Documentation / git-blame.txton commit Merge branch 'jk/make-coccicheck-detect-errors' (876eb61)
   1git-blame(1)
   2============
   3
   4NAME
   5----
   6git-blame - Show what revision and author last modified each line of a file
   7
   8SYNOPSIS
   9--------
  10[verse]
  11'git blame' [-c] [-b] [-l] [--root] [-t] [-f] [-n] [-s] [-e] [-p] [-w] [--incremental]
  12            [-L <range>] [-S <revs-file>] [-M] [-C] [-C] [-C] [--since=<date>]
  13            [--progress] [--abbrev=<n>] [<rev> | --contents <file> | --reverse <rev>..<rev>]
  14            [--] <file>
  15
  16DESCRIPTION
  17-----------
  18
  19Annotates each line in the given file with information from the revision which
  20last modified the line. Optionally, start annotating from the given revision.
  21
  22When specified one or more times, `-L` restricts annotation to the requested
  23lines.
  24
  25The origin of lines is automatically followed across whole-file
  26renames (currently there is no option to turn the rename-following
  27off). To follow lines moved from one file to another, or to follow
  28lines that were copied and pasted from another file, etc., see the
  29`-C` and `-M` options.
  30
  31The report does not tell you anything about lines which have been deleted or
  32replaced; you need to use a tool such as 'git diff' or the "pickaxe"
  33interface briefly mentioned in the following paragraph.
  34
  35Apart from supporting file annotation, Git also supports searching the
  36development history for when a code snippet occurred in a change. This makes it
  37possible to track when a code snippet was added to a file, moved or copied
  38between files, and eventually deleted or replaced. It works by searching for
  39a text string in the diff. A small example of the pickaxe interface
  40that searches for `blame_usage`:
  41
  42-----------------------------------------------------------------------------
  43$ git log --pretty=oneline -S'blame_usage'
  445040f17eba15504bad66b14a645bddd9b015ebb7 blame -S <ancestry-file>
  45ea4c7f9bf69e781dd0cd88d2bccb2bf5cc15c9a7 git-blame: Make the output
  46-----------------------------------------------------------------------------
  47
  48OPTIONS
  49-------
  50include::blame-options.txt[]
  51
  52-c::
  53        Use the same output mode as linkgit:git-annotate[1] (Default: off).
  54
  55--score-debug::
  56        Include debugging information related to the movement of
  57        lines between files (see `-C`) and lines moved within a
  58        file (see `-M`).  The first number listed is the score.
  59        This is the number of alphanumeric characters detected
  60        as having been moved between or within files.  This must be above
  61        a certain threshold for 'git blame' to consider those lines
  62        of code to have been moved.
  63
  64-f::
  65--show-name::
  66        Show the filename in the original commit.  By default
  67        the filename is shown if there is any line that came from a
  68        file with a different name, due to rename detection.
  69
  70-n::
  71--show-number::
  72        Show the line number in the original commit (Default: off).
  73
  74-s::
  75        Suppress the author name and timestamp from the output.
  76
  77-e::
  78--show-email::
  79        Show the author email instead of author name (Default: off).
  80        This can also be controlled via the `blame.showEmail` config
  81        option.
  82
  83-w::
  84        Ignore whitespace when comparing the parent's version and
  85        the child's to find where the lines came from.
  86
  87--abbrev=<n>::
  88        Instead of using the default 7+1 hexadecimal digits as the
  89        abbreviated object name, use <n>+1 digits. Note that 1 column
  90        is used for a caret to mark the boundary commit.
  91
  92include::diff-heuristic-options.txt[]
  93
  94
  95THE PORCELAIN FORMAT
  96--------------------
  97
  98In this format, each line is output after a header; the
  99header at the minimum has the first line which has:
 100
 101- 40-byte SHA-1 of the commit the line is attributed to;
 102- the line number of the line in the original file;
 103- the line number of the line in the final file;
 104- on a line that starts a group of lines from a different
 105  commit than the previous one, the number of lines in this
 106  group.  On subsequent lines this field is absent.
 107
 108This header line is followed by the following information
 109at least once for each commit:
 110
 111- the author name ("author"), email ("author-mail"), time
 112  ("author-time"), and time zone ("author-tz"); similarly
 113  for committer.
 114- the filename in the commit that the line is attributed to.
 115- the first line of the commit log message ("summary").
 116
 117The contents of the actual line is output after the above
 118header, prefixed by a TAB. This is to allow adding more
 119header elements later.
 120
 121The porcelain format generally suppresses commit information that has
 122already been seen. For example, two lines that are blamed to the same
 123commit will both be shown, but the details for that commit will be shown
 124only once. This is more efficient, but may require more state be kept by
 125the reader. The `--line-porcelain` option can be used to output full
 126commit information for each line, allowing simpler (but less efficient)
 127usage like:
 128
 129        # count the number of lines attributed to each author
 130        git blame --line-porcelain file |
 131        sed -n 's/^author //p' |
 132        sort | uniq -c | sort -rn
 133
 134
 135SPECIFYING RANGES
 136-----------------
 137
 138Unlike 'git blame' and 'git annotate' in older versions of git, the extent
 139of the annotation can be limited to both line ranges and revision
 140ranges. The `-L` option, which limits annotation to a range of lines, may be
 141specified multiple times.
 142
 143When you are interested in finding the origin for
 144lines 40-60 for file `foo`, you can use the `-L` option like so
 145(they mean the same thing -- both ask for 21 lines starting at
 146line 40):
 147
 148        git blame -L 40,60 foo
 149        git blame -L 40,+21 foo
 150
 151Also you can use a regular expression to specify the line range:
 152
 153        git blame -L '/^sub hello {/,/^}$/' foo
 154
 155which limits the annotation to the body of the `hello` subroutine.
 156
 157When you are not interested in changes older than version
 158v2.6.18, or changes older than 3 weeks, you can use revision
 159range specifiers  similar to 'git rev-list':
 160
 161        git blame v2.6.18.. -- foo
 162        git blame --since=3.weeks -- foo
 163
 164When revision range specifiers are used to limit the annotation,
 165lines that have not changed since the range boundary (either the
 166commit v2.6.18 or the most recent commit that is more than 3
 167weeks old in the above example) are blamed for that range
 168boundary commit.
 169
 170A particularly useful way is to see if an added file has lines
 171created by copy-and-paste from existing files.  Sometimes this
 172indicates that the developer was being sloppy and did not
 173refactor the code properly.  You can first find the commit that
 174introduced the file with:
 175
 176        git log --diff-filter=A --pretty=short -- foo
 177
 178and then annotate the change between the commit and its
 179parents, using `commit^!` notation:
 180
 181        git blame -C -C -f $commit^! -- foo
 182
 183
 184INCREMENTAL OUTPUT
 185------------------
 186
 187When called with `--incremental` option, the command outputs the
 188result as it is built.  The output generally will talk about
 189lines touched by more recent commits first (i.e. the lines will
 190be annotated out of order) and is meant to be used by
 191interactive viewers.
 192
 193The output format is similar to the Porcelain format, but it
 194does not contain the actual lines from the file that is being
 195annotated.
 196
 197. Each blame entry always starts with a line of:
 198
 199        <40-byte hex sha1> <sourceline> <resultline> <num_lines>
 200+
 201Line numbers count from 1.
 202
 203. The first time that a commit shows up in the stream, it has various
 204  other information about it printed out with a one-word tag at the
 205  beginning of each line describing the extra commit information (author,
 206  email, committer, dates, summary, etc.).
 207
 208. Unlike the Porcelain format, the filename information is always
 209  given and terminates the entry:
 210
 211        "filename" <whitespace-quoted-filename-goes-here>
 212+
 213and thus it is really quite easy to parse for some line- and word-oriented
 214parser (which should be quite natural for most scripting languages).
 215+
 216[NOTE]
 217For people who do parsing: to make it more robust, just ignore any
 218lines between the first and last one ("<sha1>" and "filename" lines)
 219where you do not recognize the tag words (or care about that particular
 220one) at the beginning of the "extended information" lines. That way, if
 221there is ever added information (like the commit encoding or extended
 222commit commentary), a blame viewer will not care.
 223
 224
 225MAPPING AUTHORS
 226---------------
 227
 228include::mailmap.txt[]
 229
 230
 231SEE ALSO
 232--------
 233linkgit:git-annotate[1]
 234
 235GIT
 236---
 237Part of the linkgit:git[1] suite