[thirdparty/git.git] / ws.c

/*
 * Whitespace rules
 *
 * Copyright (c) 2007 Junio C Hamano
 */

#include "cache.h"
#include "attr.h"

static struct whitespace_rule {
	const char *rule_name;
	unsigned rule_bits;
	unsigned loosens_error;
} whitespace_rule_names[] = {
	{ "trailing-space", WS_TRAILING_SPACE, 0 },
	{ "space-before-tab", WS_SPACE_BEFORE_TAB, 0 },
	{ "indent-with-non-tab", WS_INDENT_WITH_NON_TAB, 0 },
	{ "cr-at-eol", WS_CR_AT_EOL, 1 },
	{ "blank-at-eol", WS_BLANK_AT_EOL, 0 },
	{ "blank-at-eof", WS_BLANK_AT_EOF, 0 },
};

unsigned parse_whitespace_rule(const char *string)
{
	unsigned rule = WS_DEFAULT_RULE;

	while (string) {
		int i;
		size_t len;
		const char *ep;
		int negated = 0;

		string = string + strspn(string, ", \t\n\r");
		ep = strchr(string, ',');
		if (!ep)
			len = strlen(string);
		else
			len = ep - string;

		if (*string == '-') {
			negated = 1;
			string++;
			len--;
		}
		if (!len)
			break;
		for (i = 0; i < ARRAY_SIZE(whitespace_rule_names); i++) {
			if (strncmp(whitespace_rule_names[i].rule_name,
				    string, len))
				continue;
			if (negated)
				rule &= ~whitespace_rule_names[i].rule_bits;
			else
				rule |= whitespace_rule_names[i].rule_bits;
			break;
		}
		string = ep;
	}
	return rule;
}

static void setup_whitespace_attr_check(struct git_attr_check *check)
{
	static struct git_attr *attr_whitespace;

	if (!attr_whitespace)
		attr_whitespace = git_attr("whitespace");
	check[0].attr = attr_whitespace;
}

unsigned whitespace_rule(const char *pathname)
{
	struct git_attr_check attr_whitespace_rule;

	setup_whitespace_attr_check(&attr_whitespace_rule);
	if (!git_checkattr(pathname, 1, &attr_whitespace_rule)) {
		const char *value;

		value = attr_whitespace_rule.value;
		if (ATTR_TRUE(value)) {
			/* true (whitespace) */
			unsigned all_rule = 0;
			int i;
			for (i = 0; i < ARRAY_SIZE(whitespace_rule_names); i++)
				if (!whitespace_rule_names[i].loosens_error)
					all_rule |= whitespace_rule_names[i].rule_bits;
			return all_rule;
		} else if (ATTR_FALSE(value)) {
			/* false (-whitespace) */
			return 0;
		} else if (ATTR_UNSET(value)) {
			/* reset to default (!whitespace) */
			return whitespace_rule_cfg;
		} else {
			/* string */
			return parse_whitespace_rule(value);
		}
	} else {
		return whitespace_rule_cfg;
	}
}

/* The returned string should be freed by the caller. */
char *whitespace_error_string(unsigned ws)
{
	struct strbuf err = STRBUF_INIT;
	if ((ws & WS_TRAILING_SPACE) == WS_TRAILING_SPACE)
		strbuf_addstr(&err, "trailing whitespace");
	else {
		if (ws & WS_BLANK_AT_EOL)
			strbuf_addstr(&err, "trailing whitespace");
		if (ws & WS_BLANK_AT_EOF) {
			if (err.len)
				strbuf_addstr(&err, ", ");
			strbuf_addstr(&err, "new blank line at EOF");
		}
	}
	if (ws & WS_SPACE_BEFORE_TAB) {
		if (err.len)
			strbuf_addstr(&err, ", ");
		strbuf_addstr(&err, "space before tab in indent");
	}
	if (ws & WS_INDENT_WITH_NON_TAB) {
		if (err.len)
			strbuf_addstr(&err, ", ");
		strbuf_addstr(&err, "indent with spaces");
	}
	return strbuf_detach(&err, NULL);
}

/* If stream is non-NULL, emits the line after checking. */
static unsigned ws_check_emit_1(const char *line, int len, unsigned ws_rule,
				FILE *stream, const char *set,
				const char *reset, const char *ws)
{
	unsigned result = 0;
	int written = 0;
	int trailing_whitespace = -1;
	int trailing_newline = 0;
	int trailing_carriage_return = 0;
	int i;

	/* Logic is simpler if we temporarily ignore the trailing newline. */
	if (len > 0 && line[len - 1] == '\n') {
		trailing_newline = 1;
		len--;
	}
	if ((ws_rule & WS_CR_AT_EOL) &&
	    len > 0 && line[len - 1] == '\r') {
		trailing_carriage_return = 1;
		len--;
	}

	/* Check for trailing whitespace. */
	if (ws_rule & WS_BLANK_AT_EOL) {
		for (i = len - 1; i >= 0; i--) {
			if (isspace(line[i])) {
				trailing_whitespace = i;
				result |= WS_BLANK_AT_EOL;
			}
			else
				break;
		}
	}

	/* Check for space before tab in initial indent. */
	for (i = 0; i < len; i++) {
		if (line[i] == ' ')
			continue;
		if (line[i] != '\t')
			break;
		if ((ws_rule & WS_SPACE_BEFORE_TAB) && written < i) {
			result |= WS_SPACE_BEFORE_TAB;
			if (stream) {
				fputs(ws, stream);
				fwrite(line + written, i - written, 1, stream);
				fputs(reset, stream);
			}
		} else if (stream)
			fwrite(line + written, i - written, 1, stream);
		if (stream)
			fwrite(line + i, 1, 1, stream);
		written = i + 1;
	}

	/* Check for indent using non-tab. */
	if ((ws_rule & WS_INDENT_WITH_NON_TAB) && i - written >= 8) {
		result |= WS_INDENT_WITH_NON_TAB;
		if (stream) {
			fputs(ws, stream);
			fwrite(line + written, i - written, 1, stream);
			fputs(reset, stream);
		}
		written = i;
	}

	if (stream) {
		/*
		 * Now the rest of the line starts at "written".
		 * The non-highlighted part ends at "trailing_whitespace".
		 */
		if (trailing_whitespace == -1)
			trailing_whitespace = len;

		/* Emit non-highlighted (middle) segment. */
		if (trailing_whitespace - written > 0) {
			fputs(set, stream);
			fwrite(line + written,
			    trailing_whitespace - written, 1, stream);
			fputs(reset, stream);
		}

		/* Highlight errors in trailing whitespace. */
		if (trailing_whitespace != len) {
			fputs(ws, stream);
			fwrite(line + trailing_whitespace,
			    len - trailing_whitespace, 1, stream);
			fputs(reset, stream);
		}
		if (trailing_carriage_return)
			fputc('\r', stream);
		if (trailing_newline)
			fputc('\n', stream);
	}
	return result;
}

void ws_check_emit(const char *line, int len, unsigned ws_rule,
		   FILE *stream, const char *set,
		   const char *reset, const char *ws)
{
	(void)ws_check_emit_1(line, len, ws_rule, stream, set, reset, ws);
}

unsigned ws_check(const char *line, int len, unsigned ws_rule)
{
	return ws_check_emit_1(line, len, ws_rule, NULL, NULL, NULL, NULL);
}

int ws_blank_line(const char *line, int len, unsigned ws_rule)
{
	/*
	 * We _might_ want to treat CR differently from other
	 * whitespace characters when ws_rule has WS_CR_AT_EOL, but
	 * for now we just use this stupid definition.
	 */
	while (len-- > 0) {
		if (!isspace(*line))
			return 0;
		line++;
	}
	return 1;
}

/* Copy the line to the buffer while fixing whitespaces */
int ws_fix_copy(char *dst, const char *src, int len, unsigned ws_rule, int *error_count)
{
	/*
	 * len is number of bytes to be copied from src, starting
	 * at src.  Typically src[len-1] is '\n', unless this is
	 * the incomplete last line.
	 */
	int i;
	int add_nl_to_tail = 0;
	int add_cr_to_tail = 0;
	int fixed = 0;
	int last_tab_in_indent = -1;
	int last_space_in_indent = -1;
	int need_fix_leading_space = 0;
	char *buf;

	/*
	 * Strip trailing whitespace
	 */
	if (ws_rule & WS_BLANK_AT_EOL) {
		if (0 < len && src[len - 1] == '\n') {
			add_nl_to_tail = 1;
			len--;
			if (0 < len && src[len - 1] == '\r') {
				add_cr_to_tail = !!(ws_rule & WS_CR_AT_EOL);
				len--;
			}
		}
		if (0 < len && isspace(src[len - 1])) {
			while (0 < len && isspace(src[len-1]))
				len--;
			fixed = 1;
		}
	}

	/*
	 * Check leading whitespaces (indent)
	 */
	for (i = 0; i < len; i++) {
		char ch = src[i];
		if (ch == '\t') {
			last_tab_in_indent = i;
			if ((ws_rule & WS_SPACE_BEFORE_TAB) &&
			    0 <= last_space_in_indent)
			    need_fix_leading_space = 1;
		} else if (ch == ' ') {
			last_space_in_indent = i;
			if ((ws_rule & WS_INDENT_WITH_NON_TAB) &&
			    8 <= i - last_tab_in_indent)
				need_fix_leading_space = 1;
		} else
			break;
	}

	buf = dst;
	if (need_fix_leading_space) {
		/* Process indent ourselves */
		int consecutive_spaces = 0;
		int last = last_tab_in_indent + 1;

		if (ws_rule & WS_INDENT_WITH_NON_TAB) {
			/* have "last" point at one past the indent */
			if (last_tab_in_indent < last_space_in_indent)
				last = last_space_in_indent + 1;
			else
				last = last_tab_in_indent + 1;
		}

		/*
		 * between src[0..last-1], strip the funny spaces,
		 * updating them to tab as needed.
		 */
		for (i = 0; i < last; i++) {
			char ch = src[i];
			if (ch != ' ') {
				consecutive_spaces = 0;
				*dst++ = ch;
			} else {
				consecutive_spaces++;
				if (consecutive_spaces == 8) {
					*dst++ = '\t';
					consecutive_spaces = 0;
				}
			}
		}
		while (0 < consecutive_spaces--)
			*dst++ = ' ';
		len -= last;
		src += last;
		fixed = 1;
	}

	memcpy(dst, src, len);
	if (add_cr_to_tail)
		dst[len++] = '\r';
	if (add_nl_to_tail)
		dst[len++] = '\n';
	if (fixed && error_count)
		(*error_count)++;
	return dst + len - buf;
}
Commit	Line	Data
cf1b7869 JH	1	/*
	2	* Whitespace rules
	3	*
	4	* Copyright (c) 2007 Junio C Hamano
	5	*/
	6
	7	#include "cache.h"
	8	#include "attr.h"
	9
	10	static struct whitespace_rule {
	11	const char *rule_name;
	12	unsigned rule_bits;
a437900f	13	unsigned loosens_error;
cf1b7869	14	} whitespace_rule_names[] = {
a437900f JH	15	{ "trailing-space", WS_TRAILING_SPACE, 0 },
	16	{ "space-before-tab", WS_SPACE_BEFORE_TAB, 0 },
	17	{ "indent-with-non-tab", WS_INDENT_WITH_NON_TAB, 0 },
	18	{ "cr-at-eol", WS_CR_AT_EOL, 1 },
afd9db41 JH	19	{ "blank-at-eol", WS_BLANK_AT_EOL, 0 },
afd9db41 JH	20	{ "blank-at-eof", WS_BLANK_AT_EOF, 0 },
cf1b7869 JH	21	};
	22
	23	unsigned parse_whitespace_rule(const char *string)
	24	{
	25	unsigned rule = WS_DEFAULT_RULE;
	26
	27	while (string) {
	28	int i;
	29	size_t len;
	30	const char *ep;
	31	int negated = 0;
	32
	33	string = string + strspn(string, ", \t\n\r");
	34	ep = strchr(string, ',');
	35	if (!ep)
	36	len = strlen(string);
	37	else
	38	len = ep - string;
	39
	40	if (*string == '-') {
	41	negated = 1;
	42	string++;
	43	len--;
	44	}
	45	if (!len)
	46	break;
	47	for (i = 0; i < ARRAY_SIZE(whitespace_rule_names); i++) {
	48	if (strncmp(whitespace_rule_names[i].rule_name,
	49	string, len))
	50	continue;
	51	if (negated)
	52	rule &= ~whitespace_rule_names[i].rule_bits;
	53	else
	54	rule \|= whitespace_rule_names[i].rule_bits;
	55	break;
	56	}
	57	string = ep;
	58	}
	59	return rule;
	60	}
	61
	62	static void setup_whitespace_attr_check(struct git_attr_check *check)
	63	{
	64	static struct git_attr *attr_whitespace;
	65
	66	if (!attr_whitespace)
7fb0eaa2	67	attr_whitespace = git_attr("whitespace");
cf1b7869 JH	68	check[0].attr = attr_whitespace;
	69	}
	70
	71	unsigned whitespace_rule(const char *pathname)
	72	{
	73	struct git_attr_check attr_whitespace_rule;
	74
	75	setup_whitespace_attr_check(&attr_whitespace_rule);
	76	if (!git_checkattr(pathname, 1, &attr_whitespace_rule)) {
	77	const char *value;
	78
	79	value = attr_whitespace_rule.value;
	80	if (ATTR_TRUE(value)) {
	81	/* true (whitespace) */
	82	unsigned all_rule = 0;
	83	int i;
	84	for (i = 0; i < ARRAY_SIZE(whitespace_rule_names); i++)
a437900f JH	85	if (!whitespace_rule_names[i].loosens_error)
a437900f JH	86	all_rule \|= whitespace_rule_names[i].rule_bits;
cf1b7869 JH	87	return all_rule;
	88	} else if (ATTR_FALSE(value)) {
	89	/* false (-whitespace) */
	90	return 0;
	91	} else if (ATTR_UNSET(value)) {
	92	/* reset to default (!whitespace) */
	93	return whitespace_rule_cfg;
	94	} else {
	95	/* string */
	96	return parse_whitespace_rule(value);
	97	}
	98	} else {
	99	return whitespace_rule_cfg;
	100	}
	101	}
c1795bb0 WC	102
	103	/* The returned string should be freed by the caller. */
	104	char *whitespace_error_string(unsigned ws)
	105	{
f285a2d7	106	struct strbuf err = STRBUF_INIT;
aeb84b05	107	if ((ws & WS_TRAILING_SPACE) == WS_TRAILING_SPACE)
420f4f04	108	strbuf_addstr(&err, "trailing whitespace");
aeb84b05 JH	109	else {
	110	if (ws & WS_BLANK_AT_EOL)
	111	strbuf_addstr(&err, "trailing whitespace");
	112	if (ws & WS_BLANK_AT_EOF) {
	113	if (err.len)
	114	strbuf_addstr(&err, ", ");
	115	strbuf_addstr(&err, "new blank line at EOF");
	116	}
	117	}
c1795bb0 WC	118	if (ws & WS_SPACE_BEFORE_TAB) {
	119	if (err.len)
	120	strbuf_addstr(&err, ", ");
420f4f04	121	strbuf_addstr(&err, "space before tab in indent");
c1795bb0 WC	122	}
	123	if (ws & WS_INDENT_WITH_NON_TAB) {
	124	if (err.len)
	125	strbuf_addstr(&err, ", ");
420f4f04	126	strbuf_addstr(&err, "indent with spaces");
c1795bb0 WC	127	}
	128	return strbuf_detach(&err, NULL);
	129	}
	130
	131	/* If stream is non-NULL, emits the line after checking. */
8f8841e9 JH	132	static unsigned ws_check_emit_1(const char *line, int len, unsigned ws_rule,
	133	FILE stream, const char set,
	134	const char reset, const char ws)
c1795bb0 WC	135	{
c1795bb0 WC	136	unsigned result = 0;
954ecd43	137	int written = 0;
c1795bb0 WC	138	int trailing_whitespace = -1;
c1795bb0 WC	139	int trailing_newline = 0;
b2979ff5	140	int trailing_carriage_return = 0;
c1795bb0 WC	141	int i;
	142
	143	/* Logic is simpler if we temporarily ignore the trailing newline. */
	144	if (len > 0 && line[len - 1] == '\n') {
	145	trailing_newline = 1;
	146	len--;
	147	}
b2979ff5 JH	148	if ((ws_rule & WS_CR_AT_EOL) &&
	149	len > 0 && line[len - 1] == '\r') {
	150	trailing_carriage_return = 1;
	151	len--;
	152	}
c1795bb0 WC	153
c1795bb0 WC	154	/* Check for trailing whitespace. */
aeb84b05	155	if (ws_rule & WS_BLANK_AT_EOL) {
c1795bb0 WC	156	for (i = len - 1; i >= 0; i--) {
	157	if (isspace(line[i])) {
	158	trailing_whitespace = i;
aeb84b05	159	result \|= WS_BLANK_AT_EOL;
c1795bb0 WC	160	}
	161	else
	162	break;
	163	}
	164	}
	165
	166	/* Check for space before tab in initial indent. */
	167	for (i = 0; i < len; i++) {
9afa2d4a	168	if (line[i] == ' ')
1020999a	169	continue;
1020999a	170	if (line[i] != '\t')
c1795bb0	171	break;
ffe56885	172	if ((ws_rule & WS_SPACE_BEFORE_TAB) && written < i) {
1020999a	173	result \|= WS_SPACE_BEFORE_TAB;
ffe56885 BF	174	if (stream) {
	175	fputs(ws, stream);
	176	fwrite(line + written, i - written, 1, stream);
	177	fputs(reset, stream);
	178	}
	179	} else if (stream)
	180	fwrite(line + written, i - written, 1, stream);
	181	if (stream)
	182	fwrite(line + i, 1, 1, stream);
9afa2d4a	183	written = i + 1;
c1795bb0 WC	184	}
	185
	186	/* Check for indent using non-tab. */
ffe56885	187	if ((ws_rule & WS_INDENT_WITH_NON_TAB) && i - written >= 8) {
c1795bb0	188	result \|= WS_INDENT_WITH_NON_TAB;
ffe56885	189	if (stream) {
c1795bb0	190	fputs(ws, stream);
ffe56885	191	fwrite(line + written, i - written, 1, stream);
c1795bb0	192	fputs(reset, stream);
c1795bb0	193	}
ffe56885 BF	194	written = i;
ffe56885 BF	195	}
c1795bb0	196
ffe56885	197	if (stream) {
b2979ff5 JH	198	/*
	199	* Now the rest of the line starts at "written".
	200	* The non-highlighted part ends at "trailing_whitespace".
	201	*/
c1795bb0 WC	202	if (trailing_whitespace == -1)
	203	trailing_whitespace = len;
	204
	205	/* Emit non-highlighted (middle) segment. */
954ecd43	206	if (trailing_whitespace - written > 0) {
c1795bb0	207	fputs(set, stream);
954ecd43 BF	208	fwrite(line + written,
954ecd43 BF	209	trailing_whitespace - written, 1, stream);
c1795bb0 WC	210	fputs(reset, stream);
	211	}
	212
	213	/* Highlight errors in trailing whitespace. */
	214	if (trailing_whitespace != len) {
	215	fputs(ws, stream);
	216	fwrite(line + trailing_whitespace,
	217	len - trailing_whitespace, 1, stream);
	218	fputs(reset, stream);
	219	}
b2979ff5 JH	220	if (trailing_carriage_return)
b2979ff5 JH	221	fputc('\r', stream);
c1795bb0 WC	222	if (trailing_newline)
	223	fputc('\n', stream);
	224	}
	225	return result;
	226	}
fe3403c3	227
8f8841e9 JH	228	void ws_check_emit(const char *line, int len, unsigned ws_rule,
	229	FILE stream, const char set,
	230	const char reset, const char ws)
	231	{
	232	(void)ws_check_emit_1(line, len, ws_rule, stream, set, reset, ws);
	233	}
	234
	235	unsigned ws_check(const char *line, int len, unsigned ws_rule)
	236	{
	237	return ws_check_emit_1(line, len, ws_rule, NULL, NULL, NULL, NULL);
	238	}
	239
877f23cc JH	240	int ws_blank_line(const char *line, int len, unsigned ws_rule)
	241	{
	242	/*
	243	* We _might_ want to treat CR differently from other
	244	* whitespace characters when ws_rule has WS_CR_AT_EOL, but
	245	* for now we just use this stupid definition.
	246	*/
	247	while (len-- > 0) {
	248	if (!isspace(*line))
	249	return 0;
	250	line++;
	251	}
	252	return 1;
	253	}
	254
fe3403c3 JH	255	/* Copy the line to the buffer while fixing whitespaces */
	256	int ws_fix_copy(char dst, const char src, int len, unsigned ws_rule, int *error_count)
	257	{
	258	/*
	259	* len is number of bytes to be copied from src, starting
	260	* at src. Typically src[len-1] is '\n', unless this is
	261	* the incomplete last line.
	262	*/
	263	int i;
	264	int add_nl_to_tail = 0;
	265	int add_cr_to_tail = 0;
	266	int fixed = 0;
	267	int last_tab_in_indent = -1;
	268	int last_space_in_indent = -1;
	269	int need_fix_leading_space = 0;
	270	char *buf;
	271
	272	/*
	273	* Strip trailing whitespace
	274	*/
afd9db41	275	if (ws_rule & WS_BLANK_AT_EOL) {
422a82f2	276	if (0 < len && src[len - 1] == '\n') {
fe3403c3 JH	277	add_nl_to_tail = 1;
fe3403c3 JH	278	len--;
422a82f2	279	if (0 < len && src[len - 1] == '\r') {
fe3403c3 JH	280	add_cr_to_tail = !!(ws_rule & WS_CR_AT_EOL);
	281	len--;
	282	}
	283	}
	284	if (0 < len && isspace(src[len - 1])) {
	285	while (0 < len && isspace(src[len-1]))
	286	len--;
	287	fixed = 1;
	288	}
	289	}
	290
	291	/*
	292	* Check leading whitespaces (indent)
	293	*/
	294	for (i = 0; i < len; i++) {
	295	char ch = src[i];
	296	if (ch == '\t') {
	297	last_tab_in_indent = i;
	298	if ((ws_rule & WS_SPACE_BEFORE_TAB) &&
	299	0 <= last_space_in_indent)
	300	need_fix_leading_space = 1;
	301	} else if (ch == ' ') {
	302	last_space_in_indent = i;
	303	if ((ws_rule & WS_INDENT_WITH_NON_TAB) &&
	304	8 <= i - last_tab_in_indent)
	305	need_fix_leading_space = 1;
	306	} else
	307	break;
	308	}
	309
	310	buf = dst;
	311	if (need_fix_leading_space) {
	312	/* Process indent ourselves */
	313	int consecutive_spaces = 0;
	314	int last = last_tab_in_indent + 1;
	315
	316	if (ws_rule & WS_INDENT_WITH_NON_TAB) {
	317	/* have "last" point at one past the indent */
	318	if (last_tab_in_indent < last_space_in_indent)
	319	last = last_space_in_indent + 1;
	320	else
	321	last = last_tab_in_indent + 1;
	322	}
	323
	324	/*
	325	* between src[0..last-1], strip the funny spaces,
	326	* updating them to tab as needed.
	327	*/
	328	for (i = 0; i < last; i++) {
	329	char ch = src[i];
	330	if (ch != ' ') {
	331	consecutive_spaces = 0;
	332	*dst++ = ch;
	333	} else {
	334	consecutive_spaces++;
	335	if (consecutive_spaces == 8) {
	336	*dst++ = '\t';
	337	consecutive_spaces = 0;
	338	}
	339	}
	340	}
	341	while (0 < consecutive_spaces--)
	342	*dst++ = ' ';
	343	len -= last;
344	src += last;
345	fixed = 1;
346	}
347
348	memcpy(dst, src, len);
349	if (add_cr_to_tail)
350	dst[len++] = '\r';
351	if (add_nl_to_tail)
352	dst[len++] = '\n';
353	if (fixed && error_count)
354	(*error_count)++;
355	return dst + len - buf;
356	}