Mercurial > public > mercurial-scm > hg-stable
annotate mercurial/revlog.py @ 1073:7b35a980b982
[PATCH] raise exceptions with Exception subclasses
Fixed the patch. Using Exception subclasses.
(tweaked by mpm)
author | Bart Trojanowski <bart@jukie.net> |
---|---|
date | Fri, 26 Aug 2005 19:08:25 -0700 |
parents | 6d5a62a549fa |
children | 55bf5cfde69e |
rev | line source |
---|---|
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
1 # revlog.py - storage back-end for mercurial |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
2 # |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
3 # This provides efficient delta storage with O(1) retrieve and append |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
4 # and O(changes) merge between branches |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
5 # |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
6 # Copyright 2005 Matt Mackall <mpm@selenic.com> |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
7 # |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
8 # This software may be used and distributed according to the terms |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
9 # of the GNU General Public License, incorporated herein by reference. |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
10 |
208 | 11 import zlib, struct, sha, binascii, heapq |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
12 from mercurial import mdiff |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
13 |
36
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
14 def hex(node): return binascii.hexlify(node) |
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
15 def bin(node): return binascii.unhexlify(node) |
373
67081329d49a
Change the size of the short hash representation
mpm@selenic.com
parents:
370
diff
changeset
|
16 def short(node): return hex(node[:6]) |
36
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
17 |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
18 def compress(text): |
112 | 19 if not text: return text |
20 if len(text) < 44: | |
21 if text[0] == '\0': return text | |
22 return 'u' + text | |
23 bin = zlib.compress(text) | |
24 if len(bin) > len(text): | |
25 if text[0] == '\0': return text | |
26 return 'u' + text | |
27 return bin | |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
28 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
29 def decompress(bin): |
112 | 30 if not bin: return bin |
31 t = bin[0] | |
32 if t == '\0': return bin | |
33 if t == 'x': return zlib.decompress(bin) | |
34 if t == 'u': return bin[1:] | |
1073
7b35a980b982
[PATCH] raise exceptions with Exception subclasses
Bart Trojanowski <bart@jukie.net>
parents:
1062
diff
changeset
|
35 raise RevlogError("unknown compression type %s" % t) |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
36 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
37 def hash(text, p1, p2): |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
38 l = [p1, p2] |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
39 l.sort() |
531 | 40 s = sha.new(l[0]) |
41 s.update(l[1]) | |
42 s.update(text) | |
43 return s.digest() | |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
44 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
45 nullid = "\0" * 20 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
46 indexformat = ">4l20s20s20s" |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
47 |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
48 class lazyparser: |
323 | 49 def __init__(self, data, revlog): |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
50 self.data = data |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
51 self.s = struct.calcsize(indexformat) |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
52 self.l = len(data)/self.s |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
53 self.index = [None] * self.l |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
54 self.map = {nullid: -1} |
323 | 55 self.all = 0 |
56 self.revlog = revlog | |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
57 |
323 | 58 def load(self, pos=None): |
59 if self.all: return | |
60 if pos is not None: | |
61 block = pos / 1000 | |
62 i = block * 1000 | |
63 end = min(self.l, i + 1000) | |
64 else: | |
65 self.all = 1 | |
66 i = 0 | |
67 end = self.l | |
68 self.revlog.index = self.index | |
69 self.revlog.nodemap = self.map | |
515 | 70 |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
71 while i < end: |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
72 d = self.data[i * self.s: (i + 1) * self.s] |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
73 e = struct.unpack(indexformat, d) |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
74 self.index[i] = e |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
75 self.map[e[6]] = i |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
76 i += 1 |
515 | 77 |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
78 class lazyindex: |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
79 def __init__(self, parser): |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
80 self.p = parser |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
81 def __len__(self): |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
82 return len(self.p.index) |
115 | 83 def load(self, pos): |
84 self.p.load(pos) | |
85 return self.p.index[pos] | |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
86 def __getitem__(self, pos): |
115 | 87 return self.p.index[pos] or self.load(pos) |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
88 def append(self, e): |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
89 self.p.index.append(e) |
515 | 90 |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
91 class lazymap: |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
92 def __init__(self, parser): |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
93 self.p = parser |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
94 def load(self, key): |
323 | 95 if self.p.all: return |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
96 n = self.p.data.find(key) |
86
1b945e8ba67b
Friendlier exceptions for unknown node errors
mpm@selenic.com
parents:
84
diff
changeset
|
97 if n < 0: raise KeyError("node " + hex(key)) |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
98 pos = n / self.p.s |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
99 self.p.load(pos) |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
100 def __contains__(self, key): |
323 | 101 self.p.load() |
102 return key in self.p.map | |
97 | 103 def __iter__(self): |
469 | 104 yield nullid |
97 | 105 for i in xrange(self.p.l): |
106 try: | |
107 yield self.p.index[i][6] | |
108 except: | |
109 self.p.load(i) | |
110 yield self.p.index[i][6] | |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
111 def __getitem__(self, key): |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
112 try: |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
113 return self.p.map[key] |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
114 except KeyError: |
86
1b945e8ba67b
Friendlier exceptions for unknown node errors
mpm@selenic.com
parents:
84
diff
changeset
|
115 try: |
1b945e8ba67b
Friendlier exceptions for unknown node errors
mpm@selenic.com
parents:
84
diff
changeset
|
116 self.load(key) |
1b945e8ba67b
Friendlier exceptions for unknown node errors
mpm@selenic.com
parents:
84
diff
changeset
|
117 return self.p.map[key] |
1b945e8ba67b
Friendlier exceptions for unknown node errors
mpm@selenic.com
parents:
84
diff
changeset
|
118 except KeyError: |
1b945e8ba67b
Friendlier exceptions for unknown node errors
mpm@selenic.com
parents:
84
diff
changeset
|
119 raise KeyError("node " + hex(key)) |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
120 def __setitem__(self, key, val): |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
121 self.p.map[key] = val |
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
122 |
1073
7b35a980b982
[PATCH] raise exceptions with Exception subclasses
Bart Trojanowski <bart@jukie.net>
parents:
1062
diff
changeset
|
123 class RevlogError(Exception): pass |
7b35a980b982
[PATCH] raise exceptions with Exception subclasses
Bart Trojanowski <bart@jukie.net>
parents:
1062
diff
changeset
|
124 |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
125 class revlog: |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
126 def __init__(self, opener, indexfile, datafile): |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
127 self.indexfile = indexfile |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
128 self.datafile = datafile |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
129 self.opener = opener |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
130 self.cache = None |
116
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
131 |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
132 try: |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
133 i = self.opener(self.indexfile).read() |
73 | 134 except IOError: |
76
d993ebd69d28
Add lazy{parser,index,map} to speed up processing of index files
mpm@selenic.com
parents:
73
diff
changeset
|
135 i = "" |
116
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
136 |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
137 if len(i) > 10000: |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
138 # big index, let's parse it on demand |
323 | 139 parser = lazyparser(i, self) |
116
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
140 self.index = lazyindex(parser) |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
141 self.nodemap = lazymap(parser) |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
142 else: |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
143 s = struct.calcsize(indexformat) |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
144 l = len(i) / s |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
145 self.index = [None] * l |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
146 m = [None] * l |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
147 |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
148 n = 0 |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
149 for f in xrange(0, len(i), s): |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
150 # offset, size, base, linkrev, p1, p2, nodeid |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
151 e = struct.unpack(indexformat, i[f:f + s]) |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
152 m[n] = (e[6], n) |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
153 self.index[n] = e |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
154 n += 1 |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
155 |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
156 self.nodemap = dict(m) |
e484cd5ec282
Only use lazy indexing for big indices and avoid the overhead of the
mpm@selenic.com
parents:
115
diff
changeset
|
157 self.nodemap[nullid] = -1 |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
158 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
159 def tip(self): return self.node(len(self.index) - 1) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
160 def count(self): return len(self.index) |
26 | 161 def node(self, rev): return (rev < 0) and nullid or self.index[rev][6] |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
162 def rev(self, node): return self.nodemap[node] |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
163 def linkrev(self, node): return self.index[self.nodemap[node]][3] |
2 | 164 def parents(self, node): |
165 if node == nullid: return (nullid, nullid) | |
166 return self.index[self.nodemap[node]][4:6] | |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
167 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
168 def start(self, rev): return self.index[rev][0] |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
169 def length(self, rev): return self.index[rev][1] |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
170 def end(self, rev): return self.start(rev) + self.length(rev) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
171 def base(self, rev): return self.index[rev][2] |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
172 |
902 | 173 def heads(self, stop=None): |
221 | 174 p = {} |
175 h = [] | |
902 | 176 stoprev = 0 |
177 if stop and stop in self.nodemap: | |
178 stoprev = self.rev(stop) | |
179 | |
243 | 180 for r in range(self.count() - 1, -1, -1): |
221 | 181 n = self.node(r) |
182 if n not in p: | |
183 h.append(n) | |
902 | 184 if n == stop: |
185 break | |
186 if r < stoprev: | |
187 break | |
221 | 188 for pn in self.parents(n): |
189 p[pn] = 1 | |
190 return h | |
370 | 191 |
192 def children(self, node): | |
193 c = [] | |
194 p = self.rev(node) | |
195 for r in range(p + 1, self.count()): | |
196 n = self.node(r) | |
197 for pn in self.parents(n): | |
854
473c030d34a6
Fixed revlog.children.
Tristan Wibberley <tristan@wibberley.org>
parents:
655
diff
changeset
|
198 if pn == node: |
473c030d34a6
Fixed revlog.children.
Tristan Wibberley <tristan@wibberley.org>
parents:
655
diff
changeset
|
199 c.append(n) |
370 | 200 continue |
201 elif pn == nullid: | |
202 continue | |
203 return c | |
515 | 204 |
36
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
205 def lookup(self, id): |
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
206 try: |
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
207 rev = int(id) |
469 | 208 if str(rev) != id: raise ValueError |
209 if rev < 0: rev = self.count() + rev | |
476
0a338d506268
Really _call_ method revlog.count in revlog.lookup()
Thomas Arendsen Hein <thomas@intevation.de>
parents:
469
diff
changeset
|
210 if rev < 0 or rev >= self.count(): raise ValueError |
36
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
211 return self.node(rev) |
469 | 212 except (ValueError, OverflowError): |
36
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
213 c = [] |
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
214 for n in self.nodemap: |
469 | 215 if hex(n).startswith(id): |
36
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
216 c.append(n) |
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
217 if len(c) > 1: raise KeyError("Ambiguous identifier") |
67 | 218 if len(c) < 1: raise KeyError("No match found") |
36
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
219 return c[0] |
515 | 220 |
36
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
221 return None |
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
222 |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
223 def diff(self, a, b): |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
224 return mdiff.textdiff(a, b) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
225 |
73 | 226 def patches(self, t, pl): |
227 return mdiff.patches(t, pl) | |
228 | |
119
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
229 def delta(self, node): |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
230 r = self.rev(node) |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
231 b = self.base(r) |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
232 if r == b: |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
233 return self.diff(self.revision(self.node(r - 1)), |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
234 self.revision(node)) |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
235 else: |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
236 f = self.opener(self.datafile) |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
237 f.seek(self.start(r)) |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
238 data = f.read(self.length(r)) |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
239 return decompress(data) |
c7a66f9752a4
Add code to retrieve or construct a revlog delta
mpm@selenic.com
parents:
117
diff
changeset
|
240 |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
241 def revision(self, node): |
36
da28286bf6b7
Add smart node lookup by substring or by rev number
mpm@selenic.com
parents:
26
diff
changeset
|
242 if node == nullid: return "" |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
243 if self.cache and self.cache[0] == node: return self.cache[2] |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
244 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
245 text = None |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
246 rev = self.rev(node) |
117 | 247 start, length, base, link, p1, p2, node = self.index[rev] |
248 end = start + length | |
249 if base != rev: start = self.start(base) | |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
250 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
251 if self.cache and self.cache[1] >= base and self.cache[1] < rev: |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
252 base = self.cache[1] |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
253 start = self.start(base + 1) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
254 text = self.cache[2] |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
255 last = 0 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
256 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
257 f = self.opener(self.datafile) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
258 f.seek(start) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
259 data = f.read(end - start) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
260 |
651 | 261 if text is None: |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
262 last = self.length(base) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
263 text = decompress(data[:last]) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
264 |
71
47c9a869adee
Add mdiff.patches to speed up applying thousands of patches to the manifest
mpm@selenic.com
parents:
67
diff
changeset
|
265 bins = [] |
64 | 266 for r in xrange(base + 1, rev + 1): |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
267 s = self.length(r) |
71
47c9a869adee
Add mdiff.patches to speed up applying thousands of patches to the manifest
mpm@selenic.com
parents:
67
diff
changeset
|
268 bins.append(decompress(data[last:last + s])) |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
269 last = last + s |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
270 |
71
47c9a869adee
Add mdiff.patches to speed up applying thousands of patches to the manifest
mpm@selenic.com
parents:
67
diff
changeset
|
271 text = mdiff.patches(text, bins) |
47c9a869adee
Add mdiff.patches to speed up applying thousands of patches to the manifest
mpm@selenic.com
parents:
67
diff
changeset
|
272 |
26 | 273 if node != hash(text, p1, p2): |
98 | 274 raise IOError("integrity check failed on %s:%d" |
275 % (self.datafile, rev)) | |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
276 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
277 self.cache = (node, rev, text) |
515 | 278 return text |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
279 |
644 | 280 def addrevision(self, text, transaction, link, p1=None, p2=None, d=None): |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
281 if text is None: text = "" |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
282 if p1 is None: p1 = self.tip() |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
283 if p2 is None: p2 = nullid |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
284 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
285 node = hash(text, p1, p2) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
286 |
301 | 287 if node in self.nodemap: |
288 return node | |
289 | |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
290 n = self.count() |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
291 t = n - 1 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
292 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
293 if n: |
64 | 294 base = self.base(t) |
295 start = self.start(base) | |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
296 end = self.end(t) |
644 | 297 if not d: |
298 prev = self.revision(self.tip()) | |
299 d = self.diff(prev, text) | |
98 | 300 data = compress(d) |
64 | 301 dist = end - start + len(data) |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
302 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
303 # full versions are inserted when the needed deltas |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
304 # become comparable to the uncompressed text |
64 | 305 if not n or dist > len(text) * 2: |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
306 data = compress(text) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
307 base = n |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
308 else: |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
309 base = self.base(t) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
310 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
311 offset = 0 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
312 if t >= 0: |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
313 offset = self.end(t) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
314 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
315 e = (offset, len(data), base, link, p1, p2, node) |
515 | 316 |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
317 self.index.append(e) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
318 self.nodemap[node] = n |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
319 entry = struct.pack(indexformat, *e) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
320 |
26 | 321 transaction.add(self.datafile, e[0]) |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
322 self.opener(self.datafile, "a").write(data) |
41 | 323 transaction.add(self.indexfile, n * len(entry)) |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
324 self.opener(self.indexfile, "a").write(entry) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
325 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
326 self.cache = (node, n, text) |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
327 return node |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
328 |
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
329 def ancestor(self, a, b): |
147 | 330 # calculate the distance of every node from root |
331 dist = {nullid: 0} | |
332 for i in xrange(self.count()): | |
333 n = self.node(i) | |
334 p1, p2 = self.parents(n) | |
335 dist[n] = max(dist[p1], dist[p2]) + 1 | |
515 | 336 |
147 | 337 # traverse ancestors in order of decreasing distance from root |
338 def ancestors(node): | |
339 # we store negative distances because heap returns smallest member | |
340 h = [(-dist[node], node)] | |
341 seen = {} | |
342 earliest = self.count() | |
343 while h: | |
344 d, n = heapq.heappop(h) | |
345 if n not in seen: | |
346 seen[n] = 1 | |
381 | 347 r = self.rev(n) |
348 yield (-d, r, n) | |
147 | 349 for p in self.parents(n): |
350 heapq.heappush(h, (-dist[p], p)) | |
45
f2b2d5daec30
Fix recursion depth trouble with ancestor algorithm
mpm@selenic.com
parents:
41
diff
changeset
|
351 |
147 | 352 x = ancestors(a) |
353 y = ancestors(b) | |
354 lx = x.next() | |
355 ly = y.next() | |
45
f2b2d5daec30
Fix recursion depth trouble with ancestor algorithm
mpm@selenic.com
parents:
41
diff
changeset
|
356 |
147 | 357 # increment each ancestor list until it is closer to root than |
358 # the other, or they match | |
359 while 1: | |
360 if lx == ly: | |
381 | 361 return lx[2] |
147 | 362 elif lx < ly: |
363 ly = y.next() | |
364 elif lx > ly: | |
365 lx = x.next() | |
0
9117c6561b0b
Add back links from file revisions to changeset revisions
mpm@selenic.com
parents:
diff
changeset
|
366 |
46 | 367 def group(self, linkmap): |
368 # given a list of changeset revs, return a set of deltas and | |
94 | 369 # metadata corresponding to nodes. the first delta is |
46 | 370 # parent(nodes[0]) -> nodes[0] the receiver is guaranteed to |
371 # have this parent as it has all history before these | |
372 # changesets. parent is parent[0] | |
373 | |
374 revs = [] | |
375 needed = {} | |
376 | |
377 # find file nodes/revs that match changeset revs | |
378 for i in xrange(0, self.count()): | |
379 if self.index[i][3] in linkmap: | |
380 revs.append(i) | |
381 needed[i] = 1 | |
382 | |
383 # if we don't have any revisions touched by these changesets, bail | |
192 | 384 if not revs: |
385 yield struct.pack(">l", 0) | |
386 return | |
46 | 387 |
388 # add the parent of the first rev | |
389 p = self.parents(self.node(revs[0]))[0] | |
390 revs.insert(0, self.rev(p)) | |
391 | |
392 # for each delta that isn't contiguous in the log, we need to | |
393 # reconstruct the base, reconstruct the result, and then | |
394 # calculate the delta. We also need to do this where we've | |
395 # stored a full version and not a delta | |
396 for i in xrange(0, len(revs) - 1): | |
397 a, b = revs[i], revs[i + 1] | |
398 if a + 1 != b or self.base(b) == b: | |
399 for j in xrange(self.base(a), a + 1): | |
400 needed[j] = 1 | |
401 for j in xrange(self.base(b), b + 1): | |
402 needed[j] = 1 | |
403 | |
404 # calculate spans to retrieve from datafile | |
405 needed = needed.keys() | |
406 needed.sort() | |
407 spans = [] | |
192 | 408 oo = -1 |
409 ol = 0 | |
46 | 410 for n in needed: |
411 if n < 0: continue | |
412 o = self.start(n) | |
413 l = self.length(n) | |
192 | 414 if oo + ol == o: # can we merge with the previous? |
415 nl = spans[-1][2] | |
416 nl.append((n, l)) | |
417 ol += l | |
418 spans[-1] = (oo, ol, nl) | |
46 | 419 else: |
192 | 420 oo = o |
421 ol = l | |
422 spans.append((oo, ol, [(n, l)])) | |
46 | 423 |
424 # read spans in, divide up chunks | |
425 chunks = {} | |
192 | 426 for span in spans: |
46 | 427 # we reopen the file for each span to make http happy for now |
428 f = self.opener(self.datafile) | |
429 f.seek(span[0]) | |
430 data = f.read(span[1]) | |
431 | |
432 # divide up the span | |
433 pos = 0 | |
434 for r, l in span[2]: | |
192 | 435 chunks[r] = decompress(data[pos: pos + l]) |
46 | 436 pos += l |
437 | |
438 # helper to reconstruct intermediate versions | |
439 def construct(text, base, rev): | |
192 | 440 bins = [chunks[r] for r in xrange(base + 1, rev + 1)] |
71
47c9a869adee
Add mdiff.patches to speed up applying thousands of patches to the manifest
mpm@selenic.com
parents:
67
diff
changeset
|
441 return mdiff.patches(text, bins) |
46 | 442 |
443 # build deltas | |
444 deltas = [] | |
71
47c9a869adee
Add mdiff.patches to speed up applying thousands of patches to the manifest
mpm@selenic.com
parents:
67
diff
changeset
|
445 for d in xrange(0, len(revs) - 1): |
46 | 446 a, b = revs[d], revs[d + 1] |
447 n = self.node(b) | |
192 | 448 |
449 # do we need to construct a new delta? | |
46 | 450 if a + 1 != b or self.base(b) == b: |
451 if a >= 0: | |
452 base = self.base(a) | |
192 | 453 ta = chunks[self.base(a)] |
46 | 454 ta = construct(ta, base, a) |
455 else: | |
456 ta = "" | |
515 | 457 |
46 | 458 base = self.base(b) |
459 if a > base: | |
460 base = a | |
461 tb = ta | |
462 else: | |
192 | 463 tb = chunks[self.base(b)] |
46 | 464 tb = construct(tb, base, b) |
465 d = self.diff(ta, tb) | |
466 else: | |
192 | 467 d = chunks[b] |
46 | 468 |
469 p = self.parents(n) | |
470 meta = n + p[0] + p[1] + linkmap[self.linkrev(n)] | |
471 l = struct.pack(">l", len(meta) + len(d) + 4) | |
192 | 472 yield l |
473 yield meta | |
474 yield d | |
46 | 475 |
192 | 476 yield struct.pack(">l", 0) |
477 | |
1062 | 478 def addgroup(self, revs, linkmapper, transaction, unique=0): |
46 | 479 # given a set of deltas, add them to the revision log. the |
480 # first delta is against its parent, which should be in our | |
481 # log, the rest are against the previous delta. | |
482 | |
483 # track the base of the current delta log | |
484 r = self.count() | |
485 t = r - 1 | |
192 | 486 node = nullid |
515 | 487 |
655 | 488 base = prev = -1 |
653
94cdd02792b5
Fix corruption resulting from skipping parts of a revision group
Matt Mackall <mpm@selenic.com>
parents:
651
diff
changeset
|
489 start = end = measure = 0 |
46 | 490 if r: |
491 start = self.start(self.base(t)) | |
492 end = self.end(t) | |
493 measure = self.length(self.base(t)) | |
494 base = self.base(t) | |
495 prev = self.tip() | |
496 | |
497 transaction.add(self.datafile, end) | |
498 transaction.add(self.indexfile, r * struct.calcsize(indexformat)) | |
499 dfh = self.opener(self.datafile, "a") | |
500 ifh = self.opener(self.indexfile, "a") | |
501 | |
502 # loop through our set of deltas | |
192 | 503 chain = None |
504 for chunk in revs: | |
505 node, p1, p2, cs = struct.unpack("20s20s20s20s", chunk[:80]) | |
94 | 506 link = linkmapper(cs) |
77 | 507 if node in self.nodemap: |
224
ccbcc4d76f81
fix bad assumption about uniqueness of file versions
mpm@selenic.com
parents:
221
diff
changeset
|
508 # this can happen if two branches make the same change |
ccbcc4d76f81
fix bad assumption about uniqueness of file versions
mpm@selenic.com
parents:
221
diff
changeset
|
509 if unique: |
1073
7b35a980b982
[PATCH] raise exceptions with Exception subclasses
Bart Trojanowski <bart@jukie.net>
parents:
1062
diff
changeset
|
510 raise RevlogError("already have %s" % hex(node[:4])) |
653
94cdd02792b5
Fix corruption resulting from skipping parts of a revision group
Matt Mackall <mpm@selenic.com>
parents:
651
diff
changeset
|
511 chain = node |
224
ccbcc4d76f81
fix bad assumption about uniqueness of file versions
mpm@selenic.com
parents:
221
diff
changeset
|
512 continue |
192 | 513 delta = chunk[80:] |
514 | |
515 if not chain: | |
516 # retrieve the parent revision of the delta chain | |
517 chain = p1 | |
518 if not chain in self.nodemap: | |
1073
7b35a980b982
[PATCH] raise exceptions with Exception subclasses
Bart Trojanowski <bart@jukie.net>
parents:
1062
diff
changeset
|
519 raise RevlogError("unknown base %s" % short(chain[:4])) |
46 | 520 |
521 # full versions are inserted when the needed deltas become | |
522 # comparable to the uncompressed text or when the previous | |
523 # version is not the one we have a delta against. We use | |
524 # the size of the previous full rev as a proxy for the | |
525 # current size. | |
526 | |
527 if chain == prev: | |
528 cdelta = compress(delta) | |
529 | |
530 if chain != prev or (end - start + len(cdelta)) > measure * 2: | |
531 # flush our writes here so we can read it in revision | |
532 dfh.flush() | |
533 ifh.flush() | |
65
d40cc5aacc31
Fix up a bunch of bugs in the new merge code
mpm@selenic.com
parents:
64
diff
changeset
|
534 text = self.revision(chain) |
73 | 535 text = self.patches(text, [delta]) |
46 | 536 chk = self.addrevision(text, transaction, link, p1, p2) |
537 if chk != node: | |
1073
7b35a980b982
[PATCH] raise exceptions with Exception subclasses
Bart Trojanowski <bart@jukie.net>
parents:
1062
diff
changeset
|
538 raise RevlogError("consistency error adding group") |
46 | 539 measure = len(text) |
540 else: | |
541 e = (end, len(cdelta), self.base(t), link, p1, p2, node) | |
542 self.index.append(e) | |
543 self.nodemap[node] = r | |
544 dfh.write(cdelta) | |
545 ifh.write(struct.pack(indexformat, *e)) | |
546 | |
65
d40cc5aacc31
Fix up a bunch of bugs in the new merge code
mpm@selenic.com
parents:
64
diff
changeset
|
547 t, r, chain, prev = r, r + 1, node, node |
46 | 548 start = self.start(self.base(t)) |
549 end = self.end(t) | |
550 | |
551 dfh.close() | |
552 ifh.close() | |
553 return node |