view contrib/hgdiff @ 10262:eb243551cbd8 stable

copies: speed up copy detection

On some large repos, copy detection could spend > 10min using
fctx.ancestor() to determine if file revisions were actually related.
Because ancestor must traverse history to the root to determine the
GCA, it was doing a lot more work than necessary. With this
replacement, same status -r a:b takes ~3 seconds.

#!/usr/bin/env python

import os, sys, struct, stat
import difflib
import re
from optparse import OptionParser
from mercurial.bdiff import bdiff, blocks
from mercurial.mdiff import bunidiff, diffopts

VERSION="0.3"
usage = "usage: %prog [options] file1 file2"
parser = OptionParser(usage=usage)

parser.add_option("-d", "--difflib", action="store_true", default=False)
parser.add_option('-x', '--count', default=1)
parser.add_option('-c', '--context', type="int", default=3)
parser.add_option('-p', '--show-c-function', action="store_true", default=False)
parser.add_option('-w', '--ignore-all-space', action="store_true",
                  default=False)

(options, args) = parser.parse_args()

if not args:
    parser.print_help()
    sys.exit(1)

# simple utility function to put all the
# files from a directory tree into a dict
def buildlist(names, top):
    tlen = len(top)
    for root, dirs, files in os.walk(top):
        l = root[tlen + 1:]
        for x in files:
            p = os.path.join(root, x)
            st = os.lstat(p)
            if stat.S_ISREG(st.st_mode):
                names[os.path.join(l, x)] = (st.st_dev, st.st_ino)

def diff_files(file1, file2):
    if file1 is None:
        b = file(file2).read().splitlines(True)
        l1 = "--- %s\n" % (file2)
        l2 = "+++ %s\n" % (file2)
        l3 = "@@ -0,0 +1,%d @@\n" % len(b)
        l = [l1, l2, l3] + ["+" + e for e in b]
    elif file2 is None:
        a = file(file1).read().splitlines(True)
        l1 = "--- %s\n" % (file1)
        l2 = "+++ %s\n" % (file1)
        l3 = "@@ -1,%d +0,0 @@\n" % len(a)
        l = [l1, l2, l3] + ["-" + e for e in a]
    else:
        t1 = file(file1).read()
        t2 = file(file2).read()
        l1 = t1.splitlines(True)
        l2 = t2.splitlines(True)
        if options.difflib:
            l = difflib.unified_diff(l1, l2, file1, file2)
        else:
            l = bunidiff(t1, t2, l1, l2, file1, file2,
                         diffopts(context=options.context,
                                  showfunc=options.show_c_function,
                                  ignorews=options.ignore_all_space))
    for x in l:
        if x[-1] != '\n':
            x += "\n\ No newline at end of file\n"
        print x,

file1 = args[0]
file2 = args[1]

if os.path.isfile(file1) and os.path.isfile(file2):
    diff_files(file1, file2)
elif os.path.isdir(file1):
    if not os.path.isdir(file2):
        sys.stderr.write("file types don't match\n")
        sys.exit(1)

    d1 = {}
    d2 = {}

    buildlist(d1, file1)
    buildlist(d2, file2)
    keys = d1.keys()
    keys.sort()
    for x in keys:
        if x not in d2:
            f2 = None
        else:
            f2 = os.path.join(file2, x)
            st1 = d1[x]
            st2 = d2[x]
            del d2[x]
            if st1[0] == st2[0] and st1[1] == st2[1]:
                sys.stderr.write("%s is a hard link\n" % x)
                continue
        x = os.path.join(file1, x)
        diff_files(x, f2)
    keys = d2.keys()
    keys.sort()
    for x in keys:
        f1 = None
        x = os.path.join(file2, x)
        diff_files(f1, x)