[PATCH v3 13/19] streaming: get rid of `the_repository`

From: Patrick Steinhardt <ps@pks.im>
Date: Nov 23, 2025, 18:59 UTC
Message-ID: <20251123-b4-pks-odb-read-stream-v3-13-1a129182822b@pks.im>
In-Reply-To: <20251123-b4-pks-odb-read-stream-v3-0-1a129182822b@pks.im>

Subsequent commits will move the backend-specific logic of object streaming into their respective subsystems. These subsystems have gotten rid of `the_repository` already, but we still use it in two locations in the streaming subsystem.

Prepare for the move by fixing those two cases. Converting the logic in `open_istream_pack_non_delta()` is trivial as we already got the object database as input.

But for `stream_blob_to_fd()` we have to add a new parameter to make it accessible. So, as we already have to adjust all callers anyway, rename the function to `odb_stream_blob_to_fd()` to indicate it's part of the object subsystem.

Signed-off-by: Patrick Steinhardt <ps@pks.im>
---
 builtin/cat-file.c  |  2 +-
 builtin/fsck.c      |  3 ++-
 builtin/log.c       |  4 ++--
 entry.c             |  2 +-
 parallel-checkout.c |  3 ++-
 streaming.c         | 13 +++++++------
 streaming.h         | 18 +++++++++++++++++-
 7 files changed, 32 insertions(+), 13 deletions(-)

diff --git a/builtin/cat-file.c b/builtin/cat-file.c
index 983ecec837..120d626d66 100644
--- a/builtin/cat-file.c
+++ b/builtin/cat-file.c
@@ -95,7 +95,7 @@ static int filter_object(const char *path, unsigned mode,
 
 static int stream_blob(const struct object_id *oid)
 {
-	if (stream_blob_to_fd(1, oid, NULL, 0))
+	if (odb_stream_blob_to_fd(the_repository->objects, 1, oid, NULL, 0))
 		die("unable to stream %s to stdout", oid_to_hex(oid));
 	return 0;
 }
diff --git a/builtin/fsck.c b/builtin/fsck.c
index b1a650c673..1a348d43c2 100644
--- a/builtin/fsck.c
+++ b/builtin/fsck.c
@@ -340,7 +340,8 @@ static void check_unreachable_object(struct object *obj)
 			}
 			f = xfopen(filename, "w");
 			if (obj->type == OBJ_BLOB) {
-				if (stream_blob_to_fd(fileno(f), &obj->oid, NULL, 1))
+				if (odb_stream_blob_to_fd(the_repository->objects, fileno(f),
+							  &obj->oid, NULL, 1))
 					die_errno(_("could not write '%s'"), filename);
 			} else
 				fprintf(f, "%s\n", describe_object(&obj->oid));
diff --git a/builtin/log.c b/builtin/log.c
index c8319b8af3..e7b83a6e00 100644
--- a/builtin/log.c
+++ b/builtin/log.c
@@ -584,7 +584,7 @@ static int show_blob_object(const struct object_id *oid, struct rev_info *rev, c
 	fflush(rev->diffopt.file);
 	if (!rev->diffopt.flags.textconv_set_via_cmdline ||
 	    !rev->diffopt.flags.allow_textconv)
-		return stream_blob_to_fd(1, oid, NULL, 0);
+		return odb_stream_blob_to_fd(the_repository->objects, 1, oid, NULL, 0);
 
 	if (get_oid_with_context(the_repository, obj_name,
 				 GET_OID_RECORD_PATH,
@@ -594,7 +594,7 @@ static int show_blob_object(const struct object_id *oid, struct rev_info *rev, c
 	    !textconv_object(the_repository, obj_context.path,
 			     obj_context.mode, &oidc, 1, &buf, &size)) {
 		object_context_release(&obj_context);
-		return stream_blob_to_fd(1, oid, NULL, 0);
+		return odb_stream_blob_to_fd(the_repository->objects, 1, oid, NULL, 0);
 	}
 
 	if (!buf)
diff --git a/entry.c b/entry.c
index cae02eb503..38dfe670f7 100644
--- a/entry.c
+++ b/entry.c
@@ -139,7 +139,7 @@ static int streaming_write_entry(const struct cache_entry *ce, char *path,
 	if (fd < 0)
 		return -1;
 
-	result |= stream_blob_to_fd(fd, &ce->oid, filter, 1);
+	result |= odb_stream_blob_to_fd(the_repository->objects, fd, &ce->oid, filter, 1);
 	*fstat_done = fstat_checkout_output(fd, state, statbuf);
 	result |= close(fd);
 
diff --git a/parallel-checkout.c b/parallel-checkout.c
index fba6aa65a6..1cb6701b92 100644
--- a/parallel-checkout.c
+++ b/parallel-checkout.c
@@ -281,7 +281,8 @@ static int write_pc_item_to_fd(struct parallel_checkout_item *pc_item, int fd,
 
 	filter = get_stream_filter_ca(&pc_item->ca, &pc_item->ce->oid);
 	if (filter) {
-		if (stream_blob_to_fd(fd, &pc_item->ce->oid, filter, 1)) {
+		if (odb_stream_blob_to_fd(the_repository->objects, fd,
+					  &pc_item->ce->oid, filter, 1)) {
 			/* On error, reset fd to try writing without streaming */
 			if (reset_fd(fd, path))
 				return -1;
diff --git a/streaming.c b/streaming.c
index f0f7d31956..807a6e03a8 100644
--- a/streaming.c
+++ b/streaming.c
@@ -2,8 +2,6 @@
  * Copyright (c) 2011, Google Inc.
  */
 
-#define USE_THE_REPOSITORY_VARIABLE
-
 #include "git-compat-util.h"
 #include "convert.h"
 #include "environment.h"
@@ -359,7 +357,7 @@ static int open_istream_pack_non_delta(struct odb_read_stream **out,
 
 	if (packfile_store_read_object_info(odb->packfiles, oid, &oi, 0) ||
 	    oi.u.packed.is_delta ||
-	    repo_settings_get_big_file_threshold(the_repository) >= size)
+	    repo_settings_get_big_file_threshold(odb->repo) >= size)
 		return -1;
 
 	in_pack_type = unpack_object_header(oi.u.packed.pack,
@@ -518,8 +516,11 @@ struct odb_read_stream *open_istream(struct repository *r,
 	return st;
 }
 
-int stream_blob_to_fd(int fd, const struct object_id *oid, struct stream_filter *filter,
-		      int can_seek)
+int odb_stream_blob_to_fd(struct object_database *odb,
+			  int fd,
+			  const struct object_id *oid,
+			  struct stream_filter *filter,
+			  int can_seek)
 {
 	struct odb_read_stream *st;
 	enum object_type type;
@@ -527,7 +528,7 @@ int stream_blob_to_fd(int fd, const struct object_id *oid, struct stream_filter
 	ssize_t kept = 0;
 	int result = -1;
 
-	st = open_istream(the_repository, oid, &type, &sz, filter);
+	st = open_istream(odb->repo, oid, &type, &sz, filter);
 	if (!st) {
 		if (filter)
 			free_stream_filter(filter);
diff --git a/streaming.h b/streaming.h
index f5ff5d7ac9..148f6b3069 100644
--- a/streaming.h
+++ b/streaming.h
@@ -6,6 +6,7 @@
 
 #include "object.h"
 
+struct object_database;
 /* opaque */
 struct odb_read_stream;
 struct stream_filter;
@@ -16,6 +17,21 @@ struct odb_read_stream *open_istream(struct repository *, const struct object_id
 int close_istream(struct odb_read_stream *);
 ssize_t read_istream(struct odb_read_stream *, void *, size_t);
 
-int stream_blob_to_fd(int fd, const struct object_id *, struct stream_filter *, int can_seek);
+/*
+ * Look up the object by its ID and write the full contents to the file
+ * descriptor. The object must be a blob, or the function will fail. When
+ * provided, the filter is used to transform the blob contents.
+ *
+ * `can_seek` should be set to 1 in case the given file descriptor can be
+ * seek(3p)'d on. This is used to support files with holes in case a
+ * significant portion of the blob contains NUL bytes.
+ *
+ * Returns a negative error code on failure, 0 on success.
+ */
+int odb_stream_blob_to_fd(struct object_database *odb,
+			  int fd,
+			  const struct object_id *oid,
+			  struct stream_filter *filter,
+			  int can_seek);
 
 #endif /* STREAMING_H */

-- 
2.52.0.rc2.482.gaa765fefd0.dirty