revisionbranchcache: fall back to slow path if starting readonly (issue4531) Transitioning to Mercurial versions with revision branch cache could be slow as long as all operations were readonly (revset queries) and the cache would be populated but not written back. Instead, fall back to using the consistently slow path when readonly and the cache doesn't exist yet. That avoids the overhead of populating the cache without writing it back. If not readonly, it will still populate all missing entries initially. That avoids repeated writing of the cache file with small updates, and it also makes sure a fully populated cache available for the readonly operations.

     1

#!/usr/bin/env python

     2

# Dump revlogs as raw data stream

     3

# $ find .hg/store/ -name "*.i" | xargs dumprevlog > repo.dump

     4

     5

import sys

     6

from mercurial import revlog, node, util

     7

     8

for fp in (sys.stdin, sys.stdout, sys.stderr):

     9

    util.setbinary(fp)

    10

    11

for f in sys.argv[1:]:

    12

    binopen = lambda fn: open(fn, 'rb')

    13

    r = revlog.revlog(binopen, f)

    14

    print "file:", f

    15

    for i in r:

    16

        n = r.node(i)

    17

        p = r.parents(n)

    18

        d = r.revision(n)

    19

        print "node:", node.hex(n)

    20

        print "linkrev:", r.linkrev(i)

    21

        print "parents:", node.hex(p[0]), node.hex(p[1])

    22

        print "length:", len(d)

    23

        print "-start-"

    24

        print d

    25

        print "-end-"