NAME

git-diff-index - Compare a tree to the working tree or index

SYNOPSIS

  1. git diff-index [-m] [--cached] [<common diff options>] <tree-ish> [<path>…​]

DESCRIPTION

Compares the content and mode of the blobs found in a tree objectwith the corresponding tracked files in the working tree, or with thecorresponding paths in the index. When <path> arguments are present,compares only paths matching those patterns. Otherwise all trackedfiles are compared.

OPTIONS

  • -p
  • -u
  • —patch
  • Generate patch (see section on generating patches).

  • -s

  • —no-patch
  • Suppress diff output. Useful for commands like git show thatshow the patch by default, or to cancel the effect of —patch.

  • -U

  • —unified=
  • Generate diffs with lines of context instead ofthe usual three. Implies —patch.Implies -p.

  • —output=

  • Output to a specific file instead of stdout.

  • —output-indicator-new=

  • —output-indicator-old=
  • —output-indicator-context=
  • Specify the character used to indicate new, old or contextlines in the generated patch. Normally they are +, - and' ' respectively.

  • —raw

  • Generate the diff in raw format.This is the default.

  • —patch-with-raw

  • Synonym for -p —raw.

  • —indent-heuristic

  • Enable the heuristic that shifts diff hunk boundaries to make patcheseasier to read. This is the default.

  • —no-indent-heuristic

  • Disable the indent heuristic.

  • —minimal

  • Spend extra time to make sure the smallest possiblediff is produced.

  • —patience

  • Generate a diff using the "patience diff" algorithm.

  • —histogram

  • Generate a diff using the "histogram diff" algorithm.

  • —anchored=

  • Generate a diff using the "anchored diff" algorithm.

This option may be specified more than once.

If a line exists in both the source and destination, exists only once,and starts with this text, this algorithm attempts to prevent it fromappearing as a deletion or addition in the output. It uses the "patiencediff" algorithm internally.

  • —diff-algorithm={patience|minimal|histogram|myers}
  • Choose a diff algorithm. The variants are as follows:
  • default, myers
  • The basic greedy diff algorithm. Currently, this is the default.

  • minimal

  • Spend extra time to make sure the smallest possible diff isproduced.

  • patience

  • Use "patience diff" algorithm when generating patches.

  • histogram

  • This algorithm extends the patience algorithm to "supportlow-occurrence common elements".

For instance, if you configured the diff.algorithm variable to anon-default value and want to use the default one, then youhave to use —diff-algorithm=default option.

  • —stat[=[,[,]]]
  • Generate a diffstat. By default, as much space as necessarywill be used for the filename part, and the rest for the graphpart. Maximum width defaults to terminal width, or 80 columnsif not connected to a terminal, and can be overridden by<width>. The width of the filename part can be limited bygiving another width <name-width> after a comma. The widthof the graph part can be limited by using—stat-graph-width=<width> (affects all commands generatinga stat graph) or by setting diff.statGraphWidth=<width>(does not affect git format-patch).By giving a third parameter <count>, you can limit theoutput to the first <count> lines, followed by ifthere are more.

These parameters can also be set individually with —stat-width=<width>,—stat-name-width=<name-width> and —stat-count=<count>.

  • —compact-summary
  • Output a condensed summary of extended header information suchas file creations or deletions ("new" or "gone", optionally "+l"if it’s a symlink) and mode changes ("+x" or "-x" for addingor removing executable bit respectively) in diffstat. Theinformation is put between the filename part and the graphpart. Implies —stat.

  • —numstat

  • Similar to —stat, but shows number of added anddeleted lines in decimal notation and pathname withoutabbreviation, to make it more machine friendly. Forbinary files, outputs two - instead of saying0 0.

  • —shortstat

  • Output only the last line of the —stat format containing totalnumber of modified files, as well as number of added and deletedlines.

  • -X[]

  • —dirstat[=]
  • Output the distribution of relative amount of changes for eachsub-directory. The behavior of —dirstat can be customized bypassing it a comma separated list of parameters.The defaults are controlled by the diff.dirstat configurationvariable (see git-config[1]).The following parameters are available:
  • changes
  • Compute the dirstat numbers by counting the lines that have beenremoved from the source, or added to the destination. This ignoresthe amount of pure code movements within a file. In other words,rearranging lines in a file is not counted as much as other changes.This is the default behavior when no parameter is given.

  • lines

  • Compute the dirstat numbers by doing the regular line-based diffanalysis, and summing the removed/added line counts. (For binaryfiles, count 64-byte chunks instead, since binary files have nonatural concept of lines). This is a more expensive —dirstatbehavior than the changes behavior, but it does count rearrangedlines within a file as much as other changes. The resulting outputis consistent with what you get from the other —*stat options.

  • files

  • Compute the dirstat numbers by counting the number of files changed.Each changed file counts equally in the dirstat analysis. This isthe computationally cheapest —dirstat behavior, since it doesnot have to look at the file contents at all.

  • cumulative

  • Count changes in a child directory for the parent directory as well.Note that when using cumulative, the sum of the percentagesreported may exceed 100%. The default (non-cumulative) behavior canbe specified with the noncumulative parameter.

  • An integer parameter specifies a cut-off percent (3% by default).Directories contributing less than this percentage of the changesare not shown in the output.

Example: The following will count changed files, while ignoringdirectories with less than 10% of the total amount of changed files,and accumulating child directory counts in the parent directories:—dirstat=files,10,cumulative.

  • —cumulative
  • Synonym for —dirstat=cumulative

  • —dirstat-by-file[=…​]

  • Synonym for —dirstat=files,param1,param2…​

  • —summary

  • Output a condensed summary of extended header informationsuch as creations, renames and mode changes.

  • —patch-with-stat

  • Synonym for -p —stat.

  • -z

  • When —raw, —numstat, —name-only or —name-status has beengiven, do not munge pathnames and use NULs as output field terminators.

Without this option, pathnames with "unusual" characters are quoted asexplained for the configuration variable core.quotePath (seegit-config[1]).

  • —name-only
  • Show only names of changed files.

  • —name-status

  • Show only names and status of changed files. See the descriptionof the —diff-filter option on what the status letters mean.

  • —submodule[=]

  • Specify how differences in submodules are shown. When specifying—submodule=short the short format is used. This format justshows the names of the commits at the beginning and end of the range.When —submodule or —submodule=log is specified, the log_format is used. This format lists the commits in the range likegit-submodule[1] summary does. When —submodule=diffis specified, the _diff format is used. This format shows aninline diff of the changes in the submodule contents between thecommit range. Defaults to diff.submodule or the short formatif the config option is unset.

  • —color[=]

  • Show colored diff.—color (i.e. without =) is the same as —color=always. can be one of always, never, or auto.

  • —no-color

  • Turn off colored diff.It is the same as —color=never.

  • —color-moved[=]

  • Moved lines of code are colored differently.The defaults to no if the option is not givenand to zebra if the option with no mode is given.The mode must be one of:
  • no
  • Moved lines are not highlighted.

  • default

  • Is a synonym for zebra. This may change to a more sensible modein the future.

  • plain

  • Any line that is added in one location and was removedin another location will be colored with color.diff.newMoved.Similarly color.diff.oldMoved will be used for removed linesthat are added somewhere else in the diff. This mode picks up anymoved line, but it is not very useful in a review to determineif a block of code was moved without permutation.

  • blocks

  • Blocks of moved text of at least 20 alphanumeric charactersare detected greedily. The detected blocks arepainted using either the color.diff.{old,new}Moved color.Adjacent blocks cannot be told apart.

  • zebra

  • Blocks of moved text are detected as in blocks mode. The blocksare painted using either the color.diff.{old,new}Moved color orcolor.diff.{old,new}MovedAlternative. The change betweenthe two colors indicates that a new block was detected.

  • dimmed-zebra

  • Similar to zebra, but additional dimming of uninteresting partsof moved code is performed. The bordering lines of two adjacentblocks are considered interesting, the rest is uninteresting.dimmed_zebra is a deprecated synonym.
  • —no-color-moved
  • Turn off move detection. This can be used to override configurationsettings. It is the same as —color-moved=no.

  • —color-moved-ws=

  • This configures how whitespace is ignored when performing themove detection for —color-moved.These modes can be given as a comma separated list:
  • no
  • Do not ignore whitespace when performing move detection.

  • ignore-space-at-eol

  • Ignore changes in whitespace at EOL.

  • ignore-space-change

  • Ignore changes in amount of whitespace. This ignores whitespaceat line end, and considers all other sequences of one ormore whitespace characters to be equivalent.

  • ignore-all-space

  • Ignore whitespace when comparing lines. This ignores differenceseven if one line has whitespace where the other line has none.

  • allow-indentation-change

  • Initially ignore any whitespace in the move detection, thengroup the moved code blocks only into a block if the change inwhitespace is the same per line. This is incompatible with theother modes.
  • —no-color-moved-ws
  • Do not ignore whitespace when performing move detection. This can beused to override configuration settings. It is the same as—color-moved-ws=no.

  • —word-diff[=]

  • Show a word diff, using the to delimit changed words.By default, words are delimited by whitespace; see—word-diff-regex below. The defaults to plain, andmust be one of:
  • color
  • Highlight changed words using only colors. Implies —color.

  • plain

  • Show words as [-removed-] and {+added+}. Makes noattempts to escape the delimiters if they appear in the input,so the output may be ambiguous.

  • porcelain

  • Use a special line-based format intended for scriptconsumption. Added/removed/unchanged runs are printed in theusual unified diff format, starting with a +/-/ character at the beginning of the line and extending to theend of the line. Newlines in the input are represented by atilde ~ on a line of its own.

  • none

  • Disable word diff again.

Note that despite the name of the first mode, color is used tohighlight the changed parts in all modes if enabled.

  • —word-diff-regex=
  • Use to decide what a word is, instead of consideringruns of non-whitespace to be a word. Also implies—word-diff unless it was already enabled.

Every non-overlapping match of the is considered a word. Anything between these matches isconsidered whitespace and ignored(!) for the purposes of findingdifferences. You may want to append |[^[:space:]] to your regularexpression to make sure that it matches all non-whitespace characters.A match that contains a newline is silently truncated(!) at thenewline.

For example, —word-diff-regex=. will treat each character as a wordand, correspondingly, show differences character by character.

The regex can also be set via a diff driver or configuration option, seegitattributes[5] or git-config[1]. Giving it explicitlyoverrides any diff driver or configuration setting. Diff driversoverride configuration settings.

  • —color-words[=]
  • Equivalent to —word-diff=color plus (if a regex wasspecified) —word-diff-regex=<regex>.

  • —no-renames

  • Turn off rename detection, even when the configurationfile gives the default to do so.

  • —[no-]rename-empty

  • Whether to use empty blobs as rename source.

  • —check

  • Warn if changes introduce conflict markers or whitespace errors.What are considered whitespace errors is controlled by core.whitespaceconfiguration. By default, trailing whitespaces (includinglines that consist solely of whitespaces) and a space characterthat is immediately followed by a tab character inside theinitial indent of the line are considered whitespace errors.Exits with non-zero status if problems are found. Not compatiblewith —exit-code.

  • —ws-error-highlight=

  • Highlight whitespace errors in the context, old or newlines of the diff. Multiple values are separated by comma,none resets previous values, default reset the list tonew and all is a shorthand for old,new,context. Whenthis option is not given, and the configuration variablediff.wsErrorHighlight is not set, only whitespace errors innew lines are highlighted. The whitespace errors are coloredwith color.diff.whitespace.

  • —full-index

  • Instead of the first handful of characters, show the fullpre- and post-image blob object names on the "index"line when generating patch format output.

  • —binary

  • In addition to —full-index, output a binary diff thatcan be applied with git-apply. Implies —patch.

  • —abbrev[=]

  • Instead of showing the full 40-byte hexadecimal objectname in diff-raw format output and diff-tree headerlines, show only a partial prefix. This isindependent of the —full-index option above, which controlsthe diff-patch output format. Non default number ofdigits can be specified with —abbrev=<n>.

  • -B[][/]

  • —break-rewrites[=[][/]]
  • Break complete rewrite changes into pairs of delete andcreate. This serves two purposes:

It affects the way a change that amounts to a total rewrite of a filenot as a series of deletion and insertion mixed together with a veryfew lines that happen to match textually as the context, but as asingle deletion of everything old followed by a single insertion ofeverything new, and the number m controls this aspect of the -Boption (defaults to 60%). -B/70% specifies that less than 30% of theoriginal should remain in the result for Git to consider it a totalrewrite (i.e. otherwise the resulting patch will be a series ofdeletion and insertion mixed together with context lines).

When used with -M, a totally-rewritten file is also considered as thesource of a rename (usually -M only considers a file that disappearedas the source of a rename), and the number n controls this aspect ofthe -B option (defaults to 50%). -B20% specifies that a change withaddition and deletion compared to 20% or more of the file’s size areeligible for being picked up as a possible source of a rename toanother file.

  • -M[]
  • —find-renames[=]
  • Detect renames.If n is specified, it is a threshold on the similarityindex (i.e. amount of addition/deletions compared to thefile’s size). For example, -M90% means Git should consider adelete/add pair to be a rename if more than 90% of the filehasn’t changed. Without a % sign, the number is to be read asa fraction, with a decimal point before it. I.e., -M5 becomes0.5, and is thus the same as -M50%. Similarly, -M05 isthe same as -M5%. To limit detection to exact renames, use-M100%. The default similarity index is 50%.

  • -C[]

  • —find-copies[=]
  • Detect copies as well as renames. See also —find-copies-harder.If n is specified, it has the same meaning as for -M<n>.

  • —find-copies-harder

  • For performance reasons, by default, -C option finds copies onlyif the original file of the copy was modified in the samechangeset. This flag makes the commandinspect unmodified files as candidates for the source ofcopy. This is a very expensive operation for largeprojects, so use it with caution. Giving more than one-C option has the same effect.

  • -D

  • —irreversible-delete
  • Omit the preimage for deletes, i.e. print only the header but notthe diff between the preimage and /dev/null. The resulting patchis not meant to be applied with patch or git apply; this issolely for people who want to just concentrate on reviewing thetext after the change. In addition, the output obviously lacksenough information to apply such a patch in reverse, even manually,hence the name of the option.

When used together with -B, omit also the preimage in the deletion partof a delete/create pair.

  • -l
  • The -M and -C options require O(n^2) processing time where nis the number of potential rename/copy targets. Thisoption prevents rename/copy detection from running ifthe number of rename/copy targets exceeds the specifiednumber.

  • —diff-filter=[(A|C|D|M|R|T|U|X|B)…​[*]]

  • Select only files that are Added (A), Copied (C),Deleted (D), Modified (M), Renamed (R), have theirtype (i.e. regular file, symlink, submodule, …​) changed (T),are Unmerged (U), areUnknown (X), or have had their pairing Broken (B).Any combination of the filter characters (including none) can be used.When * (All-or-none) is added to the combination, allpaths are selected if there is any file that matchesother criteria in the comparison; if there is no filethat matches other criteria, nothing is selected.

Also, these upper-case letters can be downcased to exclude. E.g.—diff-filter=ad excludes added and deleted paths.

Note that not all diffs can feature all types. For instance, diffsfrom the index to the working tree can never have Added entries(because the set of paths included in the diff is limited by what is inthe index). Similarly, copied and renamed entries cannot appear ifdetection for those types is disabled.

  • -S
  • Look for differences that change the number of occurrences ofthe specified string (i.e. addition/deletion) in a file.Intended for the scripter’s use.

It is useful when you’re looking for an exact block of code (like astruct), and want to know the history of that block since it firstcame into being: use the feature iteratively to feed the interestingblock in the preimage back into -S, and keep going until you get thevery first version of the block.

Binary files are searched as well.

  • -G
  • Look for differences whose patch text contains added/removedlines that match .

To illustrate the difference between -S<regex> —pickaxe-regex and-G<regex>, consider a commit with the following diff in the samefile:

  1. + return !regexec(regexp, two->ptr, 1, &regmatch, 0);
  2. ...
  3. - hit = !regexec(regexp, mf2.ptr, 1, &regmatch, 0);

While git log -G"regexec(regexp" will show this commit, git log-S"regexec(regexp" —pickaxe-regex will not (because the number ofoccurrences of that string did not change).

Unless —text is supplied patches of binary files without a textconvfilter will be ignored.

See the pickaxe entry in gitdiffcore[7] for moreinformation.

  • —find-object=
  • Look for differences that change the number of occurrences ofthe specified object. Similar to -S, just the argument is differentin that it doesn’t search for a specific string but for a specificobject id.

The object can be a blob or a submodule commit. It implies the -t option ingit-log to also find trees.

  • —pickaxe-all
  • When -S or -G finds a change, show all the changes in thatchangeset, not just the files that contain the changein .

  • —pickaxe-regex

  • Treat the given to -S as an extended POSIX regularexpression to match.

  • -O

  • Control the order in which files appear in the output.This overrides the diff.orderFile configuration variable(see git-config[1]). To cancel diff.orderFile,use -O/dev/null.

The output order is determined by the order of glob patterns in.All files with pathnames that match the first pattern are outputfirst, all files with pathnames that match the second pattern (but notthe first) are output next, and so on.All files with pathnames that do not match any pattern are outputlast, as if there was an implicit match-all pattern at the end of thefile.If multiple pathnames have the same rank (they match the same patternbut no earlier patterns), their output order relative to each other isthe normal order.

is parsed as follows:

  • Blank lines are ignored, so they can be used as separators forreadability.

  • Lines starting with a hash ("#") are ignored, so they can be usedfor comments. Add a backslash ("\") to the beginning of thepattern if it starts with a hash.

  • Each other line contains a single pattern.

Patterns have the same syntax and semantics as patterns used forfnmatch(3) without the FNM_PATHNAME flag, except a pathname alsomatches a pattern if removing any number of the final pathnamecomponents matches the pattern. For example, the pattern "foo*bar"matches "fooasdfbar" and "foo/bar/baz/asdf" but not "foobarx".

  • -R
  • Swap two inputs; that is, show differences from index oron-disk file to tree contents.

  • —relative[=]

  • When run from a subdirectory of the project, it can betold to exclude changes outside the directory and showpathnames relative to it with this option. When you arenot in a subdirectory (e.g. in a bare repository), youcan name which subdirectory to make the output relativeto by giving a as an argument.

  • -a

  • —text
  • Treat all files as text.

  • —ignore-cr-at-eol

  • Ignore carriage-return at the end of line when doing a comparison.

  • —ignore-space-at-eol

  • Ignore changes in whitespace at EOL.

  • -b

  • —ignore-space-change
  • Ignore changes in amount of whitespace. This ignores whitespaceat line end, and considers all other sequences of one ormore whitespace characters to be equivalent.

  • -w

  • —ignore-all-space
  • Ignore whitespace when comparing lines. This ignoresdifferences even if one line has whitespace where the otherline has none.

  • —ignore-blank-lines

  • Ignore changes whose lines are all blank.

  • —inter-hunk-context=

  • Show the context between diff hunks, up to the specified numberof lines, thereby fusing hunks that are close to each other.Defaults to diff.interHunkContext or 0 if the config optionis unset.

  • -W

  • —function-context
  • Show whole surrounding functions of changes.

  • —exit-code

  • Make the program exit with codes similar to diff(1).That is, it exits with 1 if there were differences and0 means no differences.

  • —quiet

  • Disable all output of the program. Implies —exit-code.

  • —ext-diff

  • Allow an external diff helper to be executed. If you set anexternal diff driver with gitattributes[5], you needto use this option with git-log[1] and friends.

  • —no-ext-diff

  • Disallow external diff drivers.

  • —textconv

  • —no-textconv
  • Allow (or disallow) external text conversion filters to be runwhen comparing binary files. See gitattributes[5] fordetails. Because textconv filters are typically a one-wayconversion, the resulting diff is suitable for humanconsumption, but cannot be applied. For this reason, textconvfilters are enabled by default only for git-diff[1] andgit-log[1], but not for git-format-patch[1] ordiff plumbing commands.

  • —ignore-submodules[=]

  • Ignore changes to submodules in the diff generation. can beeither "none", "untracked", "dirty" or "all", which is the default.Using "none" will consider the submodule modified when it either containsuntracked or modified files or its HEAD differs from the commit recordedin the superproject and can be used to override any settings of theignore option in git-config[1] or gitmodules[5]. When"untracked" is used submodules are not considered dirty when they onlycontain untracked content (but they are still scanned for modifiedcontent). Using "dirty" ignores all changes to the work tree of submodules,only changes to the commits stored in the superproject are shown (this wasthe behavior until 1.7.0). Using "all" hides all changes to submodules.

  • —src-prefix=

  • Show the given source prefix instead of "a/".

  • —dst-prefix=

  • Show the given destination prefix instead of "b/".

  • —no-prefix

  • Do not show any source or destination prefix.

  • —line-prefix=

  • Prepend an additional prefix to every line of output.

  • —ita-invisible-in-index

  • By default entries added by "git add -N" appear as an existingempty file in "git diff" and a new file in "git diff —cached".This option makes the entry appear as a new file in "git diff"and non-existent in "git diff —cached". This option could bereverted with —ita-visible-in-index. Both options areexperimental and could be removed in future.

For more detailed explanation on these common options, see alsogitdiffcore[7].

  • The id of a tree object to diff against.

  • —cached

  • do not consider the on-disk file at all

  • -m

  • By default, files recorded in the index but not checkedout are reported as deleted. This flag makesgit diff-index say that all non-checked-out files are upto date.

Raw output format

The raw output format from "git-diff-index", "git-diff-tree","git-diff-files" and "git diff —raw" are very similar.

These commands all compare two sets of things; what iscompared differs:

  • git-diff-index
  • compares the and the files on the filesystem.

  • git-diff-index —cached

  • compares the and the index.

  • git-diff-tree [-r] […​]

  • compares the trees named by the two arguments.

  • git-diff-files […​]

  • compares the index and the files on the filesystem.

The "git-diff-tree" command begins its output by printing the hash ofwhat is being compared. After that, all the commands print one outputline per changed file.

An output line is formatted this way:

  1. in-place edit :100644 100644 bcd1234 0123456 M file0
  2. copy-edit :100644 100644 abcd123 1234567 C68 file1 file2
  3. rename-edit :100644 100644 abcd123 1234567 R86 file1 file3
  4. create :000000 100644 0000000 1234567 A file4
  5. delete :100644 000000 1234567 0000000 D file5
  6. unmerged :000000 000000 0000000 0000000 U file6

That is, from the left to the right:

  • a colon.

  • mode for "src"; 000000 if creation or unmerged.

  • a space.

  • mode for "dst"; 000000 if deletion or unmerged.

  • a space.

  • sha1 for "src"; 0{40} if creation or unmerged.

  • a space.

  • sha1 for "dst"; 0{40} if creation, unmerged or "look at work tree".

  • a space.

  • status, followed by optional "score" number.

  • a tab or a NUL when -z option is used.

  • path for "src"

  • a tab or a NUL when -z option is used; only exists for C or R.

  • path for "dst"; only exists for C or R.

  • an LF or a NUL when -z option is used, to terminate the record.

Possible status letters are:

  • A: addition of a file

  • C: copy of a file into a new one

  • D: deletion of a file

  • M: modification of the contents or mode of a file

  • R: renaming of a file

  • T: change in the type of the file

  • U: file is unmerged (you must complete the merge before it canbe committed)

  • X: "unknown" change type (most probably a bug, please report it)

Status letters C and R are always followed by a score (denoting thepercentage of similarity between the source and target of the move orcopy). Status letter M may be followed by a score (denoting thepercentage of dissimilarity) for file rewrites.

<sha1> is shown as all 0’s if a file is new on the filesystemand it is out of sync with the index.

Example:

  1. :100644 100644 5be4a4a 0000000 M file.c

Without the -z option, pathnames with "unusual" characters arequoted as explained for the configuration variable core.quotePath(see git-config[1]). Using -z the filename is outputverbatim and the line is terminated by a NUL byte.

diff format for merges

"git-diff-tree", "git-diff-files" and "git-diff —raw"can take -c or —cc optionto generate diff output also for merge commits. The output differsfrom the format described above in the following way:

  • there is a colon for each parent

  • there are more "src" modes and "src" sha1

  • status is concatenated status characters for each parent

  • no optional "score" number

  • tab-separated pathname(s) of the file

For -c and —cc, only the destination or final path is shown evenif the file was renamed on any side of history. With—combined-all-paths, the name of the path in each parent is shownfollowed by the name of the path in the merge commit.

Examples for -c and —cc without —combined-all-paths:

  1. ::100644 100644 100644 fabadb8 cc95eb0 4866510 MM desc.c
  2. ::100755 100755 100755 52b7a2d 6d1ac04 d2ac7d7 RM bar.sh
  3. ::100644 100644 100644 e07d6c5 9042e82 ee91881 RR phooey.c

Examples when —combined-all-paths added to either -c or —cc:

  1. ::100644 100644 100644 fabadb8 cc95eb0 4866510 MM desc.c desc.c desc.c
  2. ::100755 100755 100755 52b7a2d 6d1ac04 d2ac7d7 RM foo.sh bar.sh bar.sh
  3. ::100644 100644 100644 e07d6c5 9042e82 ee91881 RR fooey.c fuey.c phooey.c

Note that combined diff lists only files which were modified fromall parents.

Generating patches with -p

When "git-diff-index", "git-diff-tree", or "git-diff-files" are runwith a -p option, "git diff" without the —raw option, or"git log" with the "-p" option, theydo not produce the output described above; instead they produce apatch file. You can customize the creation of such patches via theGIT_EXTERNAL_DIFF and the GIT_DIFF_OPTS environment variables.

What the -p option produces is slightly different from the traditionaldiff format:

  • It is preceded with a "git diff" header that looks like this:
  1. diff --git a/file1 b/file2

The a/ and b/ filenames are the same unless rename/copy isinvolved. Especially, even for a creation or a deletion,/dev/null is not used in place of the a/ or b/ filenames.

When rename/copy is involved, file1 and file2 show thename of the source file of the rename/copy and the name ofthe file that rename/copy produces, respectively.

  • It is followed by one or more extended header lines:
  1. old mode <mode>
  2. new mode <mode>
  3. deleted file mode <mode>
  4. new file mode <mode>
  5. copy from <path>
  6. copy to <path>
  7. rename from <path>
  8. rename to <path>
  9. similarity index <number>
  10. dissimilarity index <number>
  11. index <hash>..<hash> <mode>

File modes are printed as 6-digit octal numbers including the file typeand file permission bits.

Path names in extended headers do not include the a/ and b/ prefixes.

The similarity index is the percentage of unchanged lines, andthe dissimilarity index is the percentage of changed lines. Itis a rounded down integer, followed by a percent sign. Thesimilarity index value of 100% is thus reserved for two equalfiles, while 100% dissimilarity means that no line from the oldfile made it into the new one.

The index line includes the SHA-1 checksum before and after the change.The is included if the file mode does not change; otherwise,separate lines indicate the old and the new mode.

  • Pathnames with "unusual" characters are quoted as explained forthe configuration variable core.quotePath (seegit-config[1]).

  • All the file1 files in the output refer to files before thecommit, and all the file2 files refer to files after the commit.It is incorrect to apply each change to each file sequentially. Forexample, this patch will swap a and b:

  1. diff --git a/a b/b
  2. rename from a
  3. rename to b
  4. diff --git a/b b/a
  5. rename from b
  6. rename to a

combined diff format

Any diff-generating command can take the -c or —cc option toproduce a combined diff when showing a merge. This is the defaultformat when showing merges with git-diff[1] orgit-show[1]. Note also that you can give the -m option to anyof these commands to force generation of diffs with individual parentsof a merge.

A combined diff format looks like this:

  1. diff --combined describe.c
  2. index fabadb8,cc95eb0..4866510
  3. --- a/describe.c
  4. +++ b/describe.c
  5. @@@ -98,20 -98,12 +98,20 @@@
  6. return (a_date > b_date) ? -1 : (a_date == b_date) ? 0 : 1;
  7. }
  8.  
  9. - static void describe(char *arg)
  10. -static void describe(struct commit *cmit, int last_one)
  11. ++static void describe(char *arg, int last_one)
  12. {
  13. + unsigned char sha1[20];
  14. + struct commit *cmit;
  15. struct commit_list *list;
  16. static int initialized = 0;
  17. struct commit_name *n;
  18.  
  19. + if (get_sha1(arg, sha1) < 0)
  20. + usage(describe_usage);
  21. + cmit = lookup_commit_reference(sha1);
  22. + if (!cmit)
  23. + usage(describe_usage);
  24. +
  25. if (!initialized) {
  26. initialized = 1;
  27. for_each_ref(get_name);
  • It is preceded with a "git diff" header, that looks likethis (when -c option is used):
  1. diff --combined file

or like this (when —cc option is used):

  1. diff --cc file
  • It is followed by one or more extended header lines(this example shows a merge with two parents):
  1. index <hash>,<hash>..<hash>
  2. mode <mode>,<mode>..<mode>
  3. new file mode <mode>
  4. deleted file mode <mode>,<mode>

The mode <mode>,<mode>..<mode> line appears only if at least one ofthe is different from the rest. Extended headers withinformation about detected contents movement (renames andcopying detection) are designed to work with diff of two and are not used by combined diff format.

  • It is followed by two-line from-file/to-file header
  1. --- a/file
  2. +++ b/file

Similar to two-line header for traditional unified diffformat, /dev/null is used to signal created or deletedfiles.

However, if the —combined-all-paths option is provided, instead of atwo-line from-file/to-file you get a N+1 line from-file/to-file header,where N is the number of parents in the merge commit

  1. --- a/file
  2. --- a/file
  3. --- a/file
  4. +++ b/file

This extended format can be useful if rename or copy detection isactive, to allow you to see the original name of the file in differentparents.

  • Chunk header format is modified to prevent people fromaccidentally feeding it to patch -p1. Combined diff formatwas created for review of merge commit changes, and was notmeant for apply. The change is similar to the change in theextended index header:

  1. @@@ <from-file-range> <from-file-range> <to-file-range> @@@

There are (number of parents + 1) @ characters in the chunkheader for combined diff format.

Unlike the traditional unified diff format, which shows twofiles A and B with a single column that has - (minus — appears in A but removed in B), + (plus — missing in A butadded to B), or " " (space — unchanged) prefix, this formatcompares two or more files file1, file2,…​ with one file X, andshows how X differs from each of fileN. One column for each offileN is prepended to the output line to note how X’s line isdifferent from it.

A - character in the column N means that the line appears infileN but it does not appear in the result. A + characterin the column N means that the line appears in the result,and fileN does not have that line (in other words, the line wasadded, from the point of view of that parent).

In the above example output, the function signature was changedfrom both files (hence two - removals from both file1 andfile2, plus ++ to mean one line that was added does not appearin either file1 or file2). Also eight other lines are the samefrom file1 but do not appear in file2 (hence prefixed with +).

When shown by git diff-tree -c, it compares the parents of amerge commit with the merge result (i.e. file1..fileN are theparents). When shown by git diff-files -c, it compares thetwo unresolved merge parents with the working tree file(i.e. file1 is stage 2 aka "our version", file2 is stage 3 aka"their version").

other diff formats

The —summary option describes newly added, deleted, renamed andcopied files. The —stat option adds diffstat(1) graph to theoutput. These options can be combined with other options, such as-p, and are meant for human consumption.

When showing a change that involves a rename or a copy, —stat outputformats the pathnames compactly by combining common prefix and suffix ofthe pathnames. For example, a change that moves arch/i386/Makefile toarch/x86/Makefile while modifying 4 lines will be shown like this:

  1. arch/{i386 => x86}/Makefile | 4 +--

The —numstat option gives the diffstat(1) information but is designedfor easier machine consumption. An entry in —numstat output lookslike this:

  1. 1 2 README
  2. 3 1 arch/{i386 => x86}/Makefile

That is, from left to right:

  • the number of added lines;

  • a tab;

  • the number of deleted lines;

  • a tab;

  • pathname (possibly with rename/copy information);

  • a newline.

When -z output option is in effect, the output is formatted this way:

  1. 1 2 README NUL
  2. 3 1 NUL arch/i386/Makefile NUL arch/x86/Makefile NUL

That is:

  • the number of added lines;

  • a tab;

  • the number of deleted lines;

  • a tab;

  • a NUL (only exists if renamed/copied);

  • pathname in preimage;

  • a NUL (only exists if renamed/copied);

  • pathname in postimage (only exists if renamed/copied);

  • a NUL.

The extra NUL before the preimage path in renamed case is to allowscripts that read the output to tell if the current record being read isa single-path record or a rename/copy record without reading ahead.After reading added and deleted lines, reading up to NUL would yieldthe pathname, but if that is NUL, the record will show two paths.

OPERATING MODES

You can choose whether you want to trust the index file entirely(using the —cached flag) or ask the diff logic to show any filesthat don’t match the stat state as being "tentatively changed". Bothof these operations are very useful indeed.

CACHED MODE

If —cached is specified, it allows you to ask:

  1. show me the differences between HEAD and the current index
  2. contents (the ones I'd write using 'git write-tree')

For example, let’s say that you have worked on your working directory, updatedsome files in the index and are ready to commit. You want to see exactlywhat you are going to commit, without having to write a new treeobject and compare it that way, and to do that, you just do

  1. git diff-index --cached HEAD

Example: let’s say I had renamed commit.c to git-commit.c, and I haddone an update-index to make that effective in the index file.git diff-files wouldn’t show anything at all, since the index filematches my working directory. But doing a git diff-index does:

  1. torvalds@ppc970:~/git> git diff-index --cached HEAD
  2. -100644 blob 4161aecc6700a2eb579e842af0b7f22b98443f74 commit.c
  3. +100644 blob 4161aecc6700a2eb579e842af0b7f22b98443f74 git-commit.c

You can see easily that the above is a rename.

In fact, git diff-index —cached should always be entirely equivalent toactually doing a git write-tree and comparing that. Except this one is muchnicer for the case where you just want to check where you are.

So doing a git diff-index —cached is basically very useful when you areasking yourself "what have I already marked for being committed, andwhat’s the difference to a previous tree".

NON-CACHED MODE

The "non-cached" mode takes a different approach, and is potentiallythe more useful of the two in that what it does can’t be emulated witha git write-tree + git diff-tree. Thus that’s the default mode.The non-cached version asks the question:

  1. show me the differences between HEAD and the currently checked out
  2. tree - index contents _and_ files that aren't up to date

which is obviously a very useful question too, since that tells you whatyou could commit. Again, the output matches the _git diff-tree -r_output to a tee, but with a twist.

The twist is that if some file doesn’t match the index, we don’t havea backing store thing for it, and we use the magic "all-zero" sha1 toshow that. So let’s say that you have edited kernel/sched.c, buthave not actually done a git update-index on it yet - there is no"object" associated with the new state, and you get:

  1. torvalds@ppc970:~/v2.6/linux> git diff-index --abbrev HEAD
  2. :100644 100664 7476bb... 000000... kernel/sched.c

i.e., it shows that the tree has changed, and that kernel/sched.c isnot up to date and may contain new stuff. The all-zero sha1 means that toget the real diff, you need to look at the object in the working directorydirectly rather than do an object-to-object diff.

NoteAs with other commands of this type, git diff-index does notactually look at the contents of the file at all. So maybekernel/sched.c hasn’t actually changed, and it’s just that youtouched it. In either case, it’s a note that you need togit update-index it to make the index be in sync.
NoteYou can have a mixture of files show up as "has been updated"and "is still dirty in the working directory" together. You can alwaystell which file is in which state, since the "has been updated" onesshow a valid sha1, and the "not in sync with the index" ones willalways have the special all-zero sha1.

GIT

Part of the git[1] suite