merge: fix a copy detection bug (issue672)
When merging rev1 and rev2, we want to search for copies that happened
in rev1 but not in rev2 and vice-versa. We were starting the search at
rev1/rev2 and then going back, stopping as soon as we reached the revno
of the ancestor, but that can miss some cases (see the new
test-issue672).
Now we calculate the revisions that are ancestors of rev1 or rev2 (but
not both) and make sure the search doesn't stop too early.
Simplified test provided by mpm, based on a test case provided by
Edward Lee.
#!/bin/sh -e
umask 027
mkdir test1
cd test1
hg init
touch a b
hg add a b
hg ci -m "added a b" -d "1000000 0"
cd ..
hg clone test1 test3
mkdir test2
cd test2
hg init
hg pull ../test1
hg co
chmod +x a
hg ci -m "chmod +x a" -d "1000000 0"
echo % the changelog should mention file a:
hg tip --template '#files#\n'
cd ../test1
echo 123 >>a
hg ci -m "a updated" -d "1000000 0"
hg pull ../test2
hg heads
hg history
hg -v merge
cd ../test3
echo 123 >>b
hg ci -m "b updated" -d "1000000 0"
hg pull ../test2
hg heads
hg history
hg -v merge
ls -l ../test[123]/a > foo
cut -b 1-10 < foo
hg debugindex .hg/store/data/a.i
hg debugindex ../test2/.hg/store/data/a.i
hg debugindex ../test1/.hg/store/data/a.i