Reduce ram used for very large inlined index files
During a clone, an inline index is not converted to a split index
file until the very end. When the conversion happens, the index
can be very large, and the inline index loading functions always load
the entire index file into ram.
This changes the revlog code to read the index in smaller chunks.
#!/bin/sh
# initial
hg init test-a
cd test-a
cat >test.txt <<"EOF"
1
2
3
EOF
hg add test.txt
hg commit -m "Initial" -d "1000000 0"
# clone
cd ..
hg clone test-a test-b
# change test-a
cd test-a
cat >test.txt <<"EOF"
one
two
three
EOF
hg commit -m "Numbers as words" -d "1000000 0"
# change test-b
cd ../test-b
cat >test.txt <<"EOF"
1
2.5
3
EOF
hg commit -m "2 -> 2.5" -d "1000000 0"
# now pull and merge from test-a
hg pull ../test-a
HGMERGE=merge hg update -m
# resolve conflict
cat >test.txt <<"EOF"
one
two-point-five
three
EOF
rm -f *.orig
hg commit -m "Merge 1" -d "1000000 0"
# change test-a again
cd ../test-a
cat >test.txt <<"EOF"
one
two-point-one
three
EOF
hg commit -m "two -> two-point-one" -d "1000000 0"
# pull and merge from test-a again
cd ../test-b
hg pull ../test-a
HGMERGE=merge hg update --debug -m
cat test.txt | sed "s% .*%%"
hg debugindex .hg/data/test.txt.i
hg log