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
echo "[extensions]" >> $HGRCPATH
echo "fetch=" >> $HGRCPATH
hg init a
echo a > a/a
hg --cwd a commit -d '1 0' -Ama
hg clone a b
hg clone a c
echo b > a/b
hg --cwd a commit -d '2 0' -Amb
hg --cwd a parents -q
echo % should pull one change
hg --cwd b fetch ../a
hg --cwd b parents -q
echo c > c/c
hg --cwd c commit -d '3 0' -Amc
hg --cwd c fetch -d '4 0' -m 'automated merge' ../a
ls c