]> git.ipfire.org Git - thirdparty/git.git/blob - Documentation/gitdiffcore.txt
diffcore-rename: avoid doing basename comparisons for irrelevant sources
[thirdparty/git.git] / Documentation / gitdiffcore.txt
1 gitdiffcore(7)
2 ==============
3
4 NAME
5 ----
6 gitdiffcore - Tweaking diff output
7
8 SYNOPSIS
9 --------
10 [verse]
11 'git diff' *
12
13 DESCRIPTION
14 -----------
15
16 The diff commands 'git diff-index', 'git diff-files', and 'git diff-tree'
17 can be told to manipulate differences they find in
18 unconventional ways before showing 'diff' output. The manipulation
19 is collectively called "diffcore transformation". This short note
20 describes what they are and how to use them to produce 'diff' output
21 that is easier to understand than the conventional kind.
22
23
24 The chain of operation
25 ----------------------
26
27 The 'git diff-{asterisk}' family works by first comparing two sets of
28 files:
29
30 - 'git diff-index' compares contents of a "tree" object and the
31 working directory (when `--cached` flag is not used) or a
32 "tree" object and the index file (when `--cached` flag is
33 used);
34
35 - 'git diff-files' compares contents of the index file and the
36 working directory;
37
38 - 'git diff-tree' compares contents of two "tree" objects;
39
40 In all of these cases, the commands themselves first optionally limit
41 the two sets of files by any pathspecs given on their command-lines,
42 and compare corresponding paths in the two resulting sets of files.
43
44 The pathspecs are used to limit the world diff operates in. They remove
45 the filepairs outside the specified sets of pathnames. E.g. If the
46 input set of filepairs included:
47
48 ------------------------------------------------
49 :100644 100644 bcd1234... 0123456... M junkfile
50 ------------------------------------------------
51
52 but the command invocation was `git diff-files myfile`, then the
53 junkfile entry would be removed from the list because only "myfile"
54 is under consideration.
55
56 The result of comparison is passed from these commands to what is
57 internally called "diffcore", in a format similar to what is output
58 when the -p option is not used. E.g.
59
60 ------------------------------------------------
61 in-place edit :100644 100644 bcd1234... 0123456... M file0
62 create :000000 100644 0000000... 1234567... A file4
63 delete :100644 000000 1234567... 0000000... D file5
64 unmerged :000000 000000 0000000... 0000000... U file6
65 ------------------------------------------------
66
67 The diffcore mechanism is fed a list of such comparison results
68 (each of which is called "filepair", although at this point each
69 of them talks about a single file), and transforms such a list
70 into another list. There are currently 5 such transformations:
71
72 - diffcore-break
73 - diffcore-rename
74 - diffcore-merge-broken
75 - diffcore-pickaxe
76 - diffcore-order
77
78 These are applied in sequence. The set of filepairs 'git diff-{asterisk}'
79 commands find are used as the input to diffcore-break, and
80 the output from diffcore-break is used as the input to the
81 next transformation. The final result is then passed to the
82 output routine and generates either diff-raw format (see Output
83 format sections of the manual for 'git diff-{asterisk}' commands) or
84 diff-patch format.
85
86
87 diffcore-break: For Splitting Up Complete Rewrites
88 --------------------------------------------------
89
90 The second transformation in the chain is diffcore-break, and is
91 controlled by the -B option to the 'git diff-{asterisk}' commands. This is
92 used to detect a filepair that represents "complete rewrite" and
93 break such filepair into two filepairs that represent delete and
94 create. E.g. If the input contained this filepair:
95
96 ------------------------------------------------
97 :100644 100644 bcd1234... 0123456... M file0
98 ------------------------------------------------
99
100 and if it detects that the file "file0" is completely rewritten,
101 it changes it to:
102
103 ------------------------------------------------
104 :100644 000000 bcd1234... 0000000... D file0
105 :000000 100644 0000000... 0123456... A file0
106 ------------------------------------------------
107
108 For the purpose of breaking a filepair, diffcore-break examines
109 the extent of changes between the contents of the files before
110 and after modification (i.e. the contents that have "bcd1234..."
111 and "0123456..." as their SHA-1 content ID, in the above
112 example). The amount of deletion of original contents and
113 insertion of new material are added together, and if it exceeds
114 the "break score", the filepair is broken into two. The break
115 score defaults to 50% of the size of the smaller of the original
116 and the result (i.e. if the edit shrinks the file, the size of
117 the result is used; if the edit lengthens the file, the size of
118 the original is used), and can be customized by giving a number
119 after "-B" option (e.g. "-B75" to tell it to use 75%).
120
121
122 diffcore-rename: For Detecting Renames and Copies
123 -------------------------------------------------
124
125 This transformation is used to detect renames and copies, and is
126 controlled by the -M option (to detect renames) and the -C option
127 (to detect copies as well) to the 'git diff-{asterisk}' commands. If the
128 input contained these filepairs:
129
130 ------------------------------------------------
131 :100644 000000 0123456... 0000000... D fileX
132 :000000 100644 0000000... 0123456... A file0
133 ------------------------------------------------
134
135 and the contents of the deleted file fileX is similar enough to
136 the contents of the created file file0, then rename detection
137 merges these filepairs and creates:
138
139 ------------------------------------------------
140 :100644 100644 0123456... 0123456... R100 fileX file0
141 ------------------------------------------------
142
143 When the "-C" option is used, the original contents of modified files,
144 and deleted files (and also unmodified files, if the
145 "--find-copies-harder" option is used) are considered as candidates
146 of the source files in rename/copy operation. If the input were like
147 these filepairs, that talk about a modified file fileY and a newly
148 created file file0:
149
150 ------------------------------------------------
151 :100644 100644 0123456... 1234567... M fileY
152 :000000 100644 0000000... bcd3456... A file0
153 ------------------------------------------------
154
155 the original contents of fileY and the resulting contents of
156 file0 are compared, and if they are similar enough, they are
157 changed to:
158
159 ------------------------------------------------
160 :100644 100644 0123456... 1234567... M fileY
161 :100644 100644 0123456... bcd3456... C100 fileY file0
162 ------------------------------------------------
163
164 In both rename and copy detection, the same "extent of changes"
165 algorithm used in diffcore-break is used to determine if two
166 files are "similar enough", and can be customized to use
167 a similarity score different from the default of 50% by giving a
168 number after the "-M" or "-C" option (e.g. "-M8" to tell it to use
169 8/10 = 80%).
170
171 Note that when rename detection is on but both copy and break
172 detection are off, rename detection adds a preliminary step that first
173 checks if files are moved across directories while keeping their
174 filename the same. If there is a file added to a directory whose
175 contents is sufficiently similar to a file with the same name that got
176 deleted from a different directory, it will mark them as renames and
177 exclude them from the later quadratic step (the one that pairwise
178 compares all unmatched files to find the "best" matches, determined by
179 the highest content similarity). So, for example, if a deleted
180 docs/ext.txt and an added docs/config/ext.txt are similar enough, they
181 will be marked as a rename and prevent an added docs/ext.md that may
182 be even more similar to the deleted docs/ext.txt from being considered
183 as the rename destination in the later step. For this reason, the
184 preliminary "match same filename" step uses a bit higher threshold to
185 mark a file pair as a rename and stop considering other candidates for
186 better matches. At most, one comparison is done per file in this
187 preliminary pass; so if there are several remaining ext.txt files
188 throughout the directory hierarchy after exact rename detection, this
189 preliminary step may be skipped for those files.
190
191 Note. When the "-C" option is used with `--find-copies-harder`
192 option, 'git diff-{asterisk}' commands feed unmodified filepairs to
193 diffcore mechanism as well as modified ones. This lets the copy
194 detector consider unmodified files as copy source candidates at
195 the expense of making it slower. Without `--find-copies-harder`,
196 'git diff-{asterisk}' commands can detect copies only if the file that was
197 copied happened to have been modified in the same changeset.
198
199
200 diffcore-merge-broken: For Putting Complete Rewrites Back Together
201 ------------------------------------------------------------------
202
203 This transformation is used to merge filepairs broken by
204 diffcore-break, and not transformed into rename/copy by
205 diffcore-rename, back into a single modification. This always
206 runs when diffcore-break is used.
207
208 For the purpose of merging broken filepairs back, it uses a
209 different "extent of changes" computation from the ones used by
210 diffcore-break and diffcore-rename. It counts only the deletion
211 from the original, and does not count insertion. If you removed
212 only 10 lines from a 100-line document, even if you added 910
213 new lines to make a new 1000-line document, you did not do a
214 complete rewrite. diffcore-break breaks such a case in order to
215 help diffcore-rename to consider such filepairs as candidate of
216 rename/copy detection, but if filepairs broken that way were not
217 matched with other filepairs to create rename/copy, then this
218 transformation merges them back into the original
219 "modification".
220
221 The "extent of changes" parameter can be tweaked from the
222 default 80% (that is, unless more than 80% of the original
223 material is deleted, the broken pairs are merged back into a
224 single modification) by giving a second number to -B option,
225 like these:
226
227 * -B50/60 (give 50% "break score" to diffcore-break, use 60%
228 for diffcore-merge-broken).
229
230 * -B/60 (the same as above, since diffcore-break defaults to 50%).
231
232 Note that earlier implementation left a broken pair as a separate
233 creation and deletion patches. This was an unnecessary hack and
234 the latest implementation always merges all the broken pairs
235 back into modifications, but the resulting patch output is
236 formatted differently for easier review in case of such
237 a complete rewrite by showing the entire contents of old version
238 prefixed with '-', followed by the entire contents of new
239 version prefixed with '+'.
240
241
242 diffcore-pickaxe: For Detecting Addition/Deletion of Specified String
243 ---------------------------------------------------------------------
244
245 This transformation limits the set of filepairs to those that change
246 specified strings between the preimage and the postimage in a certain
247 way. -S<block of text> and -G<regular expression> options are used to
248 specify different ways these strings are sought.
249
250 "-S<block of text>" detects filepairs whose preimage and postimage
251 have different number of occurrences of the specified block of text.
252 By definition, it will not detect in-file moves. Also, when a
253 changeset moves a file wholesale without affecting the interesting
254 string, diffcore-rename kicks in as usual, and `-S` omits the filepair
255 (since the number of occurrences of that string didn't change in that
256 rename-detected filepair). When used with `--pickaxe-regex`, treat
257 the <block of text> as an extended POSIX regular expression to match,
258 instead of a literal string.
259
260 "-G<regular expression>" (mnemonic: grep) detects filepairs whose
261 textual diff has an added or a deleted line that matches the given
262 regular expression. This means that it will detect in-file (or what
263 rename-detection considers the same file) moves, which is noise. The
264 implementation runs diff twice and greps, and this can be quite
265 expensive. To speed things up binary files without textconv filters
266 will be ignored.
267
268 When `-S` or `-G` are used without `--pickaxe-all`, only filepairs
269 that match their respective criterion are kept in the output. When
270 `--pickaxe-all` is used, if even one filepair matches their respective
271 criterion in a changeset, the entire changeset is kept. This behavior
272 is designed to make reviewing changes in the context of the whole
273 changeset easier.
274
275 diffcore-order: For Sorting the Output Based on Filenames
276 ---------------------------------------------------------
277
278 This is used to reorder the filepairs according to the user's
279 (or project's) taste, and is controlled by the -O option to the
280 'git diff-{asterisk}' commands.
281
282 This takes a text file each of whose lines is a shell glob
283 pattern. Filepairs that match a glob pattern on an earlier line
284 in the file are output before ones that match a later line, and
285 filepairs that do not match any glob pattern are output last.
286
287 As an example, a typical orderfile for the core Git probably
288 would look like this:
289
290 ------------------------------------------------
291 README
292 Makefile
293 Documentation
294 *.h
295 *.c
296 t
297 ------------------------------------------------
298
299 SEE ALSO
300 --------
301 linkgit:git-diff[1],
302 linkgit:git-diff-files[1],
303 linkgit:git-diff-index[1],
304 linkgit:git-diff-tree[1],
305 linkgit:git-format-patch[1],
306 linkgit:git-log[1],
307 linkgit:gitglossary[7],
308 link:user-manual.html[The Git User's Manual]
309
310 GIT
311 ---
312 Part of the linkgit:git[1] suite