httprepo: use separate handlers for HTTP and HTTPS
This is needed to keep the code in keepalive.py from sharing
the same connection between HTTP and HTTPS.
52ce0d6bc375 explains why we were using a single handler.
This should fix issue892.
# should fail - foo is not managed
foo: not copying - file is not managed
abort: no files to copy
? foo
# dry-run; print a warning that this is not a real copy; foo is added
foo has not been committed yet, so no copy data will be stored for bar.
A foo
# should print a warning that this is not a real copy; bar is added
foo has not been committed yet, so no copy data will be stored for bar.
A bar
# should print a warning that this is not a real copy; foo is added
bar has not been committed yet, so no copy data will be stored for foo.
A foo
# dry-run; should show that foo is clean
C foo
# should show copy
A bar
foo
# shouldn't show copy
# should match
rev offset length base linkrev nodeid p1 p2
0 0 5 0 0 2ed2a3912a0b 000000000000 000000000000
bar renamed from foo:2ed2a3912a0b24502043eae84ee4b279c18b90dd
# should not be renamed
bar not renamed
# should show copy
M bar
foo
# should show no parents for tip
rev offset length base linkrev nodeid p1 p2
0 0 69 0 1 6ca237634e1f 000000000000 000000000000
1 69 6 1 2 7a1ff8e75f5b 6ca237634e1f 000000000000
2 75 82 1 3 243dfe60f3d9 000000000000 000000000000
# should match
rev offset length base linkrev nodeid p1 p2
0 0 5 0 0 2ed2a3912a0b 000000000000 000000000000
1 5 7 1 2 dd12c926cf16 2ed2a3912a0b 000000000000
bar renamed from foo:dd12c926cf165e3eb4cf87b084955cb617221c17
# should show no copies
# copy --after on an added file
A baz
bar
# foo was clean:
C foo
# but it's considered modified after a copy --after --force
M foo
bar