##// END OF EJS Templates
pure parsers: properly detect corrupt index files...
Augie Fackler -
r14421:639f26ca default
parent child Browse files
Show More
@@ -1,88 +1,91
1 # parsers.py - Python implementation of parsers.c
1 # parsers.py - Python implementation of parsers.c
2 #
2 #
3 # Copyright 2009 Matt Mackall <mpm@selenic.com> and others
3 # Copyright 2009 Matt Mackall <mpm@selenic.com> and others
4 #
4 #
5 # This software may be used and distributed according to the terms of the
5 # This software may be used and distributed according to the terms of the
6 # GNU General Public License version 2 or any later version.
6 # GNU General Public License version 2 or any later version.
7
7
8 from mercurial.node import bin, nullid
8 from mercurial.node import bin, nullid
9 from mercurial import util
9 from mercurial import util
10 import struct, zlib
10 import struct, zlib
11
11
12 _pack = struct.pack
12 _pack = struct.pack
13 _unpack = struct.unpack
13 _unpack = struct.unpack
14 _compress = zlib.compress
14 _compress = zlib.compress
15 _decompress = zlib.decompress
15 _decompress = zlib.decompress
16 _sha = util.sha1
16 _sha = util.sha1
17
17
18 def parse_manifest(mfdict, fdict, lines):
18 def parse_manifest(mfdict, fdict, lines):
19 for l in lines.splitlines():
19 for l in lines.splitlines():
20 f, n = l.split('\0')
20 f, n = l.split('\0')
21 if len(n) > 40:
21 if len(n) > 40:
22 fdict[f] = n[40:]
22 fdict[f] = n[40:]
23 mfdict[f] = bin(n[:40])
23 mfdict[f] = bin(n[:40])
24 else:
24 else:
25 mfdict[f] = bin(n)
25 mfdict[f] = bin(n)
26
26
27 def parse_index2(data, inline):
27 def parse_index2(data, inline):
28 def gettype(q):
28 def gettype(q):
29 return int(q & 0xFFFF)
29 return int(q & 0xFFFF)
30
30
31 def offset_type(offset, type):
31 def offset_type(offset, type):
32 return long(long(offset) << 16 | type)
32 return long(long(offset) << 16 | type)
33
33
34 indexformatng = ">Qiiiiii20s12x"
34 indexformatng = ">Qiiiiii20s12x"
35
35
36 s = struct.calcsize(indexformatng)
36 s = struct.calcsize(indexformatng)
37 index = []
37 index = []
38 cache = None
38 cache = None
39 n = off = 0
39 n = off = 0
40
40
41 l = len(data) - s
41 l = len(data) - s
42 append = index.append
42 append = index.append
43 if inline:
43 if inline:
44 cache = (0, data)
44 cache = (0, data)
45 while off <= l:
45 while off <= l:
46 e = _unpack(indexformatng, data[off:off + s])
46 e = _unpack(indexformatng, data[off:off + s])
47 append(e)
47 append(e)
48 n += 1
48 n += 1
49 if e[1] < 0:
49 if e[1] < 0:
50 break
50 break
51 off += e[1] + s
51 off += e[1] + s
52 else:
52 else:
53 while off <= l:
53 while off <= l:
54 e = _unpack(indexformatng, data[off:off + s])
54 e = _unpack(indexformatng, data[off:off + s])
55 append(e)
55 append(e)
56 n += 1
56 n += 1
57 off += s
57 off += s
58
58
59 if off != len(data):
60 raise ValueError('corrupt index file')
61
59 if index:
62 if index:
60 e = list(index[0])
63 e = list(index[0])
61 type = gettype(e[0])
64 type = gettype(e[0])
62 e[0] = offset_type(0, type)
65 e[0] = offset_type(0, type)
63 index[0] = tuple(e)
66 index[0] = tuple(e)
64
67
65 # add the magic null revision at -1
68 # add the magic null revision at -1
66 index.append((0, 0, 0, -1, -1, -1, -1, nullid))
69 index.append((0, 0, 0, -1, -1, -1, -1, nullid))
67
70
68 return index, cache
71 return index, cache
69
72
70 def parse_dirstate(dmap, copymap, st):
73 def parse_dirstate(dmap, copymap, st):
71 parents = [st[:20], st[20: 40]]
74 parents = [st[:20], st[20: 40]]
72 # deref fields so they will be local in loop
75 # deref fields so they will be local in loop
73 format = ">cllll"
76 format = ">cllll"
74 e_size = struct.calcsize(format)
77 e_size = struct.calcsize(format)
75 pos1 = 40
78 pos1 = 40
76 l = len(st)
79 l = len(st)
77
80
78 # the inner loop
81 # the inner loop
79 while pos1 < l:
82 while pos1 < l:
80 pos2 = pos1 + e_size
83 pos2 = pos1 + e_size
81 e = _unpack(">cllll", st[pos1:pos2]) # a literal here is faster
84 e = _unpack(">cllll", st[pos1:pos2]) # a literal here is faster
82 pos1 = pos2 + e[4]
85 pos1 = pos2 + e[4]
83 f = st[pos2:pos1]
86 f = st[pos2:pos1]
84 if '\0' in f:
87 if '\0' in f:
85 f, c = f.split('\0')
88 f, c = f.split('\0')
86 copymap[f] = c
89 copymap[f] = c
87 dmap[f] = e[:4]
90 dmap[f] = e[:4]
88 return parents
91 return parents
General Comments 0
You need to be logged in to leave comments. Login now