1GIT-DIFF-FILES(1) Git Manual GIT-DIFF-FILES(1)
2
6 git-diff-files - Compares files in the working tree and the index
7
9 git diff-files [-q] [-0|-1|-2|-3|-c|--cc] [<common diff options>] [<path>...]
10
13 Compares the files in the working tree and the index. When paths are
14 specified, compares only those named paths. Otherwise all entries in
15 the index are compared. The output format is the same as for git
16 diff-index and git diff-tree.
17
19 -p, -u, --patch
20 Generate patch (see section on generating patches).
21 -U<n>, --unified=<n>
23 Generate diffs with <n> lines of context instead of the usual
24 three. Implies -p.
25 --raw
27 Generate the raw format. This is the default.
28 --patch-with-raw
30 Synonym for -p --raw.
31 --minimal
33 Spend extra time to make sure the smallest possible diff is
34 produced.
35 --patience
37 Generate a diff using the "patience diff" algorithm.
38 --histogram
40 Generate a diff using the "histogram diff" algorithm.
41 --diff-algorithm={patience|minimal|histogram|myers}
43 Choose a diff algorithm. The variants are as follows:
44 default, myers
46 The basic greedy diff algorithm. Currently, this is the
47 default.
48 minimal
50 Spend extra time to make sure the smallest possible diff is
51 produced.
52 patience
54 Use "patience diff" algorithm when generating patches.
55 histogram
57 This algorithm extends the patience algorithm to "support
58 low-occurrence common elements".
59 For instance, if you configured diff.algorithm variable to a
61 non-default value and want to use the default one, then you have to
62 use --diff-algorithm=default option.
63 --stat[=<width>[,<name-width>[,<count>]]]
65 Generate a diffstat. By default, as much space as necessary will be
66 used for the filename part, and the rest for the graph part.
67 Maximum width defaults to terminal width, or 80 columns if not
68 connected to a terminal, and can be overridden by <width>. The
69 width of the filename part can be limited by giving another width
70 <name-width> after a comma. The width of the graph part can be
71 limited by using --stat-graph-width=<width> (affects all commands
72 generating a stat graph) or by setting diff.statGraphWidth=<width>
73 (does not affect git format-patch). By giving a third parameter
74 <count>, you can limit the output to the first <count> lines,
75 followed by ... if there are more.
76 These parameters can also be set individually with
78 --stat-width=<width>, --stat-name-width=<name-width> and
79 --stat-count=<count>.
80 --numstat
82 Similar to --stat, but shows number of added and deleted lines in
83 decimal notation and pathname without abbreviation, to make it more
84 machine friendly. For binary files, outputs two - instead of saying
85 0 0.
86 --shortstat
88 Output only the last line of the --stat format containing total
89 number of modified files, as well as number of added and deleted
90 lines.
91 --dirstat[=<param1,param2,...>]
93 Output the distribution of relative amount of changes for each
94 sub-directory. The behavior of --dirstat can be customized by
95 passing it a comma separated list of parameters. The defaults are
96 controlled by the diff.dirstat configuration variable (see git-
97 config(1)). The following parameters are available:
98 changes
100 Compute the dirstat numbers by counting the lines that have
101 been removed from the source, or added to the destination. This
102 ignores the amount of pure code movements within a file. In
103 other words, rearranging lines in a file is not counted as much
104 as other changes. This is the default behavior when no
105 parameter is given.
106 lines
108 Compute the dirstat numbers by doing the regular line-based
109 diff analysis, and summing the removed/added line counts. (For
110 binary files, count 64-byte chunks instead, since binary files
111 have no natural concept of lines). This is a more expensive
112 --dirstat behavior than the changes behavior, but it does count
113 rearranged lines within a file as much as other changes. The
114 resulting output is consistent with what you get from the other
115 --*stat options.
116 files
118 Compute the dirstat numbers by counting the number of files
119 changed. Each changed file counts equally in the dirstat
120 analysis. This is the computationally cheapest --dirstat
121 behavior, since it does not have to look at the file contents
122 at all.
123 cumulative
125 Count changes in a child directory for the parent directory as
126 well. Note that when using cumulative, the sum of the
127 percentages reported may exceed 100%. The default
128 (non-cumulative) behavior can be specified with the
129 noncumulative parameter.
130 <limit>
132 An integer parameter specifies a cut-off percent (3% by
133 default). Directories contributing less than this percentage of
134 the changes are not shown in the output.
135 Example: The following will count changed files, while ignoring
137 directories with less than 10% of the total amount of changed
138 files, and accumulating child directory counts in the parent
139 directories: --dirstat=files,10,cumulative.
140 --summary
142 Output a condensed summary of extended header information such as
143 creations, renames and mode changes.
144 --patch-with-stat
146 Synonym for -p --stat.
147 -z
149 When --raw, --numstat, --name-only or --name-status has been given,
150 do not munge pathnames and use NULs as output field terminators.
151 Without this option, each pathname output will have TAB, LF, double
153 quotes, and backslash characters replaced with \t, \n, \", and \\,
154 respectively, and the pathname will be enclosed in double quotes if
155 any of those replacements occurred.
156 --name-only
158 Show only names of changed files.
159 --name-status
161 Show only names and status of changed files. See the description of
162 the --diff-filter option on what the status letters mean.
163 --submodule[=<format>]
165 Specify how differences in submodules are shown. When --submodule
166 or --submodule=log is given, the log format is used. This format
167 lists the commits in the range like git-submodule(1)summary does.
168 Omitting the --submodule option or specifying --submodule=short,
169 uses the short format. This format just shows the names of the
170 commits at the beginning and end of the range. Can be tweaked via
171 the diff.submodule configuration variable.
172 --color[=<when>]
174 Show colored diff. --color (i.e. without =<when>) is the same as
175 --color=always. <when> can be one of always, never, or auto.
176 --no-color
178 Turn off colored diff. It is the same as --color=never.
179 --word-diff[=<mode>]
181 Show a word diff, using the <mode> to delimit changed words. By
182 default, words are delimited by whitespace; see --word-diff-regex
183 below. The <mode> defaults to plain, and must be one of:
184 color
186 Highlight changed words using only colors. Implies --color.
187 plain
189 Show words as [-removed-] and {+added+}. Makes no attempts to
190 escape the delimiters if they appear in the input, so the
191 output may be ambiguous.
192 porcelain
194 Use a special line-based format intended for script
195 consumption. Added/removed/unchanged runs are printed in the
196 usual unified diff format, starting with a +/-/` ` character at
197 the beginning of the line and extending to the end of the line.
198 Newlines in the input are represented by a tilde ~ on a line of
199 its own.
200 none
202 Disable word diff again.
203 Note that despite the name of the first mode, color is used to
205 highlight the changed parts in all modes if enabled.
206 --word-diff-regex=<regex>
208 Use <regex> to decide what a word is, instead of considering runs
209 of non-whitespace to be a word. Also implies --word-diff unless it
210 was already enabled.
211 Every non-overlapping match of the <regex> is considered a word.
213 Anything between these matches is considered whitespace and
214 ignored(!) for the purposes of finding differences. You may want to
215 append |[^[:space:]] to your regular expression to make sure that
216 it matches all non-whitespace characters. A match that contains a
217 newline is silently truncated(!) at the newline.
218 The regex can also be set via a diff driver or configuration
220 option, see gitattributes(1) or git-config(1). Giving it explicitly
221 overrides any diff driver or configuration setting. Diff drivers
222 override configuration settings.
223 --color-words[=<regex>]
225 Equivalent to --word-diff=color plus (if a regex was specified)
226 --word-diff-regex=<regex>.
227 --no-renames
229 Turn off rename detection, even when the configuration file gives
230 the default to do so.
231 --check
233 Warn if changes introduce whitespace errors. What are considered
234 whitespace errors is controlled by core.whitespace configuration.
235 By default, trailing whitespaces (including lines that solely
236 consist of whitespaces) and a space character that is immediately
237 followed by a tab character inside the initial indent of the line
238 are considered whitespace errors. Exits with non-zero status if
239 problems are found. Not compatible with --exit-code.
240 --full-index
242 Instead of the first handful of characters, show the full pre- and
243 post-image blob object names on the "index" line when generating
244 patch format output.
245 --binary
247 In addition to --full-index, output a binary diff that can be
248 applied with git-apply.
249 --abbrev[=<n>]
251 Instead of showing the full 40-byte hexadecimal object name in
252 diff-raw format output and diff-tree header lines, show only a
253 partial prefix. This is independent of the --full-index option
254 above, which controls the diff-patch output format. Non default
255 number of digits can be specified with --abbrev=<n>.
256 -B[<n>][/<m>], --break-rewrites[=[<n>][/<m>]]
258 Break complete rewrite changes into pairs of delete and create.
259 This serves two purposes:
260 It affects the way a change that amounts to a total rewrite of a
262 file not as a series of deletion and insertion mixed together with
263 a very few lines that happen to match textually as the context, but
264 as a single deletion of everything old followed by a single
265 insertion of everything new, and the number m controls this aspect
266 of the -B option (defaults to 60%). -B/70% specifies that less
267 than 30% of the original should remain in the result for Git to
268 consider it a total rewrite (i.e. otherwise the resulting patch
269 will be a series of deletion and insertion mixed together with
270 context lines).
271 When used with -M, a totally-rewritten file is also considered as
273 the source of a rename (usually -M only considers a file that
274 disappeared as the source of a rename), and the number n controls
275 this aspect of the -B option (defaults to 50%). -B20% specifies
276 that a change with addition and deletion compared to 20% or more of
277 the file’s size are eligible for being picked up as a possible
278 source of a rename to another file.
279 -M[<n>], --find-renames[=<n>]
281 Detect renames. If n is specified, it is a threshold on the
282 similarity index (i.e. amount of addition/deletions compared to the
283 file’s size). For example, -M90% means Git should consider a
284 delete/add pair to be a rename if more than 90% of the file hasn’t
285 changed. Without a % sign, the number is to be read as a fraction,
286 with a decimal point before it. I.e., -M5 becomes 0.5, and is thus
287 the same as -M50%. Similarly, -M05 is the same as -M5%. To limit
288 detection to exact renames, use -M100%.
289 -C[<n>], --find-copies[=<n>]
291 Detect copies as well as renames. See also --find-copies-harder. If
292 n is specified, it has the same meaning as for -M<n>.
293 --find-copies-harder
295 For performance reasons, by default, -C option finds copies only if
296 the original file of the copy was modified in the same changeset.
297 This flag makes the command inspect unmodified files as candidates
298 for the source of copy. This is a very expensive operation for
299 large projects, so use it with caution. Giving more than one -C
300 option has the same effect.
301 -D, --irreversible-delete
303 Omit the preimage for deletes, i.e. print only the header but not
304 the diff between the preimage and /dev/null. The resulting patch is
305 not meant to be applied with patch nor git apply; this is solely
306 for people who want to just concentrate on reviewing the text after
307 the change. In addition, the output obviously lack enough
308 information to apply such a patch in reverse, even manually, hence
309 the name of the option.
310 When used together with -B, omit also the preimage in the deletion
312 part of a delete/create pair.
313 -l<num>
315 The -M and -C options require O(n^2) processing time where n is the
316 number of potential rename/copy targets. This option prevents
317 rename/copy detection from running if the number of rename/copy
318 targets exceeds the specified number.
319 --diff-filter=[(A|C|D|M|R|T|U|X|B)...[*]]
321 Select only files that are Added (A), Copied (C), Deleted (D),
322 Modified (M), Renamed (R), have their type (i.e. regular file,
323 symlink, submodule, ...) changed (T), are Unmerged (U), are Unknown
324 (X), or have had their pairing Broken (B). Any combination of the
325 filter characters (including none) can be used. When *
326 (All-or-none) is added to the combination, all paths are selected
327 if there is any file that matches other criteria in the comparison;
328 if there is no file that matches other criteria, nothing is
329 selected.
330 -S<string>
332 Look for differences that introduce or remove an instance of
333 <string>. Note that this is different than the string simply
334 appearing in diff output; see the pickaxe entry in gitdiffcore(7)
335 for more details.
336 -G<regex>
338 Look for differences whose added or removed line matches the given
339 <regex>.
340 --pickaxe-all
342 When -S or -G finds a change, show all the changes in that
343 changeset, not just the files that contain the change in <string>.
344 --pickaxe-regex
346 Make the <string> not a plain string but an extended POSIX regex to
347 match.
348 -O<orderfile>
350 Output the patch in the order specified in the <orderfile>, which
351 has one shell glob pattern per line.
352 -R
354 Swap two inputs; that is, show differences from index or on-disk
355 file to tree contents.
356 --relative[=<path>]
358 When run from a subdirectory of the project, it can be told to
359 exclude changes outside the directory and show pathnames relative
360 to it with this option. When you are not in a subdirectory (e.g. in
361 a bare repository), you can name which subdirectory to make the
362 output relative to by giving a <path> as an argument.
363 -a, --text
365 Treat all files as text.
366 --ignore-space-at-eol
368 Ignore changes in whitespace at EOL.
369 -b, --ignore-space-change
371 Ignore changes in amount of whitespace. This ignores whitespace at
372 line end, and considers all other sequences of one or more
373 whitespace characters to be equivalent.
374 -w, --ignore-all-space
376 Ignore whitespace when comparing lines. This ignores differences
377 even if one line has whitespace where the other line has none.
378 --inter-hunk-context=<lines>
380 Show the context between diff hunks, up to the specified number of
381 lines, thereby fusing hunks that are close to each other.
382 -W, --function-context
384 Show whole surrounding functions of changes.
385 --exit-code
387 Make the program exit with codes similar to diff(1). That is, it
388 exits with 1 if there were differences and 0 means no differences.
389 --quiet
391 Disable all output of the program. Implies --exit-code.
392 --ext-diff
394 Allow an external diff helper to be executed. If you set an
395 external diff driver with gitattributes(5), you need to use this
396 option with git-log(1) and friends.
397 --no-ext-diff
399 Disallow external diff drivers.
400 --textconv, --no-textconv
402 Allow (or disallow) external text conversion filters to be run when
403 comparing binary files. See gitattributes(5) for details. Because
404 textconv filters are typically a one-way conversion, the resulting
405 diff is suitable for human consumption, but cannot be applied. For
406 this reason, textconv filters are enabled by default only for git-
407 diff(1) and git-log(1), but not for git-format-patch(1) or diff
408 plumbing commands.
409 --ignore-submodules[=<when>]
411 Ignore changes to submodules in the diff generation. <when> can be
412 either "none", "untracked", "dirty" or "all", which is the default.
413 Using "none" will consider the submodule modified when it either
414 contains untracked or modified files or its HEAD differs from the
415 commit recorded in the superproject and can be used to override any
416 settings of the ignore option in git-config(1) or gitmodules(5).
417 When "untracked" is used submodules are not considered dirty when
418 they only contain untracked content (but they are still scanned for
419 modified content). Using "dirty" ignores all changes to the work
420 tree of submodules, only changes to the commits stored in the
421 superproject are shown (this was the behavior until 1.7.0). Using
422 "all" hides all changes to submodules.
423 --src-prefix=<prefix>
425 Show the given source prefix instead of "a/".
426 --dst-prefix=<prefix>
428 Show the given destination prefix instead of "b/".
429 --no-prefix
431 Do not show any source or destination prefix.
432 For more detailed explanation on these common options, see also
434 gitdiffcore(7).
435 -1 --base, -2 --ours, -3 --theirs, -0
437 Diff against the "base" version, "our branch" or "their branch"
438 respectively. With these options, diffs for merged entries are not
439 shown.
440 The default is to diff against our branch (-2) and the cleanly
442 resolved paths. The option -0 can be given to omit diff output for
443 unmerged entries and just show "Unmerged".
444 -c, --cc
446 This compares stage 2 (our branch), stage 3 (their branch) and the
447 working tree file and outputs a combined diff, similar to the way
448 diff-tree shows a merge commit with these flags.
449 -q
451 Remain silent even on nonexistent files
452
454 The raw output format from "git-diff-index", "git-diff-tree",
455 "git-diff-files" and "git diff --raw" are very similar.
456 These commands all compare two sets of things; what is compared
458 differs:
459 git-diff-index <tree-ish>
461 compares the <tree-ish> and the files on the filesystem.
462 git-diff-index --cached <tree-ish>
464 compares the <tree-ish> and the index.
465 git-diff-tree [-r] <tree-ish-1> <tree-ish-2> [<pattern>...]
467 compares the trees named by the two arguments.
468 git-diff-files [<pattern>...]
470 compares the index and the files on the filesystem.
471 The "git-diff-tree" command begins its output by printing the hash of
473 what is being compared. After that, all the commands print one output
474 line per changed file.
475 An output line is formatted this way:
477 in-place edit :100644 100644 bcd1234... 0123456... M file0
479 copy-edit :100644 100644 abcd123... 1234567... C68 file1 file2
480 rename-edit :100644 100644 abcd123... 1234567... R86 file1 file3
481 create :000000 100644 0000000... 1234567... A file4
482 delete :100644 000000 1234567... 0000000... D file5
483 unmerged :000000 000000 0000000... 0000000... U file6
484 That is, from the left to the right:
487 1. a colon.
489 2. mode for "src"; 000000 if creation or unmerged.
491 3. a space.
493 4. mode for "dst"; 000000 if deletion or unmerged.
495 5. a space.
497 6. sha1 for "src"; 0{40} if creation or unmerged.
499 7. a space.
501 8. sha1 for "dst"; 0{40} if creation, unmerged or "look at work tree".
503 9. a space.
505 10. status, followed by optional "score" number.
507 11. a tab or a NUL when -z option is used.
509 12. path for "src"
511 13. a tab or a NUL when -z option is used; only exists for C or R.
513 14. path for "dst"; only exists for C or R.
515 15. an LF or a NUL when -z option is used, to terminate the record.
517 Possible status letters are:
519 · A: addition of a file
521 · C: copy of a file into a new one
523 · D: deletion of a file
525 · M: modification of the contents or mode of a file
527 · R: renaming of a file
529 · T: change in the type of the file
531 · U: file is unmerged (you must complete the merge before it can be
533 committed)
534 · X: "unknown" change type (most probably a bug, please report it)
536 Status letters C and R are always followed by a score (denoting the
538 percentage of similarity between the source and target of the move or
539 copy), and are the only ones to be so.
540 <sha1> is shown as all 0’s if a file is new on the filesystem and it is
542 out of sync with the index.
543 Example:
545 :100644 100644 5be4a4...... 000000...... M file.c
547 When -z option is not used, TAB, LF, and backslash characters in
550 pathnames are represented as \t, \n, and \\, respectively.
551
553 "git-diff-tree", "git-diff-files" and "git-diff --raw" can take -c or
554 --cc option to generate diff output also for merge commits. The output
555 differs from the format described above in the following way:
556 1. there is a colon for each parent
558 2. there are more "src" modes and "src" sha1
560 3. status is concatenated status characters for each parent
562 4. no optional "score" number
564 5. single path, only for "dst"
566 Example:
568 ::100644 100644 100644 fabadb8... cc95eb0... 4866510... MM describe.c
570 Note that combined diff lists only files which were modified from all
573 parents.
574
576 When "git-diff-index", "git-diff-tree", or "git-diff-files" are run
577 with a -p option, "git diff" without the --raw option, or "git log"
578 with the "-p" option, they do not produce the output described above;
579 instead they produce a patch file. You can customize the creation of
580 such patches via the GIT_EXTERNAL_DIFF and the GIT_DIFF_OPTS
581 environment variables.
582 What the -p option produces is slightly different from the traditional
584 diff format:
585 1. It is preceded with a "git diff" header that looks like this:
587 diff --git a/file1 b/file2
589 The a/ and b/ filenames are the same unless rename/copy is
591 involved. Especially, even for a creation or a deletion, /dev/null
592 is not used in place of the a/ or b/ filenames.
593 When rename/copy is involved, file1 and file2 show the name of the
595 source file of the rename/copy and the name of the file that
596 rename/copy produces, respectively.
597 2. It is followed by one or more extended header lines:
599 old mode <mode>
601 new mode <mode>
602 deleted file mode <mode>
603 new file mode <mode>
604 copy from <path>
605 copy to <path>
606 rename from <path>
607 rename to <path>
608 similarity index <number>
609 dissimilarity index <number>
610 index <hash>..<hash> <mode>
611 File modes are printed as 6-digit octal numbers including the file
613 type and file permission bits.
614 Path names in extended headers do not include the a/ and b/
616 prefixes.
617 The similarity index is the percentage of unchanged lines, and the
619 dissimilarity index is the percentage of changed lines. It is a
620 rounded down integer, followed by a percent sign. The similarity
621 index value of 100% is thus reserved for two equal files, while
622 100% dissimilarity means that no line from the old file made it
623 into the new one.
624 The index line includes the SHA-1 checksum before and after the
626 change. The <mode> is included if the file mode does not change;
627 otherwise, separate lines indicate the old and the new mode.
628 3. TAB, LF, double quote and backslash characters in pathnames are
630 represented as \t, \n, \" and \\, respectively. If there is need
631 for such substitution then the whole pathname is put in double
632 quotes.
633 4. All the file1 files in the output refer to files before the commit,
635 and all the file2 files refer to files after the commit. It is
636 incorrect to apply each change to each file sequentially. For
637 example, this patch will swap a and b:
638 diff --git a/a b/b
640 rename from a
641 rename to b
642 diff --git a/b b/a
643 rename from b
644 rename to a
645
647 Any diff-generating command can take the ‘-c` or --cc option to produce
648 a combined diff when showing a merge. This is the default format when
649 showing merges with git-diff(1) or git-show(1). Note also that you can
650 give the `-m’ option to any of these commands to force generation of
651 diffs with individual parents of a merge.
652 A combined diff format looks like this:
654 diff --combined describe.c
656 index fabadb8,cc95eb0..4866510
657 --- a/describe.c
658 +++ b/describe.c
659 @@@ -98,20 -98,12 +98,20 @@@
660 return (a_date > b_date) ? -1 : (a_date == b_date) ? 0 : 1;
661 }
662 - static void describe(char *arg)
664 -static void describe(struct commit *cmit, int last_one)
665 ++static void describe(char *arg, int last_one)
666 {
667 + unsigned char sha1[20];
668 + struct commit *cmit;
669 struct commit_list *list;
670 static int initialized = 0;
671 struct commit_name *n;
672 + if (get_sha1(arg, sha1) < 0)
674 + usage(describe_usage);
675 + cmit = lookup_commit_reference(sha1);
676 + if (!cmit)
677 + usage(describe_usage);
678 +
679 if (!initialized) {
680 initialized = 1;
681 for_each_ref(get_name);
682 1. It is preceded with a "git diff" header, that looks like this (when
686 -c option is used):
687 diff --combined file
689 or like this (when --cc option is used):
691 diff --cc file
693 2. It is followed by one or more extended header lines (this example
695 shows a merge with two parents):
696 index <hash>,<hash>..<hash>
698 mode <mode>,<mode>..<mode>
699 new file mode <mode>
700 deleted file mode <mode>,<mode>
701 The mode <mode>,<mode>..<mode> line appears only if at least one of
703 the <mode> is different from the rest. Extended headers with
704 information about detected contents movement (renames and copying
705 detection) are designed to work with diff of two <tree-ish> and are
706 not used by combined diff format.
707 3. It is followed by two-line from-file/to-file header
709 --- a/file
711 +++ b/file
712 Similar to two-line header for traditional unified diff format,
714 /dev/null is used to signal created or deleted files.
715 4. Chunk header format is modified to prevent people from accidentally
717 feeding it to patch -p1. Combined diff format was created for
718 review of merge commit changes, and was not meant for apply. The
719 change is similar to the change in the extended index header:
720 @@@ <from-file-range> <from-file-range> <to-file-range> @@@
722 There are (number of parents + 1) @ characters in the chunk header
724 for combined diff format.
725 Unlike the traditional unified diff format, which shows two files A and
727 B with a single column that has - (minus — appears in A but removed in
728 B), + (plus — missing in A but added to B), or " " (space — unchanged)
729 prefix, this format compares two or more files file1, file2,... with
730 one file X, and shows how X differs from each of fileN. One column for
731 each of fileN is prepended to the output line to note how X’s line is
732 different from it.
733 A - character in the column N means that the line appears in fileN but
735 it does not appear in the result. A + character in the column N means
736 that the line appears in the result, and fileN does not have that line
737 (in other words, the line was added, from the point of view of that
738 parent).
739 In the above example output, the function signature was changed from
741 both files (hence two - removals from both file1 and file2, plus ++ to
742 mean one line that was added does not appear in either file1 nor
743 file2). Also eight other lines are the same from file1 but do not
744 appear in file2 (hence prefixed with +).
745 When shown by git diff-tree -c, it compares the parents of a merge
747 commit with the merge result (i.e. file1..fileN are the parents). When
748 shown by git diff-files -c, it compares the two unresolved merge
749 parents with the working tree file (i.e. file1 is stage 2 aka "our
750 version", file2 is stage 3 aka "their version").
751
753 The --summary option describes newly added, deleted, renamed and copied
754 files. The --stat option adds diffstat(1) graph to the output. These
755 options can be combined with other options, such as -p, and are meant
756 for human consumption.
757 When showing a change that involves a rename or a copy, --stat output
759 formats the pathnames compactly by combining common prefix and suffix
760 of the pathnames. For example, a change that moves arch/i386/Makefile
761 to arch/x86/Makefile while modifying 4 lines will be shown like this:
762 arch/{i386 => x86}/Makefile | 4 +--
764 The --numstat option gives the diffstat(1) information but is designed
767 for easier machine consumption. An entry in --numstat output looks like
768 this:
769 1 2 README
771 3 1 arch/{i386 => x86}/Makefile
772 That is, from left to right:
775 1. the number of added lines;
777 2. a tab;
779 3. the number of deleted lines;
781 4. a tab;
783 5. pathname (possibly with rename/copy information);
785 6. a newline.
787 When -z output option is in effect, the output is formatted this way:
789 1 2 README NUL
791 3 1 NUL arch/i386/Makefile NUL arch/x86/Makefile NUL
792 That is:
795 1. the number of added lines;
797 2. a tab;
799 3. the number of deleted lines;
801 4. a tab;
803 5. a NUL (only exists if renamed/copied);
805 6. pathname in preimage;
807 7. a NUL (only exists if renamed/copied);
809 8. pathname in postimage (only exists if renamed/copied);
811 9. a NUL.
813 The extra NUL before the preimage path in renamed case is to allow
815 scripts that read the output to tell if the current record being read
816 is a single-path record or a rename/copy record without reading ahead.
817 After reading added and deleted lines, reading up to NUL would yield
818 the pathname, but if that is NUL, the record will show two paths.
819
821 Part of the git(1) suite
822Git 1.8.3.1 11/19/2018 GIT-DIFF-FILES(1)