summaryrefslogtreecommitdiff
path: root/Documentation/diffcore.txt
blob: 4ade0140cc12af5ec1eb3af6f4abe6a9cc9aca91 (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
Tweaking diff output
====================
June 2005
 
 
Introduction
------------
 
The diff commands git-diff-cache, git-diff-files, and
git-diff-tree can be told to manipulate differences they find
in unconventional ways before showing diff(1) output.  The
manipulation is collectively called "diffcore transformation".
This short note describes what they are and how to use them to
produce diff outputs that are easier to understand than the
conventional kind.
 
 
The chain of operation
----------------------
 
The git-diff-* family works by first comparing two sets of
files:
 
 - git-diff-cache compares contents of a "tree" object and the
   working directory (when --cached flag is not used) or a
   "tree" object and the index file (when --cached flag is
   used);
 
 - git-diff-files compares contents of the index file and the
   working directory;
 
 - git-diff-tree compares contents of two "tree" objects.
 
In all of these cases, the commands themselves compare
corresponding paths in the two sets of files.  The result of
comparison is passed from these commands to what is internally
called "diffcore", in a format similar to what is output when
the -p option is not used.  E.g.
 
    in-place edit  :100644 100644 bcd1234... 0123456... M file0
    create         :000000 100644 0000000... 1234567... N file4
    delete         :100644 000000 1234567... 0000000... D file5
    unmerged       :000000 000000 0000000... 0000000... U file6
 
The diffcore mechanism is fed a list of such comparison results
(each of which is called "filepair", although at this point each
of them talks about a single file), and transforms such a list
into another list.  There are currently 7 such transformations:
 
 - diffcore-pathspec
 - diffcore-break
 - diffcore-rename
 - diffcore-merge-broken
 - diffcore-pickaxe
 - diffcore-order
 
These are applied in sequence.  The set of filepairs git-diff-*
commands find are used as the input to diffcore-pathspec, and
the output from diffcore-pathspec is used as the input to the
next transformation.  The final result is then passed to the
output routine and generates either diff-raw format (see Output
format sections of the manual for git-diff-* commands) or
diff-patch format.
 
 
diffcore-pathspec
-----------------
 
The first transformation in the chain is diffcore-pathspec, and
is controlled by giving the pathname parameters to the
git-diff-* commands on the command line.  The pathspec is used
to limit the world diff operates in.  It removes the filepairs
outside the specified set of pathnames.
 
Implementation note.  For performance reasons, git-diff-tree
uses the pathname parameters on the command line to cull set of
filepairs it feeds the diffcore mechanism itself, and does not
use diffcore-pathspec, but the end result is the same.
 
 
diffcore-break
--------------
 
The second transformation in the chain is diffcore-break, and is
controlled by the -B option to the git-diff-* commands.  This is
used to detect a filepair that represents "complete rewrite" and
break such filepair into two filepairs that represent delete and
create.  E.g.  If the input contained this filepair:
 
    :100644 100644 bcd1234... 0123456... M file0
 
and if it detects that the file "file0" is completely rewritten,
it changes it to:
 
    :100644 000000 bcd1234... 0000000... D file0
    :000000 100644 0000000... 0123456... N file0
 
For the purpose of breaking a filepair, diffcore-break examines
the extent of changes between the contents of the files before
and after modification (i.e. the contents that have "bcd1234..."
and "0123456..." as their SHA1 content ID, in the above
example).  The amount of deletion of original contents and
insertion of new material are added together, and if it exceeds
the "break score", the filepair is broken into two.  The break
score defaults to 50% of the size of the smaller of the original
and the result (i.e. if the edit shrinks the file, the size of
the result is used; if the edit lengthens the file, the size of
the original is used), and can be customized by giving a number
after "-B" option (e.g. "-B75" to tell it to use 75%).
 
 
diffcore-rename
---------------
 
This transformation is used to detect renames and copies, and is
controlled by the -M option (to detect renames) and the -C option
(to detect copies as well) to the git-diff-* commands.  If the
input contained these filepairs:
 
    :100644 000000 0123456... 0000000... D fileX
    :000000 100644 0000000... 0123456... N file0
 
and the contents of the deleted file fileX is similar enough to
the contents of the created file file0, then rename detection
merges these filepairs and creates:
 
    :100644 100644 0123456... 0123456... R100 fileX file0
 
When the "-C" option is used, the original contents of modified
files and contents of unchanged files are considered as
candidates of the source files in rename/copy operation, in
addition to the deleted files.  If the input were like these
filepairs, that talk about a modified file fileY and a newly
created file file0:
 
    :100644 100644 0123456... 1234567... M fileY
    :000000 100644 0000000... 0123456... N file0
 
the original contents of fileY and the resulting contents of
file0 are compared, and if they are similar enough, they are
changed to:
 
    :100644 100644 0123456... 1234567... M fileY
    :100644 100644 0123456... 0123456... C100 fileY file0
 
In both rename and copy detection, the same "extent of changes"
algorithm used in diffcore-break is used to determine if two
files are "similar enough", and can be customized to use
similarity score different from the default 50% by giving a
number after "-M" or "-C" option (e.g. "-M8" to tell it to use
8/10 = 80%).
 
Note.  When the "-C" option is used, git-diff-cache and
git-diff-file commands feed not just modified filepairs but
unmodified ones to diffcore mechanism as well.  This lets the
copy detector consider unmodified files as copy source
candidates at the expense of making it slower.  Currently
git-diff-tree does not feed unmodified filepairs even when the
"-C" option is used, so it can detect copies only if the file
that was copied happened to have been modified in the same
changeset.
 
 
diffcore-merge-broken
---------------------
 
This transformation is used to merge filepairs broken by
diffcore-break, and were not transformed into rename/copy by
diffcore-rename, back into a single modification.  This always
runs when diffcore-break is used.
 
For the purpose of merging broken filepairs back, it uses a
different "extent of changes" computation from the ones used by
diffcore-break and diffcore-rename.  It counts only the deletion
from the original, and does not count insertion.  If you removed
only 10 lines from a 100-line document, even if you added 910
new lines to make a new 1000-line document, you did not do a
complete rewrite.  diffcore-break breaks such a case in order to
help diffcore-rename to consider such filepairs as candidate of
rename/copy detection, but if filepairs broken that way were not
matched with other filepairs to create rename/copy, then this
transformation merges them back into the original
"modification".
 
The "extent of changes" parameter can be tweaked from the
default 80% (that is, unless more than 80% of the original
material is deleted, the broken pairs are merged back into a
single modification) by giving a second number to -B option,
like these:
 
	-B50/60 (give 50% "break score" to diffcore-break, use
                 60% for diffcore-merge-broken).
	-B/60   (the same as above, since diffcore-break defautls to
		 50%).
 
 
diffcore-pickaxe
----------------
 
This transformation is used to find filepairs that represent
changes that touch a specified string, and is controlled by the
-S option and the --pickaxe-all option to the git-diff-*
commands.
 
When diffcore-pickaxe is in use, it checks if there are
filepairs whose "original" side has the specified string and
whose "result" side does not.  Such a filepair represents "the
string appeared in this changeset".  It also checks for the
opposite case that loses the specified string.
 
When --pickaxe-all is not in effect, diffcore-pickaxe leaves
only such filepairs that touches the specified string in its
output.  When --pickaxe-all is used, diffcore-pickaxe leaves all
filepairs intact if there is such a filepair, or makes the
output empty otherwise.  The latter behaviour is designed to
make reviewing of the changes in the context of the whole
changeset easier.
 
 
diffcore-order
--------------
 
This is used to reorder the filepairs according to the user's
(or project's) taste, and is controlled by the -O option to the
git-diff-* commands.
 
This takes a text file each of whose line is a shell glob
pattern.  Filepairs that match a glob pattern on an earlier line
in the file are output before ones that match a later line, and
filepairs that do not match any glob pattern are output last.
 
As an example, typical orderfile for the core GIT probably
should look like this:
 
    README
    Makefile
    Documentation
    *.h
    *.c
    t