1Git(3) User Contributed Perl Documentation Git(3)
2
6 Git - Perl interface to the Git version control system
7
9 use Git;
10 my $version = Git::command_oneline('version');
12 git_cmd_try { Git::command_noisy('update-server-info') }
14 '%s failed w/ code %d';
15 my $repo = Git->repository (Directory => '/srv/git/cogito.git');
17 my @revs = $repo->command('rev-list', '--since=last monday', '--all');
20 my ($fh, $c) = $repo->command_output_pipe('rev-list', '--since=last monday', '--all');
22 my $lastrev = <$fh>; chomp $lastrev;
23 $repo->command_close_pipe($fh, $c);
24 my $lastrev = $repo->command_oneline( [ 'rev-list', '--all' ],
26 STDERR => 0 );
27 my $sha1 = $repo->hash_and_insert_object('file.txt');
29 my $tempfile = tempfile();
30 my $size = $repo->cat_blob($sha1, $tempfile);
31
33 This module provides Perl scripts easy way to interface the Git version
34 control system. The modules have an easy and well-tested way to call
35 arbitrary Git commands; in the future, the interface will also provide
36 specialized methods for doing easily operations which are not totally
37 trivial to do over the generic command interface.
38 While some commands can be executed outside of any context (e.g.
40 'version' or 'init'), most operations require a repository context,
41 which in practice means getting an instance of the Git object using the
42 repository() constructor. (In the future, we will also get a
43 new_repository() constructor.) All commands called as methods of the
44 object are then executed in the context of the repository.
45 Part of the "repository state" is also information about path to the
47 attached working copy (unless you work with a bare repository). You can
48 also navigate inside of the working copy using the "wc_chdir()" method.
49 (Note that the repository object is self-contained and will not change
50 working directory of your process.)
51 TODO: In the future, we might also do
53 my $remoterepo = $repo->remote_repository (Name => 'cogito', Branch => 'master');
55 $remoterepo ||= Git->remote_repository ('http://git.or.cz/cogito.git/');
56 my @refs = $remoterepo->refs();
57 Currently, the module merely wraps calls to external Git tools. In the
59 future, it will provide a much faster way to interact with Git by
60 linking directly to libgit. This should be completely opaque to the
61 user, though (performance increase notwithstanding).
62
64 repository ( OPTIONS )
65 repository ( DIRECTORY )
66 repository ()
67 Construct a new repository object. "OPTIONS" are passed in a hash
68 like fashion, using key and value pairs. Possible options are:
69 Repository - Path to the Git repository.
71 WorkingCopy - Path to the associated working copy; not strictly
73 required as many commands will happily crunch on a bare repository.
74 WorkingSubdir - Subdirectory in the working copy to work inside.
76 Just left undefined if you do not want to limit the scope of
77 operations.
78 Directory - Path to the Git working directory in its usual setup.
80 The ".git" directory is searched in the directory and all the
81 parent directories; if found, "WorkingCopy" is set to the directory
82 containing it and "Repository" to the ".git" directory itself. If
83 no ".git" directory was found, the "Directory" is assumed to be a
84 bare repository, "Repository" is set to point at it and
85 "WorkingCopy" is left undefined. If the $GIT_DIR environment
86 variable is set, things behave as expected as well.
87 You should not use both "Directory" and either of "Repository" and
89 "WorkingCopy" - the results of that are undefined.
90 Alternatively, a directory path may be passed as a single scalar
92 argument to the constructor; it is equivalent to setting only the
93 "Directory" option field.
94 Calling the constructor with no options whatsoever is equivalent to
96 calling it with "Directory => '.'". In general, if you are building
97 a standard porcelain command, simply doing "Git->repository()"
98 should do the right thing and setup the object to reflect exactly
99 where the user is right now.
100
102 command ( COMMAND [, ARGUMENTS... ] )
103 command ( [ COMMAND, ARGUMENTS... ], { Opt => Val ... } )
104 Execute the given Git "COMMAND" (specify it without the 'git-'
105 prefix), optionally with the specified extra "ARGUMENTS".
106 The second more elaborate form can be used if you want to further
108 adjust the command execution. Currently, only one option is
109 supported:
110 STDERR - How to deal with the command's error output. By default
112 ("undef") it is delivered to the caller's "STDERR". A false value
113 (0 or '') will cause it to be thrown away. If you want to process
114 it, you can get it in a filehandle you specify, but you must be
115 extremely careful; if the error output is not very short and you
116 want to read it in the same process as where you called
117 "command()", you are set up for a nice deadlock!
118 The method can be called without any instance or on a specified Git
120 repository (in that case the command will be run in the repository
121 context).
122 In scalar context, it returns all the command output in a single
124 string (verbatim).
125 In array context, it returns an array containing lines printed to
127 the command's stdout (without trailing newlines).
128 In both cases, the command's stdin and stderr are the same as the
130 caller's.
131 command_oneline ( COMMAND [, ARGUMENTS... ] )
133 command_oneline ( [ COMMAND, ARGUMENTS... ], { Opt => Val ... } )
134 Execute the given "COMMAND" in the same way as command() does but
135 always return a scalar string containing the first line of the
136 command's standard output.
137 command_output_pipe ( COMMAND [, ARGUMENTS... ] )
139 command_output_pipe ( [ COMMAND, ARGUMENTS... ], { Opt => Val ... } )
140 Execute the given "COMMAND" in the same way as command() does but
141 return a pipe filehandle from which the command output can be read.
142 The function can return "($pipe, $ctx)" in array context. See
144 "command_close_pipe()" for details.
145 command_input_pipe ( COMMAND [, ARGUMENTS... ] )
147 command_input_pipe ( [ COMMAND, ARGUMENTS... ], { Opt => Val ... } )
148 Execute the given "COMMAND" in the same way as
149 command_output_pipe() does but return an input pipe filehandle
150 instead; the command output is not captured.
151 The function can return "($pipe, $ctx)" in array context. See
153 "command_close_pipe()" for details.
154 command_close_pipe ( PIPE [, CTX ] )
156 Close the "PIPE" as returned from "command_*_pipe()", checking
157 whether the command finished successfully. The optional "CTX"
158 argument is required if you want to see the command name in the
159 error message, and it is the second value returned by
160 "command_*_pipe()" when called in array context. The call idiom is:
161 my ($fh, $ctx) = $r->command_output_pipe('status');
163 while (<$fh>) { ... }
164 $r->command_close_pipe($fh, $ctx);
165 Note that you should not rely on whatever actually is in "CTX";
167 currently it is simply the command name but in future the context
168 might have more complicated structure.
169 command_bidi_pipe ( COMMAND [, ARGUMENTS... ] )
171 Execute the given "COMMAND" in the same way as
172 command_output_pipe() does but return both an input pipe filehandle
173 and an output pipe filehandle.
174 The function will return "($pid, $pipe_in, $pipe_out, $ctx)". See
176 "command_close_bidi_pipe()" for details.
177 command_close_bidi_pipe ( PID, PIPE_IN, PIPE_OUT [, CTX] )
179 Close the "PIPE_IN" and "PIPE_OUT" as returned from
180 "command_bidi_pipe()", checking whether the command finished
181 successfully. The optional "CTX" argument is required if you want
182 to see the command name in the error message, and it is the fourth
183 value returned by "command_bidi_pipe()". The call idiom is:
184 my ($pid, $in, $out, $ctx) = $r->command_bidi_pipe('cat-file --batch-check');
186 print $out "000000000\n";
187 while (<$in>) { ... }
188 $r->command_close_bidi_pipe($pid, $in, $out, $ctx);
189 Note that you should not rely on whatever actually is in "CTX";
191 currently it is simply the command name but in future the context
192 might have more complicated structure.
193 "PIPE_IN" and "PIPE_OUT" may be "undef" if they have been closed
195 prior to calling this function. This may be useful in a query-
196 response type of commands where caller first writes a query and
197 later reads response, eg:
198 my ($pid, $in, $out, $ctx) = $r->command_bidi_pipe('cat-file --batch-check');
200 print $out "000000000\n";
201 close $out;
202 while (<$in>) { ... }
203 $r->command_close_bidi_pipe($pid, $in, undef, $ctx);
204 This idiom may prevent potential dead locks caused by data sent to
206 the output pipe not being flushed and thus not reaching the
207 executed command.
208 command_noisy ( COMMAND [, ARGUMENTS... ] )
210 Execute the given "COMMAND" in the same way as command() does but
211 do not capture the command output - the standard output is not
212 redirected and goes to the standard output of the caller
213 application.
214 While the method is called command_noisy(), you might want to as
216 well use it for the most silent Git commands which you know will
217 never pollute your stdout but you want to avoid the overhead of the
218 pipe setup when calling them.
219 The function returns only after the command has finished running.
221 version ()
223 Return the Git version in use.
224 exec_path ()
226 Return path to the Git sub-command executables (the same as "git
227 --exec-path"). Useful mostly only internally.
228 html_path ()
230 Return path to the Git html documentation (the same as "git
231 --html-path"). Useful mostly only internally.
232 get_tz_offset ( TIME )
234 Return the time zone offset from GMT in the form +/-HHMM where HH
235 is the number of hours from GMT and MM is the number of minutes.
236 This is the equivalent of what strftime("%z", ...) would provide on
237 a GNU platform.
238 If TIME is not supplied, the current local time is used.
240 get_record ( FILEHANDLE, INPUT_RECORD_SEPARATOR )
242 Read one record from FILEHANDLE delimited by
243 INPUT_RECORD_SEPARATOR, removing any trailing
244 INPUT_RECORD_SEPARATOR.
245 prompt ( PROMPT , ISPASSWORD )
247 Query user "PROMPT" and return answer from user.
248 Honours GIT_ASKPASS and SSH_ASKPASS environment variables for
250 querying the user. If no *_ASKPASS variable is set or an error
251 occurred, the terminal is tried as a fallback. If "ISPASSWORD" is
252 set and true, the terminal disables echo.
253 repo_path ()
255 Return path to the git repository. Must be called on a repository
256 instance.
257 wc_path ()
259 Return path to the working copy. Must be called on a repository
260 instance.
261 wc_subdir ()
263 Return path to the subdirectory inside of a working copy. Must be
264 called on a repository instance.
265 wc_chdir ( SUBDIR )
267 Change the working copy subdirectory to work within. The "SUBDIR"
268 is relative to the working copy root directory (not the current
269 subdirectory). Must be called on a repository instance attached to
270 a working copy and the directory must exist.
271 config ( VARIABLE )
273 Retrieve the configuration "VARIABLE" in the same manner as
274 "config" does. In scalar context requires the variable to be set
275 only one time (exception is thrown otherwise), in array context
276 returns allows the variable to be set multiple times and returns
277 all the values.
278 config_bool ( VARIABLE )
280 Retrieve the bool configuration "VARIABLE". The return value is
281 usable as a boolean in perl (and "undef" if it's not defined, of
282 course).
283 config_path ( VARIABLE )
285 Retrieve the path configuration "VARIABLE". The return value is an
286 expanded path or "undef" if it's not defined.
287 config_int ( VARIABLE )
289 Retrieve the integer configuration "VARIABLE". The return value is
290 simple decimal number. An optional value suffix of 'k', 'm', or
291 'g' in the config file will cause the value to be multiplied by
292 1024, 1048576 (1024^2), or 1073741824 (1024^3) prior to output. It
293 would return "undef" if configuration variable is not defined.
294 get_colorbool ( NAME )
296 Finds if color should be used for NAMEd operation from the
297 configuration, and returns boolean (true for "use color", false for
298 "do not use color").
299 get_color ( SLOT, COLOR )
301 Finds color for SLOT from the configuration, while defaulting to
302 COLOR, and returns the ANSI color escape sequence:
303 print $repo->get_color("color.interactive.prompt", "underline blue white");
305 print "some text";
306 print $repo->get_color("", "normal");
307 remote_refs ( REPOSITORY [, GROUPS [, REFGLOBS ] ] )
309 This function returns a hashref of refs stored in a given remote
310 repository. The hash is in the format "refname =\" hash>. For
311 tags, the "refname" entry contains the tag object while a
312 "refname^{}" entry gives the tagged objects.
313 "REPOSITORY" has the same meaning as the appropriate
315 "git-ls-remote" argument; either a URL or a remote name (if called
316 on a repository instance). "GROUPS" is an optional arrayref that
317 can contain 'tags' to return all the tags and/or 'heads' to return
318 all the heads. "REFGLOB" is an optional array of strings containing
319 a shell-like glob to further limit the refs returned in the hash;
320 the meaning is again the same as the appropriate "git-ls-remote"
321 argument.
322 This function may or may not be called on a repository instance. In
324 the former case, remote names as defined in the repository are
325 recognized as repository specifiers.
326 ident ( TYPE | IDENTSTR )
328 ident_person ( TYPE | IDENTSTR | IDENTARRAY )
329 This suite of functions retrieves and parses ident information, as
330 stored in the commit and tag objects or produced by "var
331 GIT_type_IDENT" (thus "TYPE" can be either author or committer;
332 case is insignificant).
333 The "ident" method retrieves the ident information from "git var"
335 and either returns it as a scalar string or as an array with the
336 fields parsed. Alternatively, it can take a prepared ident string
337 (e.g. from the commit object) and just parse it.
338 "ident_person" returns the person part of the ident - name and
340 email; it can take the same arguments as "ident" or the array
341 returned by "ident".
342 The synopsis is like:
344 my ($name, $email, $time_tz) = ident('author');
346 "$name <$email>" eq ident_person('author');
347 "$name <$email>" eq ident_person($name);
348 $time_tz =~ /^\d+ [+-]\d{4}$/;
349 hash_object ( TYPE, FILENAME )
351 Compute the SHA1 object id of the given "FILENAME" considering it
352 is of the "TYPE" object type ("blob", "commit", "tree").
353 The method can be called without any instance or on a specified Git
355 repository, it makes zero difference.
356 The function returns the SHA1 hash.
358 hash_and_insert_object ( FILENAME )
360 Compute the SHA1 object id of the given "FILENAME" and add the
361 object to the object database.
362 The function returns the SHA1 hash.
364 cat_blob ( SHA1, FILEHANDLE )
366 Prints the contents of the blob identified by "SHA1" to
367 "FILEHANDLE" and returns the number of bytes printed.
368 credential_read( FILEHANDLE )
370 Reads credential key-value pairs from "FILEHANDLE". Reading stops
371 at EOF or when an empty line is encountered. Each line must be of
372 the form "key=value" with a non-empty key. Function returns hash
373 with all read values. Any white space (other than new-line
374 character) is preserved.
375 credential_write( FILEHANDLE, CREDENTIAL_HASHREF )
377 Writes credential key-value pairs from hash referenced by
378 "CREDENTIAL_HASHREF" to "FILEHANDLE". Keys and values cannot
379 contain new-lines or NUL bytes characters, and key cannot contain
380 equal signs nor be empty (if they do Error::Simple is thrown). Any
381 white space is preserved. If value for a key is "undef", it will
382 be skipped.
383 If 'url' key exists it will be written first. (All the other key-
385 value pairs are written in sorted order but you should not depend
386 on that). Once all lines are written, an empty line is printed.
387 credential( CREDENTIAL_HASHREF [, OPERATION ] )
389 credential( CREDENTIAL_HASHREF, CODE )
390 Executes "git credential" for a given set of credentials and
391 specified operation. In both forms "CREDENTIAL_HASHREF" needs to
392 be a reference to a hash which stores credentials. Under certain
393 conditions the hash can change.
394 In the first form, "OPERATION" can be 'fill', 'approve' or
396 'reject', and function will execute corresponding "git credential"
397 sub-command. If it's omitted 'fill' is assumed. In case of 'fill'
398 the values stored in "CREDENTIAL_HASHREF" will be changed to the
399 ones returned by the "git credential fill" command. The usual
400 usage would look something like:
401 my %cred = (
403 'protocol' => 'https',
404 'host' => 'example.com',
405 'username' => 'bob'
406 );
407 Git::credential \%cred;
408 if (try_to_authenticate($cred{'username'}, $cred{'password'})) {
409 Git::credential \%cred, 'approve';
410 ... do more stuff ...
411 } else {
412 Git::credential \%cred, 'reject';
413 }
414 In the second form, "CODE" needs to be a reference to a subroutine.
416 The function will execute "git credential fill" to fill the
417 provided credential hash, then call "CODE" with
418 "CREDENTIAL_HASHREF" as the sole argument. If "CODE"'s return
419 value is defined, the function will execute "git credential
420 approve" (if return value yields true) or "git credential reject"
421 (if return value is false). If the return value is undef, nothing
422 at all is executed; this is useful, for example, if the credential
423 could neither be verified nor rejected due to an unrelated network
424 error. The return value is the same as what "CODE" returns. With
425 this form, the usage might look as follows:
426 if (Git::credential {
428 'protocol' => 'https',
429 'host' => 'example.com',
430 'username' => 'bob'
431 }, sub {
432 my $cred = shift;
433 return !!try_to_authenticate($cred->{'username'},
434 $cred->{'password'});
435 }) {
436 ... do more stuff ...
437 }
438 temp_acquire ( NAME )
440 Attempts to retrieve the temporary file mapped to the string
441 "NAME". If an associated temp file has not been created this
442 session or was closed, it is created, cached, and set for autoflush
443 and binmode.
444 Internally locks the file mapped to "NAME". This lock must be
446 released with "temp_release()" when the temp file is no longer
447 needed. Subsequent attempts to retrieve temporary files mapped to
448 the same "NAME" while still locked will cause an error. This
449 locking mechanism provides a weak guarantee and is not threadsafe.
450 It does provide some error checking to help prevent temp file refs
451 writing over one another.
452 In general, the File::Handle returned should not be closed by
454 consumers as it defeats the purpose of this caching mechanism. If
455 you need to close the temp file handle, then you should use
456 File::Temp or another temp file faculty directly. If a handle is
457 closed and then requested again, then a warning will issue.
458 temp_is_locked ( NAME )
460 Returns true if the internal lock created by a previous
461 "temp_acquire()" call with "NAME" is still in effect.
462 When temp_acquire is called on a "NAME", it internally locks the
464 temporary file mapped to "NAME". That lock will not be released
465 until "temp_release()" is called with either the original "NAME" or
466 the File::Handle that was returned from the original call to
467 temp_acquire.
468 Subsequent attempts to call "temp_acquire()" with the same "NAME"
470 will fail unless there has been an intervening "temp_release()"
471 call for that "NAME" (or its corresponding File::Handle that was
472 returned by the original "temp_acquire()" call).
473 If true is returned by "temp_is_locked()" for a "NAME", an attempt
475 to "temp_acquire()" the same "NAME" will cause an error unless
476 "temp_release" is first called on that "NAME" (or its corresponding
477 File::Handle that was returned by the original "temp_acquire()"
478 call).
479 temp_release ( NAME )
481 temp_release ( FILEHANDLE )
482 Releases a lock acquired through "temp_acquire()". Can be called
483 either with the "NAME" mapping used when acquiring the temp file or
484 with the "FILEHANDLE" referencing a locked temp file.
485 Warns if an attempt is made to release a file that is not locked.
487 The temp file will be truncated before being released. This can
489 help to reduce disk I/O where the system is smart enough to detect
490 the truncation while data is in the output buffers. Beware that
491 after the temp file is released and truncated, any operations on
492 that file may fail miserably until it is re-acquired. All contents
493 are lost between each release and acquire mapped to the same
494 string.
495 temp_reset ( FILEHANDLE )
497 Truncates and resets the position of the "FILEHANDLE".
498 temp_path ( NAME )
500 temp_path ( FILEHANDLE )
501 Returns the filename associated with the given tempfile.
502 prefix_lines ( PREFIX, STRING [, STRING... ])
504 Prefixes lines in "STRING" with "PREFIX".
505 unquote_path ( PATH )
507 Unquote a quoted path containing c-escapes as returned by ls-files
508 etc. when not using -z or when parsing the output of diff -u.
509 get_comment_line_char ( )
511 Gets the core.commentchar configuration value. The value falls-
512 back to '#' if core.commentchar is set to 'auto'.
513 comment_lines ( STRING [, STRING... ])
515 Comments lines following core.commentchar configuration.
516
518 All functions are supposed to throw Perl exceptions in case of errors.
519 See the Error module on how to catch those. Most exceptions are mere
520 Error::Simple instances.
521 However, the "command()", "command_oneline()" and "command_noisy()"
523 functions suite can throw "Git::Error::Command" exceptions as well:
524 those are thrown when the external command returns an error code and
525 contain the error code as well as access to the captured command's
526 output. The exception class provides the usual "stringify" and "value"
527 (command's exit code) methods and in addition also a "cmd_output"
528 method that returns either an array or a string with the captured
529 command output (depending on the original function call context;
530 "command_noisy()" returns "undef") and $<cmdline> which returns the
531 command and its arguments (but without proper quoting).
532 Note that the "command_*_pipe()" functions cannot throw this exception
534 since it has no idea whether the command failed or not. You will only
535 find out at the time you "close" the pipe; if you want to have that
536 automated, use "command_close_pipe()", which can throw the exception.
537 git_cmd_try { CODE } ERRMSG
539 This magical statement will automatically catch any
540 "Git::Error::Command" exceptions thrown by "CODE" and make your
541 program die with "ERRMSG" on its lips; the message will have %s
542 substituted for the command line and %d for the exit status. This
543 statement is useful mostly for producing more user-friendly error
544 messages.
545 In case of no exception caught the statement returns "CODE"'s
547 return value.
548 Note that this is the only auto-exported function.
550
552 Copyright 2006 by Petr Baudis <pasky@suse.cz>.
553 This module is free software; it may be used, copied, modified and
555 distributed under the terms of the GNU General Public Licence, either
556 version 2, or (at your option) any later version.
557perl v5.30.2 2020-04-20 Git(3)