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>]
10 [<path>...]
11
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
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 --patience
33 Generate a diff using the "patience diff" algorithm.
34 --stat[=width[,name-width]]
36 Generate a diffstat. You can override the default output width for
37 80-column terminal by --stat=width. The width of the filename part
38 can be controlled by giving another width to it separated by a
39 comma.
40 --numstat
42 Similar to --stat, but shows number of added and deleted lines in
43 decimal notation and pathname without abbreviation, to make it more
44 machine friendly. For binary files, outputs two - instead of saying
45 0 0.
46 --shortstat
48 Output only the last line of the --stat format containing total
49 number of modified files, as well as number of added and deleted
50 lines.
51 --dirstat[=limit]
53 Output the distribution of relative amount of changes (number of
54 lines added or removed) for each sub-directory. Directories with
55 changes below a cut-off percent (3% by default) are not shown. The
56 cut-off percent can be set with --dirstat=limit. Changes in a child
57 directory is not counted for the parent directory, unless
58 --cumulative is used.
59 --dirstat-by-file[=limit]
61 Same as --dirstat, but counts changed files instead of lines.
62 --summary
64 Output a condensed summary of extended header information such as
65 creations, renames and mode changes.
66 --patch-with-stat
68 Synonym for -p --stat.
69 -z
71 When --raw, --numstat, --name-only or --name-status has been given,
72 do not munge pathnames and use NULs as output field terminators.
73 Without this option, each pathname output will have TAB, LF, double
75 quotes, and backslash characters replaced with \t, \n, \", and \\,
76 respectively, and the pathname will be enclosed in double quotes if
77 any of those replacements occurred.
78 --name-only
80 Show only names of changed files.
81 --name-status
83 Show only names and status of changed files. See the description of
84 the --diff-filter option on what the status letters mean.
85 --submodule[=<format>]
87 Chose the output format for submodule differences. <format> can be
88 one of short and log. short just shows pairs of commit names, this
89 format is used when this option is not given. log is the default
90 value for this option and lists the commits in that commit range
91 like the summary option of git-submodule(1) does.
92 --color[=<when>]
94 Show colored diff. The value must be always (the default), never,
95 or auto.
96 --no-color
98 Turn off colored diff, even when the configuration file gives the
99 default to color output. Same as --color=never.
100 --color-words[=<regex>]
102 Show colored word diff, i.e., color words which have changed. By
103 default, words are separated by whitespace.
104 When a <regex> is specified, every non-overlapping match of the
106 <regex> is considered a word. Anything between these matches is
107 considered whitespace and ignored(!) for the purposes of finding
108 differences. You may want to append |[^[:space:]] to your regular
109 expression to make sure that it matches all non-whitespace
110 characters. A match that contains a newline is silently
111 truncated(!) at the newline.
112 The regex can also be set via a diff driver or configuration
114 option, see gitattributes(1) or git-config(1). Giving it explicitly
115 overrides any diff driver or configuration setting. Diff drivers
116 override configuration settings.
117 --no-renames
119 Turn off rename detection, even when the configuration file gives
120 the default to do so.
121 --check
123 Warn if changes introduce trailing whitespace or an indent that
124 uses a space before a tab. Exits with non-zero status if problems
125 are found. Not compatible with --exit-code.
126 --full-index
128 Instead of the first handful of characters, show the full pre- and
129 post-image blob object names on the "index" line when generating
130 patch format output.
131 --binary
133 In addition to --full-index, output a binary diff that can be
134 applied with git-apply.
135 --abbrev[=<n>]
137 Instead of showing the full 40-byte hexadecimal object name in
138 diff-raw format output and diff-tree header lines, show only a
139 partial prefix. This is independent of the --full-index option
140 above, which controls the diff-patch output format. Non default
141 number of digits can be specified with --abbrev=<n>.
142 -B
144 Break complete rewrite changes into pairs of delete and create.
145 -M
147 Detect renames.
148 -C
150 Detect copies as well as renames. See also --find-copies-harder.
151 --diff-filter=[ACDMRTUXB*]
153 Select only files that are Added (A), Copied (C), Deleted (D),
154 Modified (M), Renamed (R), have their type (i.e. regular file,
155 symlink, submodule, ...) changed (T), are Unmerged (U), are Unknown
156 (X), or have had their pairing Broken (B). Any combination of the
157 filter characters may be used. When * (All-or-none) is added to the
158 combination, all paths are selected if there is any file that
159 matches other criteria in the comparison; if there is no file that
160 matches other criteria, nothing is selected.
161 --find-copies-harder
163 For performance reasons, by default, -C option finds copies only if
164 the original file of the copy was modified in the same changeset.
165 This flag makes the command inspect unmodified files as candidates
166 for the source of copy. This is a very expensive operation for
167 large projects, so use it with caution. Giving more than one -C
168 option has the same effect.
169 -l<num>
171 The -M and -C options require O(n^2) processing time where n is the
172 number of potential rename/copy targets. This option prevents
173 rename/copy detection from running if the number of rename/copy
174 targets exceeds the specified number.
175 -S<string>
177 Look for differences that introduce or remove an instance of
178 <string>. Note that this is different than the string simply
179 appearing in diff output; see the pickaxe entry in gitdiffcore(7)
180 for more details.
181 --pickaxe-all
183 When -S finds a change, show all the changes in that changeset, not
184 just the files that contain the change in <string>.
185 --pickaxe-regex
187 Make the <string> not a plain string but an extended POSIX regex to
188 match.
189 -O<orderfile>
191 Output the patch in the order specified in the <orderfile>, which
192 has one shell glob pattern per line.
193 -R
195 Swap two inputs; that is, show differences from index or on-disk
196 file to tree contents.
197 --relative[=<path>]
199 When run from a subdirectory of the project, it can be told to
200 exclude changes outside the directory and show pathnames relative
201 to it with this option. When you are not in a subdirectory (e.g. in
202 a bare repository), you can name which subdirectory to make the
203 output relative to by giving a <path> as an argument.
204 -a, --text
206 Treat all files as text.
207 --ignore-space-at-eol
209 Ignore changes in whitespace at EOL.
210 -b, --ignore-space-change
212 Ignore changes in amount of whitespace. This ignores whitespace at
213 line end, and considers all other sequences of one or more
214 whitespace characters to be equivalent.
215 -w, --ignore-all-space
217 Ignore whitespace when comparing lines. This ignores differences
218 even if one line has whitespace where the other line has none.
219 --inter-hunk-context=<lines>
221 Show the context between diff hunks, up to the specified number of
222 lines, thereby fusing hunks that are close to each other.
223 --exit-code
225 Make the program exit with codes similar to diff(1). That is, it
226 exits with 1 if there were differences and 0 means no differences.
227 --quiet
229 Disable all output of the program. Implies --exit-code.
230 --ext-diff
232 Allow an external diff helper to be executed. If you set an
233 external diff driver with gitattributes(5), you need to use this
234 option with git-log(1) and friends.
235 --no-ext-diff
237 Disallow external diff drivers.
238 --ignore-submodules
240 Ignore changes to submodules in the diff generation.
241 --src-prefix=<prefix>
243 Show the given source prefix instead of "a/".
244 --dst-prefix=<prefix>
246 Show the given destination prefix instead of "b/".
247 --no-prefix
249 Do not show any source or destination prefix.
250 For more detailed explanation on these common options, see also
252 gitdiffcore(7).
253 -1 --base, -2 --ours, -3 --theirs, -0
255 Diff against the "base" version, "our branch" or "their branch"
256 respectively. With these options, diffs for merged entries are not
257 shown.
258 The default is to diff against our branch (-2) and the cleanly
260 resolved paths. The option -0 can be given to omit diff output for
261 unmerged entries and just show "Unmerged".
262 -c, --cc
264 This compares stage 2 (our branch), stage 3 (their branch) and the
265 working tree file and outputs a combined diff, similar to the way
266 diff-tree shows a merge commit with these flags.
267 -q
269 Remain silent even on nonexistent files
270
272 The raw output format from "git-diff-index", "git-diff-tree",
273 "git-diff-files" and "git diff --raw" are very similar.
274 These commands all compare two sets of things; what is compared
276 differs:
277 git-diff-index <tree-ish>
279 compares the <tree-ish> and the files on the filesystem.
280 git-diff-index --cached <tree-ish>
282 compares the <tree-ish> and the index.
283 git-diff-tree [-r] <tree-ish-1> <tree-ish-2> [<pattern>...]
285 compares the trees named by the two arguments.
286 git-diff-files [<pattern>...]
288 compares the index and the files on the filesystem.
289 The "git-diff-tree" command begins its output by printing the hash of
291 what is being compared. After that, all the commands print one output
292 line per changed file.
293 An output line is formatted this way:
295 in-place edit :100644 100644 bcd1234... 0123456... M file0
297 copy-edit :100644 100644 abcd123... 1234567... C68 file1 file2
298 rename-edit :100644 100644 abcd123... 1234567... R86 file1 file3
299 create :000000 100644 0000000... 1234567... A file4
300 delete :100644 000000 1234567... 0000000... D file5
301 unmerged :000000 000000 0000000... 0000000... U file6
302 That is, from the left to the right:
305 1. a colon.
307 2. mode for "src"; 000000 if creation or unmerged.
309 3. a space.
311 4. mode for "dst"; 000000 if deletion or unmerged.
313 5. a space.
315 6. sha1 for "src"; 0{40} if creation or unmerged.
317 7. a space.
319 8. sha1 for "dst"; 0{40} if creation, unmerged or "look at work tree".
321 9. a space.
323 10. status, followed by optional "score" number.
325 11. a tab or a NUL when -z option is used.
327 12. path for "src"
329 13. a tab or a NUL when -z option is used; only exists for C or R.
331 14. path for "dst"; only exists for C or R.
333 15. an LF or a NUL when -z option is used, to terminate the record.
335 Possible status letters are:
337 · A: addition of a file
339 · C: copy of a file into a new one
341 · D: deletion of a file
343 · M: modification of the contents or mode of a file
345 · R: renaming of a file
347 · T: change in the type of the file
349 · U: file is unmerged (you must complete the merge before it can be
351 committed)
352 · X: "unknown" change type (most probably a bug, please report it)
354 Status letters C and R are always followed by a score (denoting the
356 percentage of similarity between the source and target of the move or
357 copy), and are the only ones to be so.
358 <sha1> is shown as all 0’s if a file is new on the filesystem and it is
360 out of sync with the index.
361 Example:
363 :100644 100644 5be4a4...... 000000...... M file.c
365 When -z option is not used, TAB, LF, and backslash characters in
368 pathnames are represented as \t, \n, and \\, respectively.
369
371 "git-diff-tree", "git-diff-files" and "git-diff --raw" can take -c or
372 --cc option to generate diff output also for merge commits. The output
373 differs from the format described above in the following way:
374 1. there is a colon for each parent
376 2. there are more "src" modes and "src" sha1
378 3. status is concatenated status characters for each parent
380 4. no optional "score" number
382 5. single path, only for "dst"
384 Example:
386 ::100644 100644 100644 fabadb8... cc95eb0... 4866510... MM describe.c
388 Note that combined diff lists only files which were modified from all
391 parents.
392
394 When "git-diff-index", "git-diff-tree", or "git-diff-files" are run
395 with a -p option, "git diff" without the --raw option, or "git log"
396 with the "-p" option, they do not produce the output described above;
397 instead they produce a patch file. You can customize the creation of
398 such patches via the GIT_EXTERNAL_DIFF and the GIT_DIFF_OPTS
399 environment variables.
400 What the -p option produces is slightly different from the traditional
402 diff format.
403 1. It is preceded with a "git diff" header, that looks like this:
405 diff --git a/file1 b/file2
407 The a/ and b/ filenames are the same unless rename/copy is
409 involved. Especially, even for a creation or a deletion, /dev/null
410 is not used in place of a/ or b/ filenames.
411 When rename/copy is involved, file1 and file2 show the name of the
413 source file of the rename/copy and the name of the file that
414 rename/copy produces, respectively.
415 2. It is followed by one or more extended header lines:
417 old mode <mode>
419 new mode <mode>
420 deleted file mode <mode>
421 new file mode <mode>
422 copy from <path>
423 copy to <path>
424 rename from <path>
425 rename to <path>
426 similarity index <number>
427 dissimilarity index <number>
428 index <hash>..<hash> <mode>
429 3. TAB, LF, double quote and backslash characters in pathnames are
431 represented as \t, \n, \" and \\, respectively. If there is need
432 for such substitution then the whole pathname is put in double
433 quotes.
434 The similarity index is the percentage of unchanged lines, and the
436 dissimilarity index is the percentage of changed lines. It is a rounded
437 down integer, followed by a percent sign. The similarity index value of
438 100% is thus reserved for two equal files, while 100% dissimilarity
439 means that no line from the old file made it into the new one.
440
442 "git-diff-tree", "git-diff-files" and "git-diff" can take -c or --cc
443 option to produce combined diff. For showing a merge commit with "git
444 log -p", this is the default format; you can force showing full diff
445 with the -m option. A combined diff format looks like this:
446 diff --combined describe.c
448 index fabadb8,cc95eb0..4866510
449 --- a/describe.c
450 +++ b/describe.c
451 @@@ -98,20 -98,12 +98,20 @@@
452 return (a_date > b_date) ? -1 : (a_date == b_date) ? 0 : 1;
453 }
454 - static void describe(char *arg)
456 -static void describe(struct commit *cmit, int last_one)
457 ++static void describe(char *arg, int last_one)
458 {
459 + unsigned char sha1[20];
460 + struct commit *cmit;
461 struct commit_list *list;
462 static int initialized = 0;
463 struct commit_name *n;
464 + if (get_sha1(arg, sha1) < 0)
466 + usage(describe_usage);
467 + cmit = lookup_commit_reference(sha1);
468 + if (!cmit)
469 + usage(describe_usage);
470 +
471 if (!initialized) {
472 initialized = 1;
473 for_each_ref(get_name);
474 1. It is preceded with a "git diff" header, that looks like this (when
478 -c option is used):
479 diff --combined file
481 or like this (when --cc option is used):
483 diff --cc file
485 2. It is followed by one or more extended header lines (this example
487 shows a merge with two parents):
488 index <hash>,<hash>..<hash>
490 mode <mode>,<mode>..<mode>
491 new file mode <mode>
492 deleted file mode <mode>,<mode>
493 The mode <mode>,<mode>..<mode> line appears only if at least one of
495 the <mode> is different from the rest. Extended headers with
496 information about detected contents movement (renames and copying
497 detection) are designed to work with diff of two <tree-ish> and are
498 not used by combined diff format.
499 3. It is followed by two-line from-file/to-file header
501 --- a/file
503 +++ b/file
504 Similar to two-line header for traditional unified diff format,
506 /dev/null is used to signal created or deleted files.
507 4. Chunk header format is modified to prevent people from accidentally
509 feeding it to patch -p1. Combined diff format was created for
510 review of merge commit changes, and was not meant for apply. The
511 change is similar to the change in the extended index header:
512 @@@ <from-file-range> <from-file-range> <to-file-range> @@@
514 There are (number of parents + 1) @ characters in the chunk header
516 for combined diff format.
517 Unlike the traditional unified diff format, which shows two files A and
519 B with a single column that has - (minus — appears in A but removed in
520 B), + (plus — missing in A but added to B), or " " (space — unchanged)
521 prefix, this format compares two or more files file1, file2,... with
522 one file X, and shows how X differs from each of fileN. One column for
523 each of fileN is prepended to the output line to note how X’s line is
524 different from it.
525 A - character in the column N means that the line appears in fileN but
527 it does not appear in the result. A + character in the column N means
528 that the line appears in the result, and fileN does not have that line
529 (in other words, the line was added, from the point of view of that
530 parent).
531 In the above example output, the function signature was changed from
533 both files (hence two - removals from both file1 and file2, plus ++ to
534 mean one line that was added does not appear in either file1 nor
535 file2). Also eight other lines are the same from file1 but do not
536 appear in file2 (hence prefixed with +).
537 When shown by git diff-tree -c, it compares the parents of a merge
539 commit with the merge result (i.e. file1..fileN are the parents). When
540 shown by git diff-files -c, it compares the two unresolved merge
541 parents with the working tree file (i.e. file1 is stage 2 aka "our
542 version", file2 is stage 3 aka "their version").
543
545 The --summary option describes newly added, deleted, renamed and copied
546 files. The --stat option adds diffstat(1) graph to the output. These
547 options can be combined with other options, such as -p, and are meant
548 for human consumption.
549 When showing a change that involves a rename or a copy, --stat output
551 formats the pathnames compactly by combining common prefix and suffix
552 of the pathnames. For example, a change that moves arch/i386/Makefile
553 to arch/x86/Makefile while modifying 4 lines will be shown like this:
554 arch/{i386 => x86}/Makefile | 4 +--
556 The --numstat option gives the diffstat(1) information but is designed
559 for easier machine consumption. An entry in --numstat output looks like
560 this:
561 1 2 README
563 3 1 arch/{i386 => x86}/Makefile
564 That is, from left to right:
567 1. the number of added lines;
569 2. a tab;
571 3. the number of deleted lines;
573 4. a tab;
575 5. pathname (possibly with rename/copy information);
577 6. a newline.
579 When -z output option is in effect, the output is formatted this way:
581 1 2 README NUL
583 3 1 NUL arch/i386/Makefile NUL arch/x86/Makefile NUL
584 That is:
587 1. the number of added lines;
589 2. a tab;
591 3. the number of deleted lines;
593 4. a tab;
595 5. a NUL (only exists if renamed/copied);
597 6. pathname in preimage;
599 7. a NUL (only exists if renamed/copied);
601 8. pathname in postimage (only exists if renamed/copied);
603 9. a NUL.
605 The extra NUL before the preimage path in renamed case is to allow
607 scripts that read the output to tell if the current record being read
608 is a single-path record or a rename/copy record without reading ahead.
609 After reading added and deleted lines, reading up to NUL would yield
610 the pathname, but if that is NUL, the record will show two paths.
611
613 Written by Linus Torvalds <torvalds@osdl.org[1]>
614
616 Documentation by David Greaves, Junio C Hamano and the git-list
617 <git@vger.kernel.org[2]>.
618
620 Part of the git(1) suite
621
623 1. torvalds@osdl.org
624 mailto:torvalds@osdl.org
625 2. git@vger.kernel.org
627 mailto:git@vger.kernel.org
628Git 1.7.1 08/16/2017 GIT-DIFF-FILES(1)