revlog: break up compression of large deltas
Python's zlib apparently makes an internal copy of strings passed to
compress(). To avoid this, compress strings 1M at a time, then join
them at the end if the result would be smaller than the original.
For initial commits of large but compressible files, this cuts peak
memory usage nearly in half.
adding a
adding b
adding dir.h/foo
adding t.h
adding t/b
adding t/e.h
adding t/x
hg locate a
a
locate succeeded
hg locate NONEXISTENT
locate failed
hg locate
a
b
dir.h/foo
t.h
t/b
t/e.h
t/x
hg locate a
hg locate NONEXISTENT
hg locate relpath:NONEXISTENT
hg locate
b
dir.h/foo
t.h
t/b
t/e.h
t/x
hg locate -r 0 a
a
hg locate -r 0 NONEXISTENT
hg locate -r 0 relpath:NONEXISTENT
hg locate -r 0
a
b
dir.h/foo
t.h
t/b
t/e.h
t/x
% -I/-X with relative path should work
hg locate
b
dir.h/foo
t.h
t/b
t/e.h
t/x
hg locate -I ../t
t/b
t/e.h
t/x
hg locate t/**
t/b
t/e.h
t/x
hg locate b
../b
../t/b
hg locate *.h
../t.h
../t/e.h
hg locate path:t/x
../t/x
hg locate re:.*\.h$
../t.h
../t/e.h
hg locate -r 0 b
../b
../t/b
hg locate -r 0 *.h
../t.h
../t/e.h
hg locate -r 0 path:t/x
../t/x
hg locate -r 0 re:.*\.h$
../t.h
../t/e.h