1GIT-BLAME(1) Git Manual GIT-BLAME(1)
2
6 git-blame - Show what revision and author last modified each line of a
7 file
8
10 git blame [-c] [-b] [-l] [--root] [-t] [-f] [-n] [-s] [-e] [-p] [-w] [--incremental] [-L n,m]
11 [-S <revs-file>] [-M] [-C] [-C] [-C] [--since=<date>] [--abbrev=<n>]
12 [<rev> | --contents <file> | --reverse <rev>] [--] <file>
13
16 Annotates each line in the given file with information from the
17 revision which last modified the line. Optionally, start annotating
18 from the given revision.
19 The command can also limit the range of lines annotated.
21 The origin of lines is automatically followed across whole-file renames
23 (currently there is no option to turn the rename-following off). To
24 follow lines moved from one file to another, or to follow lines that
25 were copied and pasted from another file, etc., see the -C and -M
26 options.
27 The report does not tell you anything about lines which have been
29 deleted or replaced; you need to use a tool such as git diff or the
30 "pickaxe" interface briefly mentioned in the following paragraph.
31 Apart from supporting file annotation, Git also supports searching the
33 development history for when a code snippet occurred in a change. This
34 makes it possible to track when a code snippet was added to a file,
35 moved or copied between files, and eventually deleted or replaced. It
36 works by searching for a text string in the diff. A small example:
37 $ git log --pretty=oneline -S'blame_usage'
39 5040f17eba15504bad66b14a645bddd9b015ebb7 blame -S <ancestry-file>
40 ea4c7f9bf69e781dd0cd88d2bccb2bf5cc15c9a7 git-blame: Make the output
41
44 -b
45 Show blank SHA-1 for boundary commits. This can also be controlled
46 via the blame.blankboundary config option.
47 --root
49 Do not treat root commits as boundaries. This can also be
50 controlled via the blame.showroot config option.
51 --show-stats
53 Include additional statistics at the end of blame output.
54 -L <start>,<end>
56 Annotate only the given line range. <start> and <end> can take one
57 of these forms:
58 · number
60 If <start> or <end> is a number, it specifies an absolute line
62 number (lines count from 1).
63 · /regex/
65 This form will use the first line matching the given POSIX
67 regex. If <end> is a regex, it will search starting at the line
68 given by <start>.
69 · +offset or -offset
71 This is only valid for <end> and will specify a number of lines
73 before or after the line given by <start>.
74 -l
76 Show long rev (Default: off).
77 -t
79 Show raw timestamp (Default: off).
80 -S <revs-file>
82 Use revisions from revs-file instead of calling git-rev-list(1).
83 --reverse
85 Walk history forward instead of backward. Instead of showing the
86 revision in which a line appeared, this shows the last revision in
87 which a line has existed. This requires a range of revision like
88 START..END where the path to blame exists in START.
89 -p, --porcelain
91 Show in a format designed for machine consumption.
92 --line-porcelain
94 Show the porcelain format, but output commit information for each
95 line, not just the first time a commit is referenced. Implies
96 --porcelain.
97 --incremental
99 Show the result incrementally in a format designed for machine
100 consumption.
101 --encoding=<encoding>
103 Specifies the encoding used to output author names and commit
104 summaries. Setting it to none makes blame output unconverted data.
105 For more information see the discussion about encoding in the git-
106 log(1) manual page.
107 --contents <file>
109 When <rev> is not specified, the command annotates the changes
110 starting backwards from the working tree copy. This flag makes the
111 command pretend as if the working tree copy has the contents of the
112 named file (specify - to make the command read from the standard
113 input).
114 --date <format>
116 The value is one of the following alternatives:
117 {relative,local,default,iso,rfc,short}. If --date is not provided,
118 the value of the blame.date config variable is used. If the
119 blame.date config variable is also not set, the iso format is used.
120 For more information, See the discussion of the --date option at
121 git-log(1).
122 -M|<num>|
124 Detect moved or copied lines within a file. When a commit moves or
125 copies a block of lines (e.g. the original file has A and then B,
126 and the commit changes it to B and then A), the traditional blame
127 algorithm notices only half of the movement and typically blames
128 the lines that were moved up (i.e. B) to the parent and assigns
129 blame to the lines that were moved down (i.e. A) to the child
130 commit. With this option, both groups of lines are blamed on the
131 parent by running extra passes of inspection.
132 <num> is optional but it is the lower bound on the number of
134 alphanumeric characters that Git must detect as moving/copying
135 within a file for it to associate those lines with the parent
136 commit. The default value is 20.
137 -C|<num>|
139 In addition to -M, detect lines moved or copied from other files
140 that were modified in the same commit. This is useful when you
141 reorganize your program and move code around across files. When
142 this option is given twice, the command additionally looks for
143 copies from other files in the commit that creates the file. When
144 this option is given three times, the command additionally looks
145 for copies from other files in any commit.
146 <num> is optional but it is the lower bound on the number of
148 alphanumeric characters that Git must detect as moving/copying
149 between files for it to associate those lines with the parent
150 commit. And the default value is 40. If there are more than one -C
151 options given, the <num> argument of the last -C will take effect.
152 -h
154 Show help message.
155 -c
157 Use the same output mode as git-annotate(1) (Default: off).
158 --score-debug
160 Include debugging information related to the movement of lines
161 between files (see -C) and lines moved within a file (see -M). The
162 first number listed is the score. This is the number of
163 alphanumeric characters detected as having been moved between or
164 within files. This must be above a certain threshold for git blame
165 to consider those lines of code to have been moved.
166 -f, --show-name
168 Show the filename in the original commit. By default the filename
169 is shown if there is any line that came from a file with a
170 different name, due to rename detection.
171 -n, --show-number
173 Show the line number in the original commit (Default: off).
174 -s
176 Suppress the author name and timestamp from the output.
177 -e, --show-email
179 Show the author email instead of author name (Default: off).
180 -w
182 Ignore whitespace when comparing the parent’s version and the
183 child’s to find where the lines came from.
184 --abbrev=<n>
186 Instead of using the default 7+1 hexadecimal digits as the
187 abbreviated object name, use <n>+1 digits. Note that 1 column is
188 used for a caret to mark the boundary commit.
189
191 In this format, each line is output after a header; the header at the
192 minimum has the first line which has:
193 · 40-byte SHA-1 of the commit the line is attributed to;
195 · the line number of the line in the original file;
197 · the line number of the line in the final file;
199 · on a line that starts a group of lines from a different commit than
201 the previous one, the number of lines in this group. On subsequent
202 lines this field is absent.
203 This header line is followed by the following information at least once
205 for each commit:
206 · the author name ("author"), email ("author-mail"), time
208 ("author-time"), and timezone ("author-tz"); similarly for
209 committer.
210 · the filename in the commit that the line is attributed to.
212 · the first line of the commit log message ("summary").
214 The contents of the actual line is output after the above header,
216 prefixed by a TAB. This is to allow adding more header elements later.
217 The porcelain format generally suppresses commit information that has
219 already been seen. For example, two lines that are blamed to the same
220 commit will both be shown, but the details for that commit will be
221 shown only once. This is more efficient, but may require more state be
222 kept by the reader. The --line-porcelain option can be used to output
223 full commit information for each line, allowing simpler (but less
224 efficient) usage like:
225 # count the number of lines attributed to each author
227 git blame --line-porcelain file |
228 sed -n 's/^author //p' |
229 sort | uniq -c | sort -rn
230
232 Unlike git blame and git annotate in older versions of git, the extent
233 of the annotation can be limited to both line ranges and revision
234 ranges. When you are interested in finding the origin for lines 40-60
235 for file foo, you can use the -L option like so (they mean the same
236 thing — both ask for 21 lines starting at line 40):
237 git blame -L 40,60 foo
239 git blame -L 40,+21 foo
240 Also you can use a regular expression to specify the line range:
242 git blame -L '/^sub hello {/,/^}$/' foo
244 which limits the annotation to the body of the hello subroutine.
246 When you are not interested in changes older than version v2.6.18, or
248 changes older than 3 weeks, you can use revision range specifiers
249 similar to git rev-list:
250 git blame v2.6.18.. -- foo
252 git blame --since=3.weeks -- foo
253 When revision range specifiers are used to limit the annotation, lines
255 that have not changed since the range boundary (either the commit
256 v2.6.18 or the most recent commit that is more than 3 weeks old in the
257 above example) are blamed for that range boundary commit.
258 A particularly useful way is to see if an added file has lines created
260 by copy-and-paste from existing files. Sometimes this indicates that
261 the developer was being sloppy and did not refactor the code properly.
262 You can first find the commit that introduced the file with:
263 git log --diff-filter=A --pretty=short -- foo
265 and then annotate the change between the commit and its parents, using
267 commit^! notation:
268 git blame -C -C -f $commit^! -- foo
270
272 When called with --incremental option, the command outputs the result
273 as it is built. The output generally will talk about lines touched by
274 more recent commits first (i.e. the lines will be annotated out of
275 order) and is meant to be used by interactive viewers.
276 The output format is similar to the Porcelain format, but it does not
278 contain the actual lines from the file that is being annotated.
279 1. Each blame entry always starts with a line of:
281 <40-byte hex sha1> <sourceline> <resultline> <num_lines>
283 Line numbers count from 1.
285 2. The first time that a commit shows up in the stream, it has various
287 other information about it printed out with a one-word tag at the
288 beginning of each line describing the extra commit information
289 (author, email, committer, dates, summary, etc.).
290 3. Unlike the Porcelain format, the filename information is always
292 given and terminates the entry:
293 "filename" <whitespace-quoted-filename-goes-here>
295 and thus it is really quite easy to parse for some line- and
297 word-oriented parser (which should be quite natural for most
298 scripting languages).
299 Note
301 For people who do parsing: to make it more robust, just ignore
302 any lines between the first and last one ("<sha1>" and
303 "filename" lines) where you do not recognize the tag words (or
304 care about that particular one) at the beginning of the
305 "extended information" lines. That way, if there is ever added
306 information (like the commit encoding or extended commit
307 commentary), a blame viewer will not care.
308
310 If the file .mailmap exists at the toplevel of the repository, or at
311 the location pointed to by the mailmap.file or mailmap.blob
312 configuration options, it is used to map author and committer names and
313 email addresses to canonical real names and email addresses.
314 In the simple form, each line in the file consists of the canonical
316 real name of an author, whitespace, and an email address used in the
317 commit (enclosed by < and >) to map to the name. For example:
318 Proper Name <commit@email.xx>
320 The more complex forms are:
322 <proper@email.xx> <commit@email.xx>
324 which allows mailmap to replace only the email part of a commit, and:
326 Proper Name <proper@email.xx> <commit@email.xx>
328 which allows mailmap to replace both the name and the email of a commit
330 matching the specified commit email address, and:
331 Proper Name <proper@email.xx> Commit Name <commit@email.xx>
333 which allows mailmap to replace both the name and the email of a commit
335 matching both the specified commit name and email address.
336 Example 1: Your history contains commits by two authors, Jane and Joe,
338 whose names appear in the repository under several forms:
339 Joe Developer <joe@example.com>
341 Joe R. Developer <joe@example.com>
342 Jane Doe <jane@example.com>
343 Jane Doe <jane@laptop.(none)>
344 Jane D. <jane@desktop.(none)>
345 Now suppose that Joe wants his middle name initial used, and Jane
348 prefers her family name fully spelled out. A proper .mailmap file would
349 look like:
350 Jane Doe <jane@desktop.(none)>
352 Joe R. Developer <joe@example.com>
353 Note how there is no need for an entry for <jane@laptop.(none)>,
356 because the real name of that author is already correct.
357 Example 2: Your repository contains commits from the following authors:
359 nick1 <bugs@company.xx>
361 nick2 <bugs@company.xx>
362 nick2 <nick2@company.xx>
363 santa <me@company.xx>
364 claus <me@company.xx>
365 CTO <cto@coompany.xx>
366 Then you might want a .mailmap file that looks like:
369 <cto@company.xx> <cto@coompany.xx>
371 Some Dude <some@dude.xx> nick1 <bugs@company.xx>
372 Other Author <other@author.xx> nick2 <bugs@company.xx>
373 Other Author <other@author.xx> <nick2@company.xx>
374 Santa Claus <santa.claus@northpole.xx> <me@company.xx>
375 Use hash # for comments that are either on their own line, or after the
378 email address.
379
381 git-annotate(1)
382
384 Part of the git(1) suite
385Git 1.8.3.1 11/19/2018 GIT-BLAME(1)