Mercurial > public > mercurial-scm > hg
annotate mercurial/pure/bdiff.py @ 13741:b51bf961b3cb
wireproto: add getbundle() function
getbundle(common, heads) -> bundle
Returns the changegroup for all ancestors of heads which are not ancestors of common. For both
sets, the heads are included in the set.
Intended to eventually supercede changegroupsubset and changegroup. Uses heads of common region
to exclude unwanted changesets instead of bases of desired region, which is more useful and
easier to implement.
Designed to be extensible with new optional arguments (which will have to be guarded by
corresponding capabilities).
author | Peter Arrenbrecht <peter.arrenbrecht@gmail.com> |
---|---|
date | Wed, 23 Mar 2011 16:02:11 +0100 |
parents | 08a0f04b56bd |
children | 14fac6c0536a |
rev | line source |
---|---|
7703
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
1 # bdiff.py - Python implementation of bdiff.c |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
2 # |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
3 # Copyright 2009 Matt Mackall <mpm@selenic.com> and others |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
4 # |
8225
46293a0c7e9f
updated license to be explicit about GPL version 2
Martin Geisler <mg@lazybytes.net>
parents:
7944
diff
changeset
|
5 # This software may be used and distributed according to the terms of the |
10263 | 6 # GNU General Public License version 2 or any later version. |
7703
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
7 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
8 import struct, difflib |
7944
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
9 |
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
10 def splitnewlines(text): |
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
11 '''like str.splitlines, but only split on newlines.''' |
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
12 lines = [l + '\n' for l in text.split('\n')] |
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
13 if lines: |
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
14 if lines[-1] == '\n': |
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
15 lines.pop() |
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
16 else: |
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
17 lines[-1] = lines[-1][:-1] |
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
18 return lines |
7703
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
19 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
20 def _normalizeblocks(a, b, blocks): |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
21 prev = None |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
22 for curr in blocks: |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
23 if prev is None: |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
24 prev = curr |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
25 continue |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
26 shift = 0 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
27 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
28 a1, b1, l1 = prev |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
29 a1end = a1 + l1 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
30 b1end = b1 + l1 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
31 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
32 a2, b2, l2 = curr |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
33 a2end = a2 + l2 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
34 b2end = b2 + l2 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
35 if a1end == a2: |
10282
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
10263
diff
changeset
|
36 while (a1end + shift < a2end and |
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
10263
diff
changeset
|
37 a[a1end + shift] == b[b1end + shift]): |
7703
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
38 shift += 1 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
39 elif b1end == b2: |
10282
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
10263
diff
changeset
|
40 while (b1end + shift < b2end and |
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
10263
diff
changeset
|
41 a[a1end + shift] == b[b1end + shift]): |
7703
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
42 shift += 1 |
10282
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
10263
diff
changeset
|
43 yield a1, b1, l1 + shift |
08a0f04b56bd
many, many trivial check-code fixups
Matt Mackall <mpm@selenic.com>
parents:
10263
diff
changeset
|
44 prev = a2 + shift, b2 + shift, l2 - shift |
7703
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
45 yield prev |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
46 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
47 def bdiff(a, b): |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
48 a = str(a).splitlines(True) |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
49 b = str(b).splitlines(True) |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
50 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
51 if not a: |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
52 s = "".join(b) |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
53 return s and (struct.pack(">lll", 0, 0, len(s)) + s) |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
54 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
55 bin = [] |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
56 p = [0] |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
57 for i in a: p.append(p[-1] + len(i)) |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
58 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
59 d = difflib.SequenceMatcher(None, a, b).get_matching_blocks() |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
60 d = _normalizeblocks(a, b, d) |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
61 la = 0 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
62 lb = 0 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
63 for am, bm, size in d: |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
64 s = "".join(b[lb:bm]) |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
65 if am > la or s: |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
66 bin.append(struct.pack(">lll", p[la], p[am], len(s)) + s) |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
67 la = am + size |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
68 lb = bm + size |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
69 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
70 return "".join(bin) |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
71 |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
72 def blocks(a, b): |
7944
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
73 an = splitnewlines(a) |
e9b48afd0e78
pure/bdiff: fix circular import
Matt Mackall <mpm@selenic.com>
parents:
7703
diff
changeset
|
74 bn = splitnewlines(b) |
7703
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
75 d = difflib.SequenceMatcher(None, an, bn).get_matching_blocks() |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
76 d = _normalizeblocks(an, bn, d) |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
77 return [(i, i + n, j, j + n) for (i, j, n) in d] |
9044d3567f6d
pure Python implementation of bdiff.c
Martin Geisler <mg@daimi.au.dk>
parents:
diff
changeset
|
78 |