[git/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:1,
		exclude_default:1;
} 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 &&
				    !whitespace_rule_names[i].exclude_default)
					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;
727c3718 JH	13	unsigned loosens_error:1,
727c3718 JH	14	exclude_default:1;
cf1b7869	15	} whitespace_rule_names[] = {
a437900f JH	16	{ "trailing-space", WS_TRAILING_SPACE, 0 },
	17	{ "space-before-tab", WS_SPACE_BEFORE_TAB, 0 },
	18	{ "indent-with-non-tab", WS_INDENT_WITH_NON_TAB, 0 },
	19	{ "cr-at-eol", WS_CR_AT_EOL, 1 },
afd9db41 JH	20	{ "blank-at-eol", WS_BLANK_AT_EOL, 0 },
afd9db41 JH	21	{ "blank-at-eof", WS_BLANK_AT_EOF, 0 },
cf1b7869 JH	22	};
	23
	24	unsigned parse_whitespace_rule(const char *string)
	25	{
	26	unsigned rule = WS_DEFAULT_RULE;
	27
	28	while (string) {
	29	int i;
	30	size_t len;
	31	const char *ep;
	32	int negated = 0;
	33
	34	string = string + strspn(string, ", \t\n\r");
	35	ep = strchr(string, ',');
	36	if (!ep)
	37	len = strlen(string);
	38	else
	39	len = ep - string;
	40
	41	if (*string == '-') {
	42	negated = 1;
	43	string++;
	44	len--;
	45	}
	46	if (!len)
	47	break;
	48	for (i = 0; i < ARRAY_SIZE(whitespace_rule_names); i++) {
	49	if (strncmp(whitespace_rule_names[i].rule_name,
	50	string, len))
	51	continue;
	52	if (negated)
	53	rule &= ~whitespace_rule_names[i].rule_bits;
	54	else
	55	rule \|= whitespace_rule_names[i].rule_bits;
	56	break;
	57	}
	58	string = ep;
	59	}
	60	return rule;
	61	}
	62
	63	static void setup_whitespace_attr_check(struct git_attr_check *check)
	64	{
	65	static struct git_attr *attr_whitespace;
	66
	67	if (!attr_whitespace)
7fb0eaa2	68	attr_whitespace = git_attr("whitespace");
cf1b7869 JH	69	check[0].attr = attr_whitespace;
	70	}
	71
	72	unsigned whitespace_rule(const char *pathname)
	73	{
	74	struct git_attr_check attr_whitespace_rule;
	75
	76	setup_whitespace_attr_check(&attr_whitespace_rule);
	77	if (!git_checkattr(pathname, 1, &attr_whitespace_rule)) {
	78	const char *value;
	79
	80	value = attr_whitespace_rule.value;
	81	if (ATTR_TRUE(value)) {
	82	/* true (whitespace) */
	83	unsigned all_rule = 0;
	84	int i;
	85	for (i = 0; i < ARRAY_SIZE(whitespace_rule_names); i++)
727c3718 JH	86	if (!whitespace_rule_names[i].loosens_error &&
727c3718 JH	87	!whitespace_rule_names[i].exclude_default)
a437900f	88	all_rule \|= whitespace_rule_names[i].rule_bits;
cf1b7869 JH	89	return all_rule;
	90	} else if (ATTR_FALSE(value)) {
	91	/* false (-whitespace) */
	92	return 0;
	93	} else if (ATTR_UNSET(value)) {
	94	/* reset to default (!whitespace) */
	95	return whitespace_rule_cfg;
	96	} else {
	97	/* string */
	98	return parse_whitespace_rule(value);
	99	}
	100	} else {
	101	return whitespace_rule_cfg;
	102	}
	103	}
c1795bb0 WC	104
	105	/* The returned string should be freed by the caller. */
	106	char *whitespace_error_string(unsigned ws)
	107	{
f285a2d7	108	struct strbuf err = STRBUF_INIT;
aeb84b05	109	if ((ws & WS_TRAILING_SPACE) == WS_TRAILING_SPACE)
420f4f04	110	strbuf_addstr(&err, "trailing whitespace");
aeb84b05 JH	111	else {
	112	if (ws & WS_BLANK_AT_EOL)
	113	strbuf_addstr(&err, "trailing whitespace");
	114	if (ws & WS_BLANK_AT_EOF) {
	115	if (err.len)
	116	strbuf_addstr(&err, ", ");
	117	strbuf_addstr(&err, "new blank line at EOF");
	118	}
	119	}
c1795bb0 WC	120	if (ws & WS_SPACE_BEFORE_TAB) {
	121	if (err.len)
	122	strbuf_addstr(&err, ", ");
420f4f04	123	strbuf_addstr(&err, "space before tab in indent");
c1795bb0 WC	124	}
	125	if (ws & WS_INDENT_WITH_NON_TAB) {
	126	if (err.len)
	127	strbuf_addstr(&err, ", ");
420f4f04	128	strbuf_addstr(&err, "indent with spaces");
c1795bb0 WC	129	}
	130	return strbuf_detach(&err, NULL);
	131	}
	132
	133	/* If stream is non-NULL, emits the line after checking. */
8f8841e9 JH	134	static unsigned ws_check_emit_1(const char *line, int len, unsigned ws_rule,
	135	FILE stream, const char set,
	136	const char reset, const char ws)
c1795bb0 WC	137	{
c1795bb0 WC	138	unsigned result = 0;
954ecd43	139	int written = 0;
c1795bb0 WC	140	int trailing_whitespace = -1;
c1795bb0 WC	141	int trailing_newline = 0;
b2979ff5	142	int trailing_carriage_return = 0;
c1795bb0 WC	143	int i;
	144
	145	/* Logic is simpler if we temporarily ignore the trailing newline. */
	146	if (len > 0 && line[len - 1] == '\n') {
	147	trailing_newline = 1;
	148	len--;
	149	}
b2979ff5 JH	150	if ((ws_rule & WS_CR_AT_EOL) &&
	151	len > 0 && line[len - 1] == '\r') {
	152	trailing_carriage_return = 1;
	153	len--;
	154	}
c1795bb0 WC	155
c1795bb0 WC	156	/* Check for trailing whitespace. */
aeb84b05	157	if (ws_rule & WS_BLANK_AT_EOL) {
c1795bb0 WC	158	for (i = len - 1; i >= 0; i--) {
	159	if (isspace(line[i])) {
	160	trailing_whitespace = i;
aeb84b05	161	result \|= WS_BLANK_AT_EOL;
c1795bb0 WC	162	}
	163	else
	164	break;
	165	}
	166	}
	167
	168	/* Check for space before tab in initial indent. */
	169	for (i = 0; i < len; i++) {
9afa2d4a	170	if (line[i] == ' ')
1020999a	171	continue;
1020999a	172	if (line[i] != '\t')
c1795bb0	173	break;
ffe56885	174	if ((ws_rule & WS_SPACE_BEFORE_TAB) && written < i) {
1020999a	175	result \|= WS_SPACE_BEFORE_TAB;
ffe56885 BF	176	if (stream) {
	177	fputs(ws, stream);
	178	fwrite(line + written, i - written, 1, stream);
	179	fputs(reset, stream);
	180	}
	181	} else if (stream)
	182	fwrite(line + written, i - written, 1, stream);
	183	if (stream)
	184	fwrite(line + i, 1, 1, stream);
9afa2d4a	185	written = i + 1;
c1795bb0 WC	186	}
	187
	188	/* Check for indent using non-tab. */
ffe56885	189	if ((ws_rule & WS_INDENT_WITH_NON_TAB) && i - written >= 8) {
c1795bb0	190	result \|= WS_INDENT_WITH_NON_TAB;
ffe56885	191	if (stream) {
c1795bb0	192	fputs(ws, stream);
ffe56885	193	fwrite(line + written, i - written, 1, stream);
c1795bb0	194	fputs(reset, stream);
c1795bb0	195	}
ffe56885 BF	196	written = i;
ffe56885 BF	197	}
c1795bb0	198
ffe56885	199	if (stream) {
b2979ff5 JH	200	/*
	201	* Now the rest of the line starts at "written".
	202	* The non-highlighted part ends at "trailing_whitespace".
	203	*/
c1795bb0 WC	204	if (trailing_whitespace == -1)
	205	trailing_whitespace = len;
	206
	207	/* Emit non-highlighted (middle) segment. */
954ecd43	208	if (trailing_whitespace - written > 0) {
c1795bb0	209	fputs(set, stream);
954ecd43 BF	210	fwrite(line + written,
954ecd43 BF	211	trailing_whitespace - written, 1, stream);
c1795bb0 WC	212	fputs(reset, stream);
	213	}
	214
	215	/* Highlight errors in trailing whitespace. */
	216	if (trailing_whitespace != len) {
	217	fputs(ws, stream);
	218	fwrite(line + trailing_whitespace,
	219	len - trailing_whitespace, 1, stream);
	220	fputs(reset, stream);
	221	}
b2979ff5 JH	222	if (trailing_carriage_return)
b2979ff5 JH	223	fputc('\r', stream);
c1795bb0 WC	224	if (trailing_newline)
	225	fputc('\n', stream);
	226	}
	227	return result;
	228	}
fe3403c3	229
8f8841e9 JH	230	void ws_check_emit(const char *line, int len, unsigned ws_rule,
	231	FILE stream, const char set,
	232	const char reset, const char ws)
	233	{
	234	(void)ws_check_emit_1(line, len, ws_rule, stream, set, reset, ws);
	235	}
	236
	237	unsigned ws_check(const char *line, int len, unsigned ws_rule)
	238	{
	239	return ws_check_emit_1(line, len, ws_rule, NULL, NULL, NULL, NULL);
	240	}
	241
877f23cc JH	242	int ws_blank_line(const char *line, int len, unsigned ws_rule)
	243	{
	244	/*
	245	* We _might_ want to treat CR differently from other
	246	* whitespace characters when ws_rule has WS_CR_AT_EOL, but
	247	* for now we just use this stupid definition.
	248	*/
	249	while (len-- > 0) {
	250	if (!isspace(*line))
	251	return 0;
	252	line++;
	253	}
	254	return 1;
	255	}
	256
fe3403c3 JH	257	/* Copy the line to the buffer while fixing whitespaces */
	258	int ws_fix_copy(char dst, const char src, int len, unsigned ws_rule, int *error_count)
	259	{
	260	/*
	261	* len is number of bytes to be copied from src, starting
	262	* at src. Typically src[len-1] is '\n', unless this is
	263	* the incomplete last line.
	264	*/
	265	int i;
	266	int add_nl_to_tail = 0;
	267	int add_cr_to_tail = 0;
	268	int fixed = 0;
	269	int last_tab_in_indent = -1;
	270	int last_space_in_indent = -1;
	271	int need_fix_leading_space = 0;
	272	char *buf;
	273
	274	/*
	275	* Strip trailing whitespace
	276	*/
afd9db41	277	if (ws_rule & WS_BLANK_AT_EOL) {
422a82f2	278	if (0 < len && src[len - 1] == '\n') {
fe3403c3 JH	279	add_nl_to_tail = 1;
fe3403c3 JH	280	len--;
422a82f2	281	if (0 < len && src[len - 1] == '\r') {
fe3403c3 JH	282	add_cr_to_tail = !!(ws_rule & WS_CR_AT_EOL);
	283	len--;
	284	}
	285	}
	286	if (0 < len && isspace(src[len - 1])) {
	287	while (0 < len && isspace(src[len-1]))
	288	len--;
	289	fixed = 1;
	290	}
	291	}
	292
	293	/*
	294	* Check leading whitespaces (indent)
	295	*/
	296	for (i = 0; i < len; i++) {
	297	char ch = src[i];
	298	if (ch == '\t') {
	299	last_tab_in_indent = i;
	300	if ((ws_rule & WS_SPACE_BEFORE_TAB) &&
	301	0 <= last_space_in_indent)
	302	need_fix_leading_space = 1;
	303	} else if (ch == ' ') {
	304	last_space_in_indent = i;
	305	if ((ws_rule & WS_INDENT_WITH_NON_TAB) &&
	306	8 <= i - last_tab_in_indent)
	307	need_fix_leading_space = 1;
	308	} else
	309	break;
	310	}
	311
	312	buf = dst;
	313	if (need_fix_leading_space) {
	314	/* Process indent ourselves */
	315	int consecutive_spaces = 0;
	316	int last = last_tab_in_indent + 1;
	317
	318	if (ws_rule & WS_INDENT_WITH_NON_TAB) {
	319	/* have "last" point at one past the indent */
	320	if (last_tab_in_indent < last_space_in_indent)
	321	last = last_space_in_indent + 1;
	322	else
	323	last = last_tab_in_indent + 1;
	324	}
	325
	326	/*
	327	* between src[0..last-1], strip the funny spaces,
	328	* updating them to tab as needed.
	329	*/
	330	for (i = 0; i < last; i++) {
	331	char ch = src[i];
	332	if (ch != ' ') {
	333	consecutive_spaces = 0;
	334	*dst++ = ch;
	335	} else {
	336	consecutive_spaces++;
	337	if (consecutive_spaces == 8) {
	338	*dst++ = '\t';
	339	consecutive_spaces = 0;
	340	}
	341	}
	342	}
	343	while (0 < consecutive_spaces--)
	344	*dst++ = ' ';
	345	len -= last;
346	src += last;
347	fixed = 1;
348	}
349
350	memcpy(dst, src, len);
351	if (add_cr_to_tail)
352	dst[len++] = '\r';
353	if (add_nl_to_tail)
354	dst[len++] = '\n';
355	if (fixed && error_count)
356	(*error_count)++;
357	return dst + len - buf;
358	}