Mercurial > public > mercurial-scm > hg
annotate mercurial/utils/storageutil.py @ 39881:d63153611ed5
storageutil: extract revision number iteration
This code is a bit quirky (and possibly buggy). It will likely be used
by multiple storage backends. Let's extract it so it is reusable.
Differential Revision: https://phab.mercurial-scm.org/D4757
author | Gregory Szorc <gregory.szorc@gmail.com> |
---|---|
date | Mon, 24 Sep 2018 15:19:52 -0700 |
parents | 1b65fb4d43d6 |
children | 0e8836be9541 |
rev | line source |
---|---|
39877
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
1 # storageutil.py - Storage functionality agnostic of backend implementation. |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
2 # |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
3 # Copyright 2018 Gregory Szorc <gregory.szorc@gmail.com> |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
4 # |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
5 # This software may be used and distributed according to the terms of the |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
6 # GNU General Public License version 2 or any later version. |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
7 |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
8 from __future__ import absolute_import |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
9 |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
10 import hashlib |
39878
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
11 import re |
39877
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
12 |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
13 from ..node import ( |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
14 nullid, |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
15 ) |
39881
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
16 from .. import ( |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
17 pycompat, |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
18 ) |
39877
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
19 |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
20 _nullhash = hashlib.sha1(nullid) |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
21 |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
22 def hashrevisionsha1(text, p1, p2): |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
23 """Compute the SHA-1 for revision data and its parents. |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
24 |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
25 This hash combines both the current file contents and its history |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
26 in a manner that makes it easy to distinguish nodes with the same |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
27 content in the revision graph. |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
28 """ |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
29 # As of now, if one of the parent node is null, p2 is null |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
30 if p2 == nullid: |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
31 # deep copy of a hash is faster than creating one |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
32 s = _nullhash.copy() |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
33 s.update(p1) |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
34 else: |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
35 # none of the parent nodes are nullid |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
36 if p1 < p2: |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
37 a = p1 |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
38 b = p2 |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
39 else: |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
40 a = p2 |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
41 b = p1 |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
42 s = hashlib.sha1(a) |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
43 s.update(b) |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
44 s.update(text) |
f8eb71f9e3bd
storageutil: new module for storage primitives (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
diff
changeset
|
45 return s.digest() |
39878
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
46 |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
47 METADATA_RE = re.compile(b'\x01\n') |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
48 |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
49 def parsemeta(text): |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
50 """Parse metadata header from revision data. |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
51 |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
52 Returns a 2-tuple of (metadata, offset), where both can be None if there |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
53 is no metadata. |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
54 """ |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
55 # text can be buffer, so we can't use .startswith or .index |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
56 if text[:2] != b'\x01\n': |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
57 return None, None |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
58 s = METADATA_RE.search(text, 2).start() |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
59 mtext = text[2:s] |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
60 meta = {} |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
61 for l in mtext.splitlines(): |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
62 k, v = l.split(b': ', 1) |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
63 meta[k] = v |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
64 return meta, s + 2 |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
65 |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
66 def packmeta(meta, text): |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
67 """Add metadata to fulltext to produce revision text.""" |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
68 keys = sorted(meta) |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
69 metatext = b''.join(b'%s: %s\n' % (k, meta[k]) for k in keys) |
3e896b51aa5d
storageutil: move metadata parsing and packing from revlog (API)
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39877
diff
changeset
|
70 return b'\x01\n%s\x01\n%s' % (metatext, text) |
39879
d269ddbf54f0
storageutil: move _censoredtext() from revlog
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39878
diff
changeset
|
71 |
d269ddbf54f0
storageutil: move _censoredtext() from revlog
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39878
diff
changeset
|
72 def iscensoredtext(text): |
d269ddbf54f0
storageutil: move _censoredtext() from revlog
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39878
diff
changeset
|
73 meta = parsemeta(text)[0] |
d269ddbf54f0
storageutil: move _censoredtext() from revlog
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39878
diff
changeset
|
74 return meta and b'censored' in meta |
39880
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
75 |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
76 def filtermetadata(text): |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
77 """Extract just the revision data from source text. |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
78 |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
79 Returns ``text`` unless it has a metadata header, in which case we return |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
80 a new buffer without hte metadata. |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
81 """ |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
82 if not text.startswith(b'\x01\n'): |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
83 return text |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
84 |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
85 offset = text.index(b'\x01\n', 2) |
1b65fb4d43d6
storageutil: new function for extracting metadata-less content from text
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39879
diff
changeset
|
86 return text[offset + 2:] |
39881
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
87 |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
88 def iterrevs(storelen, start=0, stop=None): |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
89 """Iterate over revision numbers in a store.""" |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
90 step = 1 |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
91 |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
92 if stop is not None: |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
93 if start > stop: |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
94 step = -1 |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
95 stop += step |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
96 if stop > storelen: |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
97 stop = storelen |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
98 else: |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
99 stop = storelen |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
100 |
d63153611ed5
storageutil: extract revision number iteration
Gregory Szorc <gregory.szorc@gmail.com>
parents:
39880
diff
changeset
|
101 return pycompat.xrange(start, stop, step) |