[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 },
};

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++)
				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)
		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. */
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_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;
}

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_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;
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 },
cf1b7869 JH	19	};
	20
	21	unsigned parse_whitespace_rule(const char *string)
	22	{
	23	unsigned rule = WS_DEFAULT_RULE;
	24
	25	while (string) {
	26	int i;
	27	size_t len;
	28	const char *ep;
	29	int negated = 0;
	30
	31	string = string + strspn(string, ", \t\n\r");
	32	ep = strchr(string, ',');
	33	if (!ep)
	34	len = strlen(string);
	35	else
	36	len = ep - string;
	37
	38	if (*string == '-') {
	39	negated = 1;
	40	string++;
	41	len--;
	42	}
	43	if (!len)
	44	break;
	45	for (i = 0; i < ARRAY_SIZE(whitespace_rule_names); i++) {
	46	if (strncmp(whitespace_rule_names[i].rule_name,
	47	string, len))
	48	continue;
	49	if (negated)
	50	rule &= ~whitespace_rule_names[i].rule_bits;
	51	else
	52	rule \|= whitespace_rule_names[i].rule_bits;
	53	break;
	54	}
	55	string = ep;
	56	}
	57	return rule;
	58	}
	59
	60	static void setup_whitespace_attr_check(struct git_attr_check *check)
	61	{
	62	static struct git_attr *attr_whitespace;
	63
	64	if (!attr_whitespace)
	65	attr_whitespace = git_attr("whitespace", 10);
	66	check[0].attr = attr_whitespace;
	67	}
	68
	69	unsigned whitespace_rule(const char *pathname)
	70	{
	71	struct git_attr_check attr_whitespace_rule;
	72
	73	setup_whitespace_attr_check(&attr_whitespace_rule);
	74	if (!git_checkattr(pathname, 1, &attr_whitespace_rule)) {
	75	const char *value;
	76
	77	value = attr_whitespace_rule.value;
	78	if (ATTR_TRUE(value)) {
	79	/* true (whitespace) */
	80	unsigned all_rule = 0;
	81	int i;
	82	for (i = 0; i < ARRAY_SIZE(whitespace_rule_names); i++)
a437900f JH	83	if (!whitespace_rule_names[i].loosens_error)
a437900f JH	84	all_rule \|= whitespace_rule_names[i].rule_bits;
cf1b7869 JH	85	return all_rule;
	86	} else if (ATTR_FALSE(value)) {
	87	/* false (-whitespace) */
	88	return 0;
	89	} else if (ATTR_UNSET(value)) {
	90	/* reset to default (!whitespace) */
	91	return whitespace_rule_cfg;
	92	} else {
	93	/* string */
	94	return parse_whitespace_rule(value);
	95	}
	96	} else {
	97	return whitespace_rule_cfg;
	98	}
	99	}
c1795bb0 WC	100
	101	/* The returned string should be freed by the caller. */
	102	char *whitespace_error_string(unsigned ws)
	103	{
f285a2d7	104	struct strbuf err = STRBUF_INIT;
c1795bb0	105	if (ws & WS_TRAILING_SPACE)
420f4f04	106	strbuf_addstr(&err, "trailing whitespace");
c1795bb0 WC	107	if (ws & WS_SPACE_BEFORE_TAB) {
	108	if (err.len)
	109	strbuf_addstr(&err, ", ");
420f4f04	110	strbuf_addstr(&err, "space before tab in indent");
c1795bb0 WC	111	}
	112	if (ws & WS_INDENT_WITH_NON_TAB) {
	113	if (err.len)
	114	strbuf_addstr(&err, ", ");
420f4f04	115	strbuf_addstr(&err, "indent with spaces");
c1795bb0 WC	116	}
	117	return strbuf_detach(&err, NULL);
	118	}
	119
	120	/* If stream is non-NULL, emits the line after checking. */
8f8841e9 JH	121	static unsigned ws_check_emit_1(const char *line, int len, unsigned ws_rule,
	122	FILE stream, const char set,
	123	const char reset, const char ws)
c1795bb0 WC	124	{
c1795bb0 WC	125	unsigned result = 0;
954ecd43	126	int written = 0;
c1795bb0 WC	127	int trailing_whitespace = -1;
c1795bb0 WC	128	int trailing_newline = 0;
b2979ff5	129	int trailing_carriage_return = 0;
c1795bb0 WC	130	int i;
	131
	132	/* Logic is simpler if we temporarily ignore the trailing newline. */
	133	if (len > 0 && line[len - 1] == '\n') {
	134	trailing_newline = 1;
	135	len--;
	136	}
b2979ff5 JH	137	if ((ws_rule & WS_CR_AT_EOL) &&
	138	len > 0 && line[len - 1] == '\r') {
	139	trailing_carriage_return = 1;
	140	len--;
	141	}
c1795bb0 WC	142
	143	/* Check for trailing whitespace. */
	144	if (ws_rule & WS_TRAILING_SPACE) {
	145	for (i = len - 1; i >= 0; i--) {
	146	if (isspace(line[i])) {
	147	trailing_whitespace = i;
	148	result \|= WS_TRAILING_SPACE;
	149	}
	150	else
	151	break;
	152	}
	153	}
	154
	155	/* Check for space before tab in initial indent. */
	156	for (i = 0; i < len; i++) {
9afa2d4a	157	if (line[i] == ' ')
1020999a	158	continue;
1020999a	159	if (line[i] != '\t')
c1795bb0	160	break;
ffe56885	161	if ((ws_rule & WS_SPACE_BEFORE_TAB) && written < i) {
1020999a	162	result \|= WS_SPACE_BEFORE_TAB;
ffe56885 BF	163	if (stream) {
	164	fputs(ws, stream);
	165	fwrite(line + written, i - written, 1, stream);
	166	fputs(reset, stream);
	167	}
	168	} else if (stream)
	169	fwrite(line + written, i - written, 1, stream);
	170	if (stream)
	171	fwrite(line + i, 1, 1, stream);
9afa2d4a	172	written = i + 1;
c1795bb0 WC	173	}
	174
	175	/* Check for indent using non-tab. */
ffe56885	176	if ((ws_rule & WS_INDENT_WITH_NON_TAB) && i - written >= 8) {
c1795bb0	177	result \|= WS_INDENT_WITH_NON_TAB;
ffe56885	178	if (stream) {
c1795bb0	179	fputs(ws, stream);
ffe56885	180	fwrite(line + written, i - written, 1, stream);
c1795bb0	181	fputs(reset, stream);
c1795bb0	182	}
ffe56885 BF	183	written = i;
ffe56885 BF	184	}
c1795bb0	185
ffe56885	186	if (stream) {
b2979ff5 JH	187	/*
	188	* Now the rest of the line starts at "written".
	189	* The non-highlighted part ends at "trailing_whitespace".
	190	*/
c1795bb0 WC	191	if (trailing_whitespace == -1)
	192	trailing_whitespace = len;
	193
	194	/* Emit non-highlighted (middle) segment. */
954ecd43	195	if (trailing_whitespace - written > 0) {
c1795bb0	196	fputs(set, stream);
954ecd43 BF	197	fwrite(line + written,
954ecd43 BF	198	trailing_whitespace - written, 1, stream);
c1795bb0 WC	199	fputs(reset, stream);
	200	}
	201
	202	/* Highlight errors in trailing whitespace. */
	203	if (trailing_whitespace != len) {
	204	fputs(ws, stream);
	205	fwrite(line + trailing_whitespace,
	206	len - trailing_whitespace, 1, stream);
	207	fputs(reset, stream);
	208	}
b2979ff5 JH	209	if (trailing_carriage_return)
b2979ff5 JH	210	fputc('\r', stream);
c1795bb0 WC	211	if (trailing_newline)
	212	fputc('\n', stream);
	213	}
	214	return result;
	215	}
fe3403c3	216
8f8841e9 JH	217	void ws_check_emit(const char *line, int len, unsigned ws_rule,
	218	FILE stream, const char set,
	219	const char reset, const char ws)
	220	{
	221	(void)ws_check_emit_1(line, len, ws_rule, stream, set, reset, ws);
	222	}
	223
	224	unsigned ws_check(const char *line, int len, unsigned ws_rule)
	225	{
	226	return ws_check_emit_1(line, len, ws_rule, NULL, NULL, NULL, NULL);
	227	}
	228
877f23cc JH	229	int ws_blank_line(const char *line, int len, unsigned ws_rule)
	230	{
	231	/*
	232	* We _might_ want to treat CR differently from other
	233	* whitespace characters when ws_rule has WS_CR_AT_EOL, but
	234	* for now we just use this stupid definition.
	235	*/
	236	while (len-- > 0) {
	237	if (!isspace(*line))
	238	return 0;
	239	line++;
	240	}
	241	return 1;
	242	}
	243
fe3403c3 JH	244	/* Copy the line to the buffer while fixing whitespaces */
	245	int ws_fix_copy(char dst, const char src, int len, unsigned ws_rule, int *error_count)
	246	{
	247	/*
	248	* len is number of bytes to be copied from src, starting
	249	* at src. Typically src[len-1] is '\n', unless this is
	250	* the incomplete last line.
	251	*/
	252	int i;
	253	int add_nl_to_tail = 0;
	254	int add_cr_to_tail = 0;
	255	int fixed = 0;
	256	int last_tab_in_indent = -1;
	257	int last_space_in_indent = -1;
	258	int need_fix_leading_space = 0;
	259	char *buf;
	260
	261	/*
	262	* Strip trailing whitespace
	263	*/
	264	if ((ws_rule & WS_TRAILING_SPACE) &&
c6fabfaf	265	(2 <= len && isspace(src[len-2]))) {
fe3403c3 JH	266	if (src[len - 1] == '\n') {
	267	add_nl_to_tail = 1;
	268	len--;
	269	if (1 < len && src[len - 1] == '\r') {
	270	add_cr_to_tail = !!(ws_rule & WS_CR_AT_EOL);
	271	len--;
	272	}
	273	}
	274	if (0 < len && isspace(src[len - 1])) {
	275	while (0 < len && isspace(src[len-1]))
	276	len--;
	277	fixed = 1;
	278	}
	279	}
	280
	281	/*
	282	* Check leading whitespaces (indent)
	283	*/
	284	for (i = 0; i < len; i++) {
	285	char ch = src[i];
	286	if (ch == '\t') {
	287	last_tab_in_indent = i;
	288	if ((ws_rule & WS_SPACE_BEFORE_TAB) &&
	289	0 <= last_space_in_indent)
	290	need_fix_leading_space = 1;
	291	} else if (ch == ' ') {
	292	last_space_in_indent = i;
	293	if ((ws_rule & WS_INDENT_WITH_NON_TAB) &&
	294	8 <= i - last_tab_in_indent)
	295	need_fix_leading_space = 1;
	296	} else
	297	break;
	298	}
	299
	300	buf = dst;
	301	if (need_fix_leading_space) {
	302	/* Process indent ourselves */
	303	int consecutive_spaces = 0;
	304	int last = last_tab_in_indent + 1;
	305
	306	if (ws_rule & WS_INDENT_WITH_NON_TAB) {
	307	/* have "last" point at one past the indent */
	308	if (last_tab_in_indent < last_space_in_indent)
	309	last = last_space_in_indent + 1;
	310	else
	311	last = last_tab_in_indent + 1;
	312	}
	313
	314	/*
	315	* between src[0..last-1], strip the funny spaces,
	316	* updating them to tab as needed.
	317	*/
	318	for (i = 0; i < last; i++) {
	319	char ch = src[i];
	320	if (ch != ' ') {
	321	consecutive_spaces = 0;
	322	*dst++ = ch;
	323	} else {
	324	consecutive_spaces++;
	325	if (consecutive_spaces == 8) {
	326	*dst++ = '\t';
	327	consecutive_spaces = 0;
	328	}
	329	}
330	}
331	while (0 < consecutive_spaces--)
332	*dst++ = ' ';
333	len -= last;
334	src += last;
335	fixed = 1;
336	}
337
338	memcpy(dst, src, len);
339	if (add_cr_to_tail)
340	dst[len++] = '\r';
341	if (add_nl_to_tail)
342	dst[len++] = '\n';
343	if (fixed && error_count)
344	(*error_count)++;
345	return dst + len - buf;
346	}