[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;
} whitespace_rule_names[] = {
	{ "trailing-space", WS_TRAILING_SPACE },
	{ "space-before-tab", WS_SPACE_BEFORE_TAB },
	{ "indent-with-non-tab", WS_INDENT_WITH_NON_TAB },
	{ "cr-at-eol", WS_CR_AT_EOL },
};

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", 10);
	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++)
				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(&err, 0);
	if (ws & WS_TRAILING_SPACE)
		strbuf_addstr(&err, "trailing whitespace");
	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. */
unsigned check_and_emit_line(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_TRAILING_SPACE) {
		for (i = len - 1; i >= 0; i--) {
			if (isspace(line[i])) {
				trailing_whitespace = i;
				result |= WS_TRAILING_SPACE;
			}
			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;
}

/* 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_TRAILING_SPACE) &&
	    (2 <= len && isspace(src[len-2]))) {
		if (src[len - 1] == '\n') {
			add_nl_to_tail = 1;
			len--;
			if (1 < 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;
	13	} whitespace_rule_names[] = {
	14	{ "trailing-space", WS_TRAILING_SPACE },
	15	{ "space-before-tab", WS_SPACE_BEFORE_TAB },
	16	{ "indent-with-non-tab", WS_INDENT_WITH_NON_TAB },
b2979ff5	17	{ "cr-at-eol", WS_CR_AT_EOL },
cf1b7869 JH	18	};
	19
	20	unsigned parse_whitespace_rule(const char *string)
	21	{
	22	unsigned rule = WS_DEFAULT_RULE;
	23
	24	while (string) {
	25	int i;
	26	size_t len;
	27	const char *ep;
	28	int negated = 0;
	29
	30	string = string + strspn(string, ", \t\n\r");
	31	ep = strchr(string, ',');
	32	if (!ep)
	33	len = strlen(string);
	34	else
	35	len = ep - string;
	36
	37	if (*string == '-') {
	38	negated = 1;
	39	string++;
	40	len--;
	41	}
	42	if (!len)
	43	break;
	44	for (i = 0; i < ARRAY_SIZE(whitespace_rule_names); i++) {
	45	if (strncmp(whitespace_rule_names[i].rule_name,
	46	string, len))
	47	continue;
	48	if (negated)
	49	rule &= ~whitespace_rule_names[i].rule_bits;
	50	else
	51	rule \|= whitespace_rule_names[i].rule_bits;
	52	break;
	53	}
	54	string = ep;
	55	}
	56	return rule;
	57	}
	58
	59	static void setup_whitespace_attr_check(struct git_attr_check *check)
	60	{
	61	static struct git_attr *attr_whitespace;
	62
	63	if (!attr_whitespace)
	64	attr_whitespace = git_attr("whitespace", 10);
	65	check[0].attr = attr_whitespace;
	66	}
	67
	68	unsigned whitespace_rule(const char *pathname)
	69	{
	70	struct git_attr_check attr_whitespace_rule;
	71
	72	setup_whitespace_attr_check(&attr_whitespace_rule);
	73	if (!git_checkattr(pathname, 1, &attr_whitespace_rule)) {
	74	const char *value;
	75
	76	value = attr_whitespace_rule.value;
	77	if (ATTR_TRUE(value)) {
	78	/* true (whitespace) */
	79	unsigned all_rule = 0;
	80	int i;
	81	for (i = 0; i < ARRAY_SIZE(whitespace_rule_names); i++)
82	all_rule \|= whitespace_rule_names[i].rule_bits;
83	return all_rule;
84	} else if (ATTR_FALSE(value)) {
85	/* false (-whitespace) */
86	return 0;
87	} else if (ATTR_UNSET(value)) {
88	/* reset to default (!whitespace) */
89	return whitespace_rule_cfg;
90	} else {
91	/* string */
92	return parse_whitespace_rule(value);
93	}
94	} else {
95	return whitespace_rule_cfg;
96	}
97	}
c1795bb0 WC	98
	99	/* The returned string should be freed by the caller. */
	100	char *whitespace_error_string(unsigned ws)
	101	{
	102	struct strbuf err;
	103	strbuf_init(&err, 0);
	104	if (ws & WS_TRAILING_SPACE)
420f4f04	105	strbuf_addstr(&err, "trailing whitespace");
c1795bb0 WC	106	if (ws & WS_SPACE_BEFORE_TAB) {
	107	if (err.len)
	108	strbuf_addstr(&err, ", ");
420f4f04	109	strbuf_addstr(&err, "space before tab in indent");
c1795bb0 WC	110	}
	111	if (ws & WS_INDENT_WITH_NON_TAB) {
	112	if (err.len)
	113	strbuf_addstr(&err, ", ");
420f4f04	114	strbuf_addstr(&err, "indent with spaces");
c1795bb0 WC	115	}
	116	return strbuf_detach(&err, NULL);
	117	}
	118
	119	/* If stream is non-NULL, emits the line after checking. */
	120	unsigned check_and_emit_line(const char *line, int len, unsigned ws_rule,
	121	FILE stream, const char set,
	122	const char reset, const char ws)
	123	{
	124	unsigned result = 0;
954ecd43	125	int written = 0;
c1795bb0 WC	126	int trailing_whitespace = -1;
c1795bb0 WC	127	int trailing_newline = 0;
b2979ff5	128	int trailing_carriage_return = 0;
c1795bb0 WC	129	int i;
	130
	131	/* Logic is simpler if we temporarily ignore the trailing newline. */
	132	if (len > 0 && line[len - 1] == '\n') {
	133	trailing_newline = 1;
	134	len--;
	135	}
b2979ff5 JH	136	if ((ws_rule & WS_CR_AT_EOL) &&
	137	len > 0 && line[len - 1] == '\r') {
	138	trailing_carriage_return = 1;
	139	len--;
	140	}
c1795bb0 WC	141
	142	/* Check for trailing whitespace. */
	143	if (ws_rule & WS_TRAILING_SPACE) {
	144	for (i = len - 1; i >= 0; i--) {
	145	if (isspace(line[i])) {
	146	trailing_whitespace = i;
	147	result \|= WS_TRAILING_SPACE;
	148	}
	149	else
	150	break;
	151	}
	152	}
	153
	154	/* Check for space before tab in initial indent. */
	155	for (i = 0; i < len; i++) {
9afa2d4a	156	if (line[i] == ' ')
1020999a	157	continue;
1020999a	158	if (line[i] != '\t')
c1795bb0	159	break;
ffe56885	160	if ((ws_rule & WS_SPACE_BEFORE_TAB) && written < i) {
1020999a	161	result \|= WS_SPACE_BEFORE_TAB;
ffe56885 BF	162	if (stream) {
	163	fputs(ws, stream);
	164	fwrite(line + written, i - written, 1, stream);
	165	fputs(reset, stream);
	166	}
	167	} else if (stream)
	168	fwrite(line + written, i - written, 1, stream);
	169	if (stream)
	170	fwrite(line + i, 1, 1, stream);
9afa2d4a	171	written = i + 1;
c1795bb0 WC	172	}
	173
	174	/* Check for indent using non-tab. */
ffe56885	175	if ((ws_rule & WS_INDENT_WITH_NON_TAB) && i - written >= 8) {
c1795bb0	176	result \|= WS_INDENT_WITH_NON_TAB;
ffe56885	177	if (stream) {
c1795bb0	178	fputs(ws, stream);
ffe56885	179	fwrite(line + written, i - written, 1, stream);
c1795bb0	180	fputs(reset, stream);
c1795bb0	181	}
ffe56885 BF	182	written = i;
ffe56885 BF	183	}
c1795bb0	184
ffe56885	185	if (stream) {
b2979ff5 JH	186	/*
	187	* Now the rest of the line starts at "written".
	188	* The non-highlighted part ends at "trailing_whitespace".
	189	*/
c1795bb0 WC	190	if (trailing_whitespace == -1)
	191	trailing_whitespace = len;
	192
	193	/* Emit non-highlighted (middle) segment. */
954ecd43	194	if (trailing_whitespace - written > 0) {
c1795bb0	195	fputs(set, stream);
954ecd43 BF	196	fwrite(line + written,
954ecd43 BF	197	trailing_whitespace - written, 1, stream);
c1795bb0 WC	198	fputs(reset, stream);
	199	}
	200
	201	/* Highlight errors in trailing whitespace. */
	202	if (trailing_whitespace != len) {
	203	fputs(ws, stream);
	204	fwrite(line + trailing_whitespace,
	205	len - trailing_whitespace, 1, stream);
	206	fputs(reset, stream);
	207	}
b2979ff5 JH	208	if (trailing_carriage_return)
b2979ff5 JH	209	fputc('\r', stream);
c1795bb0 WC	210	if (trailing_newline)
	211	fputc('\n', stream);
	212	}
	213	return result;
	214	}
fe3403c3 JH	215
	216	/* Copy the line to the buffer while fixing whitespaces */
	217	int ws_fix_copy(char dst, const char src, int len, unsigned ws_rule, int *error_count)
	218	{
	219	/*
	220	* len is number of bytes to be copied from src, starting
	221	* at src. Typically src[len-1] is '\n', unless this is
	222	* the incomplete last line.
	223	*/
	224	int i;
	225	int add_nl_to_tail = 0;
	226	int add_cr_to_tail = 0;
	227	int fixed = 0;
	228	int last_tab_in_indent = -1;
	229	int last_space_in_indent = -1;
	230	int need_fix_leading_space = 0;
	231	char *buf;
	232
	233	/*
	234	* Strip trailing whitespace
	235	*/
	236	if ((ws_rule & WS_TRAILING_SPACE) &&
c6fabfaf	237	(2 <= len && isspace(src[len-2]))) {
fe3403c3 JH	238	if (src[len - 1] == '\n') {
	239	add_nl_to_tail = 1;
	240	len--;
	241	if (1 < len && src[len - 1] == '\r') {
	242	add_cr_to_tail = !!(ws_rule & WS_CR_AT_EOL);
	243	len--;
	244	}
	245	}
	246	if (0 < len && isspace(src[len - 1])) {
	247	while (0 < len && isspace(src[len-1]))
	248	len--;
	249	fixed = 1;
	250	}
	251	}
	252
	253	/*
	254	* Check leading whitespaces (indent)
	255	*/
	256	for (i = 0; i < len; i++) {
	257	char ch = src[i];
	258	if (ch == '\t') {
	259	last_tab_in_indent = i;
	260	if ((ws_rule & WS_SPACE_BEFORE_TAB) &&
	261	0 <= last_space_in_indent)
	262	need_fix_leading_space = 1;
	263	} else if (ch == ' ') {
	264	last_space_in_indent = i;
	265	if ((ws_rule & WS_INDENT_WITH_NON_TAB) &&
	266	8 <= i - last_tab_in_indent)
	267	need_fix_leading_space = 1;
	268	} else
	269	break;
	270	}
	271
	272	buf = dst;
	273	if (need_fix_leading_space) {
	274	/* Process indent ourselves */
	275	int consecutive_spaces = 0;
	276	int last = last_tab_in_indent + 1;
	277
	278	if (ws_rule & WS_INDENT_WITH_NON_TAB) {
	279	/* have "last" point at one past the indent */
	280	if (last_tab_in_indent < last_space_in_indent)
	281	last = last_space_in_indent + 1;
	282	else
	283	last = last_tab_in_indent + 1;
	284	}
	285
	286	/*
	287	* between src[0..last-1], strip the funny spaces,
	288	* updating them to tab as needed.
	289	*/
	290	for (i = 0; i < last; i++) {
	291	char ch = src[i];
	292	if (ch != ' ') {
	293	consecutive_spaces = 0;
	294	*dst++ = ch;
	295	} else {
	296	consecutive_spaces++;
	297	if (consecutive_spaces == 8) {
	298	*dst++ = '\t';
	299	consecutive_spaces = 0;
	300	}
	301	}
302	}
303	while (0 < consecutive_spaces--)
304	*dst++ = ' ';
305	len -= last;
306	src += last;
307	fixed = 1;
308	}
309
310	memcpy(dst, src, len);
311	if (add_cr_to_tail)
312	dst[len++] = '\r';
313	if (add_nl_to_tail)
314	dst[len++] = '\n';
315	if (fixed && error_count)
316	(*error_count)++;
317	return dst + len - buf;
318	}