convert-repo

Make compression more intelligent:
- we don't attempt to compress things under 44 bytes (empirical)
- we check whether larger objects actually compress
- we tag objects to indicate their compression
  NUL means uncompressed and starts with NUL
  x means gzipped and starts with x (handy)
  u means uncompressed, drop the u

#!/usr/bin/env python
import sys, os, sha, base64, re
from mercurial import hg

ui = hg.ui()
repo = hg.repository(ui=ui)
known = {}

def encodepath(path):
    s = sha.sha(path).digest()
    s = base64.encodestring(s)[:-3]
    s = re.sub("\+", "%", s)
    s = re.sub("/", "_", s)
    return s

for i in range(repo.changelog.count()):
    n = repo.changelog.node(i)
    changes = repo.changelog.read(n)
    for f in changes[3]:
        known[f] = 1

try:
    os.mkdir(".hg/data-new")
except:
    pass

files = known.keys()
files.sort()
for f in files:
    pb = ".hg/data/" + encodepath(f)
    pn = ".hg/data-new/" + f
    print f
    try:
        file(pn+".i", "w").write(file(pb+"i").read())
    except:
        os.makedirs(os.path.dirname(pn))
        # we actually copy the files to get nice disk layout
        file(pn+".i", "w").write(file(pb+"i").read())

    file(pn+".d", "w").write(file(pb).read())

os.rename(".hg/data", ".hg/data-old")
os.rename(".hg/data-new", ".hg/data")