-#!/usr/bin/env python
-import sys, math, struct, glob, resource
-import tempfile
-from bup import options, git, midx, _helpers
-from bup.helpers import *
+#!/bin/sh
+"""": # -*-python-*-
+bup_python="$(dirname "$0")/bup-python" || exit $?
+exec "$bup_python" "$0" ${1+"$@"}
+"""
+# end of bup preamble
+
+from __future__ import absolute_import, print_function
+from binascii import hexlify
+import glob, math, os, resource, struct, sys, tempfile
+
+from bup import options, git, midx, _helpers, xstat
+from bup.compat import argv_bytes, hexstr, range
+from bup.helpers import (Sha1, add_error, atomically_replaced_file, debug1, fdatasync,
+ handle_ctrl_c, log, mmap_readwrite, qprogress,
+ saved_errors, unlink)
+from bup.io import byte_stream, path_msg
+
PAGE_SIZE=4096
SHA_PER_PAGE=PAGE_SIZE/20.
def _group(l, count):
- for i in xrange(0, len(l), count):
+ for i in range(0, len(l), count):
yield l[i:i+count]
-
-
+
+
def max_files():
mf = min(resource.getrlimit(resource.RLIMIT_NOFILE))
if mf > 32:
def check_midx(name):
nicename = git.repo_rel(name)
- log('Checking %s.\n' % nicename)
+ log('Checking %s.\n' % path_msg(nicename))
try:
ix = git.open_idx(name)
- except git.GitError, e:
- add_error('%s: %s' % (name, e))
+ except git.GitError as e:
+ add_error('%s: %s' % (pathmsg(name), e))
return
for count,subname in enumerate(ix.idxnames):
sub = git.open_idx(os.path.join(os.path.dirname(name), subname))
if not (ecount % 1234):
qprogress(' %d/%d: %s %d/%d\r'
% (count, len(ix.idxnames),
- git.shorten_hash(subname), ecount, len(sub)))
+ git.shorten_hash(subname).decode('ascii'),
+ ecount, len(sub)))
if not sub.exists(e):
add_error("%s: %s: %s missing from idx"
- % (nicename, git.shorten_hash(subname),
- str(e).encode('hex')))
+ % (path_msg(nicename),
+ git.shorten_hash(subname).decode('ascii'),
+ hexstr(e)))
if not ix.exists(e):
add_error("%s: %s: %s missing from midx"
- % (nicename, git.shorten_hash(subname),
- str(e).encode('hex')))
+ % (path_msg(nicename),
+ git.shorten_hash(subname).decode('ascii'),
+ hexstr(e)))
prev = None
for ecount,e in enumerate(ix):
if not (ecount % 1234):
qprogress(' Ordering: %d/%d\r' % (ecount, len(ix)))
- if not e >= prev:
+ if e and prev and not e >= prev:
add_error('%s: ordering error: %s < %s'
- % (nicename,
- str(e).encode('hex'), str(prev).encode('hex')))
+ % (nicename, hexstr(e), hexstr(prev)))
prev = e
global _first
if not outfilename:
assert(outdir)
- sum = Sha1('\0'.join(infilenames)).hexdigest()
- outfilename = '%s/midx-%s.midx' % (outdir, sum)
+ sum = hexlify(Sha1(b'\0'.join(infilenames)).digest())
+ outfilename = b'%s/midx-%s.midx' % (outdir, sum)
inp = []
total = 0
allfilenames = []
- for name in infilenames:
- ix = git.open_idx(name)
- inp.append((
- ix.map,
- len(ix),
- ix.sha_ofs,
- isinstance(ix, midx.PackMidx) and ix.which_ofs or 0,
- len(allfilenames),
- ))
- for n in ix.idxnames:
- allfilenames.append(os.path.basename(n))
- total += len(ix)
- inp.sort(lambda x,y: cmp(str(y[0][y[2]:y[2]+20]),str(x[0][x[2]:x[2]+20])))
-
- if not _first: _first = outdir
- dirprefix = (_first != outdir) and git.repo_rel(outdir)+': ' or ''
- log('midx: %s%screating from %d files (%d objects).\n'
- % (dirprefix, prefixstr, len(infilenames), total))
- if (opt.auto and (total < 1024 and len(infilenames) < 3)) \
- or ((opt.auto or opt.force) and len(infilenames) < 2) \
- or (opt.force and not total):
- debug1('midx: nothing to do.\n')
- return
-
- pages = int(total/SHA_PER_PAGE) or 1
- bits = int(math.ceil(math.log(pages, 2)))
- entries = 2**bits
- debug1('midx: table size: %d (%d bits)\n' % (entries*4, bits))
-
- unlink(outfilename)
- f = open(outfilename + '.tmp', 'w+b')
- f.write('MIDX')
- f.write(struct.pack('!II', midx.MIDX_VERSION, bits))
- assert(f.tell() == 12)
-
- f.truncate(12 + 4*entries + 20*total + 4*total)
-
- fmap = mmap_readwrite(f, close=False)
-
- count = merge_into(fmap, bits, total, inp)
- del fmap
-
- f.seek(0, git.SEEK_END)
- f.write('\0'.join(allfilenames))
- f.close()
- os.rename(outfilename + '.tmp', outfilename)
-
- # this is just for testing
+ midxs = []
+ try:
+ for name in infilenames:
+ ix = git.open_idx(name)
+ midxs.append(ix)
+ inp.append((
+ ix.map,
+ len(ix),
+ ix.sha_ofs,
+ isinstance(ix, midx.PackMidx) and ix.which_ofs or 0,
+ len(allfilenames),
+ ))
+ for n in ix.idxnames:
+ allfilenames.append(os.path.basename(n))
+ total += len(ix)
+ inp.sort(reverse=True, key=lambda x: x[0][x[2] : x[2] + 20])
+
+ if not _first: _first = outdir
+ dirprefix = (_first != outdir) and git.repo_rel(outdir) + b': ' or b''
+ debug1('midx: %s%screating from %d files (%d objects).\n'
+ % (dirprefix, prefixstr, len(infilenames), total))
+ if (opt.auto and (total < 1024 and len(infilenames) < 3)) \
+ or ((opt.auto or opt.force) and len(infilenames) < 2) \
+ or (opt.force and not total):
+ debug1('midx: nothing to do.\n')
+ return
+
+ pages = int(total/SHA_PER_PAGE) or 1
+ bits = int(math.ceil(math.log(pages, 2)))
+ entries = 2**bits
+ debug1('midx: table size: %d (%d bits)\n' % (entries*4, bits))
+
+ unlink(outfilename)
+ with atomically_replaced_file(outfilename, 'wb') as f:
+ f.write(b'MIDX')
+ f.write(struct.pack('!II', midx.MIDX_VERSION, bits))
+ assert(f.tell() == 12)
+
+ f.truncate(12 + 4*entries + 20*total + 4*total)
+ f.flush()
+ fdatasync(f.fileno())
+
+ fmap = mmap_readwrite(f, close=False)
+ count = merge_into(fmap, bits, total, inp)
+ del fmap # Assume this calls msync() now.
+ f.seek(0, os.SEEK_END)
+ f.write(b'\0'.join(allfilenames))
+ finally:
+ for ix in midxs:
+ if isinstance(ix, midx.PackMidx):
+ ix.close()
+ midxs = None
+ inp = None
+
+
+ # This is just for testing (if you enable this, don't clear inp above)
if 0:
p = midx.PackMidx(outfilename)
assert(len(p.idxnames) == len(infilenames))
- print p.idxnames
+ log(repr(p.idxnames) + '\n')
assert(len(p) == total)
for pe, e in p, git.idxmerge(inp, final_progress=False):
- assert(i == pi.next())
+ pin = next(pi)
+ assert(i == pin)
assert(p.exists(i))
return total, outfilename
-def do_midx(outdir, outfilename, infilenames, prefixstr):
+def do_midx(outdir, outfilename, infilenames, prefixstr, prout):
rv = _do_midx(outdir, outfilename, infilenames, prefixstr)
if rv and opt['print']:
- print rv[1]
+ prout.write(rv[1] + b'\n')
-def do_midx_dir(path):
+def do_midx_dir(path, outfilename, prout):
already = {}
sizes = {}
if opt.force and not opt.auto:
midxs = [] # don't use existing midx files
else:
- midxs = glob.glob('%s/*.midx' % path)
+ midxs = glob.glob(b'%s/*.midx' % path)
contents = {}
for mname in midxs:
m = git.open_idx(mname)
- contents[mname] = [('%s/%s' % (path,i)) for i in m.idxnames]
+ contents[mname] = [(b'%s/%s' % (path,i)) for i in m.idxnames]
sizes[mname] = len(m)
# sort the biggest+newest midxes first, so that we can eliminate
# smaller (or older) redundant ones that come later in the list
- midxs.sort(key=lambda ix: (-sizes[ix], -os.stat(ix).st_mtime))
+ midxs.sort(key=lambda ix: (-sizes[ix], -xstat.stat(ix).st_mtime))
for mname in midxs:
any = 0
already[mname] = 1
midxs = [k for k in midxs if not already.get(k)]
- idxs = [k for k in glob.glob('%s/*.idx' % path) if not already.get(k)]
+ idxs = [k for k in glob.glob(b'%s/*.idx' % path) if not already.get(k)]
for iname in idxs:
i = git.open_idx(iname)
all.sort()
part1 = [name for sz,name in all[:len(all)-DESIRED_LWM+1]]
part2 = all[len(all)-DESIRED_LWM+1:]
- all = list(do_midx_group(path, part1)) + part2
+ all = list(do_midx_group(path, outfilename, part1)) + part2
if len(all) > DESIRED_HWM:
debug1('\nStill too many indexes (%d > %d). Merging again.\n'
% (len(all), DESIRED_HWM))
if opt['print']:
for sz,name in all:
if not existed.get(name):
- print name
+ prout.write(name + b'\n')
-def do_midx_group(outdir, infiles):
+def do_midx_group(outdir, outfilename, infiles):
groups = list(_group(infiles, opt.max_files))
gprefix = ''
for n,sublist in enumerate(groups):
if len(groups) != 1:
gprefix = 'Group %d: ' % (n+1)
- rv = _do_midx(path, None, sublist, gprefix)
+ rv = _do_midx(outdir, outfilename, sublist, gprefix)
if rv:
yield rv
o = options.Options(optspec)
(opt, flags, extra) = o.parse(sys.argv[1:])
+opt.dir = argv_bytes(opt.dir) if opt.dir else None
+opt.output = argv_bytes(opt.output) if opt.output else None
if extra and (opt.auto or opt.force):
o.fatal("you can't use -f/-a and also provide filenames")
opt.max_files = max_files()
assert(opt.max_files >= 5)
+extra = [argv_bytes(x) for x in extra]
+
if opt.check:
# check existing midx files
if extra:
paths = opt.dir and [opt.dir] or git.all_packdirs()
for path in paths:
debug1('midx: scanning %s\n' % path)
- midxes += glob.glob(os.path.join(path, '*.midx'))
+ midxes += glob.glob(os.path.join(path, b'*.midx'))
for name in midxes:
check_midx(name)
if not saved_errors:
log('All tests passed.\n')
else:
if extra:
- do_midx(git.repo('objects/pack'), opt.output, extra, '')
+ sys.stdout.flush()
+ do_midx(git.repo(b'objects/pack'), opt.output, extra, b'',
+ byte_stream(sys.stdout))
elif opt.auto or opt.force:
+ sys.stdout.flush()
paths = opt.dir and [opt.dir] or git.all_packdirs()
for path in paths:
- debug1('midx: scanning %s\n' % path)
- do_midx_dir(path)
+ debug1('midx: scanning %s\n' % path_msg(path))
+ do_midx_dir(path, opt.output, byte_stream(sys.stdout))
else:
o.fatal("you must use -f or -a or provide input filenames")