view tests/test-encoding @ 11541:ab9fa7a85dd9 stable

filelog: cmp: don't read data if hashes are identical (issue2273)

filelog.renamed() is an expensive call as it reads the filelog if p1 == nullid.
It's more efficient to first compute the hash, and to bail early if
the computed hash is the same as the stored nodeid.

'samehashes' variable is not strictly necessary, but helps for comprehension.

#!/bin/sh

hg init t
cd t

# we need a repo with some legacy latin-1 changesets
hg unbundle $TESTDIR/legacy-encoding.hg
hg co

python << EOF
f = file('latin-1', 'w'); f.write("latin-1 e' encoded: \xe9"); f.close()
f = file('utf-8', 'w'); f.write("utf-8 e' encoded: \xc3\xa9"); f.close()
f = file('latin-1-tag', 'w'); f.write("\xe9"); f.close()
EOF

echo % should fail with encoding error
echo "plain old ascii" > a
hg st
HGENCODING=ascii hg ci -l latin-1 -d "1000000 0"

echo % these should work
echo "latin-1" > a
HGENCODING=latin-1 hg ci -l latin-1 -d "1000000 0"
echo "utf-8" > a
HGENCODING=utf-8 hg ci -l utf-8 -d "1000000 0"

HGENCODING=latin-1 hg tag -d "1000000 0" `cat latin-1-tag`
HGENCODING=latin-1 hg branch `cat latin-1-tag`
HGENCODING=latin-1 hg ci -d "1000000 0" -m 'latin1 branch'
rm .hg/branch

echo "% hg log (ascii)"
hg --encoding ascii log
echo "% hg log (latin-1)"
hg --encoding latin-1 log
echo "% hg log (utf-8)"
hg --encoding utf-8 log
echo "% hg tags (ascii)"
HGENCODING=ascii hg tags
echo "% hg tags (latin-1)"
HGENCODING=latin-1 hg tags
echo "% hg tags (utf-8)"
HGENCODING=utf-8 hg tags
echo "% hg branches (ascii)"
HGENCODING=ascii hg branches
echo "% hg branches (latin-1)"
HGENCODING=latin-1 hg branches
echo "% hg branches (utf-8)"
HGENCODING=utf-8 hg branches

echo '[ui]' >> .hg/hgrc
echo 'fallbackencoding = koi8-r' >> .hg/hgrc
echo "% hg log (utf-8)"
HGENCODING=utf-8 hg log

echo "% hg log (dolphin)"
HGENCODING=dolphin hg log

HGENCODING=ascii hg branch `cat latin-1-tag`
cp latin-1-tag .hg/branch
HGENCODING=latin-1 hg ci -d "1000000 0" -m 'should fail'
exit 0