streaming: move logic to read loose objects streams into backend

Move the logic to read loose object streams into the respective
subsystem. This allows us to make a couple of function declarations
private.

Signed-off-by: Patrick Steinhardt <[email protected]>
Signed-off-by: Junio C Hamano <[email protected]>

Author: pks-t

object-file.c

@@ -234,9 +234,9 @@ static void *map_fd(int fd, const char *path, unsigned long *size)
 	return map;
 }
 
-void *odb_source_loose_map_object(struct odb_source *source,
-				  const struct object_id *oid,
-				  unsigned long *size)
+static void *odb_source_loose_map_object(struct odb_source *source,
+					 const struct object_id *oid,
+					 unsigned long *size)
 {
 	const char *p;
 	int fd = open_loose_object(source->loose, oid, &p);
@@ -246,11 +246,29 @@ void *odb_source_loose_map_object(struct odb_source *source,
 	return map_fd(fd, p, size);
 }
 
-enum unpack_loose_header_result unpack_loose_header(git_zstream *stream,
-						    unsigned char *map,
-						    unsigned long mapsize,
-						    void *buffer,
-						    unsigned long bufsiz)
+enum unpack_loose_header_result {
+	ULHR_OK,
+	ULHR_BAD,
+	ULHR_TOO_LONG,
+};
+
+/**
+ * unpack_loose_header() initializes the data stream needed to unpack
+ * a loose object header.
+ *
+ * Returns:
+ *
+ * - ULHR_OK on success
+ * - ULHR_BAD on error
+ * - ULHR_TOO_LONG if the header was too long
+ *
+ * It will only parse up to MAX_HEADER_LEN bytes.
+ */
+static enum unpack_loose_header_result unpack_loose_header(git_zstream *stream,
+							   unsigned char *map,
+							   unsigned long mapsize,
+							   void *buffer,
+							   unsigned long bufsiz)
 {
 	int status;
 
@@ -329,11 +347,18 @@ static void *unpack_loose_rest(git_zstream *stream,
 }
 
 /*
+ * parse_loose_header() parses the starting "<type> <len>\0" of an
+ * object. If it doesn't follow that format -1 is returned. To check
+ * the validity of the <type> populate the "typep" in the "struct
+ * object_info". It will be OBJ_BAD if the object type is unknown. The
+ * parsed <len> can be retrieved via "oi->sizep", and from there
+ * passed to unpack_loose_rest().
+ *
  * We used to just use "sscanf()", but that's actually way
  * too permissive for what we want to check. So do an anal
  * object header parse by hand.
  */
-int parse_loose_header(const char *hdr, struct object_info *oi)
+static int parse_loose_header(const char *hdr, struct object_info *oi)
 {
 	const char *type_buf = hdr;
 	size_t size;
@@ -1976,3 +2001,127 @@ void odb_source_loose_free(struct odb_source_loose *loose)
 	loose_object_map_clear(&loose->map);
 	free(loose);
 }
+
+struct odb_loose_read_stream {
+	struct odb_read_stream base;
+	git_zstream z;
+	enum {
+		ODB_LOOSE_READ_STREAM_INUSE,
+		ODB_LOOSE_READ_STREAM_DONE,
+		ODB_LOOSE_READ_STREAM_ERROR,
+	} z_state;
+	void *mapped;
+	unsigned long mapsize;
+	char hdr[32];
+	int hdr_avail;
+	int hdr_used;
+};
+
+static ssize_t read_istream_loose(struct odb_read_stream *_st, char *buf, size_t sz)
+{
+	struct odb_loose_read_stream *st = (struct odb_loose_read_stream *)_st;
+	size_t total_read = 0;
+
+	switch (st->z_state) {
+	case ODB_LOOSE_READ_STREAM_DONE:
+		return 0;
+	case ODB_LOOSE_READ_STREAM_ERROR:
+		return -1;
+	default:
+		break;
+	}
+
+	if (st->hdr_used < st->hdr_avail) {
+		size_t to_copy = st->hdr_avail - st->hdr_used;
+		if (sz < to_copy)
+			to_copy = sz;
+		memcpy(buf, st->hdr + st->hdr_used, to_copy);
+		st->hdr_used += to_copy;
+		total_read += to_copy;
+	}
+
+	while (total_read < sz) {
+		int status;
+
+		st->z.next_out = (unsigned char *)buf + total_read;
+		st->z.avail_out = sz - total_read;
+		status = git_inflate(&st->z, Z_FINISH);
+
+		total_read = st->z.next_out - (unsigned char *)buf;
+
+		if (status == Z_STREAM_END) {
+			git_inflate_end(&st->z);
+			st->z_state = ODB_LOOSE_READ_STREAM_DONE;
+			break;
+		}
+		if (status != Z_OK && (status != Z_BUF_ERROR || total_read < sz)) {
+			git_inflate_end(&st->z);
+			st->z_state = ODB_LOOSE_READ_STREAM_ERROR;
+			return -1;
+		}
+	}
+	return total_read;
+}
+
+static int close_istream_loose(struct odb_read_stream *_st)
+{
+	struct odb_loose_read_stream *st = (struct odb_loose_read_stream *)_st;
+	if (st->z_state == ODB_LOOSE_READ_STREAM_INUSE)
+		git_inflate_end(&st->z);
+	munmap(st->mapped, st->mapsize);
+	return 0;
+}
+
+int odb_source_loose_read_object_stream(struct odb_read_stream **out,
+					struct odb_source *source,
+					const struct object_id *oid)
+{
+	struct object_info oi = OBJECT_INFO_INIT;
+	struct odb_loose_read_stream *st;
+	unsigned long mapsize;
+	void *mapped;
+
+	mapped = odb_source_loose_map_object(source, oid, &mapsize);
+	if (!mapped)
+		return -1;
+
+	/*
+	 * Note: we must allocate this structure early even though we may still
+	 * fail. This is because we need to initialize the zlib stream, and it
+	 * is not possible to copy the stream around after the fact because it
+	 * has self-referencing pointers.
+	 */
+	CALLOC_ARRAY(st, 1);
+
+	switch (unpack_loose_header(&st->z, mapped, mapsize, st->hdr,
+				    sizeof(st->hdr))) {
+	case ULHR_OK:
+		break;
+	case ULHR_BAD:
+	case ULHR_TOO_LONG:
+		goto error;
+	}
+
+	oi.sizep = &st->base.size;
+	oi.typep = &st->base.type;
+
+	if (parse_loose_header(st->hdr, &oi) < 0 || st->base.type < 0)
+		goto error;
+
+	st->mapped = mapped;
+	st->mapsize = mapsize;
+	st->hdr_used = strlen(st->hdr) + 1;
+	st->hdr_avail = st->z.total_out;
+	st->z_state = ODB_LOOSE_READ_STREAM_INUSE;
+	st->base.close = close_istream_loose;
+	st->base.read = read_istream_loose;
+
+	*out = &st->base;
+
+	return 0;
+error:
+	git_inflate_end(&st->z);
+	munmap(st->mapped, st->mapsize);
+	free(st);
+	return -1;
+}

object-file.h

@@ -16,6 +16,8 @@ enum {
 int index_fd(struct index_state *istate, struct object_id *oid, int fd, struct stat *st, enum object_type type, const char *path, unsigned flags);
 int index_path(struct index_state *istate, struct object_id *oid, const char *path, struct stat *st, unsigned flags);
 
+struct object_info;
+struct odb_read_stream;
 struct odb_source;
 
 struct odb_source_loose {
@@ -47,9 +49,9 @@ int odb_source_loose_read_object_info(struct odb_source *source,
 				      const struct object_id *oid,
 				      struct object_info *oi, int flags);
 
-void *odb_source_loose_map_object(struct odb_source *source,
-				  const struct object_id *oid,
-				  unsigned long *size);
+int odb_source_loose_read_object_stream(struct odb_read_stream **out,
+					struct odb_source *source,
+					const struct object_id *oid);
 
 /*
  * Return true iff an object database source has a loose object
@@ -143,40 +145,6 @@ int for_each_loose_object(struct object_database *odb,
 int format_object_header(char *str, size_t size, enum object_type type,
 			 size_t objsize);
 
-/**
- * unpack_loose_header() initializes the data stream needed to unpack
- * a loose object header.
- *
- * Returns:
- *
- * - ULHR_OK on success
- * - ULHR_BAD on error
- * - ULHR_TOO_LONG if the header was too long
- *
- * It will only parse up to MAX_HEADER_LEN bytes.
- */
-enum unpack_loose_header_result {
-	ULHR_OK,
-	ULHR_BAD,
-	ULHR_TOO_LONG,
-};
-enum unpack_loose_header_result unpack_loose_header(git_zstream *stream,
-						    unsigned char *map,
-						    unsigned long mapsize,
-						    void *buffer,
-						    unsigned long bufsiz);
-
-/**
- * parse_loose_header() parses the starting "<type> <len>\0" of an
- * object. If it doesn't follow that format -1 is returned. To check
- * the validity of the <type> populate the "typep" in the "struct
- * object_info". It will be OBJ_BAD if the object type is unknown. The
- * parsed <len> can be retrieved via "oi->sizep", and from there
- * passed to unpack_loose_rest().
- */
-struct object_info;
-int parse_loose_header(const char *hdr, struct object_info *oi);
-
 int force_object_loose(struct odb_source *source,
 		       const struct object_id *oid, time_t mtime);