mirror/ git
1
0
Fork 0
mirror of https://github.com/git/git.git synced 2024-05-09 19:36:10 +02:00
Code
git/Documentation/git-blame.txt
Barret Rhoden ae3f36dea1 blame: add the ability to ignore commits and their changes 
Commits that make formatting changes or function renames are often not
interesting when blaming a file.  A user may deem such a commit as 'not
interesting' and want to ignore and its changes it when assigning blame.

For example, say a file has the following git history / rev-list:

---O---A---X---B---C---D---Y---E---F

Commits X and Y both touch a particular line, and the other commits do
not:

X: "Take a third parameter"
-MyFunc(1, 2);
+MyFunc(1, 2, 3);

Y: "Remove camelcase"
-MyFunc(1, 2, 3);
+my_func(1, 2, 3);

git-blame will blame Y for the change.  I'd like to be able to ignore Y:
both the existence of the commit as well as any changes it made.  This
differs from -S rev-list, which specifies the list of commits to
process for the blame.  We would still process Y, but just don't let the
blame 'stick.'

This patch adds the ability for users to ignore a revision with
--ignore-rev=rev, which may be repeated.  They can specify a set of
files of full object names of revs, e.g. SHA-1 hashes, one per line.  A
single file may be specified with the blame.ignoreRevFile config option
or with --ignore-rev-file=file.  Both the config option and the command
line option may be repeated multiple times.  An empty file name "" will
clear the list of revs from previously processed files.  Config options
are processed before command line options.

For a typical use case, projects will maintain the file containing
revisions for commits that perform mass reformatting, and their users
have the option to ignore all of the commits in that file.

Additionally, a user can use the --ignore-rev option for one-off
investigation.  To go back to the example above, X was a substantive
change to the function, but not the change the user is interested in.
The user inspected X, but wanted to find the previous change to that
line - perhaps a commit that introduced that function call.

To make this work, we can't simply remove all ignored commits from the
rev-list.  We need to diff the changes introduced by Y so that we can
ignore them.  We let the blames get passed to Y, just like when
processing normally.  When Y is the target, we make sure that Y does not
*keep* any blames.  Any changes that Y is responsible for get passed to
its parent.  Note we make one pass through all of the scapegoats
(parents) to attempt to pass blame normally; we don't know if we *need*
to ignore the commit until we've checked all of the parents.

The blame_entry will get passed up the tree until we find a commit that
has a diff chunk that affects those lines.

One issue is that the ignored commit *did* make some change, and there is
no general solution to finding the line in the parent commit that
corresponds to a given line in the ignored commit.  That makes it hard
to attribute a particular line within an ignored commit's diff
correctly.

For example, the parent of an ignored commit has this, say at line 11:

commit-a 11) #include "a.h"
commit-b 12) #include "b.h"

Commit X, which we will ignore, swaps these lines:

commit-X 11) #include "b.h"
commit-X 12) #include "a.h"

We can pass that blame entry to the parent, but line 11 will be
attributed to commit A, even though "include b.h" came from commit B.
The blame mechanism will be looking at the parent's view of the file at
line number 11.

ignore_blame_entry() is set up to allow alternative algorithms for
guessing per-line blames.  Any line that is not attributed to the parent
will continue to be blamed on the ignored commit as if that commit was
not ignored.  Upcoming patches have the ability to detect these lines
and mark them in the blame output.

The existing algorithm is simple: blame each line on the corresponding
line in the parent's diff chunk.  Any lines beyond that stay with the
target.

For example, the parent of an ignored commit has this, say at line 11:

commit-a 11) void new_func_1(void *x, void *y);
commit-b 12) void new_func_2(void *x, void *y);
commit-c 13) some_line_c
commit-d 14) some_line_d

After a commit 'X', we have:

commit-X 11) void new_func_1(void *x,
commit-X 12)                 void *y);
commit-X 13) void new_func_2(void *x,
commit-X 14)                 void *y);
commit-c 15) some_line_c
commit-d 16) some_line_d

Commit X nets two additionally lines: 13 and 14.  The current
guess_line_blames() algorithm will not attribute these to the parent,
whose diff chunk is only two lines - not four.

When we ignore with the current algorithm, we get:

commit-a 11) void new_func_1(void *x,
commit-b 12)                 void *y);
commit-X 13) void new_func_2(void *x,
commit-X 14)                 void *y);
commit-c 15) some_line_c
commit-d 16) some_line_d

Note that line 12 was blamed on B, though B was the commit for
new_func_2(), not new_func_1().  Even when guess_line_blames() finds a
line in the parent, it may still be incorrect.

Signed-off-by: Barret Rhoden <brho@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2019-05-16 11:36:23 +09:00

237 lines
8.0 KiB
Plaintext
Raw Blame History

git-blame(1)
============
NAME
----
git-blame - Show what revision and author last modified each line of a file
SYNOPSIS
--------
[verse]
'git blame' [-c] [-b] [-l] [--root] [-t] [-f] [-n] [-s] [-e] [-p] [-w] [--incremental]
[-L <range>] [-S <revs-file>] [-M] [-C] [-C] [-C] [--since=<date>]
[--ignore-rev <rev>] [--ignore-revs-file <file>]
[--progress] [--abbrev=<n>] [<rev> | --contents <file> | --reverse <rev>..<rev>]
[--] <file>
DESCRIPTION
-----------
Annotates each line in the given file with information from the revision which
last modified the line. Optionally, start annotating from the given revision.
When specified one or more times, `-L` restricts annotation to the requested
lines.
The origin of lines is automatically followed across whole-file
renames (currently there is no option to turn the rename-following
off). To follow lines moved from one file to another, or to follow
lines that were copied and pasted from another file, etc., see the
`-C` and `-M` options.
The report does not tell you anything about lines which have been deleted or
replaced; you need to use a tool such as 'git diff' or the "pickaxe"
interface briefly mentioned in the following paragraph.
Apart from supporting file annotation, Git also supports searching the
development history for when a code snippet occurred in a change. This makes it
possible to track when a code snippet was added to a file, moved or copied
between files, and eventually deleted or replaced. It works by searching for
a text string in the diff. A small example of the pickaxe interface
that searches for `blame_usage`:
-----------------------------------------------------------------------------
$ git log --pretty=oneline -S'blame_usage'
5040f17eba15504bad66b14a645bddd9b015ebb7 blame -S <ancestry-file>
ea4c7f9bf69e781dd0cd88d2bccb2bf5cc15c9a7 git-blame: Make the output
-----------------------------------------------------------------------------
OPTIONS
-------
include::blame-options.txt[]
-c::
Use the same output mode as linkgit:git-annotate[1] (Default: off).
--score-debug::
Include debugging information related to the movement of
lines between files (see `-C`) and lines moved within a
file (see `-M`). The first number listed is the score.
This is the number of alphanumeric characters detected
as having been moved between or within files. This must be above
a certain threshold for 'git blame' to consider those lines
of code to have been moved.
-f::
--show-name::
Show the filename in the original commit. By default
the filename is shown if there is any line that came from a
file with a different name, due to rename detection.
-n::
--show-number::
Show the line number in the original commit (Default: off).
-s::
Suppress the author name and timestamp from the output.
-e::
--show-email::
Show the author email instead of author name (Default: off).
This can also be controlled via the `blame.showEmail` config
option.
-w::
Ignore whitespace when comparing the parent's version and
the child's to find where the lines came from.
--abbrev=<n>::
Instead of using the default 7+1 hexadecimal digits as the
abbreviated object name, use <n>+1 digits. Note that 1 column
is used for a caret to mark the boundary commit.
THE PORCELAIN FORMAT
--------------------
In this format, each line is output after a header; the
header at the minimum has the first line which has:
- 40-byte SHA-1 of the commit the line is attributed to;
- the line number of the line in the original file;
- the line number of the line in the final file;
- on a line that starts a group of lines from a different
commit than the previous one, the number of lines in this
group. On subsequent lines this field is absent.
This header line is followed by the following information
at least once for each commit:
- the author name ("author"), email ("author-mail"), time
("author-time"), and time zone ("author-tz"); similarly
for committer.
- the filename in the commit that the line is attributed to.
- the first line of the commit log message ("summary").
The contents of the actual line is output after the above
header, prefixed by a TAB. This is to allow adding more
header elements later.
The porcelain format generally suppresses commit information that has
already been seen. For example, two lines that are blamed to the same
commit will both be shown, but the details for that commit will be shown
only once. This is more efficient, but may require more state be kept by
the reader. The `--line-porcelain` option can be used to output full
commit information for each line, allowing simpler (but less efficient)
usage like:
# count the number of lines attributed to each author
git blame --line-porcelain file |
sed -n 's/^author //p' |
sort | uniq -c | sort -rn
SPECIFYING RANGES
-----------------
Unlike 'git blame' and 'git annotate' in older versions of git, the extent
of the annotation can be limited to both line ranges and revision
ranges. The `-L` option, which limits annotation to a range of lines, may be
specified multiple times.
When you are interested in finding the origin for
lines 40-60 for file `foo`, you can use the `-L` option like so
(they mean the same thing -- both ask for 21 lines starting at
line 40):
git blame -L 40,60 foo
git blame -L 40,+21 foo
Also you can use a regular expression to specify the line range:
git blame -L '/^sub hello {/,/^}$/' foo
which limits the annotation to the body of the `hello` subroutine.
When you are not interested in changes older than version
v2.6.18, or changes older than 3 weeks, you can use revision
range specifiers similar to 'git rev-list':
git blame v2.6.18.. -- foo
git blame --since=3.weeks -- foo
When revision range specifiers are used to limit the annotation,
lines that have not changed since the range boundary (either the
commit v2.6.18 or the most recent commit that is more than 3
weeks old in the above example) are blamed for that range
boundary commit.
A particularly useful way is to see if an added file has lines
created by copy-and-paste from existing files. Sometimes this
indicates that the developer was being sloppy and did not
refactor the code properly. You can first find the commit that
introduced the file with:
git log --diff-filter=A --pretty=short -- foo
and then annotate the change between the commit and its
parents, using `commit^!` notation:
git blame -C -C -f $commit^! -- foo
INCREMENTAL OUTPUT
------------------
When called with `--incremental` option, the command outputs the
result as it is built. The output generally will talk about
lines touched by more recent commits first (i.e. the lines will
be annotated out of order) and is meant to be used by
interactive viewers.
The output format is similar to the Porcelain format, but it
does not contain the actual lines from the file that is being
annotated.
. Each blame entry always starts with a line of:
<40-byte hex sha1> <sourceline> <resultline> <num_lines>
+
Line numbers count from 1.
. The first time that a commit shows up in the stream, it has various
other information about it printed out with a one-word tag at the
beginning of each line describing the extra commit information (author,
email, committer, dates, summary, etc.).
. Unlike the Porcelain format, the filename information is always
given and terminates the entry:
"filename" <whitespace-quoted-filename-goes-here>
+
and thus it is really quite easy to parse for some line- and word-oriented
parser (which should be quite natural for most scripting languages).
+
[NOTE]
For people who do parsing: to make it more robust, just ignore any
lines between the first and last one ("<sha1>" and "filename" lines)
where you do not recognize the tag words (or care about that particular
one) at the beginning of the "extended information" lines. That way, if
there is ever added information (like the commit encoding or extended
commit commentary), a blame viewer will not care.
MAPPING AUTHORS
---------------
include::mailmap.txt[]
SEE ALSO
--------
linkgit:git-annotate[1]
GIT
---
Part of the linkgit:git[1] suite
View Git Blame Copy Permalink