remote: introduce config to set prefetch refs

This commit introduces a new configuration option,
remote.<name>.prefetchref, which allows users to specify specific
ref patterns to be prefetched during a git fetch --prefetch
operation.

The new option accepts a space-separated list of ref patterns.
When the --prefetch option is used with git fetch, only the refs
matching these patterns will be prefetched, instead of the
default behavior of prefetching all fetchable refs.

Example usage in .git/config:
[remote "origin"]
    prefetchref = "refs/heads/main refs/heads/feature/*"

This change allows users to optimize their prefetch operations, potentially
reducing network traffic and improving performance for large repositories
with many refs.

Signed-off-by: Shubham Kanodia <[email protected]>

Author: pastelsky

Documentation/config/remote.txt

@@ -33,6 +33,12 @@ remote.<name>.fetch::
 	The default set of "refspec" for linkgit:git-fetch[1]. See
 	linkgit:git-fetch[1].
 
+remote.<name>.prefetchref::
+    Specify the refs to be prefetched when fetching from this remote.
+    The value is a space-separated list of ref patterns (e.g., "refs/heads/master refs/heads/develop*").
+    These patterns are used as the source part of the refspecs for prefetching.
+    This can be used to optimize fetch operations by specifying exactly which refs should be prefetched.
+
 remote.<name>.push::
 	The default set of "refspec" for linkgit:git-push[1]. See
 	linkgit:git-push[1].

builtin/fetch.c

@@ -38,6 +38,7 @@
 #include "trace.h"
 #include "trace2.h"
 #include "bundle-uri.h"
+#include "wildmatch.h"
 
 #define FORCED_UPDATES_DELAY_WARNING_IN_MS (10 * 1000)
 
@@ -485,6 +486,49 @@ static void filter_prefetch_refspec(struct refspec *rs)
 	}
 }
 
+static int matches_prefetch_refs(const char *refname, const struct string_list *prefetch_refs)
+{
+	int i;
+	int has_positive = 0;
+	int matched_positive = 0;
+	int matched_negative = 0;
+
+	for (i = 0; i < prefetch_refs->nr; i++) {
+		const char *pattern = prefetch_refs->items[i].string;
+		int is_negative = (*pattern == '!');
+
+		if (is_negative)
+			pattern++;
+		else
+			has_positive = 1;
+
+		if (wildmatch(pattern, refname, 0) == 0) {
+			if (is_negative)
+				matched_negative = 1;
+			else
+				matched_positive = 1;
+		}
+	}
+
+	if (!has_positive)
+		return !matched_negative;
+
+	return matched_positive && !matched_negative;
+}
+
+
+static void ref_remove(struct ref **head, struct ref *to_remove)
+{
+	struct ref **pp, *p;
+
+	for (pp = head; (p = *pp) != NULL; pp = &p->next) {
+		if (p == to_remove) {
+			*pp = p->next;
+			return;
+		}
+	}
+}
+
 static struct ref *get_ref_map(struct remote *remote,
 			       const struct ref *remote_refs,
 			       struct refspec *rs,
@@ -502,6 +546,7 @@ static struct ref *get_ref_map(struct remote *remote,
 	int existing_refs_populated = 0;
 
 	filter_prefetch_refspec(rs);
+
 	if (remote)
 		filter_prefetch_refspec(&remote->fetch);
 
@@ -610,6 +655,22 @@ static struct ref *get_ref_map(struct remote *remote,
 	else
 		ref_map = apply_negative_refspecs(ref_map, &remote->fetch);
 
+	/**
+	 * Filter out advertised refs that we don't want to fetch during
+	 * prefetch if a prefetchref config is set
+	 */
+	if (prefetch && remote->prefetch_refs.nr) {
+		struct ref *ref, *next;
+		for (ref = ref_map; ref; ref = next) {
+			next = ref->next;
+
+			if (!matches_prefetch_refs(ref->name, &remote->prefetch_refs)) {
+					ref_remove(&ref_map, ref);
+					free_one_ref(ref);
+			}
+		}
+	}
+
 	ref_map = ref_remove_duplicates(ref_map);
 
 	for (rm = ref_map; rm; rm = rm->next) {

remote.c

@@ -141,6 +141,7 @@ static struct remote *make_remote(struct remote_state *remote_state,
 	ret->prune = -1;  /* unspecified */
 	ret->prune_tags = -1;  /* unspecified */
 	ret->name = xstrndup(name, len);
+	string_list_init_dup(&ret->prefetch_refs);
 	refspec_init(&ret->push, REFSPEC_PUSH);
 	refspec_init(&ret->fetch, REFSPEC_FETCH);
 
@@ -166,6 +167,7 @@ static void remote_clear(struct remote *remote)
 	free((char *)remote->uploadpack);
 	FREE_AND_NULL(remote->http_proxy);
 	FREE_AND_NULL(remote->http_proxy_authmethod);
+	string_list_clear(&remote->prefetch_refs, 0);
 }
 
 static void add_merge(struct branch *branch, const char *name)
@@ -456,6 +458,12 @@ static int handle_config(const char *key, const char *value,
 		remote->prune = git_config_bool(key, value);
 	else if (!strcmp(subkey, "prunetags"))
 		remote->prune_tags = git_config_bool(key, value);
+	else if (!strcmp(subkey, "prefetchref")) {
+		if (!value)
+			return config_error_nonbool(key);
+		string_list_split(&remote->prefetch_refs, value, ' ', -1);
+		return 0;
+	}
 	else if (!strcmp(subkey, "url")) {
 		if (!value)
 			return config_error_nonbool(key);

remote.h

@@ -5,6 +5,7 @@
 #include "hashmap.h"
 #include "refspec.h"
 #include "strvec.h"
+#include "string-list.h"
 
 struct option;
 struct transport_ls_refs_options;
@@ -77,6 +78,8 @@ struct remote {
 
 	struct refspec fetch;
 
+	struct string_list prefetch_refs;
+
 	/*
 	 * The setting for whether to fetch tags (as a separate rule from the
 	 * configured refspecs);

t/t7900-maintenance.sh

@@ -245,6 +245,86 @@ test_expect_success 'prefetch multiple remotes' '
 	test_subcommand git fetch remote2 $fetchargs <skip-remote1.txt
 '
 
+test_expect_success 'prefetch with positive prefetch ref patterns' '
+	test_create_repo filter-prefetch-positive &&
+	(
+		cd filter-prefetch-positive &&
+		test_commit initial &&
+		git clone . clone2 &&
+		git remote add remote2 "file://$(pwd)/clone2" &&
+
+		cd clone2 &&
+		git checkout -b feature && test_commit feature-commit-2 &&
+		git checkout -b wip/test && test_commit wip-test-commit-2 &&
+		git checkout -b topic/x && test_commit topic-x-commit-2 &&
+		git push -f origin feature wip/test topic/x&&
+		cd .. &&
+
+		git config remote.remote2.prefetchref "refs/heads/feature" &&
+		fetchargs="--prefetch --prune --no-tags --no-write-fetch-head --recurse-submodules=no --quiet" &&
+		GIT_TRACE2_EVENT="$(pwd)/prefetch-positive.txt" git maintenance run --task=prefetch 2>/dev/null &&
+		test_subcommand git fetch remote2 $fetchargs <prefetch-positive.txt &&
+
+		git rev-parse refs/prefetch/remotes/remote2/feature &&
+		test_must_fail git rev-parse refs/prefetch/remotes/remote2/wip/test &&
+		test_must_fail git rev-parse refs/prefetch/remotes/remote2/topic/x
+	)
+'
+
+test_expect_success 'prefetch with negative prefetch ref patterns' '
+	test_create_repo filter-prefetch-negative &&
+	(
+		cd filter-prefetch-negative &&
+		test_commit initial &&
+		git clone . clone3 &&
+		git remote add remote3 "file://$(pwd)/clone3" &&
+		cat .git/config &&
+
+		cd clone3 &&
+		git checkout -b feature && test_commit feature-commit-3 &&
+		git checkout -b wip/test && test_commit wip-test-commit-3 &&
+		git checkout -b topic/x && test_commit topic-x-commit-3 &&
+		git push -f origin feature wip/test topic/x &&
+		cd .. &&
+
+		git config remote.remote3.prefetchref "!refs/heads/wip/*" &&
+		fetchargs="--prefetch --prune --no-tags --no-write-fetch-head --recurse-submodules=no --quiet" &&
+		GIT_TRACE2_EVENT="$(pwd)/prefetch-negative.txt" git maintenance run --task=prefetch 2>/dev/null &&
+		test_subcommand git fetch remote3 $fetchargs <prefetch-negative.txt &&
+		git rev-parse refs/prefetch/remotes/remote3/feature &&
+		git rev-parse refs/prefetch/remotes/remote3/topic/x &&
+		test_must_fail git rev-parse refs/prefetch/remotes/remote3/wip/test
+	)
+'
+
+test_expect_success 'prefetch with positive & negative prefetch ref patterns' '
+	test_create_repo filter-prefetch-mixed &&
+	(
+		cd filter-prefetch-mixed &&
+		test_commit initial &&
+		git clone . clone4 &&
+		git remote add remote4 "file://$(pwd)/clone4" &&
+
+		cd clone4 &&
+		git checkout -b feature && test_commit feature-commit-4 &&
+		git checkout -b topic/x && test_commit topic-x-commit-4 &&
+		git checkout -b topic/y && test_commit topic-y-commit-4 &&
+		git push -f origin feature topic/x topic/y &&
+		cd .. &&
+
+		git config remote.remote4.prefetchref "refs/heads/topic/* !refs/heads/topic/y" &&
+		# git config --add remote.remote4.prefetchref "!refs/topic/y" &&
+		cat .git/config &&
+		fetchargs="--prefetch --prune --no-tags --no-write-fetch-head --recurse-submodules=no --quiet" &&
+		GIT_TRACE2_EVENT="$(pwd)/prefetch-mixed.txt" git maintenance run --task=prefetch 2>/dev/null &&
+		test_subcommand git fetch remote4 $fetchargs <prefetch-mixed.txt &&
+
+		test_must_fail git rev-parse refs/prefetch/remotes/remote4/feature &&
+		test_must_fail git rev-parse refs/prefetch/remotes/remote4/topic/y &&
+		git rev-parse refs/prefetch/remotes/remote4/topic/x
+	)
+'
+
 test_expect_success 'loose-objects task' '
 	# Repack everything so we know the state of the object dir
 	git repack -adk &&