git_repository: return actual commit

Using that repository rules now may return a non-None value,
make git_repository return the arguments that make the rule
reproducible; in particular, return the actual commit (instead
of a tag) and the date of the commit, so support shallow clones.

The added test also demonstrates how the `bazel sync` command
together with `--experimental_repository_resolved_file` can be
used to replay an earlier state of external dependencies.

Change-Id: Ifa1cfdfdb5eb299a15b9d0ec7d285dc84c0bcdc0
PiperOrigin-RevId: 200705705

3 files changed, 140 insertions, 20 deletions

diff --git a/src/test/shell/bazel/BUILD b/src/test/shell/bazel/BUILD
index 045f33ed19..0b38915e46 100644
--- a/src/test/shell/bazel/BUILD
+++ b/src/test/shell/bazel/BUILD
@@ -369,7 +369,7 @@ sh_test(
     size = "medium",
     srcs = ["workspace_resolved_test.sh"],
     data = [":test-deps"],
-    shard_count = 3,
+    shard_count = 4,
 )
 
 sh_test(
diff --git a/src/test/shell/bazel/workspace_resolved_test.sh b/src/test/shell/bazel/workspace_resolved_test.sh
index 8504bb1512..d00202275c 100755
--- a/src/test/shell/bazel/workspace_resolved_test.sh
+++ b/src/test/shell/bazel/workspace_resolved_test.sh
@@ -90,6 +90,83 @@ EOF
       || fail "Not the correct number of original attributes"
 }
 
+test_git_return_value() {
+  EXTREPODIR=`pwd`
+  export GIT_CONFIG_NOSYSTEM=YES
+
+  mkdir extgit
+  (cd extgit && git init \
+       && git config user.email 'me@example.com' \
+       && git config user.name 'E X Ample' )
+  echo Hello World > extgit/hello.txt
+  (cd extgit
+   git add .
+   git commit --author="A U Thor <author@example.com>" -m 'initial commit'
+   git tag mytag)
+
+  # Check out the external git repository at the given tag, and record
+  # the return value of the git rule.
+  mkdir tagcheckout
+  cd tagcheckout
+  cat > WORKSPACE <<EOF
+load("@bazel_tools//tools/build_defs/repo:git.bzl", "new_git_repository")
+new_git_repository(
+  name="ext",
+  remote="file://${EXTREPODIR}/extgit/.git",
+  tag="mytag",
+  build_file_content="exports_files([\"hello.txt\"])",
+)
+EOF
+  bazel sync --experimental_repository_resolved_file=../repo.bzl
+  # some of the file systems on our test machines are really slow to
+  # notice the creation of a file---even after the call to sync(1).
+  bazel shutdown; sync; sleep 10
+
+  cd ..
+  echo; cat repo.bzl; echo
+
+  # Now add an additional commit to the upstream repository and
+  # force update the tag
+  echo CHANGED > extgit/hello.txt
+  (cd extgit
+   git add .
+   git commit --author="A U Thor <author@example.com>" -m 'change hello.txt'
+   git tag -f mytag)
+
+  # Verify that the recorded resolved information is what we expect. In
+  # particular, verify that we don't get the new upstream commit.
+  mkdir analysisrepo
+  cd analysisrepo
+  cp ../repo.bzl .
+  cat > workspace.bzl <<'EOF'
+load("@bazel_tools//tools/build_defs/repo:git.bzl", "new_git_repository")
+load("//:repo.bzl", "resolved")
+
+def repo():
+    for entry in resolved:
+        if entry["original_attributes"]["name"] == "ext":
+            new_git_repository(**(entry["repositories"][0]["attributes"]))
+EOF
+  cat > WORKSPACE <<'EOF'
+load("//:workspace.bzl", "repo")
+repo()
+EOF
+  cat > BUILD <<'EOF'
+genrule(
+  name = "out",
+  outs = ["out.txt"],
+  srcs = ["@ext//:hello.txt"],
+  cmd = "cp $< $@",
+)
+EOF
+  bazel build //:out
+  grep "Hello World" `bazel info bazel-genfiles`/out.txt \
+       || fail "ext not taken at the right commit"
+  grep "CHANGED" `bazel info bazel-genfiles`/out.txt  \
+       && fail "not taking the frozen commit" || :
+}
+
+
 test_sync_calls_all() {
   mkdir sync_calls_all && cd sync_calls_all
   rm -rf fetchrepo
diff --git a/tools/build_defs/repo/git.bzl b/tools/build_defs/repo/git.bzl
index a46330542c..3663dc804f 100644
--- a/tools/build_defs/repo/git.bzl
+++ b/tools/build_defs/repo/git.bzl
@@ -80,17 +80,37 @@ set -ex
     if st.return_code:
         fail("error updating submodules %s:\n%s" % (ctx.name, st.stderr))
 
-def _new_git_repository_implementation(ctx):
-    if ((not ctx.attr.build_file and not ctx.attr.build_file_content) or
-        (ctx.attr.build_file and ctx.attr.build_file_content)):
-        fail("Exactly one of build_file and build_file_content must be provided.")
-    _clone_or_update(ctx)
-    workspace_and_buildfile(ctx)
-    patch(ctx)
-
-def _git_repository_implementation(ctx):
-    _clone_or_update(ctx)
-    patch(ctx)
+    # After the fact, determine the actual commit and its date
+    actual_commit = ctx.execute([
+        bash_exe,
+        "-c",
+        "(cd '{directory}' && git log -n 1 --pretty='format:%H')".format(
+            directory = ctx.path("."),
+        ),
+    ]).stdout
+    shallow_date = ctx.execute([
+        bash_exe,
+        "-c",
+        "(cd '{directory}' && git log -n 1 --pretty='format:%cd' --date='format:%Y-%d-%m')".format(
+            directory = ctx.path("."),
+        ),
+    ]).stdout
+    return {"commit": actual_commit, "shallow_since": shallow_date}
+
+def _update_commit(orig, keys, override):
+    # Merge the override information into the dict, resulting by taking the
+    # given keys, as well as the name, form orig (if present there).
+    result = {}
+    for key in keys:
+        if getattr(orig, key) != None:
+            result[key] = getattr(orig, key)
+    result["name"] = orig.name
+    result.update(override)
+
+    # remove tag if we found the actual commit
+    if "commit" in result:
+        result.pop("tag", None)
+    return result
 
 _common_attrs = {
     "remote": attr.string(mandatory = True),
@@ -106,19 +126,38 @@ _common_attrs = {
     "patch_cmds": attr.string_list(default = []),
 }
 
+_new_git_repository_attrs = dict(_common_attrs.items() + {
+    "build_file": attr.label(allow_single_file = True),
+    "build_file_content": attr.string(),
+    "workspace_file": attr.label(),
+    "workspace_file_content": attr.string(),
+}.items())
+
+def _new_git_repository_implementation(ctx):
+    if ((not ctx.attr.build_file and not ctx.attr.build_file_content) or
+        (ctx.attr.build_file and ctx.attr.build_file_content)):
+        fail("Exactly one of build_file and build_file_content must be provided.")
+    update = _clone_or_update(ctx)
+    workspace_and_buildfile(ctx)
+    patch(ctx)
+    return _update_commit(ctx.attr, _new_git_repository_attrs.keys(), update)
+
+def _git_repository_implementation(ctx):
+    update = _clone_or_update(ctx)
+    patch(ctx)
+    return _update_commit(ctx.attr, _common_attrs.keys(), update)
+
 new_git_repository = repository_rule(
     implementation = _new_git_repository_implementation,
-    attrs = dict(_common_attrs.items() + {
-        "build_file": attr.label(allow_single_file = True),
-        "build_file_content": attr.string(),
-        "workspace_file": attr.label(),
-        "workspace_file_content": attr.string(),
-    }.items()),
+    attrs = _new_git_repository_attrs,
 )
 """Clone an external git repository.
 
 Clones a Git repository, checks out the specified tag, or commit, and
-makes its targets available for binding.
+makes its targets available for binding. Also determine the id of the
+commit actually checkted out and its date, and return a dict with paramters
+that provide a reproducible version of this rule (which a tag not necessarily
+is).
 
 Args:
   name: A unique name for this rule.
@@ -173,7 +212,11 @@ git_repository = repository_rule(
 """Clone an external git repository.
 
 Clones a Git repository, checks out the specified tag, or commit, and
-makes its targets available for binding.
+makes its targets available for binding. Also determine the id of the
+commit actually checkted out and its date, and return a dict with paramters
+that provide a reproducible version of this rule (which a tag not necessarily
+is).
+
 
 Args:
   name: A unique name for this rule.