index : cgit

#!/bin/sh

# The snapshot page hands a revision back as an archive. Every format is held
# byte for byte to the archive git itself writes, once decompressed where cgit
# pipes through a compressor, and each compressed format needs the same tool
# on the server and here, so one missing tool skips that format on both
# grounds at once. The rest covers how a snapshot is named and found again from
# its name, with a repository of its own for each set of refs, the signatures
# kept in notes, and the requests that are refused.

test_description='Verify snapshot'
. ./setup.sh

check_snapshot_headers() {
	sed "/^$/q" tmp >headers &&
	grep "^Status: 200 OK$" headers &&
	grep "^Content-Type: $2$" headers &&
	grep "^Content-Disposition: inline; filename=\"master.$1\"$" headers
}

check_snapshot_content() {
	ls master/ >output &&
	test_line_count = 5 output &&
	grep "^5$" master/file-5 &&
	test_line_count = 1 master/file-5
}

# Lists the members of the archive in the response held in tmp.
snapshot_members() {
	strip_headers <tmp >snapshot.tar &&
	tar -tf snapshot.tar >members
}

# The format checks below compare against these, which git's own writer makes
# from the commit the snapshots are taken of.
test_expect_success 'set up the archives git writes' '
	git -C repos/foo archive --format=tar --prefix=master/ master >expect.tar &&
	git -C repos/foo archive --format=zip --prefix=master/ master >expect.zip
'

test_expect_success 'a tar snapshot is the archive git writes' '
	cgit_ok_url "foo/snapshot/master.tar" >tmp &&
	check_snapshot_headers tar application/x-tar &&
	strip_headers <tmp >master.tar &&
	test_cmp_bin expect.tar master.tar &&
	rm -rf master &&
	tar -xf master.tar &&
	check_snapshot_content
'

# Registers the check for one compressed tar format.
test_tar_snapshot() {
	ext=$1
	type=$2
	tool=$3
	prereq=$4

	test_lazy_prereq $prereq "command -v $tool"

	test_expect_success $prereq "a $ext snapshot unpacks to the tar git writes" "
		cgit_ok_url 'foo/snapshot/master.$ext' >tmp &&
		check_snapshot_headers $ext $type &&
		strip_headers <tmp >master.$ext &&
		'$tool' -t 'master.$ext' &&
		'$tool' -dc 'master.$ext' >master.tar &&
		test_cmp_bin expect.tar master.tar
	"
}

test_tar_snapshot tar.gz  application/gzip    gzip  GZIP
test_tar_snapshot tar.bz2 application/x-bzip2 bzip2 BZIP2
test_tar_snapshot tar.lz  application/x-lzip  lzip  LZIP
test_tar_snapshot tar.xz  application/x-xz    xz    XZ
test_tar_snapshot tar.zst application/zstd    zstd  ZSTD

test_expect_success 'a zip snapshot is the archive git writes' '
	cgit_ok_url "foo/snapshot/master.zip" >tmp &&
	check_snapshot_headers zip application/zip &&
	strip_headers <tmp >master.zip &&
	test_cmp_bin expect.zip master.zip
'

test_expect_success UNZIP 'a zip snapshot unzips' '
	cgit_ok_url "foo/snapshot/master.zip" >tmp &&
	strip_headers <tmp >master.zip &&
	rm -rf master &&
	"$GIT_UNZIP" master.zip &&
	check_snapshot_content
'

# The tags sit on the older commit, so a name that led to master instead
# would bring file-2 with it. On a case-insensitive filesystem a ref lookup
# for one case finds a loose file of the other. v1.0 stays loose on purpose,
# because deciding by lookups called every fresh tag ambiguous and kept its v
# until git happened to pack the refs. The rest are packed, so that a lookup
# for v2.0 cannot find V2.0. Reftable keeps no file per ref, so v1.0 is
# looked for as a file only under the files backend.
test_expect_success 'set up a repo with tags of every shape' '
	mkrepo repos/vtags 2 &&
	git -C repos/vtags update-ref refs/tags/V2.0 HEAD~1 &&
	git -C repos/vtags update-ref refs/tags/vtags-3.0 HEAD~1 &&
	git -C repos/vtags update-ref refs/tags/release/2.0 HEAD~1 &&
	git -C repos/vtags update-ref refs/heads/topic/x HEAD~1 &&
	git -C repos/vtags pack-refs --all &&
	git -C repos/vtags update-ref refs/tags/v1.0 HEAD~1 &&
	{
		! test_have_prereq REFFILES ||
		test_path_is_file repos/vtags/.git/refs/tags/v1.0
	} &&
	cgitrc_repo vtagsrc vtags snapshots=tar
'

# A tag named v1.0 sheds its v in the snapshot name unless dropping it would
# merge two tags, and the name still leads back to the tag.
test_expect_success 'a loose v tag sheds its v and the name leads back to it' '
	cgit_ok vtagsrc "url=vtags/refs/" >tmp &&
	grep "<a href=./vtags/snapshot/vtags-1.0.tar.>vtags-1.0.tar</a>" tmp &&
	! grep "vtags-v1.0.tar" tmp &&
	cgit_ok vtagsrc "url=vtags/snapshot/vtags-1.0.tar" >tmp &&
	snapshot_members &&
	grep "^vtags-1.0/file-1$" members &&
	! grep "file-2" members
'

test_expect_success 'an underscore after the prefix leads back to the tag too' '
	cgit_ok vtagsrc "url=vtags/snapshot/vtags_1.0.tar" >tmp &&
	snapshot_members &&
	grep "^vtags_1.0/file-1$" members &&
	! grep "file-2" members
'

test_expect_success 'a V tag sheds its V and the name leads back to it' '
	cgit_ok vtagsrc "url=vtags/refs/" >tmp &&
	grep "<a href=./vtags/snapshot/vtags-2.0.tar.>vtags-2.0.tar</a>" tmp &&
	cgit_ok vtagsrc "url=vtags/snapshot/vtags-2.0.tar" >tmp &&
	snapshot_members &&
	grep "^vtags-2.0/file-1$" members &&
	! grep "file-2" members
'

test_expect_success 'a tag that starts with the prefix is not prefixed again' '
	cgit_ok vtagsrc "url=vtags/refs/" >tmp &&
	grep "<a href=./vtags/snapshot/vtags-3.0.tar.>vtags-3.0.tar</a>" tmp &&
	! grep "vtags-vtags-3.0" tmp &&
	cgit_ok vtagsrc "url=vtags/snapshot/vtags-3.0.tar" >tmp &&
	snapshot_members &&
	grep "^vtags-3.0/file-1$" members &&
	! grep "file-2" members
'

test_expect_success 'a snapshot-prefix names the snapshots in its place' '
	cgitrc_repo prefixrc vtags snapshots=tar repo.snapshot-prefix=proj &&
	cgit_ok prefixrc "url=vtags/refs/" >tmp &&
	grep "<a href=./vtags/snapshot/proj-1.0.tar.>proj-1.0.tar</a>" tmp &&
	cgit_ok prefixrc "url=vtags/snapshot/proj-1.0.tar" >tmp &&
	snapshot_members &&
	grep "^proj-1.0/file-1$" members &&
	! grep "file-2" members
'

test_expect_success 'snapshots=all offers every format' '
	cgitrc_repo allrc vtags snapshots=all &&
	cgit_ok allrc "url=vtags/refs/" >tmp &&
	for suffix in tar tar.gz tar.bz2 tar.lz tar.xz tar.zst zip
	do
		grep ">vtags-1.0.$suffix</a>" tmp || return 1
	done
'

# A snapshot name is never worked back to a ref holding a slash, so such a
# ref is offered under a dashed name and passed along with it, even from a
# page whose own head is that ref. The dashed name is no ref, so the archive
# unpacks into the repository's own prefix.
test_expect_success 'a ref with a slash is linked under a dashed name' '
	cgit_ok vtagsrc "url=vtags/refs/" >tmp &&
	grep "<a href=./vtags/snapshot/vtags-release-2.0.tar?id=release/2.0.>vtags-release-2.0.tar</a>" tmp &&
	cgit_ok vtagsrc "url=vtags/commit/&h=topic/x" >tmp &&
	grep "<a href=./vtags/snapshot/vtags-topic-x.tar?id=topic/x.>vtags-topic-x.tar</a>" tmp
'

test_expect_success 'the dashed name serves the ref under the repository prefix' '
	cgit_ok vtagsrc "url=vtags/snapshot/vtags-release-2.0.tar&id=release/2.0" >tmp &&
	snapshot_members &&
	commit=$(git -C repos/vtags get-tar-commit-id <snapshot.tar) &&
	want=$(git -C repos/vtags rev-parse "release/2.0^{commit}") &&
	test -n "$want" &&
	test "$commit" = "$want" &&
	grep "^vtags/file-1$" members &&
	! grep "^vtags-release" members
'

test_expect_success 'a format the repository does not offer is refused' '
	cgit_rc vtagsrc "url=vtags/snapshot/vtags-1.0.zip" >tmp &&
	grep "^Status: 400" tmp &&
	grep "<div class=.error.>Unsupported snapshot format: vtags-1.0.zip</div>" tmp &&
	cgit_rc vtagsrc "url=vtags/snapshot/vtags-1.0.tar.xz" >tmp &&
	grep "^Status: 400" tmp &&
	grep "<div class=.error.>Unsupported snapshot format: vtags-1.0.tar.xz</div>" tmp
'

test_expect_success 'a request naming no snapshot is refused' '
	cgit_rc vtagsrc "url=vtags/snapshot/" >tmp &&
	grep "^Status: 400" tmp &&
	grep "<div class=.error.>No snapshot name specified</div>" tmp
'

test_expect_success 'an id naming no commit is refused' '
	cgit_rc vtagsrc "url=vtags/snapshot/vtags.tar&id=nosuchref" >tmp &&
	grep "^Status: 404" tmp &&
	grep "<div class=.error.>Bad object id: nosuchref</div>" tmp
'

test_expect_success 'set up a repo with a tag holding a slash' '
	mkrepo repos/names 2 &&
	git -C repos/names tag rel/1 &&
	cgitrc_repo namesrc names snapshots=tar
'

# The stem becomes the leading directory of every member of the archive, so
# a name holding a slash is refused even where it spells a ref. Such a ref
# is linked under a dashed name instead.
test_expect_success 'a snapshot name holding a slash is refused' '
	for name in rel/1 :/../../..
	do
		cgit_rc namesrc "url=names/snapshot/$name.tar" >tmp &&
		grep "^Status: 404" tmp &&
		grep "<div class=.error.>Not found</div>" tmp || return 1
	done &&
	cgit_ok namesrc "url=names/snapshot/names-master.tar" >tmp
'

# Only a ref is worked back from a file name, since the wider revision syntax
# can walk the whole history for a match. The name holds no slash, so it
# reaches that check rather than the one above.
test_expect_success 'a snapshot name in revision syntax is refused' '
	cgit_rc namesrc "url=names/snapshot/master~1.tar" >tmp &&
	grep "^Status: 404" tmp &&
	grep "<div class=.error.>Not found</div>" tmp &&
	cgit_ok namesrc "url=names/snapshot/master.tar" >tmp
'

# Two tags differing only in the case of their v keep it on both names. Two
# loose files differing in case alone cannot coexist on a case-insensitive
# filesystem, so the first is packed before the second goes in, and both are
# packed after, so that each name is looked up exactly.
test_expect_success 'set up a repo whose tags differ only in case' '
	mkrepo repos/vcase 2 &&
	git -C repos/vcase update-ref refs/tags/v1.0 HEAD &&
	git -C repos/vcase pack-refs --all &&
	git -C repos/vcase update-ref refs/tags/V1.0 HEAD~1 &&
	git -C repos/vcase pack-refs --all &&
	cgitrc_repo vcaserc vcase snapshots=tar
'

test_expect_success 'a real case collision keeps the v on both tags' '
	cgit_ok vcaserc "url=vcase/refs/" >tmp &&
	grep ">vcase-v1.0.tar</a>" tmp &&
	grep ">vcase-V1.0.tar</a>" tmp &&
	! grep ">vcase-1.0.tar</a>" tmp
'

test_expect_success 'each of those names leads back to its own tag' '
	cgit_ok vcaserc "url=vcase/snapshot/vcase-v1.0.tar" >tmp &&
	snapshot_members &&
	grep "^vcase-v1.0/file-2$" members &&
	cgit_ok vcaserc "url=vcase/snapshot/vcase-V1.0.tar" >tmp &&
	snapshot_members &&
	grep "^vcase-V1.0/file-1$" members &&
	! grep "file-2" members
'

# A name that is all suffix names no revision, though the lookup would settle
# on the branch called v that the full name finds.
test_expect_success 'set up a repo with a branch named v' '
	mkrepo repos/stem 1 &&
	git -C repos/stem branch v HEAD &&
	cgitrc_repo stemrc stem snapshots=tar
'

test_expect_success 'a snapshot name with nothing before its suffix is refused' '
	cgit_rc stemrc "url=stem/snapshot/.tar" >tmp &&
	grep "^Status: 404" tmp &&
	grep "<div class=.error.>Not found</div>" tmp &&
	cgit_ok stemrc "url=stem/snapshot/stem-v.tar" >tmp &&
	snapshot_members &&
	grep "^stem-v/file-1$" members
'

# An empty snapshot-prefix gives the members no leading directory rather than
# a leading slash.
test_expect_success 'an empty snapshot-prefix puts no slash ahead of the members' '
	cgitrc_repo noprefixrc stem snapshots=tar repo.snapshot-prefix= &&
	cgit_ok noprefixrc "url=stem/snapshot/stem.tar&id=master" >tmp &&
	snapshot_members &&
	grep "^file-1$" members &&
	! grep "^/" members
'

# A snapshot is named after the last component of the url once its .git
# suffix is gone, so a url of .git alone leaves the ref as the whole name.
test_expect_success 'set up repositories whose urls end in .git' '
	mkrepo repos/dotgit 1 &&
	git -C repos/dotgit tag v1.0 &&
	{
		echo "virtual-root=/" &&
		echo "cache-size=0" &&
		echo "snapshots=tar" &&
		echo "repo.url=.git" &&
		echo "repo.path=$PWD/repos/dotgit/.git" &&
		echo "repo.url=sub/.git" &&
		echo "repo.path=$PWD/repos/dotgit/.git"
	} >dotgitrc
'

test_expect_success 'a url of .git names a snapshot after the ref alone' '
	cgit_ok dotgitrc "url=.git/refs/" >tmp &&
	grep ">v1.0.tar<" tmp &&
	cgit_ok dotgitrc "url=.git/snapshot/v1.0.tar" >tmp &&
	snapshot_members &&
	grep "^v1.0/file-1$" members
'

test_expect_success 'a url ending in /.git names a snapshot after its directory' '
	cgit_ok dotgitrc "url=sub/.git/refs/" >tmp &&
	grep ">sub-1.0.tar<" tmp &&
	cgit_ok dotgitrc "url=sub/.git/snapshot/sub-1.0.tar" >tmp &&
	snapshot_members &&
	grep "^sub-1.0/file-1$" members
'

test -n "$CGIT_VALGRIND" || test_set_prereq NO_VALGRIND

# A server may run cgit with no PATH at all, and the compressor is then looked
# for on the default path execvp searches. Under valgrind the cgit found here
# is a wrapper script, which needs a PATH of its own.
test_expect_success GZIP,NO_VALGRIND 'a compressed snapshot is made without a PATH' '
	cgitrc_copy nopathrc &&
	cgit_bin=$(command -v cgit) &&
	(
		unset PATH &&
		CGIT_CONFIG="$PWD/nopathrc" QUERY_STRING="url=foo/snapshot/master.tar.gz" "$cgit_bin" >tmp 2>err
	) &&
	test_must_be_empty err &&
	grep "^Status: 200" tmp &&
	strip_headers <tmp >nopath.tar.gz &&
	gzip -dc nopath.tar.gz >nopath.tar &&
	test_cmp_bin expect.tar nopath.tar
'

# A link offers the snapshot of a tag, and the name on it has to lead back to
# that tag even when a branch is spelled like the tag's stem. The branch holds
# the same tree, so only the commit the archive records tells them apart.
test_expect_success 'a snapshot name leads to the tag ahead of a branch' '
	mkrepo repos/shadow 1 &&
	git -C repos/shadow tag v1.0 &&
	shadow=$(git -C repos/shadow commit-tree -m shadow "HEAD^{tree}") &&
	git -C repos/shadow update-ref refs/heads/1.0 "$shadow" &&
	cgitrc_repo shadowrc shadow snapshots=tar &&
	cgit_ok shadowrc "url=shadow/snapshot/shadow-1.0.tar" >tmp &&
	strip_headers <tmp >tag.tar &&
	commit=$(git -C repos/shadow get-tar-commit-id <tag.tar) &&
	want=$(git -C repos/shadow rev-parse "v1.0^{commit}") &&
	test -n "$want" &&
	test "$commit" = "$want"
'

# A signature is a note a pusher stored against the commit, served as it is
# under headers that keep a browser from taking it for anything else, and held
# to max-blob-size like any blob. A compressed tar with no signature of its
# own offers the one made for the plain tar it expands to.
test_expect_success 'set up a repo with signed snapshots' '
	test_create_repo repos/signed &&
	(
		cd repos/signed &&
		for tag in one two three big
		do
			git commit --allow-empty -m $tag &&
			git tag $tag || return 1
		done &&
		echo signature >sig &&
		sig=$(git hash-object -w sig) &&
		git notes --ref signatures/tar.gz add -C $sig one &&
		echo plain >plain &&
		plain=$(git hash-object -w plain) &&
		git notes --ref signatures/tar add -C $plain three &&
		big=$(test_seq 1 2000 | git hash-object -w --stdin) &&
		git notes --ref signatures/tar.gz add -C $big big
	) &&
	cgitrc_repo signedrc signed snapshots=tar.gz
'

test_expect_success 'a snapshot and its signature are linked as one unit' '
	cgit_ok signedrc "url=signed/refs/" >tmp &&
	grep "<span class=.snapshot.><a href=./signed/snapshot/signed-one.tar.gz.>signed-one.tar.gz</a> (<a href=./signed/snapshot/signed-one.tar.gz.asc.>sig</a>)</span>" tmp
'

test_expect_success 'a signature is served as the note holds it' '
	cgit_ok signedrc "url=signed/snapshot/signed-one.tar.gz.asc" >tmp &&
	sed "/^$/q" tmp >headers &&
	grep "^Content-Type: application/pgp-signature$" headers &&
	grep "^Content-Disposition: inline; filename=\"signed-one.tar.gz.asc\"$" headers &&
	grep "^X-Content-Type-Options: nosniff$" headers &&
	grep "^Content-Security-Policy: default-src .none.$" headers &&
	strip_headers <tmp >actual &&
	echo signature >expect &&
	test_cmp expect actual
'

test_expect_success 'a ref with no signature has none to offer' '
	cgit_ok signedrc "url=signed/refs/" >tmp &&
	grep "<span class=.snapshot.><a href=./signed/snapshot/signed-two.tar.gz.>signed-two.tar.gz</a></span>" tmp &&
	cgit_rc signedrc "url=signed/snapshot/signed-two.tar.gz.asc" >tmp &&
	grep "^Status: 404" tmp &&
	grep "<div class=.error.>No signature for signed-two.tar.gz</div>" tmp
'

test_expect_success 'a compressed tar offers the plain tar signature' '
	cgit_ok signedrc "url=signed/refs/" >tmp &&
	grep "<span class=.snapshot.><a href=./signed/snapshot/signed-three.tar.gz.>signed-three.tar.gz</a> (<a href=./signed/snapshot/signed-three.tar.asc.>sig</a>)</span>" tmp &&
	cgit_ok signedrc "url=signed/snapshot/signed-three.tar.asc" >tmp &&
	strip_headers <tmp >actual &&
	echo plain >expect &&
	test_cmp expect actual
'

test_expect_success 'a signature past max-blob-size is refused' '
	cgitrc_repo sigrc signed snapshots=tar.gz max-blob-size=1 &&
	cgit_rc sigrc "url=signed/snapshot/signed-big.tar.gz.asc" >tmp &&
	grep "^Status: 413" tmp &&
	cgit_ok sigrc "url=signed/snapshot/signed-one.tar.gz.asc" >tmp
'

# A tag named in UTF-8 gives a snapshot name past ASCII, which is sent as it
# is in filename and percent-encoded in filename*.
test_expect_success 'a snapshot name in UTF-8 is also sent as filename*' '
	mkrepo repos/accent 1 &&
	tag=$(printf "caf\303\251") &&
	git -C repos/accent tag "$tag" &&
	cgitrc_repo accentrc accent snapshots=tar &&
	cgit_ok accentrc "url=accent/refs/tags" >tmp &&
	test_grep "<a href=./accent/snapshot/accent-caf%C3%A9.tar.>accent-$tag.tar</a>" tmp &&
	cgit_ok accentrc "url=accent/snapshot/accent-caf%C3%A9.tar" >tmp &&
	sed "/^$/q" tmp >headers &&
	test_grep "^Content-Disposition: inline; filename=.accent-$tag.tar.; filename\*=UTF-8..accent-caf%C3%A9.tar$" headers
'

test_done