2 import sys, math, struct, glob
3 from bup import options, git
4 from bup.helpers import *
7 SHA_PER_PAGE=PAGE_SIZE/20.
11 for i in xrange(0, len(l), count):
15 def merge(idxlist, bits, table):
17 for e in git.idxmerge(idxlist):
19 prefix = git.extract_bits(e, bits)
24 def do_midx(outdir, outfilename, infilenames):
27 sum = Sha1('\0'.join(infilenames)).hexdigest()
28 outfilename = '%s/midx-%s.midx' % (outdir, sum)
32 for name in infilenames:
33 ix = git.PackIdx(name)
37 log('Merging %d indexes (%d objects).\n' % (len(infilenames), total))
38 if (not opt.force and (total < 1024 and len(infilenames) < 3)) \
39 or (opt.force and not total):
40 log('midx: nothing to do.\n')
43 pages = int(total/SHA_PER_PAGE) or 1
44 bits = int(math.ceil(math.log(pages, 2)))
46 log('Table size: %d (%d bits)\n' % (entries*4, bits))
51 os.unlink(outfilename)
54 f = open(outfilename + '.tmp', 'w+')
55 f.write('MIDX\0\0\0\2')
56 f.write(struct.pack('!I', bits))
57 assert(f.tell() == 12)
58 f.write('\0'*4*entries)
60 for e in merge(inp, bits, table):
63 f.write('\0'.join(os.path.basename(p) for p in infilenames))
66 f.write(struct.pack('!%dI' % entries, *table))
68 os.rename(outfilename + '.tmp', outfilename)
70 # this is just for testing
72 p = git.PackMidx(outfilename)
73 assert(len(p.idxnames) == len(infilenames))
75 assert(len(p) == total)
77 for i in merge(inp, total, bits, table):
78 assert(i == pi.next())
84 bup midx [options...] <idxnames...>
86 o,output= output midx filename (default: auto-generated)
87 a,auto automatically create .midx from any unindexed .idx files
88 f,force automatically create .midx from *all* .idx files
89 max-files= maximum number of idx files to open at once [500]
91 o = options.Options('bup midx', optspec)
92 (opt, flags, extra) = o.parse(sys.argv[1:])
94 if extra and (opt.auto or opt.force):
95 o.fatal("you can't use -f/-a and also provide filenames")
97 git.check_repo_or_die()
100 do_midx(git.repo('objects/pack'), opt.output, extra)
101 elif opt.auto or opt.force:
102 paths = [git.repo('objects/pack')]
103 paths += glob.glob(git.repo('index-cache/*/.'))
105 log('midx: scanning %s\n' % path)
107 idxs = glob.glob('%s/*.idx' % path)
108 for sublist in _group(idxs, opt.max_files):
109 do_midx(path, opt.output, sublist)
111 m = git.PackIdxList(path)
113 for pack in m.packs: # only .idx files without a .midx are open
114 if pack.name.endswith('.idx'):
115 needed[pack.name] = 1
117 do_midx(path, opt.output, needed.keys())
120 o.fatal("you must use -f or -a or provide input filenames")