Show More
@@ -1,286 +1,287 b'' | |||||
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 __future__ import absolute_import |
|
8 | from __future__ import absolute_import | |
9 |
|
9 | |||
10 | import struct |
|
10 | import struct | |
11 | import zlib |
|
11 | import zlib | |
12 |
|
12 | |||
13 | from ..node import nullid, nullrev |
|
13 | from ..node import nullid, nullrev | |
14 | from .. import ( |
|
14 | from .. import ( | |
15 | pycompat, |
|
15 | pycompat, | |
16 | util, |
|
16 | util, | |
17 | ) |
|
17 | ) | |
18 |
|
18 | |||
19 | from ..revlogutils import nodemap as nodemaputil |
|
19 | from ..revlogutils import nodemap as nodemaputil | |
20 |
|
20 | |||
21 | stringio = pycompat.bytesio |
|
21 | stringio = pycompat.bytesio | |
22 |
|
22 | |||
23 |
|
23 | |||
24 | _pack = struct.pack |
|
24 | _pack = struct.pack | |
25 | _unpack = struct.unpack |
|
25 | _unpack = struct.unpack | |
26 | _compress = zlib.compress |
|
26 | _compress = zlib.compress | |
27 | _decompress = zlib.decompress |
|
27 | _decompress = zlib.decompress | |
28 |
|
28 | |||
29 | # Some code below makes tuples directly because it's more convenient. However, |
|
29 | # Some code below makes tuples directly because it's more convenient. However, | |
30 | # code outside this module should always use dirstatetuple. |
|
30 | # code outside this module should always use dirstatetuple. | |
31 | def dirstatetuple(*x): |
|
31 | def dirstatetuple(*x): | |
32 | # x is a tuple |
|
32 | # x is a tuple | |
33 | return x |
|
33 | return x | |
34 |
|
34 | |||
35 |
|
35 | |||
36 | indexformatng = b">Qiiiiii20s12x" |
|
|||
37 | indexfirst = struct.calcsize(b'Q') |
|
|||
38 | sizeint = struct.calcsize(b'i') |
|
|||
39 | indexsize = struct.calcsize(indexformatng) |
|
|||
40 | nullitem = (0, 0, 0, -1, -1, -1, -1, nullid) |
|
|||
41 |
|
||||
42 |
|
||||
43 | def gettype(q): |
|
36 | def gettype(q): | |
44 | return int(q & 0xFFFF) |
|
37 | return int(q & 0xFFFF) | |
45 |
|
38 | |||
46 |
|
39 | |||
47 | def offset_type(offset, type): |
|
40 | def offset_type(offset, type): | |
48 | return int(int(offset) << 16 | type) |
|
41 | return int(int(offset) << 16 | type) | |
49 |
|
42 | |||
50 |
|
43 | |||
51 | class BaseIndexObject(object): |
|
44 | class BaseIndexObject(object): | |
|
45 | index_format = b">Qiiiiii20s12x" | |||
|
46 | big_int_size = struct.calcsize(b'Q') | |||
|
47 | int_size = struct.calcsize(b'i') | |||
|
48 | index_size = struct.calcsize(index_format) | |||
|
49 | null_item = (0, 0, 0, -1, -1, -1, -1, nullid) | |||
|
50 | ||||
52 | @property |
|
51 | @property | |
53 | def nodemap(self): |
|
52 | def nodemap(self): | |
54 | msg = b"index.nodemap is deprecated, use index.[has_node|rev|get_rev]" |
|
53 | msg = b"index.nodemap is deprecated, use index.[has_node|rev|get_rev]" | |
55 | util.nouideprecwarn(msg, b'5.3', stacklevel=2) |
|
54 | util.nouideprecwarn(msg, b'5.3', stacklevel=2) | |
56 | return self._nodemap |
|
55 | return self._nodemap | |
57 |
|
56 | |||
58 | @util.propertycache |
|
57 | @util.propertycache | |
59 | def _nodemap(self): |
|
58 | def _nodemap(self): | |
60 | nodemap = nodemaputil.NodeMap({nullid: nullrev}) |
|
59 | nodemap = nodemaputil.NodeMap({nullid: nullrev}) | |
61 | for r in range(0, len(self)): |
|
60 | for r in range(0, len(self)): | |
62 | n = self[r][7] |
|
61 | n = self[r][7] | |
63 | nodemap[n] = r |
|
62 | nodemap[n] = r | |
64 | return nodemap |
|
63 | return nodemap | |
65 |
|
64 | |||
66 | def has_node(self, node): |
|
65 | def has_node(self, node): | |
67 | """return True if the node exist in the index""" |
|
66 | """return True if the node exist in the index""" | |
68 | return node in self._nodemap |
|
67 | return node in self._nodemap | |
69 |
|
68 | |||
70 | def rev(self, node): |
|
69 | def rev(self, node): | |
71 | """return a revision for a node |
|
70 | """return a revision for a node | |
72 |
|
71 | |||
73 | If the node is unknown, raise a RevlogError""" |
|
72 | If the node is unknown, raise a RevlogError""" | |
74 | return self._nodemap[node] |
|
73 | return self._nodemap[node] | |
75 |
|
74 | |||
76 | def get_rev(self, node): |
|
75 | def get_rev(self, node): | |
77 | """return a revision for a node |
|
76 | """return a revision for a node | |
78 |
|
77 | |||
79 | If the node is unknown, return None""" |
|
78 | If the node is unknown, return None""" | |
80 | return self._nodemap.get(node) |
|
79 | return self._nodemap.get(node) | |
81 |
|
80 | |||
82 | def _stripnodes(self, start): |
|
81 | def _stripnodes(self, start): | |
83 | if '_nodemap' in vars(self): |
|
82 | if '_nodemap' in vars(self): | |
84 | for r in range(start, len(self)): |
|
83 | for r in range(start, len(self)): | |
85 | n = self[r][7] |
|
84 | n = self[r][7] | |
86 | del self._nodemap[n] |
|
85 | del self._nodemap[n] | |
87 |
|
86 | |||
88 | def clearcaches(self): |
|
87 | def clearcaches(self): | |
89 | self.__dict__.pop('_nodemap', None) |
|
88 | self.__dict__.pop('_nodemap', None) | |
90 |
|
89 | |||
91 | def __len__(self): |
|
90 | def __len__(self): | |
92 | return self._lgt + len(self._extra) |
|
91 | return self._lgt + len(self._extra) | |
93 |
|
92 | |||
94 | def append(self, tup): |
|
93 | def append(self, tup): | |
95 | if '_nodemap' in vars(self): |
|
94 | if '_nodemap' in vars(self): | |
96 | self._nodemap[tup[7]] = len(self) |
|
95 | self._nodemap[tup[7]] = len(self) | |
97 |
data = _pack(indexformat |
|
96 | data = _pack(self.index_format, *tup) | |
98 | self._extra.append(data) |
|
97 | self._extra.append(data) | |
99 |
|
98 | |||
100 | def _check_index(self, i): |
|
99 | def _check_index(self, i): | |
101 | if not isinstance(i, int): |
|
100 | if not isinstance(i, int): | |
102 | raise TypeError(b"expecting int indexes") |
|
101 | raise TypeError(b"expecting int indexes") | |
103 | if i < 0 or i >= len(self): |
|
102 | if i < 0 or i >= len(self): | |
104 | raise IndexError |
|
103 | raise IndexError | |
105 |
|
104 | |||
106 | def __getitem__(self, i): |
|
105 | def __getitem__(self, i): | |
107 | if i == -1: |
|
106 | if i == -1: | |
108 | return nullitem |
|
107 | return self.null_item | |
109 | self._check_index(i) |
|
108 | self._check_index(i) | |
110 | if i >= self._lgt: |
|
109 | if i >= self._lgt: | |
111 | data = self._extra[i - self._lgt] |
|
110 | data = self._extra[i - self._lgt] | |
112 | else: |
|
111 | else: | |
113 | index = self._calculate_index(i) |
|
112 | index = self._calculate_index(i) | |
114 | data = self._data[index : index + indexsize] |
|
113 | data = self._data[index : index + self.index_size] | |
115 |
r = _unpack(indexformat |
|
114 | r = _unpack(self.index_format, data) | |
116 | if self._lgt and i == 0: |
|
115 | if self._lgt and i == 0: | |
117 | r = (offset_type(0, gettype(r[0])),) + r[1:] |
|
116 | r = (offset_type(0, gettype(r[0])),) + r[1:] | |
118 | return r |
|
117 | return r | |
119 |
|
118 | |||
120 |
|
119 | |||
121 | class IndexObject(BaseIndexObject): |
|
120 | class IndexObject(BaseIndexObject): | |
122 | def __init__(self, data): |
|
121 | def __init__(self, data): | |
123 | assert len(data) % indexsize == 0 |
|
122 | assert len(data) % self.index_size == 0 | |
124 | self._data = data |
|
123 | self._data = data | |
125 | self._lgt = len(data) // indexsize |
|
124 | self._lgt = len(data) // self.index_size | |
126 | self._extra = [] |
|
125 | self._extra = [] | |
127 |
|
126 | |||
128 | def _calculate_index(self, i): |
|
127 | def _calculate_index(self, i): | |
129 | return i * indexsize |
|
128 | return i * self.index_size | |
130 |
|
129 | |||
131 | def __delitem__(self, i): |
|
130 | def __delitem__(self, i): | |
132 | if not isinstance(i, slice) or not i.stop == -1 or i.step is not None: |
|
131 | if not isinstance(i, slice) or not i.stop == -1 or i.step is not None: | |
133 | raise ValueError(b"deleting slices only supports a:-1 with step 1") |
|
132 | raise ValueError(b"deleting slices only supports a:-1 with step 1") | |
134 | i = i.start |
|
133 | i = i.start | |
135 | self._check_index(i) |
|
134 | self._check_index(i) | |
136 | self._stripnodes(i) |
|
135 | self._stripnodes(i) | |
137 | if i < self._lgt: |
|
136 | if i < self._lgt: | |
138 | self._data = self._data[: i * indexsize] |
|
137 | self._data = self._data[: i * self.index_size] | |
139 | self._lgt = i |
|
138 | self._lgt = i | |
140 | self._extra = [] |
|
139 | self._extra = [] | |
141 | else: |
|
140 | else: | |
142 | self._extra = self._extra[: i - self._lgt] |
|
141 | self._extra = self._extra[: i - self._lgt] | |
143 |
|
142 | |||
144 |
|
143 | |||
145 | class PersistentNodeMapIndexObject(IndexObject): |
|
144 | class PersistentNodeMapIndexObject(IndexObject): | |
146 | """a Debug oriented class to test persistent nodemap |
|
145 | """a Debug oriented class to test persistent nodemap | |
147 |
|
146 | |||
148 | We need a simple python object to test API and higher level behavior. See |
|
147 | We need a simple python object to test API and higher level behavior. See | |
149 | the Rust implementation for more serious usage. This should be used only |
|
148 | the Rust implementation for more serious usage. This should be used only | |
150 | through the dedicated `devel.persistent-nodemap` config. |
|
149 | through the dedicated `devel.persistent-nodemap` config. | |
151 | """ |
|
150 | """ | |
152 |
|
151 | |||
153 | def nodemap_data_all(self): |
|
152 | def nodemap_data_all(self): | |
154 | """Return bytes containing a full serialization of a nodemap |
|
153 | """Return bytes containing a full serialization of a nodemap | |
155 |
|
154 | |||
156 | The nodemap should be valid for the full set of revisions in the |
|
155 | The nodemap should be valid for the full set of revisions in the | |
157 | index.""" |
|
156 | index.""" | |
158 | return nodemaputil.persistent_data(self) |
|
157 | return nodemaputil.persistent_data(self) | |
159 |
|
158 | |||
160 | def nodemap_data_incremental(self): |
|
159 | def nodemap_data_incremental(self): | |
161 | """Return bytes containing a incremental update to persistent nodemap |
|
160 | """Return bytes containing a incremental update to persistent nodemap | |
162 |
|
161 | |||
163 | This containst the data for an append-only update of the data provided |
|
162 | This containst the data for an append-only update of the data provided | |
164 | in the last call to `update_nodemap_data`. |
|
163 | in the last call to `update_nodemap_data`. | |
165 | """ |
|
164 | """ | |
166 | if self._nm_root is None: |
|
165 | if self._nm_root is None: | |
167 | return None |
|
166 | return None | |
168 | docket = self._nm_docket |
|
167 | docket = self._nm_docket | |
169 | changed, data = nodemaputil.update_persistent_data( |
|
168 | changed, data = nodemaputil.update_persistent_data( | |
170 | self, self._nm_root, self._nm_max_idx, self._nm_docket.tip_rev |
|
169 | self, self._nm_root, self._nm_max_idx, self._nm_docket.tip_rev | |
171 | ) |
|
170 | ) | |
172 |
|
171 | |||
173 | self._nm_root = self._nm_max_idx = self._nm_docket = None |
|
172 | self._nm_root = self._nm_max_idx = self._nm_docket = None | |
174 | return docket, changed, data |
|
173 | return docket, changed, data | |
175 |
|
174 | |||
176 | def update_nodemap_data(self, docket, nm_data): |
|
175 | def update_nodemap_data(self, docket, nm_data): | |
177 | """provide full block of persisted binary data for a nodemap |
|
176 | """provide full block of persisted binary data for a nodemap | |
178 |
|
177 | |||
179 | The data are expected to come from disk. See `nodemap_data_all` for a |
|
178 | The data are expected to come from disk. See `nodemap_data_all` for a | |
180 | produceur of such data.""" |
|
179 | produceur of such data.""" | |
181 | if nm_data is not None: |
|
180 | if nm_data is not None: | |
182 | self._nm_root, self._nm_max_idx = nodemaputil.parse_data(nm_data) |
|
181 | self._nm_root, self._nm_max_idx = nodemaputil.parse_data(nm_data) | |
183 | if self._nm_root: |
|
182 | if self._nm_root: | |
184 | self._nm_docket = docket |
|
183 | self._nm_docket = docket | |
185 | else: |
|
184 | else: | |
186 | self._nm_root = self._nm_max_idx = self._nm_docket = None |
|
185 | self._nm_root = self._nm_max_idx = self._nm_docket = None | |
187 |
|
186 | |||
188 |
|
187 | |||
189 | class InlinedIndexObject(BaseIndexObject): |
|
188 | class InlinedIndexObject(BaseIndexObject): | |
190 | def __init__(self, data, inline=0): |
|
189 | def __init__(self, data, inline=0): | |
191 | self._data = data |
|
190 | self._data = data | |
192 | self._lgt = self._inline_scan(None) |
|
191 | self._lgt = self._inline_scan(None) | |
193 | self._inline_scan(self._lgt) |
|
192 | self._inline_scan(self._lgt) | |
194 | self._extra = [] |
|
193 | self._extra = [] | |
195 |
|
194 | |||
196 | def _inline_scan(self, lgt): |
|
195 | def _inline_scan(self, lgt): | |
197 | off = 0 |
|
196 | off = 0 | |
198 | if lgt is not None: |
|
197 | if lgt is not None: | |
199 | self._offsets = [0] * lgt |
|
198 | self._offsets = [0] * lgt | |
200 | count = 0 |
|
199 | count = 0 | |
201 | while off <= len(self._data) - indexsize: |
|
200 | while off <= len(self._data) - self.index_size: | |
|
201 | start = off + self.big_int_size | |||
202 | (s,) = struct.unpack( |
|
202 | (s,) = struct.unpack( | |
203 | b'>i', self._data[off + indexfirst : off + sizeint + indexfirst] |
|
203 | b'>i', | |
|
204 | self._data[start : start + self.int_size], | |||
204 | ) |
|
205 | ) | |
205 | if lgt is not None: |
|
206 | if lgt is not None: | |
206 | self._offsets[count] = off |
|
207 | self._offsets[count] = off | |
207 | count += 1 |
|
208 | count += 1 | |
208 | off += indexsize + s |
|
209 | off += self.index_size + s | |
209 | if off != len(self._data): |
|
210 | if off != len(self._data): | |
210 | raise ValueError(b"corrupted data") |
|
211 | raise ValueError(b"corrupted data") | |
211 | return count |
|
212 | return count | |
212 |
|
213 | |||
213 | def __delitem__(self, i): |
|
214 | def __delitem__(self, i): | |
214 | if not isinstance(i, slice) or not i.stop == -1 or i.step is not None: |
|
215 | if not isinstance(i, slice) or not i.stop == -1 or i.step is not None: | |
215 | raise ValueError(b"deleting slices only supports a:-1 with step 1") |
|
216 | raise ValueError(b"deleting slices only supports a:-1 with step 1") | |
216 | i = i.start |
|
217 | i = i.start | |
217 | self._check_index(i) |
|
218 | self._check_index(i) | |
218 | self._stripnodes(i) |
|
219 | self._stripnodes(i) | |
219 | if i < self._lgt: |
|
220 | if i < self._lgt: | |
220 | self._offsets = self._offsets[:i] |
|
221 | self._offsets = self._offsets[:i] | |
221 | self._lgt = i |
|
222 | self._lgt = i | |
222 | self._extra = [] |
|
223 | self._extra = [] | |
223 | else: |
|
224 | else: | |
224 | self._extra = self._extra[: i - self._lgt] |
|
225 | self._extra = self._extra[: i - self._lgt] | |
225 |
|
226 | |||
226 | def _calculate_index(self, i): |
|
227 | def _calculate_index(self, i): | |
227 | return self._offsets[i] |
|
228 | return self._offsets[i] | |
228 |
|
229 | |||
229 |
|
230 | |||
230 | def parse_index2(data, inline): |
|
231 | def parse_index2(data, inline): | |
231 | if not inline: |
|
232 | if not inline: | |
232 | return IndexObject(data), None |
|
233 | return IndexObject(data), None | |
233 | return InlinedIndexObject(data, inline), (0, data) |
|
234 | return InlinedIndexObject(data, inline), (0, data) | |
234 |
|
235 | |||
235 |
|
236 | |||
236 | def parse_index_devel_nodemap(data, inline): |
|
237 | def parse_index_devel_nodemap(data, inline): | |
237 | """like parse_index2, but alway return a PersistentNodeMapIndexObject""" |
|
238 | """like parse_index2, but alway return a PersistentNodeMapIndexObject""" | |
238 | return PersistentNodeMapIndexObject(data), None |
|
239 | return PersistentNodeMapIndexObject(data), None | |
239 |
|
240 | |||
240 |
|
241 | |||
241 | def parse_dirstate(dmap, copymap, st): |
|
242 | def parse_dirstate(dmap, copymap, st): | |
242 | parents = [st[:20], st[20:40]] |
|
243 | parents = [st[:20], st[20:40]] | |
243 | # dereference fields so they will be local in loop |
|
244 | # dereference fields so they will be local in loop | |
244 | format = b">cllll" |
|
245 | format = b">cllll" | |
245 | e_size = struct.calcsize(format) |
|
246 | e_size = struct.calcsize(format) | |
246 | pos1 = 40 |
|
247 | pos1 = 40 | |
247 | l = len(st) |
|
248 | l = len(st) | |
248 |
|
249 | |||
249 | # the inner loop |
|
250 | # the inner loop | |
250 | while pos1 < l: |
|
251 | while pos1 < l: | |
251 | pos2 = pos1 + e_size |
|
252 | pos2 = pos1 + e_size | |
252 | e = _unpack(b">cllll", st[pos1:pos2]) # a literal here is faster |
|
253 | e = _unpack(b">cllll", st[pos1:pos2]) # a literal here is faster | |
253 | pos1 = pos2 + e[4] |
|
254 | pos1 = pos2 + e[4] | |
254 | f = st[pos2:pos1] |
|
255 | f = st[pos2:pos1] | |
255 | if b'\0' in f: |
|
256 | if b'\0' in f: | |
256 | f, c = f.split(b'\0') |
|
257 | f, c = f.split(b'\0') | |
257 | copymap[f] = c |
|
258 | copymap[f] = c | |
258 | dmap[f] = e[:4] |
|
259 | dmap[f] = e[:4] | |
259 | return parents |
|
260 | return parents | |
260 |
|
261 | |||
261 |
|
262 | |||
262 | def pack_dirstate(dmap, copymap, pl, now): |
|
263 | def pack_dirstate(dmap, copymap, pl, now): | |
263 | now = int(now) |
|
264 | now = int(now) | |
264 | cs = stringio() |
|
265 | cs = stringio() | |
265 | write = cs.write |
|
266 | write = cs.write | |
266 | write(b"".join(pl)) |
|
267 | write(b"".join(pl)) | |
267 | for f, e in pycompat.iteritems(dmap): |
|
268 | for f, e in pycompat.iteritems(dmap): | |
268 | if e[0] == b'n' and e[3] == now: |
|
269 | if e[0] == b'n' and e[3] == now: | |
269 | # The file was last modified "simultaneously" with the current |
|
270 | # The file was last modified "simultaneously" with the current | |
270 | # write to dirstate (i.e. within the same second for file- |
|
271 | # write to dirstate (i.e. within the same second for file- | |
271 | # systems with a granularity of 1 sec). This commonly happens |
|
272 | # systems with a granularity of 1 sec). This commonly happens | |
272 | # for at least a couple of files on 'update'. |
|
273 | # for at least a couple of files on 'update'. | |
273 | # The user could change the file without changing its size |
|
274 | # The user could change the file without changing its size | |
274 | # within the same second. Invalidate the file's mtime in |
|
275 | # within the same second. Invalidate the file's mtime in | |
275 | # dirstate, forcing future 'status' calls to compare the |
|
276 | # dirstate, forcing future 'status' calls to compare the | |
276 | # contents of the file if the size is the same. This prevents |
|
277 | # contents of the file if the size is the same. This prevents | |
277 | # mistakenly treating such files as clean. |
|
278 | # mistakenly treating such files as clean. | |
278 | e = dirstatetuple(e[0], e[1], e[2], -1) |
|
279 | e = dirstatetuple(e[0], e[1], e[2], -1) | |
279 | dmap[f] = e |
|
280 | dmap[f] = e | |
280 |
|
281 | |||
281 | if f in copymap: |
|
282 | if f in copymap: | |
282 | f = b"%s\0%s" % (f, copymap[f]) |
|
283 | f = b"%s\0%s" % (f, copymap[f]) | |
283 | e = _pack(b">cllll", e[0], e[1], e[2], e[3], len(f)) |
|
284 | e = _pack(b">cllll", e[0], e[1], e[2], e[3], len(f)) | |
284 | write(e) |
|
285 | write(e) | |
285 | write(f) |
|
286 | write(f) | |
286 | return cs.getvalue() |
|
287 | return cs.getvalue() |
General Comments 0
You need to be logged in to leave comments.
Login now