view tests/test-convert-filemap @ 12252:4481f8a93c7a stable

convert/darcs: handle non-ASCII metadata in darcs changelog (issue2354)

Given a commit author or message with non-ASCII characters in a darcs
repo, convert would raise a UnicodeEncodeError when adding changesets
to the hg changelog.

This happened because etree returns back unicode objects for any text
it can't encode into ASCII. convert was passing these objects to
changelog.add(), which would then attempt encoding.fromlocal() on
them.

This patch ensures converter_source.recode() is called on each piece
of commit data returned by etree.

(Also note that darcs is currently encoding agnostic and will print
out whatever is in a patch's metadata byte-for-byte, even in the XML
changelog.)

#!/bin/sh

HGMERGE=true; export HGMERGE

echo '[extensions]' >> $HGRCPATH
echo 'graphlog =' >> $HGRCPATH
echo 'convert =' >> $HGRCPATH

glog()
{
    hg glog --template '{rev} "{desc}" files: {files}\n' "$@"
}

hg init source
cd source

echo foo > foo
echo baz > baz
mkdir -p dir/subdir
echo dir/file >> dir/file
echo dir/file2 >> dir/file2
echo dir/subdir/file3 >> dir/subdir/file3
echo dir/subdir/file4 >> dir/subdir/file4
hg ci -d '0 0' -qAm '0: add foo baz dir/'

echo bar > bar
echo quux > quux
hg copy foo copied
hg ci -d '1 0' -qAm '1: add bar quux; copy foo to copied'

echo >> foo
hg ci -d '2 0' -m '2: change foo'

hg up -qC 1
echo >> bar
echo >> quux
hg ci -d '3 0' -m '3: change bar quux'

hg up -qC 2
hg merge -qr 3
echo >> bar
echo >> baz
hg ci -d '4 0' -m '4: first merge; change bar baz'

echo >> bar
echo 1 >> baz
echo >> quux
hg ci -d '5 0' -m '5: change bar baz quux'

hg up -qC 4
echo >> foo
echo 2 >> baz
hg ci -d '6 0' -m '6: change foo baz'

hg up -qC 5
hg merge -qr 6
echo >> bar
hg ci -d '7 0' -m '7: second merge; change bar'

echo >> foo
hg ci -m '8: change foo'

glog

echo '% final file versions in this repo:'
hg manifest --debug
hg debugrename copied
echo

cd ..

splitrepo()
{
    msg="$1"
    files="$2"
    opts=$3
    echo "% $files: $msg"
    prefix=`echo "$files" | sed -e 's/ /-/g'`
    fmap="$prefix.fmap"
    repo="$prefix.repo"
    for i in $files; do
	echo "include $i" >> "$fmap"
    done
    hg -q convert $opts --filemap "$fmap" --datesort source "$repo"
    hg up -q -R "$repo"
    glog -R "$repo"
    hg -R "$repo" manifest --debug
}

splitrepo 'skip unwanted merges; use 1st parent in 1st merge, 2nd in 2nd' foo

splitrepo 'merges are not merges anymore' bar

splitrepo '1st merge is not a merge anymore; 2nd still is' baz

splitrepo 'we add additional merges when they are interesting' 'foo quux'

splitrepo 'partial conversion' 'bar quux' '-r 3'
splitrepo 'complete the partial conversion' 'bar quux'

rm -r foo.repo
splitrepo 'partial conversion' 'foo' '-r 3'
splitrepo 'complete the partial conversion' 'foo'

splitrepo 'copied file; source not included in new repo' copied
hg --cwd copied.repo debugrename copied

splitrepo 'copied file; source included in new repo' 'foo copied'
hg --cwd foo-copied.repo debugrename copied

cat > renames.fmap <<EOF
include dir
exclude dir/file2
rename dir dir2
include foo
include copied
rename foo foo2
rename copied copied2
exclude dir/subdir
include dir/subdir/file3
EOF
hg -q convert --filemap renames.fmap --datesort source renames.repo
hg up -q -R renames.repo
glog -R renames.repo
hg -R renames.repo manifest --debug
hg --cwd renames.repo debugrename copied2
echo 'copied:'
hg --cwd source cat copied
echo 'copied2:'
hg --cwd renames.repo cat copied2