changegroup: avoid large copies
- handle chunk headers separately rather than prepending them to
(potentially large) chunks
- break large chunks into 1M pieces for compression
- don't prepend file metadata onto (potentially large) file data
#!/bin/sh
hg init
echo foo > a
hg add a
hg commit -m "1" -d "1000000 0"
echo bar > b
hg add b
hg remove a
echo "%%% should show a removed and b added"
hg status
echo "reverting..."
hg revert --all
echo "%%% should show b unknown and a back to normal"
hg status
rm b
hg co -C 0
echo foo-a > a
hg commit -m "2a" -d "1000000 0"
hg co -C 0
echo foo-b > a
hg commit -m "2b" -d "1000000 0"
HGMERGE=true hg merge 1
echo "%%% should show foo-b"
cat a
echo bar > b
hg add b
rm a
hg remove a
echo "%%% should show a removed and b added"
hg status
echo "%%% revert should fail"
hg revert --all
echo "%%% revert should be ok now"
hg revert -r2 --all
echo "%%% should show b unknown and a marked modified (merged)"
hg status
echo "%%% should show foo-b"
cat a