode == 'debug' %]
563              I'm confused.
564           [% ELSIF more > less %]
565              That's more or less correct.
566           [% END %]
567
568       The "and", "or" and "not" operator are provided as aliases for "&&",
569       "||" and "!", respectively.  Unlike Perl, which treats "and", "or" and
570       "not" as separate, lower-precedence versions of the other operators,
571       the Template Toolkit performs a straightforward substitution of "and"
572       for "&&", and so on.  That means that "and", "or" and "not" have the
573       same operator precedence as "&&", "||" and "!".
574
575   SWITCH / CASE
576       The "SWITCH" / "CASE" construct can be used to perform a multi-way
577       conditional test.  The "SWITCH" directive expects an expression which
578       is first evaluated and then compared against each CASE statement in
579       turn.  Each "CASE" directive should contain a single value or a list of
580       values which should match.  "CASE" may also be left blank or written as
581       "[% CASE DEFAULT %]" to specify a default match.  Only one "CASE"
582       matches, there is no drop-through between "CASE" statements.
583
584           [% SWITCH myvar %]
585           [%   CASE 'value1' %]
586                  ...
587           [%   CASE ['value2', 'value3'] %]   # multiple values
588                  ...
589           [%   CASE myhash.keys %]            # ditto
590                  ...
591           [%   CASE %]                        # default
592                  ...
593           [% END %]
594

Loop Processing

596   FOREACH
597       The "FOREACH" directive will iterate through the items in a list,
598       processing the enclosed block for each one.
599
600           [% foo   = 'Foo'
601              items = [ 'one', 'two', 'three' ]
602           %]
603
604           Things:
605           [% FOREACH thing IN [ foo 'Bar' "$foo Baz" ] %]
606              * [% thing %]
607           [% END %]
608
609           Items:
610           [% FOREACH i IN items %]
611              * [% i %]
612           [% END %]
613
614           Stuff:
615           [% stuff = [ foo "$foo Bar" ] %]
616           [% FOREACH s IN stuff %]
617              * [% s %]
618           [% END %]
619
620       output:
621
622           Things:
623             * Foo
624             * Bar
625             * Foo Baz
626
627           Items:
628             * one
629             * two
630             * three
631
632           Stuff:
633             * Foo
634             * Foo Bar
635
636       You can use also use "=" instead of "IN" if you prefer.
637
638           [% FOREACH i = items %]
639
640       When the "FOREACH" directive is used without specifying a target
641       variable, any iterated values which are hash references will be
642       automatically imported.
643
644           [% userlist = [
645               { id => 'tom',   name => 'Thomas'  },
646               { id => 'dick',  name => 'Richard'  },
647               { id => 'larry', name => 'Lawrence' },
648              ]
649           %]
650
651           [% FOREACH user IN userlist %]
652              [% user.id %] [% user.name %]
653           [% END %]
654
655       short form:
656
657           [% FOREACH userlist %]
658              [% id %] [% name %]
659           [% END %]
660
661       Note that this particular usage creates a localised variable context to
662       prevent the imported hash keys from overwriting any existing variables.
663       The imported definitions and any other variables defined in such a
664       "FOREACH" loop will be lost at the end of the loop, when the previous
665       context and variable values are restored.
666
667       However, under normal operation, the loop variable remains in scope
668       after the "FOREACH" loop has ended (caveat: overwriting any variable
669       previously in scope). This is useful as the loop variable is secretly
670       an iterator object (see below) and can be used to analyse the last
671       entry processed by the loop.
672
673       The "FOREACH" directive can also be used to iterate through the entries
674       in a hash array.  Each entry in the hash is returned in sorted order
675       (based on the key) as a hash array containing 'key' and 'value' items.
676
677           [% users = {
678                tom   => 'Thomas',
679                dick  => 'Richard',
680                larry => 'Lawrence',
681              }
682           %]
683
684           [% FOREACH u IN users %]
685              * [% u.key %] : [% u.value %]
686           [% END %]
687
688       Output:
689
690              * dick : Richard
691              * larry : Lawrence
692              * tom : Thomas
693
694       The "NEXT" directive starts the next iteration in the "FOREACH" loop.
695
696           [% FOREACH user IN userlist %]
697              [% NEXT IF user.isguest %]
698              Name: [% user.name %]    Email: [% user.email %]
699           [% END %]
700
701       The "LAST" directive can be used to prematurely exit the loop.  "BREAK"
702       is also provided as an alias for "LAST".
703
704           [% FOREACH match IN results.nsort('score').reverse %]
705              [% LAST IF match.score < 50 %]
706              [% match.score %] : [% match.url %]
707           [% END %]
708
709       The "FOREACH" directive is implemented using the Template::Iterator
710       module.  A reference to the iterator object for a "FOREACH" directive
711       is implicitly available in the "loop" variable.  The following methods
712       can be called on the "loop" iterator.
713
714           size()      number of elements in the list
715           max()       index number of last element (size - 1)
716           index()     index of current iteration from 0 to max()
717           count()     iteration counter from 1 to size() (i.e. index() + 1)
718           first()     true if the current iteration is the first
719           last()      true if the current iteration is the last
720           prev()      return the previous item in the list
721           next()      return the next item in the list
722
723       See Template::Iterator for further details.
724
725       Example:
726
727           [% FOREACH item IN [ 'foo', 'bar', 'baz' ] -%]
728              [%- "<ul>\n" IF loop.first %]
729              <li>[% loop.count %]/[% loop.size %]: [% item %]
730              [%- "</ul>\n" IF loop.last %]
731           [% END %]
732
733       Output:
734
735           <ul>
736           <li>1/3: foo
737           <li>2/3: bar
738           <li>3/3: baz
739           </ul>
740
741       Nested loops will work as expected, with the "loop" variable correctly
742       referencing the innermost loop and being restored to any previous value
743       (i.e. an outer loop) at the end of the loop.
744
745           [% FOREACH group IN grouplist;
746                # loop => group iterator
747                "Groups:\n" IF loop.first;
748
749                FOREACH user IN group.userlist;
750                   # loop => user iterator
751                   "$loop.count: $user.name\n";
752                END;
753
754                # loop => group iterator
755                "End of Groups\n" IF loop.last;
756              END
757           %]
758
759       The "iterator" plugin can also be used to explicitly create an iterator
760       object.  This can be useful within nested loops where you need to keep
761       a reference to the outer iterator within the inner loop.  The iterator
762       plugin effectively allows you to create an iterator by a name other
763       than "loop".  See Template::Plugin::Iterator for further details.
764
765           [% USE giter = iterator(grouplist) %]
766
767           [% FOREACH group IN giter %]
768              [% FOREACH user IN group.userlist %]
769                    user #[% loop.count %] in
770                    group [% giter.count %] is
771                    named [% user.name %]
772              [% END %]
773           [% END %]
774
775   WHILE
776       The "WHILE" directive can be used to repeatedly process a template
777       block while a conditional expression evaluates true.  The expression
778       may be arbitrarily complex as per "IF" / "UNLESS".
779
780           [% WHILE total < 100 %]
781              ...
782              [% total = calculate_new_total %]
783           [% END %]
784
785       An assignment can be enclosed in parenthesis to evaluate the assigned
786       value.
787
788           [% WHILE (user = get_next_user_record) %]
789              [% user.name %]
790           [% END %]
791
792       The "NEXT" directive can be used to start the next iteration of a
793       "WHILE" loop and "BREAK" can be used to exit the loop, both as per
794       "FOREACH".
795
796       The Template Toolkit uses a failsafe counter to prevent runaway "WHILE"
797       loops which would otherwise never terminate.  If the loop exceeds 1000
798       iterations then an "undef" exception will be thrown, reporting the
799       error:
800
801           WHILE loop terminated (> 1000 iterations)
802
803       The $Template::Directive::WHILE_MAX variable controls this behaviour
804       and can be set to a higher value if necessary.
805

Filters, Plugins, Macros and Perl

807   FILTER
808       The "FILTER" directive can be used to post-process the output of a
809       block.  A number of standard filters are provided with the Template
810       Toolkit.  The "html" filter, for example, escapes the '<', '>' and '&'
811       characters to prevent them from being interpreted as HTML tags or
812       entity reference markers.
813
814           [% FILTER html %]
815              HTML text may have < and > characters embedded
816              which you want converted to the correct HTML entities.
817           [% END %]
818
819       output:
820
821              HTML text may have &lt; and &gt; characters embedded
822              which you want converted to the correct HTML entities.
823
824       The "FILTER" directive can also follow various other non-block
825       directives.  For example:
826
827           [% INCLUDE mytext FILTER html %]
828
829       The "|" character can also be used as an alias for "FILTER".
830
831           [% INCLUDE mytext | html %]
832
833       Multiple filters can be chained together and will be called in
834       sequence.
835
836           [% INCLUDE mytext FILTER html FILTER html_para %]
837
838       or
839
840           [% INCLUDE mytext | html | html_para %]
841
842       Filters come in two flavours, known as 'static' or 'dynamic'.  A static
843       filter is a simple subroutine which accepts a text string as the only
844       argument and returns the modified text.  The "html" filter is an
845       example of a static filter, implemented as:
846
847           sub html_filter {
848               my $text = shift;
849               for ($text) {
850                   s/&/&amp;/g;
851                   s/</&lt;/g;
852                   s/>/&gt;/g;
853               }
854               return $text;
855           }
856
857       Dynamic filters can accept arguments which are specified when the
858       filter is called from a template.  The "repeat" filter is such an
859       example, accepting a numerical argument which specifies the number of
860       times that the input text should be repeated.
861
862           [% FILTER repeat(3) %]blah [% END %]
863
864       output:
865
866           blah blah blah
867
868       These are implemented as filter 'factories'.  The factory subroutine is
869       passed a reference to the current Template::Context object along with
870       any additional arguments specified.  It should then return a subroutine
871       reference (e.g. a closure) which implements the filter.  The "repeat"
872       filter factory is implemented like this:
873
874           sub repeat_filter_factory {
875               my ($context, $iter) = @_;
876               $iter = 1 unless defined $iter;
877
878               return sub {
879                   my $text = shift;
880                   $text = '' unless defined $text;
881                   return join('\n', $text) x $iter;
882               }
883           }
884
885       The "FILTERS" option, described in Template::Manual::Config, allows
886       custom filters to be defined when a Template object is instantiated.
887       The define_filter() method allows further filters to be defined at any
888       time.
889
890       When using a filter, it is possible to assign an alias to it for
891       further use.  This is most useful for dynamic filters that you want to
892       re-use with the same configuration.
893
894           [% FILTER echo = repeat(2) %]
895           Is there anybody out there?
896           [% END %]
897
898           [% FILTER echo %]
899           Mother, should I build a wall?
900           [% END %]
901
902       Output:
903
904           Is there anybody out there?
905           Is there anybody out there?
906
907           Mother, should I build a wall?
908           Mother, should I build a wall?
909
910       The "FILTER" directive automatically quotes the name of the filter.  As
911       with "INCLUDE" et al, you can use a variable to provide the name of the
912       filter, prefixed by "$".
913
914           [% myfilter = 'html' %]
915           [% FILTER $myfilter %]      # same as [% FILTER html %]
916              ...
917           [% END %]
918
919       A template variable can also be used to define a static filter
920       subroutine.  However, the Template Toolkit will automatically call any
921       subroutine bound to a variable and use the value returned.  Thus, the
922       above example could be implemented as:
923
924           my $vars = {
925               myfilter => sub { return 'html' },
926           };
927
928       template:
929
930           [% FILTER $myfilter %]      # same as [% FILTER html %]
931              ...
932           [% END %]
933
934       To define a template variable that evaluates to a subroutine reference
935       that can be used by the "FILTER" directive, you should create a
936       subroutine that, when called automatically by the Template Toolkit,
937       returns another subroutine reference which can then be used to perform
938       the filter operation.  Note that only static filters can be implemented
939       in this way.
940
941           my $vars = {
942               myfilter => sub { \&my_filter_sub },
943           };
944
945           sub my_filter_sub {
946               my $text = shift;
947               # do something
948               return $text;
949           }
950
951       template:
952
953           [% FILTER $myfilter %]
954              ...
955           [% END %]
956
957       Alternately, you can bless a subroutine reference into a class (any
958       class will do) to fool the Template Toolkit into thinking it's an
959       object rather than a subroutine.  This will then bypass the automatic
960       "call-a-subroutine-to-return-a-value" magic.
961
962           my $vars = {
963               myfilter => bless(\&my_filter_sub, 'anything_you_like'),
964           };
965
966       template:
967
968           [% FILTER $myfilter %]
969              ...
970           [% END %]
971
972       Filters bound to template variables remain local to the variable
973       context in which they are defined. That is, if you define a filter in a
974       "PERL" block within a template that is loaded via "INCLUDE", then the
975       filter definition will only exist until the end of that template when
976       the stash is delocalised, restoring the previous variable state. If you
977       want to define a filter which persists for the lifetime of the
978       processor, or define additional dynamic filter factories, then you can
979       call the define_filter() method on the current Template::Context
980       object.
981
982       See Template::Manual::Filters for a complete list of available filters,
983       their descriptions and examples of use.
984
985   USE
986       The "USE" directive can be used to load and initialise "plugin"
987       extension modules.
988
989           [% USE myplugin %]
990
991       A plugin is a regular Perl module that conforms to a particular object-
992       oriented interface, allowing it to be loaded into and used
993       automatically by the Template Toolkit.  For details of this interface
994       and information on writing plugins, consult Template::Plugin.
995
996       A number of standard plugins are included with the Template Toolkit
997       (see below and Template::Manual::Plugins).  The names of these standard
998       plugins are case insensitive.
999
1000           [% USE CGI   %]        # => Template::Plugin::CGI
1001           [% USE Cgi   %]        # => Template::Plugin::CGI
1002           [% USE cgi   %]        # => Template::Plugin::CGI
1003
1004       You can also define further plugins using the "PLUGINS" option.
1005
1006           my $tt = Template->new({
1007               PLUGINS => {
1008                   foo => 'My::Plugin::Foo',
1009                   bar => 'My::Plugin::Bar',
1010               },
1011           });
1012
1013       The recommended convention is to specify these plugin names in lower
1014       case.  The Template Toolkit first looks for an exact case-sensitive
1015       match and then tries the lower case conversion of the name specified.
1016
1017           [% USE Foo %]      # look for 'Foo' then 'foo'
1018
1019       If you define all your "PLUGINS" with lower case names then they will
1020       be located regardless of how the user specifies the name in the "USE"
1021       directive.  If, on the other hand, you define your "PLUGINS" with upper
1022       or mixed case names then the name specified in the "USE" directive must
1023       match the case exactly.
1024
1025       If the plugin isn't defined in either the standard plugins
1026       ($Template::Plugins::STD_PLUGINS) or via the "PLUGINS" option, then the
1027       "PLUGIN_BASE" is searched.
1028
1029       In this case the plugin name is case-sensitive.  It is appended to each
1030       of the "PLUGIN_BASE" module namespaces in turn (default:
1031       "Template::Plugin") to construct a full module name which it attempts
1032       to locate and load.  Any periods, '"."', in the name will be converted
1033       to '"::"'.
1034
1035           [% USE MyPlugin %]     #  => Template::Plugin::MyPlugin
1036           [% USE Foo.Bar  %]     #  => Template::Plugin::Foo::Bar
1037
1038       The "LOAD_PERL" option (disabled by default) provides a further way by
1039       which external Perl modules may be loaded.  If a regular Perl module
1040       (i.e. not a "Template::Plugin::*" or other module relative to some
1041       "PLUGIN_BASE") supports an object-oriented interface and a "new()"
1042       constructor then it can be loaded and instantiated automatically.  The
1043       following trivial example shows how the IO::File module might be used.
1044
1045           [% USE file = IO.File('/tmp/mydata') %]
1046
1047           [% WHILE (line = file.getline) %]
1048              <!-- [% line %] -->
1049           [% END %]
1050
1051       Any additional parameters supplied in parenthesis after the plugin name
1052       will be also be passed to the "new()" constructor.  A reference to the
1053       current Template::Context object is passed as the first parameter.
1054
1055           [% USE MyPlugin('foo', 123) %]
1056
1057       equivalent to:
1058
1059           Template::Plugin::MyPlugin->new($context, 'foo', 123);
1060
1061       The only exception to this is when a module is loaded via the
1062       "LOAD_PERL" option.  In this case the $context reference is not passed
1063       to the "new()" constructor.  This is based on the assumption that the
1064       module is a regular Perl module rather than a Template Toolkit plugin
1065       so isn't expecting a context reference and wouldn't know what to do
1066       with it anyway.
1067
1068       Named parameters may also be specified.  These are collated into a hash
1069       which is passed by reference as the last parameter to the constructor,
1070       as per the general code calling interface.
1071
1072           [% USE url('/cgi-bin/foo', mode='submit', debug=1) %]
1073
1074       equivalent to:
1075
1076           Template::Plugin::URL->new(
1077               $context,
1078               '/cgi-bin/foo'
1079               { mode => 'submit', debug => 1 }
1080           );
1081
1082       The plugin may represent any data type; a simple variable, hash, list
1083       or code reference, but in the general case it will be an object
1084       reference.  Methods can be called on the object (or the relevant
1085       members of the specific data type) in the usual way:
1086
1087           [% USE table(mydata, rows=3) %]
1088
1089           [% FOREACH row IN table.rows %]
1090              <tr>
1091              [% FOREACH item IN row %]
1092               <td>[% item %]</td>
1093              [% END %]
1094              </tr>
1095           [% END %]
1096
1097       An alternative name may be provided for the plugin by which it can be
1098       referenced:
1099
1100           [% USE scores = table(myscores, cols=5) %]
1101
1102           [% FOREACH row IN scores.rows %]
1103              ...
1104           [% END %]
1105
1106       You can use this approach to create multiple plugin objects with
1107       different configurations.  This example shows how the format plugin is
1108       used to create sub-routines bound to variables for formatting text as
1109       per "printf()".
1110
1111           [% USE bold = format('<b>%s</b>') %]
1112           [% USE ital = format('<i>%s</i>') %]
1113           [% bold('This is bold')   %]
1114           [% ital('This is italic') %]
1115
1116       Output:
1117
1118           <b>This is bold</b>
1119           <i>This is italic</i>
1120
1121       This next example shows how the URL plugin can be used to build dynamic
1122       URLs from a base part and optional query parameters.
1123
1124           [% USE mycgi = URL('/cgi-bin/foo.pl', debug=1) %]
1125           <a href="[% mycgi %]">...
1126           <a href="[% mycgi(mode='submit') %]"...
1127
1128       Output:
1129
1130           <a href="/cgi-bin/foo.pl?debug=1">...
1131           <a href="/cgi-bin/foo.pl?mode=submit&debug=1">...
1132
1133       The CGI plugin is an example of one which delegates to another Perl
1134       module. In this case, to Lincoln Stein's "CGI" module.  All of the
1135       methods provided by the "CGI" module are available via the plugin.
1136
1137           [% USE CGI;
1138              CGI.start_form;
1139              CGI.checkbox_group( name   = 'colours',
1140                                  values = [ 'red' 'green' 'blue' ] );
1141              CGI.popup_menu( name   = 'items',
1142                              values = [ 'foo' 'bar' 'baz' ] );
1143              CGI.end_form
1144           %]
1145
1146       See Template::Manual::Plugins for more information on the plugins
1147       distributed with the toolkit or available from CPAN.
1148
1149   MACRO
1150       The "MACRO" directive allows you to define a directive or directive
1151       block which is then evaluated each time the macro is called.
1152
1153           [% MACRO header INCLUDE header %]
1154
1155       Calling the macro as:
1156
1157           [% header %]
1158
1159       is then equivalent to:
1160
1161           [% INCLUDE header %]
1162
1163       Macros can be passed named parameters when called.  These values remain
1164       local to the macro.
1165
1166           [% header(title='Hello World') %]
1167
1168       equivalent to:
1169
1170           [% INCLUDE header title='Hello World' %]
1171
1172       A "MACRO" definition may include parameter names.  Values passed to the
1173       macros are then mapped to these local variables.  Other named
1174       parameters may follow these.
1175
1176           [% MACRO header(title) INCLUDE header %]
1177           [% header('Hello World') %]
1178           [% header('Hello World', bgcol='#123456') %]
1179
1180       equivalent to:
1181
1182           [% INCLUDE header title='Hello World' %]
1183           [% INCLUDE header title='Hello World' bgcol='#123456' %]
1184
1185       Here's another example, defining a macro for display numbers in comma-
1186       delimited groups of 3, using the chunk and join virtual method.
1187
1188           [% MACRO number(n) GET n.chunk(-3).join(',') %]
1189           [% number(1234567) %]    # 1,234,567
1190
1191       A "MACRO" may precede any directive and must conform to the structure
1192       of the directive.
1193
1194           [% MACRO header IF frames %]
1195              [% INCLUDE frames/header %]
1196           [% ELSE %]
1197              [% INCLUDE header %]
1198           [% END %]
1199
1200           [% header %]
1201
1202       A "MACRO" may also be defined as an anonymous "BLOCK".  The block will
1203       be evaluated each time the macro is called.
1204
1205           [% MACRO header BLOCK %]
1206              ...content...
1207           [% END %]
1208
1209           [% header %]
1210
1211       If you've got the "EVAL_PERL" option set, then you can even define a
1212       "MACRO" as a "PERL" block (see below):
1213
1214           [% MACRO triple(n) PERL %]
1215                my $n = $stash->get('n');
1216                print $n * 3;
1217           [% END -%]
1218
1219   PERL
1220       (for the advanced reader)
1221
1222       The "PERL" directive is used to mark the start of a block which
1223       contains Perl code for evaluation.  The "EVAL_PERL" option must be
1224       enabled for Perl code to be evaluated or a "perl" exception will be
1225       thrown with the message '"EVAL_PERL not set"'.
1226
1227       Perl code is evaluated in the "Template::Perl" package.  The $context
1228       package variable contains a reference to the current Template::Context
1229       object.  This can be used to access the functionality of the Template
1230       Toolkit to process other templates, load plugins, filters, etc.  See
1231       Template::Context for further details.
1232
1233           [% PERL %]
1234              print $context->include('myfile');
1235           [% END %]
1236
1237       The $stash variable contains a reference to the top-level stash object
1238       which manages template variables.  Through this, variable values can be
1239       retrieved and updated.  See Template::Stash for further details.
1240
1241           [% PERL %]
1242              $stash->set(foo => 'bar');
1243              print "foo value: ", $stash->get('foo');
1244           [% END %]
1245
1246       Output:
1247
1248           foo value: bar
1249
1250       Output is generated from the "PERL" block by calling "print()".  Note
1251       that the "Template::Perl::PERLOUT" handle is selected (tied to an
1252       output buffer) instead of "STDOUT".
1253
1254           [% PERL %]
1255              print "foo\n";                           # OK
1256              print PERLOUT "bar\n";                   # OK, same as above
1257              print Template::Perl::PERLOUT "baz\n";   # OK, same as above
1258              print STDOUT "qux\n";                    # WRONG!
1259           [% END %]
1260
1261       The "PERL" block may contain other template directives.  These are
1262       processed before the Perl code is evaluated.
1263
1264           [% name = 'Fred Smith' %]
1265
1266           [% PERL %]
1267              print "[% name %]\n";
1268           [% END %]
1269
1270       Thus, the Perl code in the above example is evaluated as:
1271
1272           print "Fred Smith\n";
1273
1274       Exceptions may be thrown from within "PERL" blocks using "die()".  They
1275       will be correctly caught by enclosing "TRY" blocks.
1276
1277           [% TRY %]
1278              [% PERL %]
1279                 die "nothing to live for\n";
1280              [% END %]
1281           [% CATCH %]
1282              error: [% error.info %]
1283           [% END %]
1284
1285       output:
1286              error: nothing to live for
1287
1288   RAWPERL
1289       (for the very advanced reader)
1290
1291       The Template Toolkit parser reads a source template and generates the
1292       text of a Perl subroutine as output.  It then uses "eval()" to evaluate
1293       it into a subroutine reference.  This subroutine is then called to
1294       process the template, passing a reference to the current
1295       Template::Context object through which the functionality of the
1296       Template Toolkit can be accessed.  The subroutine reference can be
1297       cached, allowing the template to be processed repeatedly without
1298       requiring any further parsing.
1299
1300       For example, a template such as:
1301
1302           [% PROCESS header %]
1303           The [% animal %] sat on the [% location %]
1304           [% PROCESS footer %]
1305
1306       is converted into the following Perl subroutine definition:
1307
1308           sub {
1309               my $context = shift;
1310               my $stash   = $context->stash;
1311               my $output  = '';
1312               my $error;
1313
1314               eval { BLOCK: {
1315                   $output .=  $context->process('header');
1316                   $output .=  "The ";
1317                   $output .=  $stash->get('animal');
1318                   $output .=  " sat on the ";
1319                   $output .=  $stash->get('location');
1320                   $output .=  $context->process('footer');
1321                   $output .=  "\n";
1322               } };
1323               if ($@) {
1324                   $error = $context->catch($@, \$output);
1325                   die $error unless $error->type eq 'return';
1326               }
1327
1328               return $output;
1329           }
1330
1331       To examine the Perl code generated, such as in the above example, set
1332       the $Template::Parser::DEBUG package variable to any true value.  You
1333       can also set the $Template::Directive::PRETTY variable true to have the
1334       code formatted in a readable manner for human consumption.  The source
1335       code for each generated template subroutine will be printed to "STDERR"
1336       on compilation (i.e. the first time a template is used).
1337
1338           $Template::Parser::DEBUG = 1;
1339           $Template::Directive::PRETTY = 1;
1340
1341           $template->process($file, $vars)
1342               || die $template->error(), "\n";
1343
1344       The "PERL" ... "END" construct allows Perl code to be embedded into a
1345       template when the "EVAL_PERL" option is set.  It is evaluated at
1346       "runtime" using "eval()" each time the template subroutine is called.
1347       This is inherently flexible, but not as efficient as it could be,
1348       especially in a persistent server environment where a template may be
1349       processed many times.
1350
1351       The "RAWPERL" directive allows you to write Perl code that is
1352       integrated directly into the generated Perl subroutine text.  It is
1353       evaluated once at compile time and is stored in cached form as part of
1354       the compiled template subroutine.  This makes "RAWPERL" blocks more
1355       efficient than "PERL" blocks.
1356
1357       The downside is that you must code much closer to the metal. For
1358       example, in a "PERL" block you can call print() to generate some
1359       output. "RAWPERL" blocks don't afford such luxury. The code is inserted
1360       directly into the generated subroutine text and should conform to the
1361       convention of appending to the $output variable.
1362
1363           [% PROCESS  header %]
1364
1365           [% RAWPERL %]
1366              $output .= "Some output\n";
1367              ...
1368              $output .= "Some more output\n";
1369           [% END %]
1370
1371       The critical section of the generated subroutine for this example would
1372       then look something like:
1373
1374           ...
1375           eval { BLOCK: {
1376               $output .=  $context->process('header');
1377               $output .=  "\n";
1378               $output .= "Some output\n";
1379               ...
1380               $output .= "Some more output\n";
1381               $output .=  "\n";
1382           } };
1383           ...
1384
1385       As with "PERL" blocks, the $context and $stash references are pre-
1386       defined and available for use within "RAWPERL" code.
1387

Exception Handling and Flow Control

1389   TRY / THROW / CATCH / FINAL
1390       (more advanced material)
1391
1392       The Template Toolkit supports fully functional, nested exception
1393       handling.  The "TRY" directive introduces an exception handling scope
1394       which continues until the matching "END" directive.  Any errors that
1395       occur within that block will be caught and can be handled by one of the
1396       "CATCH" blocks defined.
1397
1398           [% TRY %]
1399              ...blah...blah...
1400              [% CALL somecode %]
1401              ...etc...
1402              [% INCLUDE someblock %]
1403              ...and so on...
1404           [% CATCH %]
1405              An error occurred!
1406           [% END %]
1407
1408       Errors are raised as exceptions (objects of the Template::Exception
1409       class) which contain two fields: "type" and "info". The exception
1410       "type" is used to indicate the kind of error that occurred. It is a
1411       simple text string which can contain letters, numbers, '"_"' or '"."'.
1412       The "info" field contains an error message indicating what actually
1413       went wrong. Within a catch block, the exception object is aliased to
1414       the "error" variable. You can access the "type" and "info" fields
1415       directly.
1416
1417           [% mydsn = 'dbi:MySQL:foobar' %]
1418           ...
1419
1420           [% TRY %]
1421              [% USE DBI(mydsn) %]
1422           [% CATCH %]
1423              ERROR! Type: [% error.type %]
1424                     Info: [% error.info %]
1425           [% END %]
1426
1427       output (assuming a non-existent database called '"foobar"'):
1428
1429           ERROR!  Type: DBI
1430                   Info: Unknown database "foobar"
1431
1432       The "error" variable can also be specified by itself and will return a
1433       string of the form ""$type error - $info"".
1434
1435           ...
1436           [% CATCH %]
1437           ERROR: [% error %]
1438           [% END %]
1439
1440       Output:
1441
1442           ERROR: DBI error - Unknown database "foobar"
1443
1444       Each "CATCH" block may be specified with a particular exception type
1445       denoting the kind of error that it should catch.  Multiple "CATCH"
1446       blocks can be provided to handle different types of exception that may
1447       be thrown in the "TRY" block.  A "CATCH" block specified without any
1448       type, as in the previous example, is a default handler which will catch
1449       any otherwise uncaught exceptions.  This can also be specified as "[%
1450       CATCH DEFAULT %]".
1451
1452           [% TRY %]
1453              [% INCLUDE myfile %]
1454              [% USE DBI(mydsn) %]
1455              [% CALL somecode %]
1456           [% CATCH file %]
1457              File Error! [% error.info %]
1458           [% CATCH DBI %]
1459              [% INCLUDE database/error.html %]
1460           [% CATCH %]
1461              [% error %]
1462           [% END %]
1463
1464       Remember that you can specify multiple directives within a single tag,
1465       each delimited by '";"'.  So the above example can be written more
1466       concisely as:
1467
1468           [% TRY;
1469                  INCLUDE myfile;
1470                  USE DBI(mydsn);
1471                  CALL somecode;
1472              CATCH file;
1473                  "File Error! $error.info";
1474              CATCH DBI;
1475                  INCLUDE database/error.html;
1476              CATCH;
1477                  error;
1478              END
1479           %]
1480
1481       The "DBI" plugin throws exceptions of the "DBI" type (in case that
1482       wasn't already obvious).  The other specific exception caught here is
1483       of the "file" type.
1484
1485       A "file" exception is automatically thrown by the Template Toolkit when
1486       it can't find a file, or fails to load, parse or process a file that
1487       has been requested by an "INCLUDE", "PROCESS", "INSERT" or "WRAPPER"
1488       directive.  If "myfile" can't be found in the example above, the "[%
1489       INCLUDE myfile %]" directive will raise a "file" exception which is
1490       then caught by the "[% CATCH file %]" block.  The output generated
1491       would be:
1492
1493           File Error! myfile: not found
1494
1495       Note that the "DEFAULT" option (disabled by default) allows you to
1496       specify a default file to be used any time a template file can't be
1497       found. This will prevent file exceptions from ever being raised when a
1498       non-existent file is requested (unless, of course, the "DEFAULT" file
1499       your specify doesn't exist).  Errors encountered once the file has been
1500       found (i.e. read error, parse error) will be raised as file exceptions
1501       as per usual.
1502
1503       Uncaught exceptions (i.e. if the "TRY" block doesn't have a type
1504       specific or default "CATCH" handler) may be caught by enclosing "TRY"
1505       blocks which can be nested indefinitely across multiple templates. If
1506       the error isn't caught at any level then processing will stop and the
1507       Template process() method will return a false value to the caller. The
1508       relevant Template::Exception object can be retrieved by calling the
1509       error() method.
1510
1511           [% TRY %]
1512              ...
1513              [% TRY %]
1514                 [% INCLUDE $user.header %]
1515              [% CATCH file %]
1516                 [% INCLUDE header %]
1517              [% END %]
1518              ...
1519           [% CATCH DBI %]
1520              [% INCLUDE database/error.html %]
1521           [% END %]
1522
1523       In this example, the inner "TRY" block is used to ensure that the first
1524       "INCLUDE" directive works as expected.  We're using a variable to
1525       provide the name of the template we want to include, "user.header", and
1526       it's possible this contains the name of a non-existent template, or
1527       perhaps one containing invalid template directives.  If the "INCLUDE"
1528       fails with a "file" error then we "CATCH" it in the inner block and
1529       "INCLUDE" the default "header" file instead.  Any "DBI" errors that
1530       occur within the scope of the outer "TRY" block will be caught in the
1531       relevant "CATCH" block, causing the "database/error.html" template to
1532       be processed.  Note that included templates inherit all currently
1533       defined template variable so these error files can quite happily access
1534       the <error> variable to retrieve information about the currently caught
1535       exception.  For example, the "database/error.html" template might look
1536       like this:
1537
1538           <h2>Database Error</h2>
1539           A database error has occurred: [% error.info %]
1540
1541       You can also specify a "FINAL" block.  This is always processed
1542       regardless of the outcome of the "TRY" and/or "CATCH" blocks.  If an
1543       exception is uncaught then the "FINAL" block is processed before
1544       jumping to the enclosing block or returning to the caller.
1545
1546           [% TRY %]
1547              ...
1548           [% CATCH this %]
1549              ...
1550           [% CATCH that %]
1551              ...
1552           [% FINAL %]
1553              All done!
1554           [% END %]
1555
1556       The output from the "TRY" block is left intact up to the point where an
1557       exception occurs.  For example, this template:
1558
1559           [% TRY %]
1560              This gets printed
1561              [% THROW food 'carrots' %]
1562              This doesn't
1563           [% CATCH food %]
1564              culinary delights: [% error.info %]
1565           [% END %]
1566
1567       generates the following output:
1568
1569           This gets printed
1570           culinary delights: carrots
1571
1572       The "CLEAR" directive can be used in a "CATCH" or "FINAL" block to
1573       clear any output created in the "TRY" block.
1574
1575           [% TRY %]
1576              This gets printed
1577              [% THROW food 'carrots' %]
1578              This doesn't
1579           [% CATCH food %]
1580              [% CLEAR %]
1581              culinary delights: [% error.info %]
1582           [% END %]
1583
1584       Output:
1585
1586           culinary delights: carrots
1587
1588       Exception types are hierarchical, with each level being separated by
1589       the familiar dot operator.  A "DBI.connect" exception is a more
1590       specific kind of "DBI" error.  Similarly, an "example.error.barf" is a
1591       more specific kind of "example.error" type which itself is also a
1592       "example" error.
1593
1594       A "CATCH" handler that specifies a general exception type (such as
1595       "DBI" or "example.error") will also catch more specific types that have
1596       the same prefix as long as a more specific handler isn't defined.  Note
1597       that the order in which "CATCH" handlers are defined is irrelevant; a
1598       more specific handler will always catch an exception in preference to a
1599       more generic or default one.
1600
1601           [% TRY %]
1602              ...
1603           [% CATCH DBI ;
1604                INCLUDE database/error.html ;
1605              CATCH DBI.connect ;
1606                INCLUDE database/connect.html ;
1607              CATCH ;
1608                INCLUDE error.html ;
1609              END
1610           %]
1611
1612       In this example, a "DBI.connect" error has it's own handler, a more
1613       general "DBI" block is used for all other "DBI" or "DBI.*" errors and a
1614       default handler catches everything else.
1615
1616       Exceptions can be raised in a template using the "THROW" directive.
1617       The first parameter is the exception type which doesn't need to be
1618       quoted (but can be, it's the same as "INCLUDE") followed by the
1619       relevant error message which can be any regular value such as a quoted
1620       string, variable, etc.
1621
1622           [% THROW food "Missing ingredients: $recipe.error" %]
1623           [% THROW user.login 'no user id: please login' %]
1624           [% THROW $myerror.type "My Error: $myerror.info" %]
1625
1626       It's also possible to specify additional positional or named parameters
1627       to the "THROW" directive if you want to pass more than just a simple
1628       message back as the error info field.
1629
1630           [% THROW food 'eggs' 'flour' msg='Missing Ingredients' %]
1631
1632       In this case, the error "info" field will be a hash array containing
1633       the named arguments and an "args" item which contains a list of the
1634       positional arguments.
1635
1636           type => 'food',
1637           info => {
1638               msg  => 'Missing Ingredients',
1639               args => ['eggs', 'flour'],
1640           }
1641
1642       In addition to specifying individual positional arguments as "[%
1643       error.info.args.n %]", the "info" hash contains keys directly pointing
1644       to the positional arguments, as a convenient shortcut.
1645
1646           [% error.info.0 %]   # same as [% error.info.args.0 %]
1647
1648       Exceptions can also be thrown from Perl code which you've bound to
1649       template variables, or defined as a plugin or other extension.  To
1650       raise an exception, call "die()" passing a reference to a
1651       Template::Exception object as the argument.  This will then be caught
1652       by any enclosing "TRY" blocks from where the code was called.
1653
1654           use Template::Exception;
1655           ...
1656           my $vars = {
1657               foo => sub {
1658                   # ... do something ...
1659                   die Template::Exception->new('myerr.naughty',
1660                                                'Bad, bad error');
1661               },
1662           };
1663
1664       Template:
1665
1666           [% TRY %]
1667              [% foo %]
1668           [% CATCH myerr ;
1669                "Error: $error" ;
1670              END
1671           %]
1672
1673       Output:
1674
1675           Error: myerr.naughty error - Bad, bad error
1676
1677       The "info" field can also be a reference to another object or data
1678       structure, if required.
1679
1680           die Template::Exception->new('myerror', {
1681               module => 'foo.pl',
1682               errors => [ 'bad permissions', 'naughty boy' ],
1683           });
1684
1685       Later, in a template:
1686
1687           [% TRY %]
1688              ...
1689           [% CATCH myerror %]
1690              [% error.info.errors.size or 'no';
1691                 error.info.errors.size == 1 ? ' error' : ' errors' %]
1692              in [% error.info.module %]:
1693                 [% error.info.errors.join(', ') %].
1694           [% END %]
1695
1696       Generating the output:
1697
1698              2 errors in foo.pl:
1699                 bad permissions, naughty boy.
1700
1701       You can also call "die()" with a single string, as is common in much
1702       existing Perl code.  This will automatically be converted to an
1703       exception of the '"undef"' type (that's the literal string '"undef"',
1704       not the undefined value).  If the string isn't terminated with a
1705       newline then Perl will append the familiar " at $file line $line"
1706       message.
1707
1708           sub foo {
1709               # ... do something ...
1710               die "I'm sorry, Dave, I can't do that\n";
1711           }
1712
1713       If you're writing a plugin, or some extension code that has the current
1714       Template::Context in scope (you can safely skip this section if this
1715       means nothing to you) then you can also raise an exception by calling
1716       the context throw() method. You can pass it an Template::Exception
1717       object reference, a pair of "($type, $info)" parameters or just an
1718       $info string to create an exception of '"undef"' type.
1719
1720           $context->throw($e);            # exception object
1721           $context->throw('Denied');      # 'undef' type
1722           $context->throw('user.passwd', 'Bad Password');
1723
1724   NEXT
1725       The "NEXT" directive can be used to start the next iteration of a
1726       "FOREACH" or "WHILE" loop.
1727
1728           [% FOREACH user IN users %]
1729              [% NEXT IF user.isguest %]
1730              Name: [% user.name %]    Email: [% user.email %]
1731           [% END %]
1732
1733   LAST
1734       The "LAST" directive can be used to prematurely exit a "FOREACH" or
1735       "WHILE" loop.
1736
1737           [% FOREACH user IN users %]
1738              Name: [% user.name %]    Email: [% user.email %]
1739              [% LAST IF some.condition %]
1740           [% END %]
1741
1742       "BREAK" can also be used as an alias for "LAST".
1743
1744   RETURN
1745       The "RETURN" directive can be used to stop processing the current
1746       template and return to the template from which it was called, resuming
1747       processing at the point immediately after the "INCLUDE", "PROCESS" or
1748       "WRAPPER" directive. If there is no enclosing template then the
1749       Template process() method will return to the calling code with a true
1750       value.
1751
1752           Before
1753           [% INCLUDE half_wit %]
1754           After
1755
1756           [% BLOCK half_wit %]
1757           This is just half...
1758           [% RETURN %]
1759           ...a complete block
1760           [% END %]
1761
1762       Output:
1763
1764           Before
1765           This is just half...
1766           After
1767
1768   STOP
1769       The "STOP" directive can be used to indicate that the processor should
1770       stop gracefully without processing any more of the template document.
1771       This is a planned stop and the Template process() method will return a
1772       true value to the caller. This indicates that the template was
1773       processed successfully according to the directives within it.
1774
1775           [% IF something.terrible.happened %]
1776              [% INCLUDE fatal/error.html %]
1777              [% STOP %]
1778           [% END %]
1779
1780           [% TRY %]
1781              [% USE DBI(mydsn) %]
1782              ...
1783           [% CATCH DBI.connect %]
1784              <h1>Cannot connect to the database: [% error.info %]</h1>
1785              <p>
1786                We apologise for the inconvenience.
1787              </p>
1788              [% INCLUDE footer %]
1789              [% STOP %]
1790           [% END %]
1791
1792   CLEAR
1793       The "CLEAR" directive can be used to clear the output buffer for the
1794       current enclosing block.   It is most commonly used to clear the output
1795       generated from a "TRY" block up to the point where the error occurred.
1796
1797           [% TRY %]
1798              blah blah blah            # this is normally left intact
1799              [% THROW some 'error' %]  # up to the point of error
1800              ...
1801           [% CATCH %]
1802              [% CLEAR %]               # clear the TRY output
1803              [% error %]               # print error string
1804           [% END %]
1805

Miscellaneous

1807   META
1808       The "META" directive allows simple metadata items to be defined within
1809       a template. These are evaluated when the template is parsed and as such
1810       may only contain simple values (e.g. it's not possible to interpolate
1811       other variables values into "META" variables).
1812
1813           [% META
1814              title   = 'The Cat in the Hat'
1815              author  = 'Dr. Seuss'
1816              version = 1.23
1817           %]
1818
1819       The "template" variable contains a reference to the main template being
1820       processed.  These metadata items may be retrieved as attributes of the
1821       template.
1822
1823           <h1>[% template.title %]</h1>
1824           <h2>[% template.author %]</h2>
1825
1826       The "name" and "modtime" metadata items are automatically defined for
1827       each template to contain its name and modification time in seconds
1828       since the epoch.
1829
1830           [% USE date %]              # use Date plugin to format time
1831           ...
1832           [% template.name %] last modified
1833           at [% date.format(template.modtime) %]
1834
1835       The "PRE_PROCESS" and "POST_PROCESS" options allow common headers and
1836       footers to be added to all templates.  The "template" reference is
1837       correctly defined when these templates are processed, allowing headers
1838       and footers to reference metadata items from the main template.
1839
1840           $template = Template->new({
1841               PRE_PROCESS  => 'header',
1842               POST_PROCESS => 'footer',
1843           });
1844
1845           $template->process('cat_in_hat');
1846
1847       header:
1848
1849           <html>
1850             <head>
1851               <title>[% template.title %]</title>
1852             </head>
1853             <body>
1854
1855       cat_in_hat:
1856
1857           [% META
1858                title   = 'The Cat in the Hat'
1859                author  = 'Dr. Seuss'
1860                version = 1.23
1861                year    = 2000
1862           %]
1863
1864               The cat in the hat sat on the mat.
1865
1866       footer:
1867
1868               <hr>
1869               &copy; [% template.year %] [% template.author %]
1870             </body>
1871           </html>
1872
1873       The output generated from the above example is:
1874
1875           <html>
1876             <head>
1877               <title>The Cat in the Hat</title>
1878             </head>
1879             <body>
1880               The cat in the hat sat on the mat.
1881               <hr>
1882               &copy; 2000 Dr. Seuss
1883             </body>
1884           </html>
1885
1886   TAGS
1887       The "TAGS" directive can be used to set the "START_TAG" and "END_TAG"
1888       values on a per-template file basis.
1889
1890           [% TAGS <+ +> %]
1891
1892           <+ INCLUDE header +>
1893
1894       The TAGS directive may also be used to set a named "TAG_STYLE"
1895
1896           [% TAGS html %]
1897           <!-- INCLUDE header -->
1898
1899       See the TAGS and TAG_STYLE configuration options for further details.
1900
1901   DEBUG
1902       The "DEBUG" directive can be used to enable or disable directive debug
1903       messages within a template.  The "DEBUG" configuration option must be
1904       set to include "DEBUG_DIRS" for the "DEBUG" directives to have any
1905       effect.  If "DEBUG_DIRS" is not set then the parser will automatically
1906       ignore and remove any "DEBUG" directives.
1907
1908       The "DEBUG" directive can be used with an "on" or "off" parameter to
1909       enable or disable directive debugging messages from that point forward.
1910       When enabled, the output of each directive in the generated output will
1911       be prefixed by a comment indicate the file, line and original directive
1912       text.
1913
1914           [% DEBUG on %]
1915           directive debugging is on (assuming DEBUG option is set true)
1916           [% DEBUG off %]
1917           directive debugging is off
1918
1919       The "format" parameter can be used to change the format of the
1920       debugging message.
1921
1922           [% DEBUG format '<!-- $file line $line : [% $text %] -->' %]
1923
1924
1925
1926perl v5.34.0                      2022-01-21   Template::Manual::Directives(3)
Impressum