view mercurial/pure/diffhelpers.py @ 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.

# diffhelpers.py - pure Python implementation of diffhelpers.c
#
# Copyright 2009 Matt Mackall <[e-mail]> and others
#
# This software may be used and distributed according to the terms of the
# GNU General Public License version 2, incorporated herein by reference.

def addlines(fp, hunk, lena, lenb, a, b):
    while True:
        todoa = lena - len(a)
        todob = lenb - len(b)
        num = max(todoa, todob)
        if num == 0:
            break
        for i in xrange(num):
            s = fp.readline()
            c = s[0]
            if s == "\\ No newline at end of file\n":
                fix_newline(hunk, a, b)
                continue
            if c == "\n":
                # Some patches may be missing the control char
                # on empty lines. Supply a leading space.
                s = " \n"
            hunk.append(s)
            if c == "+":
                b.append(s[1:])
            elif c == "-":
                a.append(s)
            else:
                b.append(s[1:])
                a.append(s)
    return 0

def fix_newline(hunk, a, b):
    l = hunk[-1]
    c = l[0]
    hline = l[:-1]

    if c == " " or c == "+":
        b[-1] = l[1:-1]
    if c == " " or c == "-":
        a[-1] = hline
    hunk[-1] = hline
    return 0


def testhunk(a, b, bstart):
    alen = len(a)
    blen = len(b)
    if alen > blen - bstart:
        return -1
    for i in xrange(alen):
        if a[i][1:] != b[i + bstart]:
            return -1
    return 0