|
1 #!/usr/bin/env python |
|
2 # |
|
3 # Copyright (C) 2007 Brendan Cully <brendan@kublai.com> |
|
4 # |
|
5 # This software may be used and distributed according to the terms |
|
6 # of the GNU General Public License, incorporated herein by reference. |
|
7 |
|
8 import os, sys |
|
9 |
|
10 class ConfigError(Exception): pass |
|
11 |
|
12 def usage(): |
|
13 print """relink <source> <destination> |
|
14 Recreate hard links between source and destination repositories""" |
|
15 |
|
16 class Config: |
|
17 def __init__(self, args): |
|
18 if len(args) != 3: |
|
19 raise ConfigError("wrong number of arguments") |
|
20 self.src = os.path.abspath(args[1]) |
|
21 self.dst = os.path.abspath(args[2]) |
|
22 for d in (self.src, self.dst): |
|
23 if not os.path.exists(os.path.join(d, '.hg')): |
|
24 raise ConfigError("%s: not a mercurial repository" % d) |
|
25 |
|
26 try: |
|
27 cfg = Config(sys.argv) |
|
28 except ConfigError, inst: |
|
29 print str(inst) |
|
30 usage() |
|
31 sys.exit(1) |
|
32 |
|
33 def collect(src): |
|
34 seplen = len(os.path.sep) |
|
35 candidates = [] |
|
36 for dirpath, dirnames, filenames in os.walk(src): |
|
37 relpath = dirpath[len(src) + seplen:] |
|
38 for filename in filenames: |
|
39 if not (filename.endswith('.i') or filename.endswith('.d')): |
|
40 continue |
|
41 st = os.stat(os.path.join(dirpath, filename)) |
|
42 candidates.append((os.path.join(relpath, filename), st)) |
|
43 |
|
44 return candidates |
|
45 |
|
46 def prune(candidates, dst): |
|
47 targets = [] |
|
48 for fn, st in candidates: |
|
49 tgt = os.path.join(dst, fn) |
|
50 try: |
|
51 ts = os.stat(tgt) |
|
52 except OSError: |
|
53 # Destination doesn't have this file? |
|
54 continue |
|
55 if st.st_ino == ts.st_ino: |
|
56 continue |
|
57 if st.st_dev != ts.st_dev: |
|
58 raise Exception('Source and destination are on different devices') |
|
59 if st.st_size != ts.st_size: |
|
60 continue |
|
61 targets.append((fn, ts.st_size)) |
|
62 |
|
63 return targets |
|
64 |
|
65 def relink(src, dst, files): |
|
66 CHUNKLEN = 65536 |
|
67 relinked = 0 |
|
68 savedbytes = 0 |
|
69 |
|
70 for f, sz in files: |
|
71 source = os.path.join(src, f) |
|
72 tgt = os.path.join(dst, f) |
|
73 sfp = file(source) |
|
74 dfp = file(tgt) |
|
75 sin = sfp.read(CHUNKLEN) |
|
76 while sin: |
|
77 din = dfp.read(CHUNKLEN) |
|
78 if sin != din: |
|
79 break |
|
80 sin = sfp.read(CHUNKLEN) |
|
81 if sin: |
|
82 continue |
|
83 try: |
|
84 os.rename(tgt, tgt + '.bak') |
|
85 try: |
|
86 os.link(source, tgt) |
|
87 except OSError: |
|
88 os.rename(tgt + '.bak', tgt) |
|
89 raise |
|
90 print 'Relinked %s' % f |
|
91 relinked += 1 |
|
92 savedbytes += sz |
|
93 os.remove(tgt + '.bak') |
|
94 except OSError, inst: |
|
95 print '%s: %s' % (tgt, str(inst)) |
|
96 |
|
97 print 'Relinked %d files (%d bytes reclaimed)' % (relinked, savedbytes) |
|
98 |
|
99 src = os.path.join(cfg.src, '.hg') |
|
100 dst = os.path.join(cfg.dst, '.hg') |
|
101 candidates = collect(src) |
|
102 targets = prune(candidates, dst) |
|
103 relink(src, dst, targets) |