[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 },
	{ "blank-at-eof", WS_BLANK_AT_EOF },
};

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");
	}
	if (ws & WS_BLANK_AT_EOF) {
		if (err.len)
			strbuf_addstr(&err, ", ");
		strbuf_addstr(&err, "new blank line at EOF");
	}
	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;
	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 },
77b15bbd	18	{ "blank-at-eof", WS_BLANK_AT_EOF },
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++)
83	all_rule \|= whitespace_rule_names[i].rule_bits;
84	return all_rule;
85	} else if (ATTR_FALSE(value)) {
86	/* false (-whitespace) */
87	return 0;
88	} else if (ATTR_UNSET(value)) {
89	/* reset to default (!whitespace) */
90	return whitespace_rule_cfg;
91	} else {
92	/* string */
93	return parse_whitespace_rule(value);
94	}
95	} else {
96	return whitespace_rule_cfg;
97	}
98	}
c1795bb0 WC	99
	100	/* The returned string should be freed by the caller. */
	101	char *whitespace_error_string(unsigned ws)
	102	{
	103	struct strbuf err;
	104	strbuf_init(&err, 0);
	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	116	}
77b15bbd JH	117	if (ws & WS_BLANK_AT_EOF) {
	118	if (err.len)
	119	strbuf_addstr(&err, ", ");
	120	strbuf_addstr(&err, "new blank line at EOF");
	121	}
c1795bb0 WC	122	return strbuf_detach(&err, NULL);
	123	}
	124
	125	/* If stream is non-NULL, emits the line after checking. */
8f8841e9 JH	126	static unsigned ws_check_emit_1(const char *line, int len, unsigned ws_rule,
	127	FILE stream, const char set,
	128	const char reset, const char ws)
c1795bb0 WC	129	{
c1795bb0 WC	130	unsigned result = 0;
954ecd43	131	int written = 0;
c1795bb0 WC	132	int trailing_whitespace = -1;
c1795bb0 WC	133	int trailing_newline = 0;
b2979ff5	134	int trailing_carriage_return = 0;
c1795bb0 WC	135	int i;
	136
	137	/* Logic is simpler if we temporarily ignore the trailing newline. */
	138	if (len > 0 && line[len - 1] == '\n') {
	139	trailing_newline = 1;
	140	len--;
	141	}
b2979ff5 JH	142	if ((ws_rule & WS_CR_AT_EOL) &&
	143	len > 0 && line[len - 1] == '\r') {
	144	trailing_carriage_return = 1;
	145	len--;
	146	}
c1795bb0 WC	147
	148	/* Check for trailing whitespace. */
	149	if (ws_rule & WS_TRAILING_SPACE) {
	150	for (i = len - 1; i >= 0; i--) {
	151	if (isspace(line[i])) {
	152	trailing_whitespace = i;
	153	result \|= WS_TRAILING_SPACE;
	154	}
	155	else
	156	break;
	157	}
	158	}
	159
	160	/* Check for space before tab in initial indent. */
	161	for (i = 0; i < len; i++) {
9afa2d4a	162	if (line[i] == ' ')
1020999a	163	continue;
1020999a	164	if (line[i] != '\t')
c1795bb0	165	break;
ffe56885	166	if ((ws_rule & WS_SPACE_BEFORE_TAB) && written < i) {
1020999a	167	result \|= WS_SPACE_BEFORE_TAB;
ffe56885 BF	168	if (stream) {
	169	fputs(ws, stream);
	170	fwrite(line + written, i - written, 1, stream);
	171	fputs(reset, stream);
	172	}
	173	} else if (stream)
	174	fwrite(line + written, i - written, 1, stream);
	175	if (stream)
	176	fwrite(line + i, 1, 1, stream);
9afa2d4a	177	written = i + 1;
c1795bb0 WC	178	}
	179
	180	/* Check for indent using non-tab. */
ffe56885	181	if ((ws_rule & WS_INDENT_WITH_NON_TAB) && i - written >= 8) {
c1795bb0	182	result \|= WS_INDENT_WITH_NON_TAB;
ffe56885	183	if (stream) {
c1795bb0	184	fputs(ws, stream);
ffe56885	185	fwrite(line + written, i - written, 1, stream);
c1795bb0	186	fputs(reset, stream);
c1795bb0	187	}
ffe56885 BF	188	written = i;
ffe56885 BF	189	}
c1795bb0	190
ffe56885	191	if (stream) {
b2979ff5 JH	192	/*
	193	* Now the rest of the line starts at "written".
	194	* The non-highlighted part ends at "trailing_whitespace".
	195	*/
c1795bb0 WC	196	if (trailing_whitespace == -1)
	197	trailing_whitespace = len;
	198
	199	/* Emit non-highlighted (middle) segment. */
954ecd43	200	if (trailing_whitespace - written > 0) {
c1795bb0	201	fputs(set, stream);
954ecd43 BF	202	fwrite(line + written,
954ecd43 BF	203	trailing_whitespace - written, 1, stream);
c1795bb0 WC	204	fputs(reset, stream);
	205	}
	206
	207	/* Highlight errors in trailing whitespace. */
	208	if (trailing_whitespace != len) {
	209	fputs(ws, stream);
	210	fwrite(line + trailing_whitespace,
	211	len - trailing_whitespace, 1, stream);
	212	fputs(reset, stream);
	213	}
b2979ff5 JH	214	if (trailing_carriage_return)
b2979ff5 JH	215	fputc('\r', stream);
c1795bb0 WC	216	if (trailing_newline)
	217	fputc('\n', stream);
	218	}
	219	return result;
	220	}
fe3403c3	221
8f8841e9 JH	222	void ws_check_emit(const char *line, int len, unsigned ws_rule,
	223	FILE stream, const char set,
	224	const char reset, const char ws)
	225	{
	226	(void)ws_check_emit_1(line, len, ws_rule, stream, set, reset, ws);
	227	}
	228
	229	unsigned ws_check(const char *line, int len, unsigned ws_rule)
	230	{
	231	return ws_check_emit_1(line, len, ws_rule, NULL, NULL, NULL, NULL);
	232	}
	233
877f23cc JH	234	int ws_blank_line(const char *line, int len, unsigned ws_rule)
	235	{
	236	/*
	237	* We _might_ want to treat CR differently from other
	238	* whitespace characters when ws_rule has WS_CR_AT_EOL, but
	239	* for now we just use this stupid definition.
	240	*/
	241	while (len-- > 0) {
	242	if (!isspace(*line))
	243	return 0;
	244	line++;
	245	}
	246	return 1;
	247	}
	248
fe3403c3 JH	249	/* Copy the line to the buffer while fixing whitespaces */
	250	int ws_fix_copy(char dst, const char src, int len, unsigned ws_rule, int *error_count)
	251	{
	252	/*
	253	* len is number of bytes to be copied from src, starting
	254	* at src. Typically src[len-1] is '\n', unless this is
	255	* the incomplete last line.
	256	*/
	257	int i;
	258	int add_nl_to_tail = 0;
	259	int add_cr_to_tail = 0;
	260	int fixed = 0;
	261	int last_tab_in_indent = -1;
	262	int last_space_in_indent = -1;
	263	int need_fix_leading_space = 0;
	264	char *buf;
	265
	266	/*
	267	* Strip trailing whitespace
	268	*/
	269	if ((ws_rule & WS_TRAILING_SPACE) &&
c6fabfaf	270	(2 <= len && isspace(src[len-2]))) {
fe3403c3 JH	271	if (src[len - 1] == '\n') {
	272	add_nl_to_tail = 1;
	273	len--;
	274	if (1 < len && src[len - 1] == '\r') {
	275	add_cr_to_tail = !!(ws_rule & WS_CR_AT_EOL);
	276	len--;
	277	}
	278	}
	279	if (0 < len && isspace(src[len - 1])) {
	280	while (0 < len && isspace(src[len-1]))
	281	len--;
	282	fixed = 1;
	283	}
	284	}
	285
	286	/*
	287	* Check leading whitespaces (indent)
	288	*/
	289	for (i = 0; i < len; i++) {
	290	char ch = src[i];
	291	if (ch == '\t') {
	292	last_tab_in_indent = i;
	293	if ((ws_rule & WS_SPACE_BEFORE_TAB) &&
	294	0 <= last_space_in_indent)
	295	need_fix_leading_space = 1;
	296	} else if (ch == ' ') {
	297	last_space_in_indent = i;
	298	if ((ws_rule & WS_INDENT_WITH_NON_TAB) &&
	299	8 <= i - last_tab_in_indent)
	300	need_fix_leading_space = 1;
	301	} else
	302	break;
	303	}
	304
	305	buf = dst;
	306	if (need_fix_leading_space) {
	307	/* Process indent ourselves */
	308	int consecutive_spaces = 0;
	309	int last = last_tab_in_indent + 1;
	310
	311	if (ws_rule & WS_INDENT_WITH_NON_TAB) {
	312	/* have "last" point at one past the indent */
	313	if (last_tab_in_indent < last_space_in_indent)
	314	last = last_space_in_indent + 1;
	315	else
	316	last = last_tab_in_indent + 1;
	317	}
	318
	319	/*
	320	* between src[0..last-1], strip the funny spaces,
	321	* updating them to tab as needed.
	322	*/
	323	for (i = 0; i < last; i++) {
	324	char ch = src[i];
	325	if (ch != ' ') {
	326	consecutive_spaces = 0;
	327	*dst++ = ch;
	328	} else {
	329	consecutive_spaces++;
	330	if (consecutive_spaces == 8) {
	331	*dst++ = '\t';
	332	consecutive_spaces = 0;
	333	}
	334	}
335	}
336	while (0 < consecutive_spaces--)
337	*dst++ = ' ';
338	len -= last;
339	src += last;
340	fixed = 1;
341	}
342
343	memcpy(dst, src, len);
344	if (add_cr_to_tail)
345	dst[len++] = '\r';
346	if (add_nl_to_tail)
347	dst[len++] = '\n';
348	if (fixed && error_count)
349	(*error_count)++;
350	return dst + len - buf;
351	}