Merge branch 'js/filter-options-should-use-plain-int'

Update the protocol message specification to allow only the limited
use of scaled quantities.  This is ensure potential compatibility
issues will not go out of hand.

* js/filter-options-should-use-plain-int:
  filter-options: expand scaled numbers
  tree:<depth>: skip some trees even when collecting omits
  list-objects-filter: teach tree:# how to handle >0

Documentation/rev-list-options.txt

@@ -730,8 +730,13 @@ specification contained in <path>.
 +
 The form '--filter=tree:<depth>' omits all blobs and trees whose depth
 from the root tree is >= <depth> (minimum depth if an object is located
-at multiple depths in the commits traversed). Currently, only <depth>=0
-is supported, which omits all blobs and trees.
+at multiple depths in the commits traversed). <depth>=0 will not include
+any trees or blobs unless included explicitly in the command-line (or
+standard input when --stdin is used). <depth>=1 will include only the
+tree and blobs which are referenced directly by a commit reachable from
+<commit> or an explicitly-given object. <depth>=2 is like <depth>=1
+while also including trees and blobs one more level removed from an
+explicitly-given commit or tree.
 
 --no-filter::
 	Turn off any previous `--filter=` argument.

Documentation/technical/protocol-v2.txt

@@ -296,7 +296,13 @@ included in the client's request:
 	Request that various objects from the packfile be omitted
 	using one of several filtering techniques. These are intended
 	for use with partial clone and partial fetch operations. See
-	`rev-list` for possible "filter-spec" values.
+	`rev-list` for possible "filter-spec" values. When communicating
+	with other processes, senders SHOULD translate scaled integers
+	(e.g. "1k") into a fully-expanded form (e.g. "1024") to aid
+	interoperability with older receivers that may not understand
+	newly-invented scaling suffixes. However, receivers SHOULD
+	accept the following suffixes: 'k', 'm', and 'g' for 1024,
+	1048576, and 1073741824, respectively.
 
 If the 'ref-in-want' feature is advertised, the following argument can
 be included in the client's request as well as the potential addition of

builtin/clone.c

@@ -1136,9 +1136,13 @@ int cmd_clone(int argc, const char **argv, const char *prefix)
 				     option_upload_pack);
 
 	if (filter_options.choice) {
+		struct strbuf expanded_filter_spec = STRBUF_INIT;
+		expand_list_objects_filter_spec(&filter_options,
+						&expanded_filter_spec);
 		transport_set_option(transport, TRANS_OPT_LIST_OBJECTS_FILTER,
-				     filter_options.filter_spec);
+				     expanded_filter_spec.buf);
 		transport_set_option(transport, TRANS_OPT_FROM_PROMISOR, "1");
+		strbuf_release(&expanded_filter_spec);
 	}
 
 	if (transport->smart_options && !deepen && !filter_options.choice)

builtin/fetch.c

@@ -1165,6 +1165,7 @@ static void add_negotiation_tips(struct git_transport_options *smart_options)
 static struct transport *prepare_transport(struct remote *remote, int deepen)
 {
 	struct transport *transport;
+
 	transport = transport_get(remote, NULL);
 	transport_set_verbosity(transport, verbosity, progress);
 	transport->family = family;
@@ -1184,9 +1185,13 @@ static struct transport *prepare_transport(struct remote *remote, int deepen)
 	if (update_shallow)
 		set_option(transport, TRANS_OPT_UPDATE_SHALLOW, "yes");
 	if (filter_options.choice) {
+		struct strbuf expanded_filter_spec = STRBUF_INIT;
+		expand_list_objects_filter_spec(&filter_options,
+						&expanded_filter_spec);
 		set_option(transport, TRANS_OPT_LIST_OBJECTS_FILTER,
-			   filter_options.filter_spec);
+			   expanded_filter_spec.buf);
 		set_option(transport, TRANS_OPT_FROM_PROMISOR, "1");
+		strbuf_release(&expanded_filter_spec);
 	}
 	if (negotiation_tip.nr) {
 		if (transport->smart_options)

fetch-pack.c

@@ -329,9 +329,14 @@ static int find_common(struct fetch_negotiator *negotiator,
 			packet_buf_write(&req_buf, "deepen-not %s", s->string);
 		}
 	}
-	if (server_supports_filtering && args->filter_options.choice)
+	if (server_supports_filtering && args->filter_options.choice) {
+		struct strbuf expanded_filter_spec = STRBUF_INIT;
+		expand_list_objects_filter_spec(&args->filter_options,
+						&expanded_filter_spec);
 		packet_buf_write(&req_buf, "filter %s",
-				 args->filter_options.filter_spec);
+				 expanded_filter_spec.buf);
+		strbuf_release(&expanded_filter_spec);
+	}
 	packet_buf_flush(&req_buf);
 	state_len = req_buf.len;
 
@@ -1122,9 +1127,13 @@ static int send_fetch_request(struct fetch_negotiator *negotiator, int fd_out,
 	/* Add filter */
 	if (server_supports_feature("fetch", "filter", 0) &&
 	    args->filter_options.choice) {
+		struct strbuf expanded_filter_spec = STRBUF_INIT;
 		print_verbose(args, _("Server supports filter"));
+		expand_list_objects_filter_spec(&args->filter_options,
+						&expanded_filter_spec);
 		packet_buf_write(&req_buf, "filter %s",
-				 args->filter_options.filter_spec);
+				 expanded_filter_spec.buf);
+		strbuf_release(&expanded_filter_spec);
 	} else if (args->filter_options.choice) {
 		warning("filtering not recognized by server, ignoring");
 	}

list-objects-filter-options.c

@@ -18,8 +18,9 @@
  * See Documentation/rev-list-options.txt for allowed values for <arg>.
  *
  * Capture the given arg as the "filter_spec".  This can be forwarded to
- * subordinate commands when necessary.  We also "intern" the arg for
- * the convenience of the current command.
+ * subordinate commands when necessary (although it's better to pass it through
+ * expand_list_objects_filter_spec() first).  We also "intern" the arg for the
+ * convenience of the current command.
  */
 static int gently_parse_list_objects_filter(
 	struct list_objects_filter_options *filter_options,
@@ -50,16 +51,15 @@ static int gently_parse_list_objects_filter(
 		}
 
 	} else if (skip_prefix(arg, "tree:", &v0)) {
-		unsigned long depth;
-		if (!git_parse_ulong(v0, &depth) || depth != 0) {
+		if (!git_parse_ulong(v0, &filter_options->tree_exclude_depth)) {
 			if (errbuf) {
 				strbuf_addstr(
 					errbuf,
-					_("only 'tree:0' is supported"));
+					_("expected 'tree:<depth>'"));
 			}
 			return 1;
 		}
-		filter_options->choice = LOFC_TREE_NONE;
+		filter_options->choice = LOFC_TREE_DEPTH;
 		return 0;
 
 	} else if (skip_prefix(arg, "sparse:oid=", &v0)) {
@@ -112,6 +112,21 @@ int opt_parse_list_objects_filter(const struct option *opt,
 	return parse_list_objects_filter(filter_options, arg);
 }
 
+void expand_list_objects_filter_spec(
+	const struct list_objects_filter_options *filter,
+	struct strbuf *expanded_spec)
+{
+	strbuf_init(expanded_spec, strlen(filter->filter_spec));
+	if (filter->choice == LOFC_BLOB_LIMIT)
+		strbuf_addf(expanded_spec, "blob:limit=%lu",
+			    filter->blob_limit_value);
+	else if (filter->choice == LOFC_TREE_DEPTH)
+		strbuf_addf(expanded_spec, "tree:%lu",
+			    filter->tree_exclude_depth);
+	else
+		strbuf_addstr(expanded_spec, filter->filter_spec);
+}
+
 void list_objects_filter_release(
 	struct list_objects_filter_options *filter_options)
 {

list-objects-filter-options.h

@@ -2,6 +2,7 @@
 #define LIST_OBJECTS_FILTER_OPTIONS_H
 
 #include "parse-options.h"
+#include "strbuf.h"
 
 /*
  * The list of defined filters for list-objects.
@@ -10,7 +11,7 @@ enum list_objects_filter_choice {
 	LOFC_DISABLED = 0,
 	LOFC_BLOB_NONE,
 	LOFC_BLOB_LIMIT,
-	LOFC_TREE_NONE,
+	LOFC_TREE_DEPTH,
 	LOFC_SPARSE_OID,
 	LOFC_SPARSE_PATH,
 	LOFC__COUNT /* must be last */
@@ -20,8 +21,9 @@ struct list_objects_filter_options {
 	/*
 	 * 'filter_spec' is the raw argument value given on the command line
 	 * or protocol request.  (The part after the "--keyword=".)  For
-	 * commands that launch filtering sub-processes, this value should be
-	 * passed to them as received by the current process.
+	 * commands that launch filtering sub-processes, or for communication
+	 * over the network, don't use this value; use the result of
+	 * expand_list_objects_filter_spec() instead.
 	 */
 	char *filter_spec;
 
@@ -44,6 +46,7 @@ struct list_objects_filter_options {
 	struct object_id *sparse_oid_value;
 	char *sparse_path_value;
 	unsigned long blob_limit_value;
+	unsigned long tree_exclude_depth;
 };
 
 /* Normalized command line arguments */
@@ -61,6 +64,17 @@ int opt_parse_list_objects_filter(const struct option *opt,
 	  N_("object filtering"), 0, \
 	  opt_parse_list_objects_filter }
 
+/*
+ * Translates abbreviated numbers in the filter's filter_spec into their
+ * fully-expanded forms (e.g., "limit:blob=1k" becomes "limit:blob=1024").
+ *
+ * This form should be used instead of the raw filter_spec field when
+ * communicating with a remote process or subprocess.
+ */
+void expand_list_objects_filter_spec(
+	const struct list_objects_filter_options *filter,
+	struct strbuf *expanded_spec);
+
 void list_objects_filter_release(
 	struct list_objects_filter_options *filter_options);
 

list-objects-filter.c

@@ -10,6 +10,7 @@
 #include "list-objects.h"
 #include "list-objects-filter.h"
 #include "list-objects-filter-options.h"
+#include "oidmap.h"
 #include "oidset.h"
 #include "object-store.h"
 
@@ -84,11 +85,44 @@ static void *filter_blobs_none__init(
  * A filter for list-objects to omit ALL trees and blobs from the traversal.
  * Can OPTIONALLY collect a list of the omitted OIDs.
  */
-struct filter_trees_none_data {
+struct filter_trees_depth_data {
 	struct oidset *omits;
+
+	/*
+	 * Maps trees to the minimum depth at which they were seen. It is not
+	 * necessary to re-traverse a tree at deeper or equal depths than it has
+	 * already been traversed.
+	 *
+	 * We can't use LOFR_MARK_SEEN for tree objects since this will prevent
+	 * it from being traversed at shallower depths.
+	 */
+	struct oidmap seen_at_depth;
+
+	unsigned long exclude_depth;
+	unsigned long current_depth;
 };
 
-static enum list_objects_filter_result filter_trees_none(
+struct seen_map_entry {
+	struct oidmap_entry base;
+	size_t depth;
+};
+
+/* Returns 1 if the oid was in the omits set before it was invoked. */
+static int filter_trees_update_omits(
+	struct object *obj,
+	struct filter_trees_depth_data *filter_data,
+	int include_it)
+{
+	if (!filter_data->omits)
+		return 0;
+
+	if (include_it)
+		return oidset_remove(filter_data->omits, &obj->oid);
+	else
+		return oidset_insert(filter_data->omits, &obj->oid);
+}
+
+static enum list_objects_filter_result filter_trees_depth(
 	struct repository *r,
 	enum list_objects_filter_situation filter_situation,
 	struct object *obj,
@@ -96,43 +130,91 @@ static enum list_objects_filter_result filter_trees_none(
 	const char *filename,
 	void *filter_data_)
 {
-	struct filter_trees_none_data *filter_data = filter_data_;
+	struct filter_trees_depth_data *filter_data = filter_data_;
+	struct seen_map_entry *seen_info;
+	int include_it = filter_data->current_depth <
+		filter_data->exclude_depth;
+	int filter_res;
+	int already_seen;
+
+	/*
+	 * Note that we do not use _MARK_SEEN in order to allow re-traversal in
+	 * case we encounter a tree or blob again at a shallower depth.
+	 */
 
 	switch (filter_situation) {
 	default:
 		BUG("unknown filter_situation: %d", filter_situation);
 
-	case LOFS_BEGIN_TREE:
-	case LOFS_BLOB:
-		if (filter_data->omits) {
-			oidset_insert(filter_data->omits, &obj->oid);
-			/* _MARK_SEEN but not _DO_SHOW (hard omit) */
-			return LOFR_MARK_SEEN;
-		} else {
-			/*
-			 * Not collecting omits so no need to to traverse tree.
-			 */
-			return LOFR_SKIP_TREE | LOFR_MARK_SEEN;
-		}
-
 	case LOFS_END_TREE:
 		assert(obj->type == OBJ_TREE);
+		filter_data->current_depth--;
 		return LOFR_ZERO;
 
+	case LOFS_BLOB:
+		filter_trees_update_omits(obj, filter_data, include_it);
+		return include_it ? LOFR_MARK_SEEN | LOFR_DO_SHOW : LOFR_ZERO;
+
+	case LOFS_BEGIN_TREE:
+		seen_info = oidmap_get(
+			&filter_data->seen_at_depth, &obj->oid);
+		if (!seen_info) {
+			seen_info = xcalloc(1, sizeof(*seen_info));
+			oidcpy(&seen_info->base.oid, &obj->oid);
+			seen_info->depth = filter_data->current_depth;
+			oidmap_put(&filter_data->seen_at_depth, seen_info);
+			already_seen = 0;
+		} else {
+			already_seen =
+				filter_data->current_depth >= seen_info->depth;
+		}
+
+		if (already_seen) {
+			filter_res = LOFR_SKIP_TREE;
+		} else {
+			int been_omitted = filter_trees_update_omits(
+				obj, filter_data, include_it);
+			seen_info->depth = filter_data->current_depth;
+
+			if (include_it)
+				filter_res = LOFR_DO_SHOW;
+			else if (filter_data->omits && !been_omitted)
+				/*
+				 * Must update omit information of children
+				 * recursively; they have not been omitted yet.
+				 */
+				filter_res = LOFR_ZERO;
+			else
+				filter_res = LOFR_SKIP_TREE;
+		}
+
+		filter_data->current_depth++;
+		return filter_res;
 	}
 }
 
-static void* filter_trees_none__init(
+static void filter_trees_free(void *filter_data) {
+	struct filter_trees_depth_data *d = filter_data;
+	if (!d)
+		return;
+	oidmap_free(&d->seen_at_depth, 1);
+	free(d);
+}
+
+static void *filter_trees_depth__init(
 	struct oidset *omitted,
 	struct list_objects_filter_options *filter_options,
 	filter_object_fn *filter_fn,
 	filter_free_fn *filter_free_fn)
 {
-	struct filter_trees_none_data *d = xcalloc(1, sizeof(*d));
+	struct filter_trees_depth_data *d = xcalloc(1, sizeof(*d));
 	d->omits = omitted;
+	oidmap_init(&d->seen_at_depth, 0);
+	d->exclude_depth = filter_options->tree_exclude_depth;
+	d->current_depth = 0;
 
-	*filter_fn = filter_trees_none;
-	*filter_free_fn = free;
+	*filter_fn = filter_trees_depth;
+	*filter_free_fn = filter_trees_free;
 	return d;
 }
 
@@ -430,7 +512,7 @@ static filter_init_fn s_filters[] = {
 	NULL,
 	filter_blobs_none__init,
 	filter_blobs_limit__init,
-	filter_trees_none__init,
+	filter_trees_depth__init,
 	filter_sparse_oid__init,
 	filter_sparse_path__init,
 };

t/t6112-rev-list-filters-objects.sh

@@ -283,7 +283,7 @@ test_expect_success 'verify tree:0 includes trees in "filtered" output' '
 
 # Make sure tree:0 does not iterate through any trees.
 
-test_expect_success 'filter a GIANT tree through tree:0' '
+test_expect_success 'verify skipping tree iteration when not collecting omits' '
 	GIT_TRACE=1 git -C r3 rev-list \
 		--objects --filter=tree:0 HEAD 2>filter_trace &&
 	grep "Skipping contents of tree [.][.][.]" filter_trace >actual &&
@@ -294,6 +294,126 @@ test_expect_success 'filter a GIANT tree through tree:0' '
 	! grep "Skipping contents of tree [^.]" filter_trace
 '
 
+# Test tree:# filters.
+
+expect_has () {
+	commit=$1 &&
+	name=$2 &&
+
+	hash=$(git -C r3 rev-parse $commit:$name) &&
+	grep "^$hash $name$" actual
+}
+
+test_expect_success 'verify tree:1 includes root trees' '
+	git -C r3 rev-list --objects --filter=tree:1 HEAD >actual &&
+
+	# We should get two root directories and two commits.
+	expect_has HEAD "" &&
+	expect_has HEAD~1 ""  &&
+	test_line_count = 4 actual
+'
+
+test_expect_success 'verify tree:2 includes root trees and immediate children' '
+	git -C r3 rev-list --objects --filter=tree:2 HEAD >actual &&
+
+	expect_has HEAD "" &&
+	expect_has HEAD~1 "" &&
+	expect_has HEAD dir1 &&
+	expect_has HEAD pattern &&
+	expect_has HEAD sparse1 &&
+	expect_has HEAD sparse2 &&
+
+	# There are also 2 commit objects
+	test_line_count = 8 actual
+'
+
+test_expect_success 'verify tree:3 includes everything expected' '
+	git -C r3 rev-list --objects --filter=tree:3 HEAD >actual &&
+
+	expect_has HEAD "" &&
+	expect_has HEAD~1 "" &&
+	expect_has HEAD dir1 &&
+	expect_has HEAD dir1/sparse1 &&
+	expect_has HEAD dir1/sparse2 &&
+	expect_has HEAD pattern &&
+	expect_has HEAD sparse1 &&
+	expect_has HEAD sparse2 &&
+
+	# There are also 2 commit objects
+	test_line_count = 10 actual
+'
+
+# Test provisional omit collection logic with a repo that has objects appearing
+# at multiple depths - first deeper than the filter's threshold, then shallow.
+
+test_expect_success 'setup r4' '
+	git init r4 &&
+
+	echo foo > r4/foo &&
+	mkdir r4/subdir &&
+	echo bar > r4/subdir/bar &&
+
+	mkdir r4/filt &&
+	cp -r r4/foo r4/subdir r4/filt &&
+
+	git -C r4 add foo subdir filt &&
+	git -C r4 commit -m "commit msg"
+'
+
+expect_has_with_different_name () {
+	repo=$1 &&
+	name=$2 &&
+
+	hash=$(git -C $repo rev-parse HEAD:$name) &&
+	! grep "^$hash $name$" actual &&
+	grep "^$hash " actual &&
+	! grep "~$hash" actual
+}
+
+test_expect_success 'test tree:# filter provisional omit for blob and tree' '
+	git -C r4 rev-list --objects --filter-print-omitted --filter=tree:2 \
+		HEAD >actual &&
+	expect_has_with_different_name r4 filt/foo &&
+	expect_has_with_different_name r4 filt/subdir
+'
+
+test_expect_success 'verify skipping tree iteration when collecting omits' '
+	GIT_TRACE=1 git -C r4 rev-list --filter-print-omitted \
+		--objects --filter=tree:0 HEAD 2>filter_trace &&
+	grep "^Skipping contents of tree " filter_trace >actual &&
+
+	echo "Skipping contents of tree subdir/..." >expect &&
+	test_cmp expect actual
+'
+
+# Test tree:<depth> where a tree is iterated to twice - once where a subentry is
+# too deep to be included, and again where the blob inside it is shallow enough
+# to be included. This makes sure we don't use LOFR_MARK_SEEN incorrectly (we
+# can't use it because a tree can be iterated over again at a lower depth).
+
+test_expect_success 'tree:<depth> where we iterate over tree at two levels' '
+	git init r5 &&
+
+	mkdir -p r5/a/subdir/b &&
+	echo foo > r5/a/subdir/b/foo &&
+
+	mkdir -p r5/subdir/b &&
+	echo foo > r5/subdir/b/foo &&
+
+	git -C r5 add a subdir &&
+	git -C r5 commit -m "commit msg" &&
+
+	git -C r5 rev-list --objects --filter=tree:4 HEAD >actual &&
+	expect_has_with_different_name r5 a/subdir/b/foo
+'
+
+test_expect_success 'tree:<depth> which filters out blob but given as arg' '
+	blob_hash=$(git -C r4 rev-parse HEAD:subdir/bar) &&
+
+	git -C r4 rev-list --objects --filter=tree:1 HEAD $blob_hash >actual &&
+	grep ^$blob_hash actual
+'
+
 # Delete some loose objects and use rev-list, but WITHOUT any filtering.
 # This models previously omitted objects that we did not receive.
 
@@ -324,4 +444,21 @@ test_expect_success 'rev-list W/ missing=allow-any' '
 	git -C r1 rev-list --quiet --missing=allow-any --objects HEAD
 '
 
+# Test expansion of filter specs.
+
+test_expect_success 'expand blob limit in protocol' '
+	git -C r2 config --local uploadpack.allowfilter 1 &&
+	GIT_TRACE_PACKET="$(pwd)/trace" git -c protocol.version=2 clone \
+		--filter=blob:limit=1k "file://$(pwd)/r2" limit &&
+	! grep "blob:limit=1k" trace &&
+	grep "blob:limit=1024" trace
+'
+
+test_expect_success 'expand tree depth limit in protocol' '
+	GIT_TRACE_PACKET="$(pwd)/tree_trace" git -c protocol.version=2 clone \
+		--filter=tree:0k "file://$(pwd)/r2" tree &&
+	! grep "tree:0k" tree_trace &&
+	grep "tree:0" tree_trace
+'
+
 test_done

transport-helper.c

@@ -679,10 +679,15 @@ static int fetch(struct transport *transport,
 	if (data->transport_options.update_shallow)
 		set_helper_option(transport, "update-shallow", "true");
 
-	if (data->transport_options.filter_options.choice)
-		set_helper_option(
-			transport, "filter",
-			data->transport_options.filter_options.filter_spec);
+	if (data->transport_options.filter_options.choice) {
+		struct strbuf expanded_filter_spec = STRBUF_INIT;
+		expand_list_objects_filter_spec(
+			&data->transport_options.filter_options,
+			&expanded_filter_spec);
+		set_helper_option(transport, "filter",
+				  expanded_filter_spec.buf);
+		strbuf_release(&expanded_filter_spec);
+	}
 
 	if (data->transport_options.negotiation_tips)
 		warning("Ignoring --negotiation-tip because the protocol does not support it.");

upload-pack.c

@@ -139,14 +139,17 @@ static void create_pack_file(const struct object_array *have_obj,
 	if (use_include_tag)
 		argv_array_push(&pack_objects.args, "--include-tag");
 	if (filter_options.filter_spec) {
+		struct strbuf expanded_filter_spec = STRBUF_INIT;
+		expand_list_objects_filter_spec(&filter_options,
+						&expanded_filter_spec);
 		if (pack_objects.use_shell) {
 			struct strbuf buf = STRBUF_INIT;
-			sq_quote_buf(&buf, filter_options.filter_spec);
+			sq_quote_buf(&buf, expanded_filter_spec.buf);
 			argv_array_pushf(&pack_objects.args, "--filter=%s", buf.buf);
 			strbuf_release(&buf);
 		} else {
 			argv_array_pushf(&pack_objects.args, "--filter=%s",
-					 filter_options.filter_spec);
+					 expanded_filter_spec.buf);
 		}
 	}