[thirdparty/git.git] / vcs-svn / svndump.c

/*
 * Parse and rearrange a svnadmin dump.
 * Create the dump with:
 * svnadmin dump --incremental -r<startrev>:<endrev> <repository> >outfile
 *
 * Licensed under a two-clause BSD-style license.
 * See LICENSE for details.
 */

#include "cache.h"
#include "repo_tree.h"
#include "fast_export.h"
#include "line_buffer.h"
#include "strbuf.h"
#include "svndump.h"

/*
 * Compare start of string to literal of equal length;
 * must be guarded by length test.
 */
#define constcmp(s, ref) memcmp(s, ref, sizeof(ref) - 1)

#define REPORT_FILENO 3

#define NODEACT_REPLACE 4
#define NODEACT_DELETE 3
#define NODEACT_ADD 2
#define NODEACT_CHANGE 1
#define NODEACT_UNKNOWN 0

/* States: */
#define DUMP_CTX 0	/* dump metadata */
#define REV_CTX  1	/* revision metadata */
#define NODE_CTX 2	/* node metadata */
#define INTERNODE_CTX 3	/* between nodes */

#define DATE_RFC2822_LEN 31

static struct line_buffer input = LINE_BUFFER_INIT;

static struct {
	uint32_t action, srcRev, type;
	off_t prop_length, text_length;
	struct strbuf src, dst;
	uint32_t text_delta, prop_delta;
} node_ctx;

static struct {
	uint32_t revision;
	unsigned long timestamp;
	struct strbuf log, author;
} rev_ctx;

static struct {
	uint32_t version;
	struct strbuf uuid, url;
} dump_ctx;

static void reset_node_ctx(char *fname)
{
	node_ctx.type = 0;
	node_ctx.action = NODEACT_UNKNOWN;
	node_ctx.prop_length = -1;
	node_ctx.text_length = -1;
	strbuf_reset(&node_ctx.src);
	node_ctx.srcRev = 0;
	strbuf_reset(&node_ctx.dst);
	if (fname)
		strbuf_addstr(&node_ctx.dst, fname);
	node_ctx.text_delta = 0;
	node_ctx.prop_delta = 0;
}

static void reset_rev_ctx(uint32_t revision)
{
	rev_ctx.revision = revision;
	rev_ctx.timestamp = 0;
	strbuf_reset(&rev_ctx.log);
	strbuf_reset(&rev_ctx.author);
}

static void reset_dump_ctx(const char *url)
{
	strbuf_reset(&dump_ctx.url);
	if (url)
		strbuf_addstr(&dump_ctx.url, url);
	dump_ctx.version = 1;
	strbuf_reset(&dump_ctx.uuid);
}

static void handle_property(const struct strbuf *key_buf,
				struct strbuf *val,
				uint32_t *type_set)
{
	const char *key = key_buf->buf;
	size_t keylen = key_buf->len;

	switch (keylen + 1) {
	case sizeof("svn:log"):
		if (constcmp(key, "svn:log"))
			break;
		if (!val)
			die("invalid dump: unsets svn:log");
		strbuf_swap(&rev_ctx.log, val);
		break;
	case sizeof("svn:author"):
		if (constcmp(key, "svn:author"))
			break;
		if (!val)
			strbuf_reset(&rev_ctx.author);
		else
			strbuf_swap(&rev_ctx.author, val);
		break;
	case sizeof("svn:date"):
		if (constcmp(key, "svn:date"))
			break;
		if (!val)
			die("invalid dump: unsets svn:date");
		if (parse_date_basic(val->buf, &rev_ctx.timestamp, NULL))
			warning("invalid timestamp: %s", val->buf);
		break;
	case sizeof("svn:executable"):
	case sizeof("svn:special"):
		if (keylen == strlen("svn:executable") &&
		    constcmp(key, "svn:executable"))
			break;
		if (keylen == strlen("svn:special") &&
		    constcmp(key, "svn:special"))
			break;
		if (*type_set) {
			if (!val)
				return;
			die("invalid dump: sets type twice");
		}
		if (!val) {
			node_ctx.type = REPO_MODE_BLB;
			return;
		}
		*type_set = 1;
		node_ctx.type = keylen == strlen("svn:executable") ?
				REPO_MODE_EXE :
				REPO_MODE_LNK;
	}
}

static void die_short_read(void)
{
	if (buffer_ferror(&input))
		die_errno("error reading dump file");
	die("invalid dump: unexpected end of file");
}

static void read_props(void)
{
	static struct strbuf key = STRBUF_INIT;
	static struct strbuf val = STRBUF_INIT;
	const char *t;
	/*
	 * NEEDSWORK: to support simple mode changes like
	 *	K 11
	 *	svn:special
	 *	V 1
	 *	*
	 *	D 14
	 *	svn:executable
	 * we keep track of whether a mode has been set and reset to
	 * plain file only if not.  We should be keeping track of the
	 * symlink and executable bits separately instead.
	 */
	uint32_t type_set = 0;
	while ((t = buffer_read_line(&input)) && strcmp(t, "PROPS-END")) {
		uint32_t len;
		const char type = t[0];
		int ch;

		if (!type || t[1] != ' ')
			die("invalid property line: %s", t);
		len = atoi(&t[2]);
		strbuf_reset(&val);
		buffer_read_binary(&input, &val, len);
		if (val.len < len)
			die_short_read();

		/* Discard trailing newline. */
		ch = buffer_read_char(&input);
		if (ch == EOF)
			die_short_read();
		if (ch != '\n')
			die("invalid dump: expected newline after %s", val.buf);

		switch (type) {
		case 'K':
			strbuf_swap(&key, &val);
			continue;
		case 'D':
			handle_property(&val, NULL, &type_set);
			continue;
		case 'V':
			handle_property(&key, &val, &type_set);
			strbuf_reset(&key);
			continue;
		default:
			die("invalid property line: %s", t);
		}
	}
}

static void handle_node(void)
{
	const uint32_t type = node_ctx.type;
	const int have_props = node_ctx.prop_length != -1;
	const int have_text = node_ctx.text_length != -1;
	/*
	 * Old text for this node:
	 *  NULL	- directory or bug
	 *  empty_blob	- empty
	 *  "<dataref>"	- data retrievable from fast-import
	 */
	static const char *const empty_blob = "::empty::";
	const char *old_data = NULL;
	uint32_t old_mode = REPO_MODE_BLB;

	if (node_ctx.action == NODEACT_DELETE) {
		if (have_text || have_props || node_ctx.srcRev)
			die("invalid dump: deletion node has "
				"copyfrom info, text, or properties");
		repo_delete(node_ctx.dst.buf);
		return;
	}
	if (node_ctx.action == NODEACT_REPLACE) {
		repo_delete(node_ctx.dst.buf);
		node_ctx.action = NODEACT_ADD;
	}
	if (node_ctx.srcRev) {
		repo_copy(node_ctx.srcRev, node_ctx.src.buf, node_ctx.dst.buf);
		if (node_ctx.action == NODEACT_ADD)
			node_ctx.action = NODEACT_CHANGE;
	}
	if (have_text && type == REPO_MODE_DIR)
		die("invalid dump: directories cannot have text attached");

	/*
	 * Find old content (old_data) and decide on the new mode.
	 */
	if (node_ctx.action == NODEACT_CHANGE && !*node_ctx.dst.buf) {
		if (type != REPO_MODE_DIR)
			die("invalid dump: root of tree is not a regular file");
		old_data = NULL;
	} else if (node_ctx.action == NODEACT_CHANGE) {
		uint32_t mode;
		old_data = repo_read_path(node_ctx.dst.buf, &mode);
		if (mode == REPO_MODE_DIR && type != REPO_MODE_DIR)
			die("invalid dump: cannot modify a directory into a file");
		if (mode != REPO_MODE_DIR && type == REPO_MODE_DIR)
			die("invalid dump: cannot modify a file into a directory");
		node_ctx.type = mode;
		old_mode = mode;
	} else if (node_ctx.action == NODEACT_ADD) {
		if (type == REPO_MODE_DIR)
			old_data = NULL;
		else if (have_text)
			old_data = empty_blob;
		else
			die("invalid dump: adds node without text");
	} else {
		die("invalid dump: Node-path block lacks Node-action");
	}

	/*
	 * Adjust mode to reflect properties.
	 */
	if (have_props) {
		if (!node_ctx.prop_delta)
			node_ctx.type = type;
		if (node_ctx.prop_length)
			read_props();
	}

	/*
	 * Save the result.
	 */
	if (type == REPO_MODE_DIR)	/* directories are not tracked. */
		return;
	assert(old_data);
	if (old_data == empty_blob)
		/* For the fast_export_* functions, NULL means empty. */
		old_data = NULL;
	if (!have_text) {
		fast_export_modify(node_ctx.dst.buf, node_ctx.type, old_data);
		return;
	}
	if (!node_ctx.text_delta) {
		fast_export_modify(node_ctx.dst.buf, node_ctx.type, "inline");
		fast_export_data(node_ctx.type, node_ctx.text_length, &input);
		return;
	}
	fast_export_modify(node_ctx.dst.buf, node_ctx.type, "inline");
	fast_export_blob_delta(node_ctx.type, old_mode, old_data,
				node_ctx.text_length, &input);
}

static void begin_revision(void)
{
	if (!rev_ctx.revision)	/* revision 0 gets no git commit. */
		return;
	fast_export_begin_commit(rev_ctx.revision, rev_ctx.author.buf,
		&rev_ctx.log, dump_ctx.uuid.buf, dump_ctx.url.buf,
		rev_ctx.timestamp);
}

static void end_revision(void)
{
	if (rev_ctx.revision)
		fast_export_end_commit(rev_ctx.revision);
}

void svndump_read(const char *url)
{
	char *val;
	char *t;
	uint32_t active_ctx = DUMP_CTX;
	uint32_t len;

	reset_dump_ctx(url);
	while ((t = buffer_read_line(&input))) {
		val = strchr(t, ':');
		if (!val)
			continue;
		val++;
		if (*val != ' ')
			continue;
		val++;

		/* strlen(key) + 1 */
		switch (val - t - 1) {
		case sizeof("SVN-fs-dump-format-version"):
			if (constcmp(t, "SVN-fs-dump-format-version"))
				continue;
			dump_ctx.version = atoi(val);
			if (dump_ctx.version > 3)
				die("expected svn dump format version <= 3, found %"PRIu32,
				    dump_ctx.version);
			break;
		case sizeof("UUID"):
			if (constcmp(t, "UUID"))
				continue;
			strbuf_reset(&dump_ctx.uuid);
			strbuf_addstr(&dump_ctx.uuid, val);
			break;
		case sizeof("Revision-number"):
			if (constcmp(t, "Revision-number"))
				continue;
			if (active_ctx == NODE_CTX)
				handle_node();
			if (active_ctx == REV_CTX)
				begin_revision();
			if (active_ctx != DUMP_CTX)
				end_revision();
			active_ctx = REV_CTX;
			reset_rev_ctx(atoi(val));
			break;
		case sizeof("Node-path"):
			if (constcmp(t, "Node-"))
				continue;
			if (!constcmp(t + strlen("Node-"), "path")) {
				if (active_ctx == NODE_CTX)
					handle_node();
				if (active_ctx == REV_CTX)
					begin_revision();
				active_ctx = NODE_CTX;
				reset_node_ctx(val);
				break;
			}
			if (constcmp(t + strlen("Node-"), "kind"))
				continue;
			if (!strcmp(val, "dir"))
				node_ctx.type = REPO_MODE_DIR;
			else if (!strcmp(val, "file"))
				node_ctx.type = REPO_MODE_BLB;
			else
				fprintf(stderr, "Unknown node-kind: %s\n", val);
			break;
		case sizeof("Node-action"):
			if (constcmp(t, "Node-action"))
				continue;
			if (!strcmp(val, "delete")) {
				node_ctx.action = NODEACT_DELETE;
			} else if (!strcmp(val, "add")) {
				node_ctx.action = NODEACT_ADD;
			} else if (!strcmp(val, "change")) {
				node_ctx.action = NODEACT_CHANGE;
			} else if (!strcmp(val, "replace")) {
				node_ctx.action = NODEACT_REPLACE;
			} else {
				fprintf(stderr, "Unknown node-action: %s\n", val);
				node_ctx.action = NODEACT_UNKNOWN;
			}
			break;
		case sizeof("Node-copyfrom-path"):
			if (constcmp(t, "Node-copyfrom-path"))
				continue;
			strbuf_reset(&node_ctx.src);
			strbuf_addstr(&node_ctx.src, val);
			break;
		case sizeof("Node-copyfrom-rev"):
			if (constcmp(t, "Node-copyfrom-rev"))
				continue;
			node_ctx.srcRev = atoi(val);
			break;
		case sizeof("Text-content-length"):
			if (constcmp(t, "Text") && constcmp(t, "Prop"))
				continue;
			if (constcmp(t + 4, "-content-length"))
				continue;
			{
				char *end;
				uintmax_t len;

				len = strtoumax(val, &end, 10);
				if (!isdigit(*val) || *end)
					die("invalid dump: non-numeric length %s", val);
				if (len > maximum_signed_value_of_type(off_t))
					die("unrepresentable length in dump: %s", val);

				if (*t == 'T')
					node_ctx.text_length = (off_t) len;
				else
					node_ctx.prop_length = (off_t) len;
				break;
			}
		case sizeof("Text-delta"):
			if (!constcmp(t, "Text-delta")) {
				node_ctx.text_delta = !strcmp(val, "true");
				break;
			}
			if (constcmp(t, "Prop-delta"))
				continue;
			node_ctx.prop_delta = !strcmp(val, "true");
			break;
		case sizeof("Content-length"):
			if (constcmp(t, "Content-length"))
				continue;
			len = atoi(val);
			t = buffer_read_line(&input);
			if (!t)
				die_short_read();
			if (*t)
				die("invalid dump: expected blank line after content length header");
			if (active_ctx == REV_CTX) {
				read_props();
			} else if (active_ctx == NODE_CTX) {
				handle_node();
				active_ctx = INTERNODE_CTX;
			} else {
				fprintf(stderr, "Unexpected content length header: %"PRIu32"\n", len);
				if (buffer_skip_bytes(&input, len) != len)
					die_short_read();
			}
		}
	}
	if (buffer_ferror(&input))
		die_short_read();
	if (active_ctx == NODE_CTX)
		handle_node();
	if (active_ctx == REV_CTX)
		begin_revision();
	if (active_ctx != DUMP_CTX)
		end_revision();
}

int svndump_init(const char *filename)
{
	if (buffer_init(&input, filename))
		return error("cannot open %s: %s", filename, strerror(errno));
	fast_export_init(REPORT_FILENO);
	strbuf_init(&dump_ctx.uuid, 4096);
	strbuf_init(&dump_ctx.url, 4096);
	strbuf_init(&rev_ctx.log, 4096);
	strbuf_init(&rev_ctx.author, 4096);
	strbuf_init(&node_ctx.src, 4096);
	strbuf_init(&node_ctx.dst, 4096);
	reset_dump_ctx(NULL);
	reset_rev_ctx(0);
	reset_node_ctx(NULL);
	return 0;
}

void svndump_deinit(void)
{
	fast_export_deinit();
	reset_dump_ctx(NULL);
	reset_rev_ctx(0);
	reset_node_ctx(NULL);
	strbuf_release(&rev_ctx.log);
	strbuf_release(&node_ctx.src);
	strbuf_release(&node_ctx.dst);
	if (buffer_deinit(&input))
		fprintf(stderr, "Input error\n");
	if (ferror(stdout))
		fprintf(stderr, "Output error\n");
}

void svndump_reset(void)
{
	strbuf_release(&dump_ctx.uuid);
	strbuf_release(&dump_ctx.url);
	strbuf_release(&rev_ctx.log);
	strbuf_release(&rev_ctx.author);
}
Commit	Line	Data
21746aa3 DB	1	/*
	2	* Parse and rearrange a svnadmin dump.
	3	* Create the dump with:
	4	* svnadmin dump --incremental -r<startrev>:<endrev> <repository> >outfile
	5	*
	6	* Licensed under a two-clause BSD-style license.
	7	* See LICENSE for details.
	8	*/
	9
	10	#include "cache.h"
	11	#include "repo_tree.h"
	12	#include "fast_export.h"
	13	#include "line_buffer.h"
dce33c9c	14	#include "strbuf.h"
c5147722	15	#include "svndump.h"
21746aa3	16
044ad290 DB	17	/*
	18	* Compare start of string to literal of equal length;
	19	* must be guarded by length test.
	20	*/
	21	#define constcmp(s, ref) memcmp(s, ref, sizeof(ref) - 1)
	22
41529bbc DB	23	#define REPORT_FILENO 3
41529bbc DB	24
21746aa3 DB	25	#define NODEACT_REPLACE 4
	26	#define NODEACT_DELETE 3
	27	#define NODEACT_ADD 2
	28	#define NODEACT_CHANGE 1
	29	#define NODEACT_UNKNOWN 0
	30
7e11902c JN	31	/* States: */
	32	#define DUMP_CTX 0 /* dump metadata */
	33	#define REV_CTX 1 /* revision metadata */
	34	#define NODE_CTX 2 /* node metadata */
	35	#define INTERNODE_CTX 3 /* between nodes */
21746aa3	36
21746aa3 DB	37	#define DATE_RFC2822_LEN 31
21746aa3 DB	38
e5e45ca1 JN	39	static struct line_buffer input = LINE_BUFFER_INIT;
e5e45ca1 JN	40
21746aa3	41	static struct {
e32b79cb JN	42	uint32_t action, srcRev, type;
e32b79cb JN	43	off_t prop_length, text_length;
03087971	44	struct strbuf src, dst;
1f05d07c	45	uint32_t text_delta, prop_delta;
21746aa3 DB	46	} node_ctx;
	47
	48	static struct {
7c5817d3	49	uint32_t revision;
21746aa3	50	unsigned long timestamp;
7c5817d3	51	struct strbuf log, author;
21746aa3 DB	52	} rev_ctx;
	53
	54	static struct {
7c5817d3 DB	55	uint32_t version;
7c5817d3 DB	56	struct strbuf uuid, url;
21746aa3 DB	57	} dump_ctx;
21746aa3 DB	58
21746aa3 DB	59	static void reset_node_ctx(char *fname)
	60	{
	61	node_ctx.type = 0;
	62	node_ctx.action = NODEACT_UNKNOWN;
e32b79cb	63	node_ctx.prop_length = -1;
150f7546	64	node_ctx.text_length = -1;
03087971	65	strbuf_reset(&node_ctx.src);
21746aa3	66	node_ctx.srcRev = 0;
03087971 DB	67	strbuf_reset(&node_ctx.dst);
	68	if (fname)
	69	strbuf_addstr(&node_ctx.dst, fname);
1f05d07c DB	70	node_ctx.text_delta = 0;
1f05d07c DB	71	node_ctx.prop_delta = 0;
21746aa3 DB	72	}
	73
	74	static void reset_rev_ctx(uint32_t revision)
	75	{
	76	rev_ctx.revision = revision;
	77	rev_ctx.timestamp = 0;
dce33c9c	78	strbuf_reset(&rev_ctx.log);
7c5817d3	79	strbuf_reset(&rev_ctx.author);
21746aa3 DB	80	}
21746aa3 DB	81
7c5817d3	82	static void reset_dump_ctx(const char *url)
21746aa3	83	{
7c5817d3 DB	84	strbuf_reset(&dump_ctx.url);
	85	if (url)
	86	strbuf_addstr(&dump_ctx.url, url);
b3e5bce1	87	dump_ctx.version = 1;
7c5817d3	88	strbuf_reset(&dump_ctx.uuid);
21746aa3 DB	89	}
21746aa3 DB	90
044ad290	91	static void handle_property(const struct strbuf *key_buf,
4c3169b0	92	struct strbuf *val,
6b01b676	93	uint32_t *type_set)
2a48afe1	94	{
044ad290 DB	95	const char *key = key_buf->buf;
	96	size_t keylen = key_buf->len;
	97
	98	switch (keylen + 1) {
	99	case sizeof("svn:log"):
	100	if (constcmp(key, "svn:log"))
	101	break;
6b01b676 DB	102	if (!val)
6b01b676 DB	103	die("invalid dump: unsets svn:log");
4c3169b0	104	strbuf_swap(&rev_ctx.log, val);
044ad290 DB	105	break;
	106	case sizeof("svn:author"):
	107	if (constcmp(key, "svn:author"))
	108	break;
4c3169b0 JN	109	if (!val)
	110	strbuf_reset(&rev_ctx.author);
	111	else
	112	strbuf_swap(&rev_ctx.author, val);
044ad290 DB	113	break;
	114	case sizeof("svn:date"):
	115	if (constcmp(key, "svn:date"))
	116	break;
6b01b676 DB	117	if (!val)
6b01b676 DB	118	die("invalid dump: unsets svn:date");
4c3169b0 JN	119	if (parse_date_basic(val->buf, &rev_ctx.timestamp, NULL))
4c3169b0 JN	120	warning("invalid timestamp: %s", val->buf);
044ad290 DB	121	break;
	122	case sizeof("svn:executable"):
	123	case sizeof("svn:special"):
	124	if (keylen == strlen("svn:executable") &&
	125	constcmp(key, "svn:executable"))
	126	break;
	127	if (keylen == strlen("svn:special") &&
	128	constcmp(key, "svn:special"))
	129	break;
6b01b676 DB	130	if (*type_set) {
	131	if (!val)
	132	return;
	133	die("invalid dump: sets type twice");
	134	}
	135	if (!val) {
	136	node_ctx.type = REPO_MODE_BLB;
	137	return;
	138	}
	139	*type_set = 1;
044ad290	140	node_ctx.type = keylen == strlen("svn:executable") ?
6b01b676 DB	141	REPO_MODE_EXE :
6b01b676 DB	142	REPO_MODE_LNK;
2a48afe1	143	}
21746aa3 DB	144	}
21746aa3 DB	145
c9d1c8ba JN	146	static void die_short_read(void)
	147	{
	148	if (buffer_ferror(&input))
	149	die_errno("error reading dump file");
	150	die("invalid dump: unexpected end of file");
	151	}
	152
21746aa3 DB	153	static void read_props(void)
21746aa3 DB	154	{
044ad290	155	static struct strbuf key = STRBUF_INIT;
e7d04ee1	156	static struct strbuf val = STRBUF_INIT;
6263c06d	157	const char *t;
6b01b676 DB	158	/*
	159	* NEEDSWORK: to support simple mode changes like
	160	* K 11
	161	* svn:special
	162	* V 1
	163	* *
	164	* D 14
	165	* svn:executable
	166	* we keep track of whether a mode has been set and reset to
	167	* plain file only if not. We should be keeping track of the
	168	* symlink and executable bits separately instead.
	169	*/
	170	uint32_t type_set = 0;
e5e45ca1	171	while ((t = buffer_read_line(&input)) && strcmp(t, "PROPS-END")) {
6263c06d	172	uint32_t len;
6263c06d	173	const char type = t[0];
c9d1c8ba	174	int ch;
6263c06d JN	175
6263c06d JN	176	if (!type \|\| t[1] != ' ')
82247e9b	177	die("invalid property line: %s", t);
6263c06d	178	len = atoi(&t[2]);
e7d04ee1 JN	179	strbuf_reset(&val);
	180	buffer_read_binary(&input, &val, len);
	181	if (val.len < len)
c9d1c8ba JN	182	die_short_read();
	183
	184	/* Discard trailing newline. */
	185	ch = buffer_read_char(&input);
	186	if (ch == EOF)
	187	die_short_read();
	188	if (ch != '\n')
e7d04ee1	189	die("invalid dump: expected newline after %s", val.buf);
6263c06d JN	190
	191	switch (type) {
	192	case 'K':
e7d04ee1 JN	193	strbuf_swap(&key, &val);
e7d04ee1 JN	194	continue;
6b01b676	195	case 'D':
4c3169b0	196	handle_property(&val, NULL, &type_set);
e7d04ee1	197	continue;
6263c06d	198	case 'V':
4c3169b0	199	handle_property(&key, &val, &type_set);
044ad290	200	strbuf_reset(&key);
6263c06d JN	201	continue;
6263c06d JN	202	default:
82247e9b	203	die("invalid property line: %s", t);
21746aa3 DB	204	}
	205	}
	206	}
	207
	208	static void handle_node(void)
	209	{
1c7bb316	210	const uint32_t type = node_ctx.type;
e32b79cb	211	const int have_props = node_ctx.prop_length != -1;
150f7546	212	const int have_text = node_ctx.text_length != -1;
723b7a27 JN	213	/*
	214	* Old text for this node:
	215	* NULL - directory or bug
	216	* empty_blob - empty
	217	* "<dataref>" - data retrievable from fast-import
	218	*/
	219	static const char *const empty_blob = "::empty::";
	220	const char *old_data = NULL;
7a75e661	221	uint32_t old_mode = REPO_MODE_BLB;
21746aa3 DB	222
21746aa3 DB	223	if (node_ctx.action == NODEACT_DELETE) {
5a38b186	224	if (have_text \|\| have_props \|\| node_ctx.srcRev)
5af8fae2 JN	225	die("invalid dump: deletion node has "
5af8fae2 JN	226	"copyfrom info, text, or properties");
c19d653c	227	repo_delete(node_ctx.dst.buf);
9e113988	228	return;
5af8fae2	229	}
6ee4a9be	230	if (node_ctx.action == NODEACT_REPLACE) {
03087971	231	repo_delete(node_ctx.dst.buf);
6ee4a9be JN	232	node_ctx.action = NODEACT_ADD;
6ee4a9be JN	233	}
1c7bb316	234	if (node_ctx.srcRev) {
03087971	235	repo_copy(node_ctx.srcRev, node_ctx.src.buf, node_ctx.dst.buf);
414e569e JN	236	if (node_ctx.action == NODEACT_ADD)
414e569e JN	237	node_ctx.action = NODEACT_CHANGE;
1c7bb316	238	}
5a38b186	239	if (have_text && type == REPO_MODE_DIR)
462e1f51	240	die("invalid dump: directories cannot have text attached");
5a38b186 JN	241
5a38b186 JN	242	/*
723b7a27	243	* Find old content (old_data) and decide on the new mode.
5a38b186	244	*/
03087971	245	if (node_ctx.action == NODEACT_CHANGE && !*node_ctx.dst.buf) {
9e8c5321 JN	246	if (type != REPO_MODE_DIR)
9e8c5321 JN	247	die("invalid dump: root of tree is not a regular file");
723b7a27	248	old_data = NULL;
9e8c5321	249	} else if (node_ctx.action == NODEACT_CHANGE) {
5a38b186	250	uint32_t mode;
c19d653c	251	old_data = repo_read_path(node_ctx.dst.buf, &mode);
c7dbf35e JN	252	if (mode == REPO_MODE_DIR && type != REPO_MODE_DIR)
	253	die("invalid dump: cannot modify a directory into a file");
	254	if (mode != REPO_MODE_DIR && type == REPO_MODE_DIR)
	255	die("invalid dump: cannot modify a file into a directory");
	256	node_ctx.type = mode;
7a75e661	257	old_mode = mode;
21746aa3	258	} else if (node_ctx.action == NODEACT_ADD) {
723b7a27 JN	259	if (type == REPO_MODE_DIR)
	260	old_data = NULL;
	261	else if (have_text)
	262	old_data = empty_blob;
	263	else
c7dbf35e	264	die("invalid dump: adds node without text");
c7dbf35e	265	} else {
414e569e	266	die("invalid dump: Node-path block lacks Node-action");
21746aa3	267	}
5a38b186 JN	268
	269	/*
	270	* Adjust mode to reflect properties.
	271	*/
1c7bb316	272	if (have_props) {
6b01b676 DB	273	if (!node_ctx.prop_delta)
6b01b676 DB	274	node_ctx.type = type;
e32b79cb	275	if (node_ctx.prop_length)
1c7bb316	276	read_props();
21746aa3	277	}
5a38b186 JN	278
	279	/*
	280	* Save the result.
	281	*/
723b7a27 JN	282	if (type == REPO_MODE_DIR) /* directories are not tracked. */
	283	return;
	284	assert(old_data);
	285	if (old_data == empty_blob)
	286	/* For the fast_export_* functions, NULL means empty. */
	287	old_data = NULL;
	288	if (!have_text) {
03087971	289	fast_export_modify(node_ctx.dst.buf, node_ctx.type, old_data);
723b7a27 JN	290	return;
723b7a27 JN	291	}
7a75e661 DB	292	if (!node_ctx.text_delta) {
7a75e661 DB	293	fast_export_modify(node_ctx.dst.buf, node_ctx.type, "inline");
150f7546	294	fast_export_data(node_ctx.type, node_ctx.text_length, &input);
7a75e661 DB	295	return;
7a75e661 DB	296	}
03087971	297	fast_export_modify(node_ctx.dst.buf, node_ctx.type, "inline");
7a75e661	298	fast_export_blob_delta(node_ctx.type, old_mode, old_data,
150f7546	299	node_ctx.text_length, &input);
21746aa3 DB	300	}
21746aa3 DB	301
7e11902c JN	302	static void begin_revision(void)
	303	{
	304	if (!rev_ctx.revision) /* revision 0 gets no git commit. */
	305	return;
9ecfa8ae JN	306	fast_export_begin_commit(rev_ctx.revision, rev_ctx.author.buf,
	307	&rev_ctx.log, dump_ctx.uuid.buf, dump_ctx.url.buf,
	308	rev_ctx.timestamp);
21746aa3 DB	309	}
21746aa3 DB	310
7e11902c	311	static void end_revision(void)
21746aa3 DB	312	{
21746aa3 DB	313	if (rev_ctx.revision)
723b7a27	314	fast_export_end_commit(rev_ctx.revision);
21746aa3 DB	315	}
	316
	317	void svndump_read(const char *url)
	318	{
	319	char *val;
	320	char *t;
	321	uint32_t active_ctx = DUMP_CTX;
	322	uint32_t len;
21746aa3	323
7c5817d3	324	reset_dump_ctx(url);
e5e45ca1	325	while ((t = buffer_read_line(&input))) {
f1602054	326	val = strchr(t, ':');
21746aa3 DB	327	if (!val)
21746aa3 DB	328	continue;
f1602054 DB	329	val++;
	330	if (*val != ' ')
	331	continue;
	332	val++;
21746aa3	333
90c0a3cf DB	334	/* strlen(key) + 1 */
	335	switch (val - t - 1) {
	336	case sizeof("SVN-fs-dump-format-version"):
	337	if (constcmp(t, "SVN-fs-dump-format-version"))
	338	continue;
b3e5bce1	339	dump_ctx.version = atoi(val);
1f05d07c	340	if (dump_ctx.version > 3)
a62bbf8f	341	die("expected svn dump format version <= 3, found %"PRIu32,
b3e5bce1	342	dump_ctx.version);
90c0a3cf DB	343	break;
	344	case sizeof("UUID"):
	345	if (constcmp(t, "UUID"))
	346	continue;
7c5817d3 DB	347	strbuf_reset(&dump_ctx.uuid);
7c5817d3 DB	348	strbuf_addstr(&dump_ctx.uuid, val);
90c0a3cf DB	349	break;
	350	case sizeof("Revision-number"):
	351	if (constcmp(t, "Revision-number"))
	352	continue;
21746aa3 DB	353	if (active_ctx == NODE_CTX)
21746aa3 DB	354	handle_node();
7e11902c JN	355	if (active_ctx == REV_CTX)
7e11902c JN	356	begin_revision();
21746aa3	357	if (active_ctx != DUMP_CTX)
7e11902c	358	end_revision();
21746aa3 DB	359	active_ctx = REV_CTX;
21746aa3 DB	360	reset_rev_ctx(atoi(val));
90c0a3cf DB	361	break;
90c0a3cf DB	362	case sizeof("Node-path"):
d8d8708b	363	if (constcmp(t, "Node-"))
90c0a3cf DB	364	continue;
	365	if (!constcmp(t + strlen("Node-"), "path")) {
	366	if (active_ctx == NODE_CTX)
	367	handle_node();
9ecfa8ae JN	368	if (active_ctx == REV_CTX)
9ecfa8ae JN	369	begin_revision();
90c0a3cf DB	370	active_ctx = NODE_CTX;
	371	reset_node_ctx(val);
	372	break;
	373	}
	374	if (constcmp(t + strlen("Node-"), "kind"))
	375	continue;
21746aa3 DB	376	if (!strcmp(val, "dir"))
	377	node_ctx.type = REPO_MODE_DIR;
	378	else if (!strcmp(val, "file"))
	379	node_ctx.type = REPO_MODE_BLB;
	380	else
	381	fprintf(stderr, "Unknown node-kind: %s\n", val);
90c0a3cf DB	382	break;
	383	case sizeof("Node-action"):
	384	if (constcmp(t, "Node-action"))
	385	continue;
21746aa3 DB	386	if (!strcmp(val, "delete")) {
	387	node_ctx.action = NODEACT_DELETE;
	388	} else if (!strcmp(val, "add")) {
	389	node_ctx.action = NODEACT_ADD;
	390	} else if (!strcmp(val, "change")) {
	391	node_ctx.action = NODEACT_CHANGE;
	392	} else if (!strcmp(val, "replace")) {
	393	node_ctx.action = NODEACT_REPLACE;
	394	} else {
	395	fprintf(stderr, "Unknown node-action: %s\n", val);
	396	node_ctx.action = NODEACT_UNKNOWN;
	397	}
90c0a3cf DB	398	break;
	399	case sizeof("Node-copyfrom-path"):
	400	if (constcmp(t, "Node-copyfrom-path"))
	401	continue;
03087971 DB	402	strbuf_reset(&node_ctx.src);
03087971 DB	403	strbuf_addstr(&node_ctx.src, val);
90c0a3cf DB	404	break;
	405	case sizeof("Node-copyfrom-rev"):
	406	if (constcmp(t, "Node-copyfrom-rev"))
	407	continue;
21746aa3	408	node_ctx.srcRev = atoi(val);
90c0a3cf DB	409	break;
90c0a3cf DB	410	case sizeof("Text-content-length"):
e32b79cb JN	411	if (constcmp(t, "Text") && constcmp(t, "Prop"))
	412	continue;
	413	if (constcmp(t + 4, "-content-length"))
	414	continue;
	415	{
150f7546	416	char *end;
e32b79cb	417	uintmax_t len;
150f7546	418
e32b79cb	419	len = strtoumax(val, &end, 10);
150f7546 JN	420	if (!isdigit(val) \|\| end)
150f7546 JN	421	die("invalid dump: non-numeric length %s", val);
e32b79cb	422	if (len > maximum_signed_value_of_type(off_t))
150f7546	423	die("unrepresentable length in dump: %s", val);
e32b79cb JN	424
	425	if (*t == 'T')
	426	node_ctx.text_length = (off_t) len;
	427	else
	428	node_ctx.prop_length = (off_t) len;
90c0a3cf DB	429	break;
90c0a3cf DB	430	}
90c0a3cf DB	431	case sizeof("Text-delta"):
	432	if (!constcmp(t, "Text-delta")) {
	433	node_ctx.text_delta = !strcmp(val, "true");
	434	break;
	435	}
	436	if (constcmp(t, "Prop-delta"))
	437	continue;
1f05d07c	438	node_ctx.prop_delta = !strcmp(val, "true");
90c0a3cf DB	439	break;
	440	case sizeof("Content-length"):
	441	if (constcmp(t, "Content-length"))
	442	continue;
21746aa3	443	len = atoi(val);
c9d1c8ba JN	444	t = buffer_read_line(&input);
	445	if (!t)
	446	die_short_read();
	447	if (*t)
	448	die("invalid dump: expected blank line after content length header");
21746aa3 DB	449	if (active_ctx == REV_CTX) {
	450	read_props();
	451	} else if (active_ctx == NODE_CTX) {
	452	handle_node();
7e11902c	453	active_ctx = INTERNODE_CTX;
21746aa3	454	} else {
5418d96d	455	fprintf(stderr, "Unexpected content length header: %"PRIu32"\n", len);
c9d1c8ba JN	456	if (buffer_skip_bytes(&input, len) != len)
c9d1c8ba JN	457	die_short_read();
21746aa3 DB	458	}
	459	}
	460	}
c9d1c8ba JN	461	if (buffer_ferror(&input))
c9d1c8ba JN	462	die_short_read();
21746aa3 DB	463	if (active_ctx == NODE_CTX)
21746aa3 DB	464	handle_node();
7e11902c JN	465	if (active_ctx == REV_CTX)
7e11902c JN	466	begin_revision();
21746aa3	467	if (active_ctx != DUMP_CTX)
7e11902c	468	end_revision();
21746aa3 DB	469	}
21746aa3 DB	470
5c28a8b0	471	int svndump_init(const char *filename)
21746aa3	472	{
a62bbf8f	473	if (buffer_init(&input, filename))
5c28a8b0	474	return error("cannot open %s: %s", filename, strerror(errno));
41529bbc	475	fast_export_init(REPORT_FILENO);
7c5817d3 DB	476	strbuf_init(&dump_ctx.uuid, 4096);
7c5817d3 DB	477	strbuf_init(&dump_ctx.url, 4096);
dce33c9c	478	strbuf_init(&rev_ctx.log, 4096);
7c5817d3	479	strbuf_init(&rev_ctx.author, 4096);
03087971 DB	480	strbuf_init(&node_ctx.src, 4096);
03087971 DB	481	strbuf_init(&node_ctx.dst, 4096);
7c5817d3	482	reset_dump_ctx(NULL);
21746aa3 DB	483	reset_rev_ctx(0);
21746aa3 DB	484	reset_node_ctx(NULL);
5c28a8b0	485	return 0;
21746aa3 DB	486	}
	487
	488	void svndump_deinit(void)
	489	{
41529bbc	490	fast_export_deinit();
7c5817d3	491	reset_dump_ctx(NULL);
21746aa3 DB	492	reset_rev_ctx(0);
21746aa3 DB	493	reset_node_ctx(NULL);
dce33c9c	494	strbuf_release(&rev_ctx.log);
03087971 DB	495	strbuf_release(&node_ctx.src);
03087971 DB	496	strbuf_release(&node_ctx.dst);
e5e45ca1	497	if (buffer_deinit(&input))
21746aa3 DB	498	fprintf(stderr, "Input error\n");
	499	if (ferror(stdout))
	500	fprintf(stderr, "Output error\n");
	501	}
	502
	503	void svndump_reset(void)
	504	{
7c5817d3 DB	505	strbuf_release(&dump_ctx.uuid);
	506	strbuf_release(&dump_ctx.url);
	507	strbuf_release(&rev_ctx.log);
	508	strbuf_release(&rev_ctx.author);
21746aa3	509	}