Show More
@@ -1,1114 +1,1117 b'' | |||||
1 | # store.py - repository store handling for Mercurial |
|
1 | # store.py - repository store handling for Mercurial | |
2 | # |
|
2 | # | |
3 | # Copyright 2008 Olivia Mackall <olivia@selenic.com> |
|
3 | # Copyright 2008 Olivia Mackall <olivia@selenic.com> | |
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 | import collections |
|
8 | import collections | |
9 | import functools |
|
9 | import functools | |
10 | import os |
|
10 | import os | |
11 | import re |
|
11 | import re | |
12 | import stat |
|
12 | import stat | |
13 | from typing import Generator |
|
13 | from typing import Generator | |
14 |
|
14 | |||
15 | from .i18n import _ |
|
15 | from .i18n import _ | |
16 | from .pycompat import getattr |
|
16 | from .pycompat import getattr | |
17 | from .thirdparty import attr |
|
17 | from .thirdparty import attr | |
18 | from .node import hex |
|
18 | from .node import hex | |
19 | from . import ( |
|
19 | from . import ( | |
20 | changelog, |
|
20 | changelog, | |
21 | error, |
|
21 | error, | |
22 | filelog, |
|
22 | filelog, | |
23 | manifest, |
|
23 | manifest, | |
24 | policy, |
|
24 | policy, | |
25 | pycompat, |
|
25 | pycompat, | |
26 | util, |
|
26 | util, | |
27 | vfs as vfsmod, |
|
27 | vfs as vfsmod, | |
28 | ) |
|
28 | ) | |
29 | from .utils import hashutil |
|
29 | from .utils import hashutil | |
30 |
|
30 | |||
31 | parsers = policy.importmod('parsers') |
|
31 | parsers = policy.importmod('parsers') | |
32 | # how much bytes should be read from fncache in one read |
|
32 | # how much bytes should be read from fncache in one read | |
33 | # It is done to prevent loading large fncache files into memory |
|
33 | # It is done to prevent loading large fncache files into memory | |
34 | fncache_chunksize = 10 ** 6 |
|
34 | fncache_chunksize = 10 ** 6 | |
35 |
|
35 | |||
36 |
|
36 | |||
37 | def _match_tracked_entry(entry, matcher): |
|
37 | def _match_tracked_entry(entry, matcher): | |
38 | """parses a fncache entry and returns whether the entry is tracking a path |
|
38 | """parses a fncache entry and returns whether the entry is tracking a path | |
39 | matched by matcher or not. |
|
39 | matched by matcher or not. | |
40 |
|
40 | |||
41 | If matcher is None, returns True""" |
|
41 | If matcher is None, returns True""" | |
42 |
|
42 | |||
43 | if matcher is None: |
|
43 | if matcher is None: | |
44 | return True |
|
44 | return True | |
45 | if entry.is_filelog: |
|
45 | if entry.is_filelog: | |
46 | return matcher(entry.target_id) |
|
46 | return matcher(entry.target_id) | |
47 | elif entry.is_manifestlog: |
|
47 | elif entry.is_manifestlog: | |
48 | return matcher.visitdir(entry.target_id.rstrip(b'/')) |
|
48 | return matcher.visitdir(entry.target_id.rstrip(b'/')) | |
49 | raise error.ProgrammingError(b"cannot process entry %r" % entry) |
|
49 | raise error.ProgrammingError(b"cannot process entry %r" % entry) | |
50 |
|
50 | |||
51 |
|
51 | |||
52 | # This avoids a collision between a file named foo and a dir named |
|
52 | # This avoids a collision between a file named foo and a dir named | |
53 | # foo.i or foo.d |
|
53 | # foo.i or foo.d | |
54 | def _encodedir(path): |
|
54 | def _encodedir(path): | |
55 | """ |
|
55 | """ | |
56 | >>> _encodedir(b'data/foo.i') |
|
56 | >>> _encodedir(b'data/foo.i') | |
57 | 'data/foo.i' |
|
57 | 'data/foo.i' | |
58 | >>> _encodedir(b'data/foo.i/bla.i') |
|
58 | >>> _encodedir(b'data/foo.i/bla.i') | |
59 | 'data/foo.i.hg/bla.i' |
|
59 | 'data/foo.i.hg/bla.i' | |
60 | >>> _encodedir(b'data/foo.i.hg/bla.i') |
|
60 | >>> _encodedir(b'data/foo.i.hg/bla.i') | |
61 | 'data/foo.i.hg.hg/bla.i' |
|
61 | 'data/foo.i.hg.hg/bla.i' | |
62 | >>> _encodedir(b'data/foo.i\\ndata/foo.i/bla.i\\ndata/foo.i.hg/bla.i\\n') |
|
62 | >>> _encodedir(b'data/foo.i\\ndata/foo.i/bla.i\\ndata/foo.i.hg/bla.i\\n') | |
63 | 'data/foo.i\\ndata/foo.i.hg/bla.i\\ndata/foo.i.hg.hg/bla.i\\n' |
|
63 | 'data/foo.i\\ndata/foo.i.hg/bla.i\\ndata/foo.i.hg.hg/bla.i\\n' | |
64 | """ |
|
64 | """ | |
65 | return ( |
|
65 | return ( | |
66 | path.replace(b".hg/", b".hg.hg/") |
|
66 | path.replace(b".hg/", b".hg.hg/") | |
67 | .replace(b".i/", b".i.hg/") |
|
67 | .replace(b".i/", b".i.hg/") | |
68 | .replace(b".d/", b".d.hg/") |
|
68 | .replace(b".d/", b".d.hg/") | |
69 | ) |
|
69 | ) | |
70 |
|
70 | |||
71 |
|
71 | |||
72 | encodedir = getattr(parsers, 'encodedir', _encodedir) |
|
72 | encodedir = getattr(parsers, 'encodedir', _encodedir) | |
73 |
|
73 | |||
74 |
|
74 | |||
75 | def decodedir(path): |
|
75 | def decodedir(path): | |
76 | """ |
|
76 | """ | |
77 | >>> decodedir(b'data/foo.i') |
|
77 | >>> decodedir(b'data/foo.i') | |
78 | 'data/foo.i' |
|
78 | 'data/foo.i' | |
79 | >>> decodedir(b'data/foo.i.hg/bla.i') |
|
79 | >>> decodedir(b'data/foo.i.hg/bla.i') | |
80 | 'data/foo.i/bla.i' |
|
80 | 'data/foo.i/bla.i' | |
81 | >>> decodedir(b'data/foo.i.hg.hg/bla.i') |
|
81 | >>> decodedir(b'data/foo.i.hg.hg/bla.i') | |
82 | 'data/foo.i.hg/bla.i' |
|
82 | 'data/foo.i.hg/bla.i' | |
83 | """ |
|
83 | """ | |
84 | if b".hg/" not in path: |
|
84 | if b".hg/" not in path: | |
85 | return path |
|
85 | return path | |
86 | return ( |
|
86 | return ( | |
87 | path.replace(b".d.hg/", b".d/") |
|
87 | path.replace(b".d.hg/", b".d/") | |
88 | .replace(b".i.hg/", b".i/") |
|
88 | .replace(b".i.hg/", b".i/") | |
89 | .replace(b".hg.hg/", b".hg/") |
|
89 | .replace(b".hg.hg/", b".hg/") | |
90 | ) |
|
90 | ) | |
91 |
|
91 | |||
92 |
|
92 | |||
93 | def _reserved(): |
|
93 | def _reserved(): | |
94 | """characters that are problematic for filesystems |
|
94 | """characters that are problematic for filesystems | |
95 |
|
95 | |||
96 | * ascii escapes (0..31) |
|
96 | * ascii escapes (0..31) | |
97 | * ascii hi (126..255) |
|
97 | * ascii hi (126..255) | |
98 | * windows specials |
|
98 | * windows specials | |
99 |
|
99 | |||
100 | these characters will be escaped by encodefunctions |
|
100 | these characters will be escaped by encodefunctions | |
101 | """ |
|
101 | """ | |
102 | winreserved = [ord(x) for x in u'\\:*?"<>|'] |
|
102 | winreserved = [ord(x) for x in u'\\:*?"<>|'] | |
103 | for x in range(32): |
|
103 | for x in range(32): | |
104 | yield x |
|
104 | yield x | |
105 | for x in range(126, 256): |
|
105 | for x in range(126, 256): | |
106 | yield x |
|
106 | yield x | |
107 | for x in winreserved: |
|
107 | for x in winreserved: | |
108 | yield x |
|
108 | yield x | |
109 |
|
109 | |||
110 |
|
110 | |||
111 | def _buildencodefun(): |
|
111 | def _buildencodefun(): | |
112 | """ |
|
112 | """ | |
113 | >>> enc, dec = _buildencodefun() |
|
113 | >>> enc, dec = _buildencodefun() | |
114 |
|
114 | |||
115 | >>> enc(b'nothing/special.txt') |
|
115 | >>> enc(b'nothing/special.txt') | |
116 | 'nothing/special.txt' |
|
116 | 'nothing/special.txt' | |
117 | >>> dec(b'nothing/special.txt') |
|
117 | >>> dec(b'nothing/special.txt') | |
118 | 'nothing/special.txt' |
|
118 | 'nothing/special.txt' | |
119 |
|
119 | |||
120 | >>> enc(b'HELLO') |
|
120 | >>> enc(b'HELLO') | |
121 | '_h_e_l_l_o' |
|
121 | '_h_e_l_l_o' | |
122 | >>> dec(b'_h_e_l_l_o') |
|
122 | >>> dec(b'_h_e_l_l_o') | |
123 | 'HELLO' |
|
123 | 'HELLO' | |
124 |
|
124 | |||
125 | >>> enc(b'hello:world?') |
|
125 | >>> enc(b'hello:world?') | |
126 | 'hello~3aworld~3f' |
|
126 | 'hello~3aworld~3f' | |
127 | >>> dec(b'hello~3aworld~3f') |
|
127 | >>> dec(b'hello~3aworld~3f') | |
128 | 'hello:world?' |
|
128 | 'hello:world?' | |
129 |
|
129 | |||
130 | >>> enc(b'the\\x07quick\\xADshot') |
|
130 | >>> enc(b'the\\x07quick\\xADshot') | |
131 | 'the~07quick~adshot' |
|
131 | 'the~07quick~adshot' | |
132 | >>> dec(b'the~07quick~adshot') |
|
132 | >>> dec(b'the~07quick~adshot') | |
133 | 'the\\x07quick\\xadshot' |
|
133 | 'the\\x07quick\\xadshot' | |
134 | """ |
|
134 | """ | |
135 | e = b'_' |
|
135 | e = b'_' | |
136 | xchr = pycompat.bytechr |
|
136 | xchr = pycompat.bytechr | |
137 | asciistr = list(map(xchr, range(127))) |
|
137 | asciistr = list(map(xchr, range(127))) | |
138 | capitals = list(range(ord(b"A"), ord(b"Z") + 1)) |
|
138 | capitals = list(range(ord(b"A"), ord(b"Z") + 1)) | |
139 |
|
139 | |||
140 | cmap = {x: x for x in asciistr} |
|
140 | cmap = {x: x for x in asciistr} | |
141 | for x in _reserved(): |
|
141 | for x in _reserved(): | |
142 | cmap[xchr(x)] = b"~%02x" % x |
|
142 | cmap[xchr(x)] = b"~%02x" % x | |
143 | for x in capitals + [ord(e)]: |
|
143 | for x in capitals + [ord(e)]: | |
144 | cmap[xchr(x)] = e + xchr(x).lower() |
|
144 | cmap[xchr(x)] = e + xchr(x).lower() | |
145 |
|
145 | |||
146 | dmap = {} |
|
146 | dmap = {} | |
147 | for k, v in cmap.items(): |
|
147 | for k, v in cmap.items(): | |
148 | dmap[v] = k |
|
148 | dmap[v] = k | |
149 |
|
149 | |||
150 | def decode(s): |
|
150 | def decode(s): | |
151 | i = 0 |
|
151 | i = 0 | |
152 | while i < len(s): |
|
152 | while i < len(s): | |
153 | for l in range(1, 4): |
|
153 | for l in range(1, 4): | |
154 | try: |
|
154 | try: | |
155 | yield dmap[s[i : i + l]] |
|
155 | yield dmap[s[i : i + l]] | |
156 | i += l |
|
156 | i += l | |
157 | break |
|
157 | break | |
158 | except KeyError: |
|
158 | except KeyError: | |
159 | pass |
|
159 | pass | |
160 | else: |
|
160 | else: | |
161 | raise KeyError |
|
161 | raise KeyError | |
162 |
|
162 | |||
163 | return ( |
|
163 | return ( | |
164 | lambda s: b''.join([cmap[s[c : c + 1]] for c in range(len(s))]), |
|
164 | lambda s: b''.join([cmap[s[c : c + 1]] for c in range(len(s))]), | |
165 | lambda s: b''.join(list(decode(s))), |
|
165 | lambda s: b''.join(list(decode(s))), | |
166 | ) |
|
166 | ) | |
167 |
|
167 | |||
168 |
|
168 | |||
169 | _encodefname, _decodefname = _buildencodefun() |
|
169 | _encodefname, _decodefname = _buildencodefun() | |
170 |
|
170 | |||
171 |
|
171 | |||
172 | def encodefilename(s): |
|
172 | def encodefilename(s): | |
173 | """ |
|
173 | """ | |
174 | >>> encodefilename(b'foo.i/bar.d/bla.hg/hi:world?/HELLO') |
|
174 | >>> encodefilename(b'foo.i/bar.d/bla.hg/hi:world?/HELLO') | |
175 | 'foo.i.hg/bar.d.hg/bla.hg.hg/hi~3aworld~3f/_h_e_l_l_o' |
|
175 | 'foo.i.hg/bar.d.hg/bla.hg.hg/hi~3aworld~3f/_h_e_l_l_o' | |
176 | """ |
|
176 | """ | |
177 | return _encodefname(encodedir(s)) |
|
177 | return _encodefname(encodedir(s)) | |
178 |
|
178 | |||
179 |
|
179 | |||
180 | def decodefilename(s): |
|
180 | def decodefilename(s): | |
181 | """ |
|
181 | """ | |
182 | >>> decodefilename(b'foo.i.hg/bar.d.hg/bla.hg.hg/hi~3aworld~3f/_h_e_l_l_o') |
|
182 | >>> decodefilename(b'foo.i.hg/bar.d.hg/bla.hg.hg/hi~3aworld~3f/_h_e_l_l_o') | |
183 | 'foo.i/bar.d/bla.hg/hi:world?/HELLO' |
|
183 | 'foo.i/bar.d/bla.hg/hi:world?/HELLO' | |
184 | """ |
|
184 | """ | |
185 | return decodedir(_decodefname(s)) |
|
185 | return decodedir(_decodefname(s)) | |
186 |
|
186 | |||
187 |
|
187 | |||
188 | def _buildlowerencodefun(): |
|
188 | def _buildlowerencodefun(): | |
189 | """ |
|
189 | """ | |
190 | >>> f = _buildlowerencodefun() |
|
190 | >>> f = _buildlowerencodefun() | |
191 | >>> f(b'nothing/special.txt') |
|
191 | >>> f(b'nothing/special.txt') | |
192 | 'nothing/special.txt' |
|
192 | 'nothing/special.txt' | |
193 | >>> f(b'HELLO') |
|
193 | >>> f(b'HELLO') | |
194 | 'hello' |
|
194 | 'hello' | |
195 | >>> f(b'hello:world?') |
|
195 | >>> f(b'hello:world?') | |
196 | 'hello~3aworld~3f' |
|
196 | 'hello~3aworld~3f' | |
197 | >>> f(b'the\\x07quick\\xADshot') |
|
197 | >>> f(b'the\\x07quick\\xADshot') | |
198 | 'the~07quick~adshot' |
|
198 | 'the~07quick~adshot' | |
199 | """ |
|
199 | """ | |
200 | xchr = pycompat.bytechr |
|
200 | xchr = pycompat.bytechr | |
201 | cmap = {xchr(x): xchr(x) for x in range(127)} |
|
201 | cmap = {xchr(x): xchr(x) for x in range(127)} | |
202 | for x in _reserved(): |
|
202 | for x in _reserved(): | |
203 | cmap[xchr(x)] = b"~%02x" % x |
|
203 | cmap[xchr(x)] = b"~%02x" % x | |
204 | for x in range(ord(b"A"), ord(b"Z") + 1): |
|
204 | for x in range(ord(b"A"), ord(b"Z") + 1): | |
205 | cmap[xchr(x)] = xchr(x).lower() |
|
205 | cmap[xchr(x)] = xchr(x).lower() | |
206 |
|
206 | |||
207 | def lowerencode(s): |
|
207 | def lowerencode(s): | |
208 | return b"".join([cmap[c] for c in pycompat.iterbytestr(s)]) |
|
208 | return b"".join([cmap[c] for c in pycompat.iterbytestr(s)]) | |
209 |
|
209 | |||
210 | return lowerencode |
|
210 | return lowerencode | |
211 |
|
211 | |||
212 |
|
212 | |||
213 | lowerencode = getattr(parsers, 'lowerencode', None) or _buildlowerencodefun() |
|
213 | lowerencode = getattr(parsers, 'lowerencode', None) or _buildlowerencodefun() | |
214 |
|
214 | |||
215 | # Windows reserved names: con, prn, aux, nul, com1..com9, lpt1..lpt9 |
|
215 | # Windows reserved names: con, prn, aux, nul, com1..com9, lpt1..lpt9 | |
216 | _winres3 = (b'aux', b'con', b'prn', b'nul') # length 3 |
|
216 | _winres3 = (b'aux', b'con', b'prn', b'nul') # length 3 | |
217 | _winres4 = (b'com', b'lpt') # length 4 (with trailing 1..9) |
|
217 | _winres4 = (b'com', b'lpt') # length 4 (with trailing 1..9) | |
218 |
|
218 | |||
219 |
|
219 | |||
220 | def _auxencode(path, dotencode): |
|
220 | def _auxencode(path, dotencode): | |
221 | """ |
|
221 | """ | |
222 | Encodes filenames containing names reserved by Windows or which end in |
|
222 | Encodes filenames containing names reserved by Windows or which end in | |
223 | period or space. Does not touch other single reserved characters c. |
|
223 | period or space. Does not touch other single reserved characters c. | |
224 | Specifically, c in '\\:*?"<>|' or ord(c) <= 31 are *not* encoded here. |
|
224 | Specifically, c in '\\:*?"<>|' or ord(c) <= 31 are *not* encoded here. | |
225 | Additionally encodes space or period at the beginning, if dotencode is |
|
225 | Additionally encodes space or period at the beginning, if dotencode is | |
226 | True. Parameter path is assumed to be all lowercase. |
|
226 | True. Parameter path is assumed to be all lowercase. | |
227 | A segment only needs encoding if a reserved name appears as a |
|
227 | A segment only needs encoding if a reserved name appears as a | |
228 | basename (e.g. "aux", "aux.foo"). A directory or file named "foo.aux" |
|
228 | basename (e.g. "aux", "aux.foo"). A directory or file named "foo.aux" | |
229 | doesn't need encoding. |
|
229 | doesn't need encoding. | |
230 |
|
230 | |||
231 | >>> s = b'.foo/aux.txt/txt.aux/con/prn/nul/foo.' |
|
231 | >>> s = b'.foo/aux.txt/txt.aux/con/prn/nul/foo.' | |
232 | >>> _auxencode(s.split(b'/'), True) |
|
232 | >>> _auxencode(s.split(b'/'), True) | |
233 | ['~2efoo', 'au~78.txt', 'txt.aux', 'co~6e', 'pr~6e', 'nu~6c', 'foo~2e'] |
|
233 | ['~2efoo', 'au~78.txt', 'txt.aux', 'co~6e', 'pr~6e', 'nu~6c', 'foo~2e'] | |
234 | >>> s = b'.com1com2/lpt9.lpt4.lpt1/conprn/com0/lpt0/foo.' |
|
234 | >>> s = b'.com1com2/lpt9.lpt4.lpt1/conprn/com0/lpt0/foo.' | |
235 | >>> _auxencode(s.split(b'/'), False) |
|
235 | >>> _auxencode(s.split(b'/'), False) | |
236 | ['.com1com2', 'lp~749.lpt4.lpt1', 'conprn', 'com0', 'lpt0', 'foo~2e'] |
|
236 | ['.com1com2', 'lp~749.lpt4.lpt1', 'conprn', 'com0', 'lpt0', 'foo~2e'] | |
237 | >>> _auxencode([b'foo. '], True) |
|
237 | >>> _auxencode([b'foo. '], True) | |
238 | ['foo.~20'] |
|
238 | ['foo.~20'] | |
239 | >>> _auxencode([b' .foo'], True) |
|
239 | >>> _auxencode([b' .foo'], True) | |
240 | ['~20.foo'] |
|
240 | ['~20.foo'] | |
241 | """ |
|
241 | """ | |
242 | for i, n in enumerate(path): |
|
242 | for i, n in enumerate(path): | |
243 | if not n: |
|
243 | if not n: | |
244 | continue |
|
244 | continue | |
245 | if dotencode and n[0] in b'. ': |
|
245 | if dotencode and n[0] in b'. ': | |
246 | n = b"~%02x" % ord(n[0:1]) + n[1:] |
|
246 | n = b"~%02x" % ord(n[0:1]) + n[1:] | |
247 | path[i] = n |
|
247 | path[i] = n | |
248 | else: |
|
248 | else: | |
249 | l = n.find(b'.') |
|
249 | l = n.find(b'.') | |
250 | if l == -1: |
|
250 | if l == -1: | |
251 | l = len(n) |
|
251 | l = len(n) | |
252 | if (l == 3 and n[:3] in _winres3) or ( |
|
252 | if (l == 3 and n[:3] in _winres3) or ( | |
253 | l == 4 |
|
253 | l == 4 | |
254 | and n[3:4] <= b'9' |
|
254 | and n[3:4] <= b'9' | |
255 | and n[3:4] >= b'1' |
|
255 | and n[3:4] >= b'1' | |
256 | and n[:3] in _winres4 |
|
256 | and n[:3] in _winres4 | |
257 | ): |
|
257 | ): | |
258 | # encode third letter ('aux' -> 'au~78') |
|
258 | # encode third letter ('aux' -> 'au~78') | |
259 | ec = b"~%02x" % ord(n[2:3]) |
|
259 | ec = b"~%02x" % ord(n[2:3]) | |
260 | n = n[0:2] + ec + n[3:] |
|
260 | n = n[0:2] + ec + n[3:] | |
261 | path[i] = n |
|
261 | path[i] = n | |
262 | if n[-1] in b'. ': |
|
262 | if n[-1] in b'. ': | |
263 | # encode last period or space ('foo...' -> 'foo..~2e') |
|
263 | # encode last period or space ('foo...' -> 'foo..~2e') | |
264 | path[i] = n[:-1] + b"~%02x" % ord(n[-1:]) |
|
264 | path[i] = n[:-1] + b"~%02x" % ord(n[-1:]) | |
265 | return path |
|
265 | return path | |
266 |
|
266 | |||
267 |
|
267 | |||
268 | _maxstorepathlen = 120 |
|
268 | _maxstorepathlen = 120 | |
269 | _dirprefixlen = 8 |
|
269 | _dirprefixlen = 8 | |
270 | _maxshortdirslen = 8 * (_dirprefixlen + 1) - 4 |
|
270 | _maxshortdirslen = 8 * (_dirprefixlen + 1) - 4 | |
271 |
|
271 | |||
272 |
|
272 | |||
273 | def _hashencode(path, dotencode): |
|
273 | def _hashencode(path, dotencode): | |
274 | digest = hex(hashutil.sha1(path).digest()) |
|
274 | digest = hex(hashutil.sha1(path).digest()) | |
275 | le = lowerencode(path[5:]).split(b'/') # skips prefix 'data/' or 'meta/' |
|
275 | le = lowerencode(path[5:]).split(b'/') # skips prefix 'data/' or 'meta/' | |
276 | parts = _auxencode(le, dotencode) |
|
276 | parts = _auxencode(le, dotencode) | |
277 | basename = parts[-1] |
|
277 | basename = parts[-1] | |
278 | _root, ext = os.path.splitext(basename) |
|
278 | _root, ext = os.path.splitext(basename) | |
279 | sdirs = [] |
|
279 | sdirs = [] | |
280 | sdirslen = 0 |
|
280 | sdirslen = 0 | |
281 | for p in parts[:-1]: |
|
281 | for p in parts[:-1]: | |
282 | d = p[:_dirprefixlen] |
|
282 | d = p[:_dirprefixlen] | |
283 | if d[-1] in b'. ': |
|
283 | if d[-1] in b'. ': | |
284 | # Windows can't access dirs ending in period or space |
|
284 | # Windows can't access dirs ending in period or space | |
285 | d = d[:-1] + b'_' |
|
285 | d = d[:-1] + b'_' | |
286 | if sdirslen == 0: |
|
286 | if sdirslen == 0: | |
287 | t = len(d) |
|
287 | t = len(d) | |
288 | else: |
|
288 | else: | |
289 | t = sdirslen + 1 + len(d) |
|
289 | t = sdirslen + 1 + len(d) | |
290 | if t > _maxshortdirslen: |
|
290 | if t > _maxshortdirslen: | |
291 | break |
|
291 | break | |
292 | sdirs.append(d) |
|
292 | sdirs.append(d) | |
293 | sdirslen = t |
|
293 | sdirslen = t | |
294 | dirs = b'/'.join(sdirs) |
|
294 | dirs = b'/'.join(sdirs) | |
295 | if len(dirs) > 0: |
|
295 | if len(dirs) > 0: | |
296 | dirs += b'/' |
|
296 | dirs += b'/' | |
297 | res = b'dh/' + dirs + digest + ext |
|
297 | res = b'dh/' + dirs + digest + ext | |
298 | spaceleft = _maxstorepathlen - len(res) |
|
298 | spaceleft = _maxstorepathlen - len(res) | |
299 | if spaceleft > 0: |
|
299 | if spaceleft > 0: | |
300 | filler = basename[:spaceleft] |
|
300 | filler = basename[:spaceleft] | |
301 | res = b'dh/' + dirs + filler + digest + ext |
|
301 | res = b'dh/' + dirs + filler + digest + ext | |
302 | return res |
|
302 | return res | |
303 |
|
303 | |||
304 |
|
304 | |||
305 | def _hybridencode(path, dotencode): |
|
305 | def _hybridencode(path, dotencode): | |
306 | """encodes path with a length limit |
|
306 | """encodes path with a length limit | |
307 |
|
307 | |||
308 | Encodes all paths that begin with 'data/', according to the following. |
|
308 | Encodes all paths that begin with 'data/', according to the following. | |
309 |
|
309 | |||
310 | Default encoding (reversible): |
|
310 | Default encoding (reversible): | |
311 |
|
311 | |||
312 | Encodes all uppercase letters 'X' as '_x'. All reserved or illegal |
|
312 | Encodes all uppercase letters 'X' as '_x'. All reserved or illegal | |
313 | characters are encoded as '~xx', where xx is the two digit hex code |
|
313 | characters are encoded as '~xx', where xx is the two digit hex code | |
314 | of the character (see encodefilename). |
|
314 | of the character (see encodefilename). | |
315 | Relevant path components consisting of Windows reserved filenames are |
|
315 | Relevant path components consisting of Windows reserved filenames are | |
316 | masked by encoding the third character ('aux' -> 'au~78', see _auxencode). |
|
316 | masked by encoding the third character ('aux' -> 'au~78', see _auxencode). | |
317 |
|
317 | |||
318 | Hashed encoding (not reversible): |
|
318 | Hashed encoding (not reversible): | |
319 |
|
319 | |||
320 | If the default-encoded path is longer than _maxstorepathlen, a |
|
320 | If the default-encoded path is longer than _maxstorepathlen, a | |
321 | non-reversible hybrid hashing of the path is done instead. |
|
321 | non-reversible hybrid hashing of the path is done instead. | |
322 | This encoding uses up to _dirprefixlen characters of all directory |
|
322 | This encoding uses up to _dirprefixlen characters of all directory | |
323 | levels of the lowerencoded path, but not more levels than can fit into |
|
323 | levels of the lowerencoded path, but not more levels than can fit into | |
324 | _maxshortdirslen. |
|
324 | _maxshortdirslen. | |
325 | Then follows the filler followed by the sha digest of the full path. |
|
325 | Then follows the filler followed by the sha digest of the full path. | |
326 | The filler is the beginning of the basename of the lowerencoded path |
|
326 | The filler is the beginning of the basename of the lowerencoded path | |
327 | (the basename is everything after the last path separator). The filler |
|
327 | (the basename is everything after the last path separator). The filler | |
328 | is as long as possible, filling in characters from the basename until |
|
328 | is as long as possible, filling in characters from the basename until | |
329 | the encoded path has _maxstorepathlen characters (or all chars of the |
|
329 | the encoded path has _maxstorepathlen characters (or all chars of the | |
330 | basename have been taken). |
|
330 | basename have been taken). | |
331 | The extension (e.g. '.i' or '.d') is preserved. |
|
331 | The extension (e.g. '.i' or '.d') is preserved. | |
332 |
|
332 | |||
333 | The string 'data/' at the beginning is replaced with 'dh/', if the hashed |
|
333 | The string 'data/' at the beginning is replaced with 'dh/', if the hashed | |
334 | encoding was used. |
|
334 | encoding was used. | |
335 | """ |
|
335 | """ | |
336 | path = encodedir(path) |
|
336 | path = encodedir(path) | |
337 | ef = _encodefname(path).split(b'/') |
|
337 | ef = _encodefname(path).split(b'/') | |
338 | res = b'/'.join(_auxencode(ef, dotencode)) |
|
338 | res = b'/'.join(_auxencode(ef, dotencode)) | |
339 | if len(res) > _maxstorepathlen: |
|
339 | if len(res) > _maxstorepathlen: | |
340 | res = _hashencode(path, dotencode) |
|
340 | res = _hashencode(path, dotencode) | |
341 | return res |
|
341 | return res | |
342 |
|
342 | |||
343 |
|
343 | |||
344 | def _pathencode(path): |
|
344 | def _pathencode(path): | |
345 | de = encodedir(path) |
|
345 | de = encodedir(path) | |
346 | if len(path) > _maxstorepathlen: |
|
346 | if len(path) > _maxstorepathlen: | |
347 | return _hashencode(de, True) |
|
347 | return _hashencode(de, True) | |
348 | ef = _encodefname(de).split(b'/') |
|
348 | ef = _encodefname(de).split(b'/') | |
349 | res = b'/'.join(_auxencode(ef, True)) |
|
349 | res = b'/'.join(_auxencode(ef, True)) | |
350 | if len(res) > _maxstorepathlen: |
|
350 | if len(res) > _maxstorepathlen: | |
351 | return _hashencode(de, True) |
|
351 | return _hashencode(de, True) | |
352 | return res |
|
352 | return res | |
353 |
|
353 | |||
354 |
|
354 | |||
355 | _pathencode = getattr(parsers, 'pathencode', _pathencode) |
|
355 | _pathencode = getattr(parsers, 'pathencode', _pathencode) | |
356 |
|
356 | |||
357 |
|
357 | |||
358 | def _plainhybridencode(f): |
|
358 | def _plainhybridencode(f): | |
359 | return _hybridencode(f, False) |
|
359 | return _hybridencode(f, False) | |
360 |
|
360 | |||
361 |
|
361 | |||
362 | def _calcmode(vfs): |
|
362 | def _calcmode(vfs): | |
363 | try: |
|
363 | try: | |
364 | # files in .hg/ will be created using this mode |
|
364 | # files in .hg/ will be created using this mode | |
365 | mode = vfs.stat().st_mode |
|
365 | mode = vfs.stat().st_mode | |
366 | # avoid some useless chmods |
|
366 | # avoid some useless chmods | |
367 | if (0o777 & ~util.umask) == (0o777 & mode): |
|
367 | if (0o777 & ~util.umask) == (0o777 & mode): | |
368 | mode = None |
|
368 | mode = None | |
369 | except OSError: |
|
369 | except OSError: | |
370 | mode = None |
|
370 | mode = None | |
371 | return mode |
|
371 | return mode | |
372 |
|
372 | |||
373 |
|
373 | |||
374 | _data = [ |
|
374 | _data = [ | |
375 | b'bookmarks', |
|
375 | b'bookmarks', | |
376 | b'narrowspec', |
|
376 | b'narrowspec', | |
377 | b'data', |
|
377 | b'data', | |
378 | b'meta', |
|
378 | b'meta', | |
379 | b'00manifest.d', |
|
379 | b'00manifest.d', | |
380 | b'00manifest.i', |
|
380 | b'00manifest.i', | |
381 | b'00changelog.d', |
|
381 | b'00changelog.d', | |
382 | b'00changelog.i', |
|
382 | b'00changelog.i', | |
383 | b'phaseroots', |
|
383 | b'phaseroots', | |
384 | b'obsstore', |
|
384 | b'obsstore', | |
385 | b'requires', |
|
385 | b'requires', | |
386 | ] |
|
386 | ] | |
387 |
|
387 | |||
388 | REVLOG_FILES_MAIN_EXT = (b'.i',) |
|
388 | REVLOG_FILES_MAIN_EXT = (b'.i',) | |
389 | REVLOG_FILES_OTHER_EXT = ( |
|
389 | REVLOG_FILES_OTHER_EXT = ( | |
390 | b'.idx', |
|
390 | b'.idx', | |
391 | b'.d', |
|
391 | b'.d', | |
392 | b'.dat', |
|
392 | b'.dat', | |
393 | b'.n', |
|
393 | b'.n', | |
394 | b'.nd', |
|
394 | b'.nd', | |
395 | b'.sda', |
|
395 | b'.sda', | |
396 | ) |
|
396 | ) | |
397 | # file extension that also use a `-SOMELONGIDHASH.ext` form |
|
397 | # file extension that also use a `-SOMELONGIDHASH.ext` form | |
398 | REVLOG_FILES_LONG_EXT = ( |
|
398 | REVLOG_FILES_LONG_EXT = ( | |
399 | b'.nd', |
|
399 | b'.nd', | |
400 | b'.idx', |
|
400 | b'.idx', | |
401 | b'.dat', |
|
401 | b'.dat', | |
402 | b'.sda', |
|
402 | b'.sda', | |
403 | ) |
|
403 | ) | |
404 | # files that are "volatile" and might change between listing and streaming |
|
404 | # files that are "volatile" and might change between listing and streaming | |
405 | # |
|
405 | # | |
406 | # note: the ".nd" file are nodemap data and won't "change" but they might be |
|
406 | # note: the ".nd" file are nodemap data and won't "change" but they might be | |
407 | # deleted. |
|
407 | # deleted. | |
408 | REVLOG_FILES_VOLATILE_EXT = (b'.n', b'.nd') |
|
408 | REVLOG_FILES_VOLATILE_EXT = (b'.n', b'.nd') | |
409 |
|
409 | |||
410 | # some exception to the above matching |
|
410 | # some exception to the above matching | |
411 | # |
|
411 | # | |
412 | # XXX This is currently not in use because of issue6542 |
|
412 | # XXX This is currently not in use because of issue6542 | |
413 | EXCLUDED = re.compile(br'.*undo\.[^/]+\.(nd?|i)$') |
|
413 | EXCLUDED = re.compile(br'.*undo\.[^/]+\.(nd?|i)$') | |
414 |
|
414 | |||
415 |
|
415 | |||
416 | def is_revlog(f, kind, st): |
|
416 | def is_revlog(f, kind, st): | |
417 | if kind != stat.S_IFREG: |
|
417 | if kind != stat.S_IFREG: | |
418 | return None |
|
418 | return None | |
419 | return revlog_type(f) |
|
419 | return revlog_type(f) | |
420 |
|
420 | |||
421 |
|
421 | |||
422 | def revlog_type(f): |
|
422 | def revlog_type(f): | |
423 | # XXX we need to filter `undo.` created by the transaction here, however |
|
423 | # XXX we need to filter `undo.` created by the transaction here, however | |
424 | # being naive about it also filter revlog for `undo.*` files, leading to |
|
424 | # being naive about it also filter revlog for `undo.*` files, leading to | |
425 | # issue6542. So we no longer use EXCLUDED. |
|
425 | # issue6542. So we no longer use EXCLUDED. | |
426 | if f.endswith(REVLOG_FILES_MAIN_EXT): |
|
426 | if f.endswith(REVLOG_FILES_MAIN_EXT): | |
427 | return FILEFLAGS_REVLOG_MAIN |
|
427 | return FILEFLAGS_REVLOG_MAIN | |
428 | elif f.endswith(REVLOG_FILES_OTHER_EXT): |
|
428 | elif f.endswith(REVLOG_FILES_OTHER_EXT): | |
429 | t = FILETYPE_FILELOG_OTHER |
|
429 | t = FILETYPE_FILELOG_OTHER | |
430 | if f.endswith(REVLOG_FILES_VOLATILE_EXT): |
|
430 | if f.endswith(REVLOG_FILES_VOLATILE_EXT): | |
431 | t |= FILEFLAGS_VOLATILE |
|
431 | t |= FILEFLAGS_VOLATILE | |
432 | return t |
|
432 | return t | |
433 | return None |
|
433 | return None | |
434 |
|
434 | |||
435 |
|
435 | |||
436 | # the file is part of changelog data |
|
436 | # the file is part of changelog data | |
437 | FILEFLAGS_CHANGELOG = 1 << 13 |
|
437 | FILEFLAGS_CHANGELOG = 1 << 13 | |
438 | # the file is part of manifest data |
|
438 | # the file is part of manifest data | |
439 | FILEFLAGS_MANIFESTLOG = 1 << 12 |
|
439 | FILEFLAGS_MANIFESTLOG = 1 << 12 | |
440 | # the file is part of filelog data |
|
440 | # the file is part of filelog data | |
441 | FILEFLAGS_FILELOG = 1 << 11 |
|
441 | FILEFLAGS_FILELOG = 1 << 11 | |
442 | # file that are not directly part of a revlog |
|
442 | # file that are not directly part of a revlog | |
443 | FILEFLAGS_OTHER = 1 << 10 |
|
443 | FILEFLAGS_OTHER = 1 << 10 | |
444 |
|
444 | |||
445 | # the main entry point for a revlog |
|
445 | # the main entry point for a revlog | |
446 | FILEFLAGS_REVLOG_MAIN = 1 << 1 |
|
446 | FILEFLAGS_REVLOG_MAIN = 1 << 1 | |
447 | # a secondary file for a revlog |
|
447 | # a secondary file for a revlog | |
448 | FILEFLAGS_REVLOG_OTHER = 1 << 0 |
|
448 | FILEFLAGS_REVLOG_OTHER = 1 << 0 | |
449 |
|
449 | |||
450 | # files that are "volatile" and might change between listing and streaming |
|
450 | # files that are "volatile" and might change between listing and streaming | |
451 | FILEFLAGS_VOLATILE = 1 << 20 |
|
451 | FILEFLAGS_VOLATILE = 1 << 20 | |
452 |
|
452 | |||
453 | FILETYPE_CHANGELOG_MAIN = FILEFLAGS_CHANGELOG | FILEFLAGS_REVLOG_MAIN |
|
453 | FILETYPE_CHANGELOG_MAIN = FILEFLAGS_CHANGELOG | FILEFLAGS_REVLOG_MAIN | |
454 | FILETYPE_CHANGELOG_OTHER = FILEFLAGS_CHANGELOG | FILEFLAGS_REVLOG_OTHER |
|
454 | FILETYPE_CHANGELOG_OTHER = FILEFLAGS_CHANGELOG | FILEFLAGS_REVLOG_OTHER | |
455 | FILETYPE_MANIFESTLOG_MAIN = FILEFLAGS_MANIFESTLOG | FILEFLAGS_REVLOG_MAIN |
|
455 | FILETYPE_MANIFESTLOG_MAIN = FILEFLAGS_MANIFESTLOG | FILEFLAGS_REVLOG_MAIN | |
456 | FILETYPE_MANIFESTLOG_OTHER = FILEFLAGS_MANIFESTLOG | FILEFLAGS_REVLOG_OTHER |
|
456 | FILETYPE_MANIFESTLOG_OTHER = FILEFLAGS_MANIFESTLOG | FILEFLAGS_REVLOG_OTHER | |
457 | FILETYPE_FILELOG_MAIN = FILEFLAGS_FILELOG | FILEFLAGS_REVLOG_MAIN |
|
457 | FILETYPE_FILELOG_MAIN = FILEFLAGS_FILELOG | FILEFLAGS_REVLOG_MAIN | |
458 | FILETYPE_FILELOG_OTHER = FILEFLAGS_FILELOG | FILEFLAGS_REVLOG_OTHER |
|
458 | FILETYPE_FILELOG_OTHER = FILEFLAGS_FILELOG | FILEFLAGS_REVLOG_OTHER | |
459 | FILETYPE_OTHER = FILEFLAGS_OTHER |
|
459 | FILETYPE_OTHER = FILEFLAGS_OTHER | |
460 |
|
460 | |||
461 |
|
461 | |||
462 | @attr.s(slots=True, init=False) |
|
462 | @attr.s(slots=True, init=False) | |
463 | class BaseStoreEntry: |
|
463 | class BaseStoreEntry: | |
464 | """An entry in the store |
|
464 | """An entry in the store | |
465 |
|
465 | |||
466 | This is returned by `store.walk` and represent some data in the store.""" |
|
466 | This is returned by `store.walk` and represent some data in the store.""" | |
467 |
|
467 | |||
468 |
|
468 | |||
469 | @attr.s(slots=True, init=False) |
|
469 | @attr.s(slots=True, init=False) | |
470 | class SimpleStoreEntry(BaseStoreEntry): |
|
470 | class SimpleStoreEntry(BaseStoreEntry): | |
471 | """A generic entry in the store""" |
|
471 | """A generic entry in the store""" | |
472 |
|
472 | |||
473 | is_revlog = False |
|
473 | is_revlog = False | |
474 |
|
474 | |||
475 | _entry_path = attr.ib() |
|
475 | _entry_path = attr.ib() | |
476 | _is_volatile = attr.ib(default=False) |
|
476 | _is_volatile = attr.ib(default=False) | |
477 | _file_size = attr.ib(default=None) |
|
477 | _file_size = attr.ib(default=None) | |
478 |
|
478 | |||
479 | def __init__( |
|
479 | def __init__( | |
480 | self, |
|
480 | self, | |
481 | entry_path, |
|
481 | entry_path, | |
482 | is_volatile=False, |
|
482 | is_volatile=False, | |
483 | file_size=None, |
|
483 | file_size=None, | |
484 | ): |
|
484 | ): | |
485 | super().__init__() |
|
485 | super().__init__() | |
486 | self._entry_path = entry_path |
|
486 | self._entry_path = entry_path | |
487 | self._is_volatile = is_volatile |
|
487 | self._is_volatile = is_volatile | |
488 | self._file_size = file_size |
|
488 | self._file_size = file_size | |
489 |
|
489 | |||
490 | def files(self): |
|
490 | def files(self): | |
491 | return [ |
|
491 | return [ | |
492 | StoreFile( |
|
492 | StoreFile( | |
493 | unencoded_path=self._entry_path, |
|
493 | unencoded_path=self._entry_path, | |
494 | file_size=self._file_size, |
|
494 | file_size=self._file_size, | |
495 | is_volatile=self._is_volatile, |
|
495 | is_volatile=self._is_volatile, | |
496 | ) |
|
496 | ) | |
497 | ] |
|
497 | ] | |
498 |
|
498 | |||
499 |
|
499 | |||
500 | @attr.s(slots=True, init=False) |
|
500 | @attr.s(slots=True, init=False) | |
501 | class RevlogStoreEntry(BaseStoreEntry): |
|
501 | class RevlogStoreEntry(BaseStoreEntry): | |
502 | """A revlog entry in the store""" |
|
502 | """A revlog entry in the store""" | |
503 |
|
503 | |||
504 | is_revlog = True |
|
504 | is_revlog = True | |
505 |
|
505 | |||
506 | revlog_type = attr.ib(default=None) |
|
506 | revlog_type = attr.ib(default=None) | |
507 | target_id = attr.ib(default=None) |
|
507 | target_id = attr.ib(default=None) | |
508 | _path_prefix = attr.ib(default=None) |
|
508 | _path_prefix = attr.ib(default=None) | |
509 | _details = attr.ib(default=None) |
|
509 | _details = attr.ib(default=None) | |
510 |
|
510 | |||
511 | def __init__( |
|
511 | def __init__( | |
512 | self, |
|
512 | self, | |
513 | revlog_type, |
|
513 | revlog_type, | |
514 | path_prefix, |
|
514 | path_prefix, | |
515 | target_id, |
|
515 | target_id, | |
516 | details, |
|
516 | details, | |
517 | ): |
|
517 | ): | |
518 | super().__init__() |
|
518 | super().__init__() | |
519 | self.revlog_type = revlog_type |
|
519 | self.revlog_type = revlog_type | |
520 | self.target_id = target_id |
|
520 | self.target_id = target_id | |
521 | self._path_prefix = path_prefix |
|
521 | self._path_prefix = path_prefix | |
522 | assert b'.i' in details, (path_prefix, details) |
|
522 | assert b'.i' in details, (path_prefix, details) | |
523 | self._details = details |
|
523 | self._details = details | |
524 |
|
524 | |||
525 | @property |
|
525 | @property | |
526 | def is_changelog(self): |
|
526 | def is_changelog(self): | |
527 | return self.revlog_type & FILEFLAGS_CHANGELOG |
|
527 | return self.revlog_type & FILEFLAGS_CHANGELOG | |
528 |
|
528 | |||
529 | @property |
|
529 | @property | |
530 | def is_manifestlog(self): |
|
530 | def is_manifestlog(self): | |
531 | return self.revlog_type & FILEFLAGS_MANIFESTLOG |
|
531 | return self.revlog_type & FILEFLAGS_MANIFESTLOG | |
532 |
|
532 | |||
533 | @property |
|
533 | @property | |
534 | def is_filelog(self): |
|
534 | def is_filelog(self): | |
535 | return self.revlog_type & FILEFLAGS_FILELOG |
|
535 | return self.revlog_type & FILEFLAGS_FILELOG | |
536 |
|
536 | |||
537 | def main_file_path(self): |
|
537 | def main_file_path(self): | |
538 | """unencoded path of the main revlog file""" |
|
538 | """unencoded path of the main revlog file""" | |
539 | return self._path_prefix + b'.i' |
|
539 | return self._path_prefix + b'.i' | |
540 |
|
540 | |||
541 | def files(self): |
|
541 | def files(self): | |
542 | files = [] |
|
542 | files = [] | |
543 | for ext in sorted(self._details, key=_ext_key): |
|
543 | for ext in sorted(self._details, key=_ext_key): | |
544 | path = self._path_prefix + ext |
|
544 | path = self._path_prefix + ext | |
545 | data = self._details[ext] |
|
545 | data = self._details[ext] | |
546 | files.append(StoreFile(unencoded_path=path, **data)) |
|
546 | files.append(StoreFile(unencoded_path=path, **data)) | |
547 | return files |
|
547 | return files | |
548 |
|
548 | |||
549 | def get_revlog_instance(self, repo): |
|
549 | def get_revlog_instance(self, repo): | |
550 | """Obtain a revlog instance from this store entry |
|
550 | """Obtain a revlog instance from this store entry | |
551 |
|
551 | |||
552 | An instance of the appropriate class is returned. |
|
552 | An instance of the appropriate class is returned. | |
553 | """ |
|
553 | """ | |
554 | if self.is_changelog: |
|
554 | if self.is_changelog: | |
555 | return changelog.changelog(repo.svfs) |
|
555 | return changelog.changelog(repo.svfs) | |
556 | elif self.is_manifestlog: |
|
556 | elif self.is_manifestlog: | |
557 | mandir = self.target_id |
|
557 | mandir = self.target_id | |
558 | return manifest.manifestrevlog( |
|
558 | return manifest.manifestrevlog( | |
559 | repo.nodeconstants, repo.svfs, tree=mandir |
|
559 | repo.nodeconstants, repo.svfs, tree=mandir | |
560 | ) |
|
560 | ) | |
561 | else: |
|
561 | else: | |
562 | return filelog.filelog(repo.svfs, self.target_id) |
|
562 | return filelog.filelog(repo.svfs, self.target_id) | |
563 |
|
563 | |||
564 |
|
564 | |||
565 | @attr.s(slots=True) |
|
565 | @attr.s(slots=True) | |
566 | class StoreFile: |
|
566 | class StoreFile: | |
567 | """a file matching an entry""" |
|
567 | """a file matching an entry""" | |
568 |
|
568 | |||
569 | unencoded_path = attr.ib() |
|
569 | unencoded_path = attr.ib() | |
570 | _file_size = attr.ib(default=None) |
|
570 | _file_size = attr.ib(default=None) | |
571 | is_volatile = attr.ib(default=False) |
|
571 | is_volatile = attr.ib(default=False) | |
572 |
|
572 | |||
573 | def file_size(self, vfs): |
|
573 | def file_size(self, vfs): | |
574 | if self._file_size is not None: |
|
574 | if self._file_size is not None: | |
575 | return self._file_size |
|
575 | return self._file_size | |
576 | try: |
|
576 | try: | |
577 | return vfs.stat(self.unencoded_path).st_size |
|
577 | return vfs.stat(self.unencoded_path).st_size | |
578 | except FileNotFoundError: |
|
578 | except FileNotFoundError: | |
579 | return 0 |
|
579 | return 0 | |
580 |
|
580 | |||
581 |
|
581 | |||
582 | def _gather_revlog(files_data): |
|
582 | def _gather_revlog(files_data): | |
583 | """group files per revlog prefix |
|
583 | """group files per revlog prefix | |
584 |
|
584 | |||
585 | The returns a two level nested dict. The top level key is the revlog prefix |
|
585 | The returns a two level nested dict. The top level key is the revlog prefix | |
586 | without extension, the second level is all the file "suffix" that were |
|
586 | without extension, the second level is all the file "suffix" that were | |
587 | seen for this revlog and arbitrary file data as value. |
|
587 | seen for this revlog and arbitrary file data as value. | |
588 | """ |
|
588 | """ | |
589 | revlogs = collections.defaultdict(dict) |
|
589 | revlogs = collections.defaultdict(dict) | |
590 | for u, value in files_data: |
|
590 | for u, value in files_data: | |
591 | name, ext = _split_revlog_ext(u) |
|
591 | name, ext = _split_revlog_ext(u) | |
592 | revlogs[name][ext] = value |
|
592 | revlogs[name][ext] = value | |
593 | return sorted(revlogs.items()) |
|
593 | return sorted(revlogs.items()) | |
594 |
|
594 | |||
595 |
|
595 | |||
596 | def _split_revlog_ext(filename): |
|
596 | def _split_revlog_ext(filename): | |
597 | """split the revlog file prefix from the variable extension""" |
|
597 | """split the revlog file prefix from the variable extension""" | |
598 | if filename.endswith(REVLOG_FILES_LONG_EXT): |
|
598 | if filename.endswith(REVLOG_FILES_LONG_EXT): | |
599 | char = b'-' |
|
599 | char = b'-' | |
600 | else: |
|
600 | else: | |
601 | char = b'.' |
|
601 | char = b'.' | |
602 | idx = filename.rfind(char) |
|
602 | idx = filename.rfind(char) | |
603 | return filename[:idx], filename[idx:] |
|
603 | return filename[:idx], filename[idx:] | |
604 |
|
604 | |||
605 |
|
605 | |||
606 | def _ext_key(ext): |
|
606 | def _ext_key(ext): | |
607 | """a key to order revlog suffix |
|
607 | """a key to order revlog suffix | |
608 |
|
608 | |||
609 | important to issue .i after other entry.""" |
|
609 | important to issue .i after other entry.""" | |
610 | # the only important part of this order is to keep the `.i` last. |
|
610 | # the only important part of this order is to keep the `.i` last. | |
611 | if ext.endswith(b'.n'): |
|
611 | if ext.endswith(b'.n'): | |
612 | return (0, ext) |
|
612 | return (0, ext) | |
613 | elif ext.endswith(b'.nd'): |
|
613 | elif ext.endswith(b'.nd'): | |
614 | return (10, ext) |
|
614 | return (10, ext) | |
615 | elif ext.endswith(b'.d'): |
|
615 | elif ext.endswith(b'.d'): | |
616 | return (20, ext) |
|
616 | return (20, ext) | |
617 | elif ext.endswith(b'.i'): |
|
617 | elif ext.endswith(b'.i'): | |
618 | return (50, ext) |
|
618 | return (50, ext) | |
619 | else: |
|
619 | else: | |
620 | return (40, ext) |
|
620 | return (40, ext) | |
621 |
|
621 | |||
622 |
|
622 | |||
623 | class basicstore: |
|
623 | class basicstore: | |
624 | '''base class for local repository stores''' |
|
624 | '''base class for local repository stores''' | |
625 |
|
625 | |||
626 | def __init__(self, path, vfstype): |
|
626 | def __init__(self, path, vfstype): | |
627 | vfs = vfstype(path) |
|
627 | vfs = vfstype(path) | |
628 | self.path = vfs.base |
|
628 | self.path = vfs.base | |
629 | self.createmode = _calcmode(vfs) |
|
629 | self.createmode = _calcmode(vfs) | |
630 | vfs.createmode = self.createmode |
|
630 | vfs.createmode = self.createmode | |
631 | self.rawvfs = vfs |
|
631 | self.rawvfs = vfs | |
632 | self.vfs = vfsmod.filtervfs(vfs, encodedir) |
|
632 | self.vfs = vfsmod.filtervfs(vfs, encodedir) | |
633 | self.opener = self.vfs |
|
633 | self.opener = self.vfs | |
634 |
|
634 | |||
635 | def join(self, f): |
|
635 | def join(self, f): | |
636 | return self.path + b'/' + encodedir(f) |
|
636 | return self.path + b'/' + encodedir(f) | |
637 |
|
637 | |||
638 | def _walk(self, relpath, recurse, undecodable=None): |
|
638 | def _walk(self, relpath, recurse, undecodable=None): | |
639 | '''yields (revlog_type, unencoded, size)''' |
|
639 | '''yields (revlog_type, unencoded, size)''' | |
640 | path = self.path |
|
640 | path = self.path | |
641 | if relpath: |
|
641 | if relpath: | |
642 | path += b'/' + relpath |
|
642 | path += b'/' + relpath | |
643 | striplen = len(self.path) + 1 |
|
643 | striplen = len(self.path) + 1 | |
644 | l = [] |
|
644 | l = [] | |
645 | if self.rawvfs.isdir(path): |
|
645 | if self.rawvfs.isdir(path): | |
646 | visit = [path] |
|
646 | visit = [path] | |
647 | readdir = self.rawvfs.readdir |
|
647 | readdir = self.rawvfs.readdir | |
648 | while visit: |
|
648 | while visit: | |
649 | p = visit.pop() |
|
649 | p = visit.pop() | |
650 | for f, kind, st in readdir(p, stat=True): |
|
650 | for f, kind, st in readdir(p, stat=True): | |
651 | fp = p + b'/' + f |
|
651 | fp = p + b'/' + f | |
652 | rl_type = is_revlog(f, kind, st) |
|
652 | rl_type = is_revlog(f, kind, st) | |
653 | if rl_type is not None: |
|
653 | if rl_type is not None: | |
654 | n = util.pconvert(fp[striplen:]) |
|
654 | n = util.pconvert(fp[striplen:]) | |
655 | l.append((decodedir(n), (rl_type, st.st_size))) |
|
655 | l.append((decodedir(n), (rl_type, st.st_size))) | |
656 | elif kind == stat.S_IFDIR and recurse: |
|
656 | elif kind == stat.S_IFDIR and recurse: | |
657 | visit.append(fp) |
|
657 | visit.append(fp) | |
658 |
|
658 | |||
659 | l.sort() |
|
659 | l.sort() | |
660 | return l |
|
660 | return l | |
661 |
|
661 | |||
662 | def changelog(self, trypending, concurrencychecker=None): |
|
662 | def changelog(self, trypending, concurrencychecker=None): | |
663 | return changelog.changelog( |
|
663 | return changelog.changelog( | |
664 | self.vfs, |
|
664 | self.vfs, | |
665 | trypending=trypending, |
|
665 | trypending=trypending, | |
666 | concurrencychecker=concurrencychecker, |
|
666 | concurrencychecker=concurrencychecker, | |
667 | ) |
|
667 | ) | |
668 |
|
668 | |||
669 | def manifestlog(self, repo, storenarrowmatch): |
|
669 | def manifestlog(self, repo, storenarrowmatch): | |
670 | rootstore = manifest.manifestrevlog(repo.nodeconstants, self.vfs) |
|
670 | rootstore = manifest.manifestrevlog(repo.nodeconstants, self.vfs) | |
671 | return manifest.manifestlog(self.vfs, repo, rootstore, storenarrowmatch) |
|
671 | return manifest.manifestlog(self.vfs, repo, rootstore, storenarrowmatch) | |
672 |
|
672 | |||
673 | def data_entries( |
|
673 | def data_entries( | |
674 | self, matcher=None, undecodable=None |
|
674 | self, matcher=None, undecodable=None | |
675 | ) -> Generator[BaseStoreEntry, None, None]: |
|
675 | ) -> Generator[BaseStoreEntry, None, None]: | |
676 | """Like walk, but excluding the changelog and root manifest. |
|
676 | """Like walk, but excluding the changelog and root manifest. | |
677 |
|
677 | |||
678 | When [undecodable] is None, revlogs names that can't be |
|
678 | When [undecodable] is None, revlogs names that can't be | |
679 | decoded cause an exception. When it is provided, it should |
|
679 | decoded cause an exception. When it is provided, it should | |
680 | be a list and the filenames that can't be decoded are added |
|
680 | be a list and the filenames that can't be decoded are added | |
681 | to it instead. This is very rarely needed.""" |
|
681 | to it instead. This is very rarely needed.""" | |
682 | dirs = [ |
|
682 | dirs = [ | |
683 | (b'data', FILEFLAGS_FILELOG), |
|
683 | (b'data', FILEFLAGS_FILELOG, False), | |
684 | (b'meta', FILEFLAGS_MANIFESTLOG), |
|
684 | (b'meta', FILEFLAGS_MANIFESTLOG, True), | |
685 | ] |
|
685 | ] | |
686 | for base_dir, rl_type in dirs: |
|
686 | for base_dir, rl_type, strip_filename in dirs: | |
687 | files = self._walk(base_dir, True, undecodable=undecodable) |
|
687 | files = self._walk(base_dir, True, undecodable=undecodable) | |
688 | files = (f for f in files if f[1][0] is not None) |
|
688 | files = (f for f in files if f[1][0] is not None) | |
689 | for revlog, details in _gather_revlog(files): |
|
689 | for revlog, details in _gather_revlog(files): | |
690 | file_details = {} |
|
690 | file_details = {} | |
691 | revlog_target_id = revlog.split(b'/', 1)[1] |
|
691 | revlog_target_id = revlog.split(b'/', 1)[1] | |
|
692 | if strip_filename and b'/' in revlog: | |||
|
693 | revlog_target_id = revlog_target_id.rsplit(b'/', 1)[0] | |||
|
694 | revlog_target_id += b'/' | |||
692 | for ext, (t, s) in sorted(details.items()): |
|
695 | for ext, (t, s) in sorted(details.items()): | |
693 | file_details[ext] = { |
|
696 | file_details[ext] = { | |
694 | 'is_volatile': bool(t & FILEFLAGS_VOLATILE), |
|
697 | 'is_volatile': bool(t & FILEFLAGS_VOLATILE), | |
695 | 'file_size': s, |
|
698 | 'file_size': s, | |
696 | } |
|
699 | } | |
697 | yield RevlogStoreEntry( |
|
700 | yield RevlogStoreEntry( | |
698 | path_prefix=revlog, |
|
701 | path_prefix=revlog, | |
699 | revlog_type=rl_type, |
|
702 | revlog_type=rl_type, | |
700 | target_id=revlog_target_id, |
|
703 | target_id=revlog_target_id, | |
701 | details=file_details, |
|
704 | details=file_details, | |
702 | ) |
|
705 | ) | |
703 |
|
706 | |||
704 | def top_entries( |
|
707 | def top_entries( | |
705 | self, phase=False, obsolescence=False |
|
708 | self, phase=False, obsolescence=False | |
706 | ) -> Generator[BaseStoreEntry, None, None]: |
|
709 | ) -> Generator[BaseStoreEntry, None, None]: | |
707 | if phase and self.vfs.exists(b'phaseroots'): |
|
710 | if phase and self.vfs.exists(b'phaseroots'): | |
708 | yield SimpleStoreEntry( |
|
711 | yield SimpleStoreEntry( | |
709 | entry_path=b'phaseroots', |
|
712 | entry_path=b'phaseroots', | |
710 | is_volatile=True, |
|
713 | is_volatile=True, | |
711 | ) |
|
714 | ) | |
712 |
|
715 | |||
713 | if obsolescence and self.vfs.exists(b'obsstore'): |
|
716 | if obsolescence and self.vfs.exists(b'obsstore'): | |
714 | # XXX if we had the file size it could be non-volatile |
|
717 | # XXX if we had the file size it could be non-volatile | |
715 | yield SimpleStoreEntry( |
|
718 | yield SimpleStoreEntry( | |
716 | entry_path=b'obsstore', |
|
719 | entry_path=b'obsstore', | |
717 | is_volatile=True, |
|
720 | is_volatile=True, | |
718 | ) |
|
721 | ) | |
719 |
|
722 | |||
720 | files = reversed(self._walk(b'', False)) |
|
723 | files = reversed(self._walk(b'', False)) | |
721 |
|
724 | |||
722 | changelogs = collections.defaultdict(dict) |
|
725 | changelogs = collections.defaultdict(dict) | |
723 | manifestlogs = collections.defaultdict(dict) |
|
726 | manifestlogs = collections.defaultdict(dict) | |
724 |
|
727 | |||
725 | for u, (t, s) in files: |
|
728 | for u, (t, s) in files: | |
726 | if u.startswith(b'00changelog'): |
|
729 | if u.startswith(b'00changelog'): | |
727 | name, ext = _split_revlog_ext(u) |
|
730 | name, ext = _split_revlog_ext(u) | |
728 | changelogs[name][ext] = (t, s) |
|
731 | changelogs[name][ext] = (t, s) | |
729 | elif u.startswith(b'00manifest'): |
|
732 | elif u.startswith(b'00manifest'): | |
730 | name, ext = _split_revlog_ext(u) |
|
733 | name, ext = _split_revlog_ext(u) | |
731 | manifestlogs[name][ext] = (t, s) |
|
734 | manifestlogs[name][ext] = (t, s) | |
732 | else: |
|
735 | else: | |
733 | yield SimpleStoreEntry( |
|
736 | yield SimpleStoreEntry( | |
734 | entry_path=u, |
|
737 | entry_path=u, | |
735 | is_volatile=bool(t & FILEFLAGS_VOLATILE), |
|
738 | is_volatile=bool(t & FILEFLAGS_VOLATILE), | |
736 | file_size=s, |
|
739 | file_size=s, | |
737 | ) |
|
740 | ) | |
738 | # yield manifest before changelog |
|
741 | # yield manifest before changelog | |
739 | top_rl = [ |
|
742 | top_rl = [ | |
740 | (manifestlogs, FILEFLAGS_MANIFESTLOG), |
|
743 | (manifestlogs, FILEFLAGS_MANIFESTLOG), | |
741 | (changelogs, FILEFLAGS_CHANGELOG), |
|
744 | (changelogs, FILEFLAGS_CHANGELOG), | |
742 | ] |
|
745 | ] | |
743 | assert len(manifestlogs) <= 1 |
|
746 | assert len(manifestlogs) <= 1 | |
744 | assert len(changelogs) <= 1 |
|
747 | assert len(changelogs) <= 1 | |
745 | for data, revlog_type in top_rl: |
|
748 | for data, revlog_type in top_rl: | |
746 | for revlog, details in sorted(data.items()): |
|
749 | for revlog, details in sorted(data.items()): | |
747 | file_details = {} |
|
750 | file_details = {} | |
748 | for ext, (t, s) in details.items(): |
|
751 | for ext, (t, s) in details.items(): | |
749 | file_details[ext] = { |
|
752 | file_details[ext] = { | |
750 | 'is_volatile': bool(t & FILEFLAGS_VOLATILE), |
|
753 | 'is_volatile': bool(t & FILEFLAGS_VOLATILE), | |
751 | 'file_size': s, |
|
754 | 'file_size': s, | |
752 | } |
|
755 | } | |
753 | yield RevlogStoreEntry( |
|
756 | yield RevlogStoreEntry( | |
754 | path_prefix=revlog, |
|
757 | path_prefix=revlog, | |
755 | revlog_type=revlog_type, |
|
758 | revlog_type=revlog_type, | |
756 | target_id=b'', |
|
759 | target_id=b'', | |
757 | details=file_details, |
|
760 | details=file_details, | |
758 | ) |
|
761 | ) | |
759 |
|
762 | |||
760 | def walk( |
|
763 | def walk( | |
761 | self, matcher=None, phase=False, obsolescence=False |
|
764 | self, matcher=None, phase=False, obsolescence=False | |
762 | ) -> Generator[BaseStoreEntry, None, None]: |
|
765 | ) -> Generator[BaseStoreEntry, None, None]: | |
763 | """return files related to data storage (ie: revlogs) |
|
766 | """return files related to data storage (ie: revlogs) | |
764 |
|
767 | |||
765 | yields instance from BaseStoreEntry subclasses |
|
768 | yields instance from BaseStoreEntry subclasses | |
766 |
|
769 | |||
767 | if a matcher is passed, storage files of only those tracked paths |
|
770 | if a matcher is passed, storage files of only those tracked paths | |
768 | are passed with matches the matcher |
|
771 | are passed with matches the matcher | |
769 | """ |
|
772 | """ | |
770 | # yield data files first |
|
773 | # yield data files first | |
771 | for x in self.data_entries(matcher): |
|
774 | for x in self.data_entries(matcher): | |
772 | yield x |
|
775 | yield x | |
773 | for x in self.top_entries(phase=phase, obsolescence=obsolescence): |
|
776 | for x in self.top_entries(phase=phase, obsolescence=obsolescence): | |
774 | yield x |
|
777 | yield x | |
775 |
|
778 | |||
776 | def copylist(self): |
|
779 | def copylist(self): | |
777 | return _data |
|
780 | return _data | |
778 |
|
781 | |||
779 | def write(self, tr): |
|
782 | def write(self, tr): | |
780 | pass |
|
783 | pass | |
781 |
|
784 | |||
782 | def invalidatecaches(self): |
|
785 | def invalidatecaches(self): | |
783 | pass |
|
786 | pass | |
784 |
|
787 | |||
785 | def markremoved(self, fn): |
|
788 | def markremoved(self, fn): | |
786 | pass |
|
789 | pass | |
787 |
|
790 | |||
788 | def __contains__(self, path): |
|
791 | def __contains__(self, path): | |
789 | '''Checks if the store contains path''' |
|
792 | '''Checks if the store contains path''' | |
790 | path = b"/".join((b"data", path)) |
|
793 | path = b"/".join((b"data", path)) | |
791 | # file? |
|
794 | # file? | |
792 | if self.vfs.exists(path + b".i"): |
|
795 | if self.vfs.exists(path + b".i"): | |
793 | return True |
|
796 | return True | |
794 | # dir? |
|
797 | # dir? | |
795 | if not path.endswith(b"/"): |
|
798 | if not path.endswith(b"/"): | |
796 | path = path + b"/" |
|
799 | path = path + b"/" | |
797 | return self.vfs.exists(path) |
|
800 | return self.vfs.exists(path) | |
798 |
|
801 | |||
799 |
|
802 | |||
800 | class encodedstore(basicstore): |
|
803 | class encodedstore(basicstore): | |
801 | def __init__(self, path, vfstype): |
|
804 | def __init__(self, path, vfstype): | |
802 | vfs = vfstype(path + b'/store') |
|
805 | vfs = vfstype(path + b'/store') | |
803 | self.path = vfs.base |
|
806 | self.path = vfs.base | |
804 | self.createmode = _calcmode(vfs) |
|
807 | self.createmode = _calcmode(vfs) | |
805 | vfs.createmode = self.createmode |
|
808 | vfs.createmode = self.createmode | |
806 | self.rawvfs = vfs |
|
809 | self.rawvfs = vfs | |
807 | self.vfs = vfsmod.filtervfs(vfs, encodefilename) |
|
810 | self.vfs = vfsmod.filtervfs(vfs, encodefilename) | |
808 | self.opener = self.vfs |
|
811 | self.opener = self.vfs | |
809 |
|
812 | |||
810 | def _walk(self, relpath, recurse, undecodable=None): |
|
813 | def _walk(self, relpath, recurse, undecodable=None): | |
811 | old = super()._walk(relpath, recurse) |
|
814 | old = super()._walk(relpath, recurse) | |
812 | new = [] |
|
815 | new = [] | |
813 | for f1, value in old: |
|
816 | for f1, value in old: | |
814 | try: |
|
817 | try: | |
815 | f2 = decodefilename(f1) |
|
818 | f2 = decodefilename(f1) | |
816 | except KeyError: |
|
819 | except KeyError: | |
817 | if undecodable is None: |
|
820 | if undecodable is None: | |
818 | msg = _(b'undecodable revlog name %s') % f1 |
|
821 | msg = _(b'undecodable revlog name %s') % f1 | |
819 | raise error.StorageError(msg) |
|
822 | raise error.StorageError(msg) | |
820 | else: |
|
823 | else: | |
821 | undecodable.append(f1) |
|
824 | undecodable.append(f1) | |
822 | continue |
|
825 | continue | |
823 | new.append((f2, value)) |
|
826 | new.append((f2, value)) | |
824 | return new |
|
827 | return new | |
825 |
|
828 | |||
826 | def data_entries( |
|
829 | def data_entries( | |
827 | self, matcher=None, undecodable=None |
|
830 | self, matcher=None, undecodable=None | |
828 | ) -> Generator[BaseStoreEntry, None, None]: |
|
831 | ) -> Generator[BaseStoreEntry, None, None]: | |
829 | entries = super(encodedstore, self).data_entries( |
|
832 | entries = super(encodedstore, self).data_entries( | |
830 | undecodable=undecodable |
|
833 | undecodable=undecodable | |
831 | ) |
|
834 | ) | |
832 | for entry in entries: |
|
835 | for entry in entries: | |
833 | if _match_tracked_entry(entry, matcher): |
|
836 | if _match_tracked_entry(entry, matcher): | |
834 | yield entry |
|
837 | yield entry | |
835 |
|
838 | |||
836 | def join(self, f): |
|
839 | def join(self, f): | |
837 | return self.path + b'/' + encodefilename(f) |
|
840 | return self.path + b'/' + encodefilename(f) | |
838 |
|
841 | |||
839 | def copylist(self): |
|
842 | def copylist(self): | |
840 | return [b'requires', b'00changelog.i'] + [b'store/' + f for f in _data] |
|
843 | return [b'requires', b'00changelog.i'] + [b'store/' + f for f in _data] | |
841 |
|
844 | |||
842 |
|
845 | |||
843 | class fncache: |
|
846 | class fncache: | |
844 | # the filename used to be partially encoded |
|
847 | # the filename used to be partially encoded | |
845 | # hence the encodedir/decodedir dance |
|
848 | # hence the encodedir/decodedir dance | |
846 | def __init__(self, vfs): |
|
849 | def __init__(self, vfs): | |
847 | self.vfs = vfs |
|
850 | self.vfs = vfs | |
848 | self._ignores = set() |
|
851 | self._ignores = set() | |
849 | self.entries = None |
|
852 | self.entries = None | |
850 | self._dirty = False |
|
853 | self._dirty = False | |
851 | # set of new additions to fncache |
|
854 | # set of new additions to fncache | |
852 | self.addls = set() |
|
855 | self.addls = set() | |
853 |
|
856 | |||
854 | def ensureloaded(self, warn=None): |
|
857 | def ensureloaded(self, warn=None): | |
855 | """read the fncache file if not already read. |
|
858 | """read the fncache file if not already read. | |
856 |
|
859 | |||
857 | If the file on disk is corrupted, raise. If warn is provided, |
|
860 | If the file on disk is corrupted, raise. If warn is provided, | |
858 | warn and keep going instead.""" |
|
861 | warn and keep going instead.""" | |
859 | if self.entries is None: |
|
862 | if self.entries is None: | |
860 | self._load(warn) |
|
863 | self._load(warn) | |
861 |
|
864 | |||
862 | def _load(self, warn=None): |
|
865 | def _load(self, warn=None): | |
863 | '''fill the entries from the fncache file''' |
|
866 | '''fill the entries from the fncache file''' | |
864 | self._dirty = False |
|
867 | self._dirty = False | |
865 | try: |
|
868 | try: | |
866 | fp = self.vfs(b'fncache', mode=b'rb') |
|
869 | fp = self.vfs(b'fncache', mode=b'rb') | |
867 | except IOError: |
|
870 | except IOError: | |
868 | # skip nonexistent file |
|
871 | # skip nonexistent file | |
869 | self.entries = set() |
|
872 | self.entries = set() | |
870 | return |
|
873 | return | |
871 |
|
874 | |||
872 | self.entries = set() |
|
875 | self.entries = set() | |
873 | chunk = b'' |
|
876 | chunk = b'' | |
874 | for c in iter(functools.partial(fp.read, fncache_chunksize), b''): |
|
877 | for c in iter(functools.partial(fp.read, fncache_chunksize), b''): | |
875 | chunk += c |
|
878 | chunk += c | |
876 | try: |
|
879 | try: | |
877 | p = chunk.rindex(b'\n') |
|
880 | p = chunk.rindex(b'\n') | |
878 | self.entries.update(decodedir(chunk[: p + 1]).splitlines()) |
|
881 | self.entries.update(decodedir(chunk[: p + 1]).splitlines()) | |
879 | chunk = chunk[p + 1 :] |
|
882 | chunk = chunk[p + 1 :] | |
880 | except ValueError: |
|
883 | except ValueError: | |
881 | # substring '\n' not found, maybe the entry is bigger than the |
|
884 | # substring '\n' not found, maybe the entry is bigger than the | |
882 | # chunksize, so let's keep iterating |
|
885 | # chunksize, so let's keep iterating | |
883 | pass |
|
886 | pass | |
884 |
|
887 | |||
885 | if chunk: |
|
888 | if chunk: | |
886 | msg = _(b"fncache does not ends with a newline") |
|
889 | msg = _(b"fncache does not ends with a newline") | |
887 | if warn: |
|
890 | if warn: | |
888 | warn(msg + b'\n') |
|
891 | warn(msg + b'\n') | |
889 | else: |
|
892 | else: | |
890 | raise error.Abort( |
|
893 | raise error.Abort( | |
891 | msg, |
|
894 | msg, | |
892 | hint=_( |
|
895 | hint=_( | |
893 | b"use 'hg debugrebuildfncache' to " |
|
896 | b"use 'hg debugrebuildfncache' to " | |
894 | b"rebuild the fncache" |
|
897 | b"rebuild the fncache" | |
895 | ), |
|
898 | ), | |
896 | ) |
|
899 | ) | |
897 | self._checkentries(fp, warn) |
|
900 | self._checkentries(fp, warn) | |
898 | fp.close() |
|
901 | fp.close() | |
899 |
|
902 | |||
900 | def _checkentries(self, fp, warn): |
|
903 | def _checkentries(self, fp, warn): | |
901 | """make sure there is no empty string in entries""" |
|
904 | """make sure there is no empty string in entries""" | |
902 | if b'' in self.entries: |
|
905 | if b'' in self.entries: | |
903 | fp.seek(0) |
|
906 | fp.seek(0) | |
904 | for n, line in enumerate(fp): |
|
907 | for n, line in enumerate(fp): | |
905 | if not line.rstrip(b'\n'): |
|
908 | if not line.rstrip(b'\n'): | |
906 | t = _(b'invalid entry in fncache, line %d') % (n + 1) |
|
909 | t = _(b'invalid entry in fncache, line %d') % (n + 1) | |
907 | if warn: |
|
910 | if warn: | |
908 | warn(t + b'\n') |
|
911 | warn(t + b'\n') | |
909 | else: |
|
912 | else: | |
910 | raise error.Abort(t) |
|
913 | raise error.Abort(t) | |
911 |
|
914 | |||
912 | def write(self, tr): |
|
915 | def write(self, tr): | |
913 | if self._dirty: |
|
916 | if self._dirty: | |
914 | assert self.entries is not None |
|
917 | assert self.entries is not None | |
915 | self.entries = self.entries | self.addls |
|
918 | self.entries = self.entries | self.addls | |
916 | self.addls = set() |
|
919 | self.addls = set() | |
917 | tr.addbackup(b'fncache') |
|
920 | tr.addbackup(b'fncache') | |
918 | fp = self.vfs(b'fncache', mode=b'wb', atomictemp=True) |
|
921 | fp = self.vfs(b'fncache', mode=b'wb', atomictemp=True) | |
919 | if self.entries: |
|
922 | if self.entries: | |
920 | fp.write(encodedir(b'\n'.join(self.entries) + b'\n')) |
|
923 | fp.write(encodedir(b'\n'.join(self.entries) + b'\n')) | |
921 | fp.close() |
|
924 | fp.close() | |
922 | self._dirty = False |
|
925 | self._dirty = False | |
923 | if self.addls: |
|
926 | if self.addls: | |
924 | # if we have just new entries, let's append them to the fncache |
|
927 | # if we have just new entries, let's append them to the fncache | |
925 | tr.addbackup(b'fncache') |
|
928 | tr.addbackup(b'fncache') | |
926 | fp = self.vfs(b'fncache', mode=b'ab', atomictemp=True) |
|
929 | fp = self.vfs(b'fncache', mode=b'ab', atomictemp=True) | |
927 | if self.addls: |
|
930 | if self.addls: | |
928 | fp.write(encodedir(b'\n'.join(self.addls) + b'\n')) |
|
931 | fp.write(encodedir(b'\n'.join(self.addls) + b'\n')) | |
929 | fp.close() |
|
932 | fp.close() | |
930 | self.entries = None |
|
933 | self.entries = None | |
931 | self.addls = set() |
|
934 | self.addls = set() | |
932 |
|
935 | |||
933 | def addignore(self, fn): |
|
936 | def addignore(self, fn): | |
934 | self._ignores.add(fn) |
|
937 | self._ignores.add(fn) | |
935 |
|
938 | |||
936 | def add(self, fn): |
|
939 | def add(self, fn): | |
937 | if fn in self._ignores: |
|
940 | if fn in self._ignores: | |
938 | return |
|
941 | return | |
939 | if self.entries is None: |
|
942 | if self.entries is None: | |
940 | self._load() |
|
943 | self._load() | |
941 | if fn not in self.entries: |
|
944 | if fn not in self.entries: | |
942 | self.addls.add(fn) |
|
945 | self.addls.add(fn) | |
943 |
|
946 | |||
944 | def remove(self, fn): |
|
947 | def remove(self, fn): | |
945 | if self.entries is None: |
|
948 | if self.entries is None: | |
946 | self._load() |
|
949 | self._load() | |
947 | if fn in self.addls: |
|
950 | if fn in self.addls: | |
948 | self.addls.remove(fn) |
|
951 | self.addls.remove(fn) | |
949 | return |
|
952 | return | |
950 | try: |
|
953 | try: | |
951 | self.entries.remove(fn) |
|
954 | self.entries.remove(fn) | |
952 | self._dirty = True |
|
955 | self._dirty = True | |
953 | except KeyError: |
|
956 | except KeyError: | |
954 | pass |
|
957 | pass | |
955 |
|
958 | |||
956 | def __contains__(self, fn): |
|
959 | def __contains__(self, fn): | |
957 | if fn in self.addls: |
|
960 | if fn in self.addls: | |
958 | return True |
|
961 | return True | |
959 | if self.entries is None: |
|
962 | if self.entries is None: | |
960 | self._load() |
|
963 | self._load() | |
961 | return fn in self.entries |
|
964 | return fn in self.entries | |
962 |
|
965 | |||
963 | def __iter__(self): |
|
966 | def __iter__(self): | |
964 | if self.entries is None: |
|
967 | if self.entries is None: | |
965 | self._load() |
|
968 | self._load() | |
966 | return iter(self.entries | self.addls) |
|
969 | return iter(self.entries | self.addls) | |
967 |
|
970 | |||
968 |
|
971 | |||
969 | class _fncachevfs(vfsmod.proxyvfs): |
|
972 | class _fncachevfs(vfsmod.proxyvfs): | |
970 | def __init__(self, vfs, fnc, encode): |
|
973 | def __init__(self, vfs, fnc, encode): | |
971 | vfsmod.proxyvfs.__init__(self, vfs) |
|
974 | vfsmod.proxyvfs.__init__(self, vfs) | |
972 | self.fncache = fnc |
|
975 | self.fncache = fnc | |
973 | self.encode = encode |
|
976 | self.encode = encode | |
974 |
|
977 | |||
975 | def __call__(self, path, mode=b'r', *args, **kw): |
|
978 | def __call__(self, path, mode=b'r', *args, **kw): | |
976 | encoded = self.encode(path) |
|
979 | encoded = self.encode(path) | |
977 | if ( |
|
980 | if ( | |
978 | mode not in (b'r', b'rb') |
|
981 | mode not in (b'r', b'rb') | |
979 | and (path.startswith(b'data/') or path.startswith(b'meta/')) |
|
982 | and (path.startswith(b'data/') or path.startswith(b'meta/')) | |
980 | and revlog_type(path) is not None |
|
983 | and revlog_type(path) is not None | |
981 | ): |
|
984 | ): | |
982 | # do not trigger a fncache load when adding a file that already is |
|
985 | # do not trigger a fncache load when adding a file that already is | |
983 | # known to exist. |
|
986 | # known to exist. | |
984 | notload = self.fncache.entries is None and self.vfs.exists(encoded) |
|
987 | notload = self.fncache.entries is None and self.vfs.exists(encoded) | |
985 | if notload and b'r+' in mode and not self.vfs.stat(encoded).st_size: |
|
988 | if notload and b'r+' in mode and not self.vfs.stat(encoded).st_size: | |
986 | # when appending to an existing file, if the file has size zero, |
|
989 | # when appending to an existing file, if the file has size zero, | |
987 | # it should be considered as missing. Such zero-size files are |
|
990 | # it should be considered as missing. Such zero-size files are | |
988 | # the result of truncation when a transaction is aborted. |
|
991 | # the result of truncation when a transaction is aborted. | |
989 | notload = False |
|
992 | notload = False | |
990 | if not notload: |
|
993 | if not notload: | |
991 | self.fncache.add(path) |
|
994 | self.fncache.add(path) | |
992 | return self.vfs(encoded, mode, *args, **kw) |
|
995 | return self.vfs(encoded, mode, *args, **kw) | |
993 |
|
996 | |||
994 | def join(self, path): |
|
997 | def join(self, path): | |
995 | if path: |
|
998 | if path: | |
996 | return self.vfs.join(self.encode(path)) |
|
999 | return self.vfs.join(self.encode(path)) | |
997 | else: |
|
1000 | else: | |
998 | return self.vfs.join(path) |
|
1001 | return self.vfs.join(path) | |
999 |
|
1002 | |||
1000 | def register_file(self, path): |
|
1003 | def register_file(self, path): | |
1001 | """generic hook point to lets fncache steer its stew""" |
|
1004 | """generic hook point to lets fncache steer its stew""" | |
1002 | if path.startswith(b'data/') or path.startswith(b'meta/'): |
|
1005 | if path.startswith(b'data/') or path.startswith(b'meta/'): | |
1003 | self.fncache.add(path) |
|
1006 | self.fncache.add(path) | |
1004 |
|
1007 | |||
1005 |
|
1008 | |||
1006 | class fncachestore(basicstore): |
|
1009 | class fncachestore(basicstore): | |
1007 | def __init__(self, path, vfstype, dotencode): |
|
1010 | def __init__(self, path, vfstype, dotencode): | |
1008 | if dotencode: |
|
1011 | if dotencode: | |
1009 | encode = _pathencode |
|
1012 | encode = _pathencode | |
1010 | else: |
|
1013 | else: | |
1011 | encode = _plainhybridencode |
|
1014 | encode = _plainhybridencode | |
1012 | self.encode = encode |
|
1015 | self.encode = encode | |
1013 | vfs = vfstype(path + b'/store') |
|
1016 | vfs = vfstype(path + b'/store') | |
1014 | self.path = vfs.base |
|
1017 | self.path = vfs.base | |
1015 | self.pathsep = self.path + b'/' |
|
1018 | self.pathsep = self.path + b'/' | |
1016 | self.createmode = _calcmode(vfs) |
|
1019 | self.createmode = _calcmode(vfs) | |
1017 | vfs.createmode = self.createmode |
|
1020 | vfs.createmode = self.createmode | |
1018 | self.rawvfs = vfs |
|
1021 | self.rawvfs = vfs | |
1019 | fnc = fncache(vfs) |
|
1022 | fnc = fncache(vfs) | |
1020 | self.fncache = fnc |
|
1023 | self.fncache = fnc | |
1021 | self.vfs = _fncachevfs(vfs, fnc, encode) |
|
1024 | self.vfs = _fncachevfs(vfs, fnc, encode) | |
1022 | self.opener = self.vfs |
|
1025 | self.opener = self.vfs | |
1023 |
|
1026 | |||
1024 | def join(self, f): |
|
1027 | def join(self, f): | |
1025 | return self.pathsep + self.encode(f) |
|
1028 | return self.pathsep + self.encode(f) | |
1026 |
|
1029 | |||
1027 | def getsize(self, path): |
|
1030 | def getsize(self, path): | |
1028 | return self.rawvfs.stat(path).st_size |
|
1031 | return self.rawvfs.stat(path).st_size | |
1029 |
|
1032 | |||
1030 | def data_entries( |
|
1033 | def data_entries( | |
1031 | self, matcher=None, undecodable=None |
|
1034 | self, matcher=None, undecodable=None | |
1032 | ) -> Generator[BaseStoreEntry, None, None]: |
|
1035 | ) -> Generator[BaseStoreEntry, None, None]: | |
1033 | files = ((f, revlog_type(f)) for f in self.fncache) |
|
1036 | files = ((f, revlog_type(f)) for f in self.fncache) | |
1034 | # Note: all files in fncache should be revlog related, However the |
|
1037 | # Note: all files in fncache should be revlog related, However the | |
1035 | # fncache might contains such file added by previous version of |
|
1038 | # fncache might contains such file added by previous version of | |
1036 | # Mercurial. |
|
1039 | # Mercurial. | |
1037 | files = (f for f in files if f[1] is not None) |
|
1040 | files = (f for f in files if f[1] is not None) | |
1038 | by_revlog = _gather_revlog(files) |
|
1041 | by_revlog = _gather_revlog(files) | |
1039 | for revlog, details in by_revlog: |
|
1042 | for revlog, details in by_revlog: | |
1040 | file_details = {} |
|
1043 | file_details = {} | |
1041 | if revlog.startswith(b'data/'): |
|
1044 | if revlog.startswith(b'data/'): | |
1042 | rl_type = FILEFLAGS_FILELOG |
|
1045 | rl_type = FILEFLAGS_FILELOG | |
1043 | revlog_target_id = revlog.split(b'/', 1)[1] |
|
1046 | revlog_target_id = revlog.split(b'/', 1)[1] | |
1044 | elif revlog.startswith(b'meta/'): |
|
1047 | elif revlog.startswith(b'meta/'): | |
1045 | rl_type = FILEFLAGS_MANIFESTLOG |
|
1048 | rl_type = FILEFLAGS_MANIFESTLOG | |
1046 | # drop the initial directory and the `00manifest` file part |
|
1049 | # drop the initial directory and the `00manifest` file part | |
1047 | tmp = revlog.split(b'/', 1)[1] |
|
1050 | tmp = revlog.split(b'/', 1)[1] | |
1048 | revlog_target_id = tmp.rsplit(b'/', 1)[0] + b'/' |
|
1051 | revlog_target_id = tmp.rsplit(b'/', 1)[0] + b'/' | |
1049 | else: |
|
1052 | else: | |
1050 | # unreachable |
|
1053 | # unreachable | |
1051 | assert False, revlog |
|
1054 | assert False, revlog | |
1052 | for ext, t in details.items(): |
|
1055 | for ext, t in details.items(): | |
1053 | file_details[ext] = { |
|
1056 | file_details[ext] = { | |
1054 | 'is_volatile': bool(t & FILEFLAGS_VOLATILE), |
|
1057 | 'is_volatile': bool(t & FILEFLAGS_VOLATILE), | |
1055 | } |
|
1058 | } | |
1056 | entry = RevlogStoreEntry( |
|
1059 | entry = RevlogStoreEntry( | |
1057 | path_prefix=revlog, |
|
1060 | path_prefix=revlog, | |
1058 | revlog_type=rl_type, |
|
1061 | revlog_type=rl_type, | |
1059 | target_id=revlog_target_id, |
|
1062 | target_id=revlog_target_id, | |
1060 | details=file_details, |
|
1063 | details=file_details, | |
1061 | ) |
|
1064 | ) | |
1062 | if _match_tracked_entry(entry, matcher): |
|
1065 | if _match_tracked_entry(entry, matcher): | |
1063 | yield entry |
|
1066 | yield entry | |
1064 |
|
1067 | |||
1065 | def copylist(self): |
|
1068 | def copylist(self): | |
1066 | d = ( |
|
1069 | d = ( | |
1067 | b'bookmarks', |
|
1070 | b'bookmarks', | |
1068 | b'narrowspec', |
|
1071 | b'narrowspec', | |
1069 | b'data', |
|
1072 | b'data', | |
1070 | b'meta', |
|
1073 | b'meta', | |
1071 | b'dh', |
|
1074 | b'dh', | |
1072 | b'fncache', |
|
1075 | b'fncache', | |
1073 | b'phaseroots', |
|
1076 | b'phaseroots', | |
1074 | b'obsstore', |
|
1077 | b'obsstore', | |
1075 | b'00manifest.d', |
|
1078 | b'00manifest.d', | |
1076 | b'00manifest.i', |
|
1079 | b'00manifest.i', | |
1077 | b'00changelog.d', |
|
1080 | b'00changelog.d', | |
1078 | b'00changelog.i', |
|
1081 | b'00changelog.i', | |
1079 | b'requires', |
|
1082 | b'requires', | |
1080 | ) |
|
1083 | ) | |
1081 | return [b'requires', b'00changelog.i'] + [b'store/' + f for f in d] |
|
1084 | return [b'requires', b'00changelog.i'] + [b'store/' + f for f in d] | |
1082 |
|
1085 | |||
1083 | def write(self, tr): |
|
1086 | def write(self, tr): | |
1084 | self.fncache.write(tr) |
|
1087 | self.fncache.write(tr) | |
1085 |
|
1088 | |||
1086 | def invalidatecaches(self): |
|
1089 | def invalidatecaches(self): | |
1087 | self.fncache.entries = None |
|
1090 | self.fncache.entries = None | |
1088 | self.fncache.addls = set() |
|
1091 | self.fncache.addls = set() | |
1089 |
|
1092 | |||
1090 | def markremoved(self, fn): |
|
1093 | def markremoved(self, fn): | |
1091 | self.fncache.remove(fn) |
|
1094 | self.fncache.remove(fn) | |
1092 |
|
1095 | |||
1093 | def _exists(self, f): |
|
1096 | def _exists(self, f): | |
1094 | ef = self.encode(f) |
|
1097 | ef = self.encode(f) | |
1095 | try: |
|
1098 | try: | |
1096 | self.getsize(ef) |
|
1099 | self.getsize(ef) | |
1097 | return True |
|
1100 | return True | |
1098 | except FileNotFoundError: |
|
1101 | except FileNotFoundError: | |
1099 | return False |
|
1102 | return False | |
1100 |
|
1103 | |||
1101 | def __contains__(self, path): |
|
1104 | def __contains__(self, path): | |
1102 | '''Checks if the store contains path''' |
|
1105 | '''Checks if the store contains path''' | |
1103 | path = b"/".join((b"data", path)) |
|
1106 | path = b"/".join((b"data", path)) | |
1104 | # check for files (exact match) |
|
1107 | # check for files (exact match) | |
1105 | e = path + b'.i' |
|
1108 | e = path + b'.i' | |
1106 | if e in self.fncache and self._exists(e): |
|
1109 | if e in self.fncache and self._exists(e): | |
1107 | return True |
|
1110 | return True | |
1108 | # now check for directories (prefix match) |
|
1111 | # now check for directories (prefix match) | |
1109 | if not path.endswith(b'/'): |
|
1112 | if not path.endswith(b'/'): | |
1110 | path += b'/' |
|
1113 | path += b'/' | |
1111 | for e in self.fncache: |
|
1114 | for e in self.fncache: | |
1112 | if e.startswith(path) and self._exists(e): |
|
1115 | if e.startswith(path) and self._exists(e): | |
1113 | return True |
|
1116 | return True | |
1114 | return False |
|
1117 | return False |
General Comments 0
You need to be logged in to leave comments.
Login now