[thirdparty/git.git] / grep.h

#ifndef GREP_H
#define GREP_H
#include "color.h"
#ifdef USE_LIBPCRE2
#define PCRE2_CODE_UNIT_WIDTH 8
#include <pcre2.h>
#if (PCRE2_MAJOR >= 10 && PCRE2_MINOR >= 36) || PCRE2_MAJOR >= 11
#define GIT_PCRE2_VERSION_10_36_OR_HIGHER
#endif
#if (PCRE2_MAJOR >= 10 && PCRE2_MINOR >= 34) || PCRE2_MAJOR >= 11
#define GIT_PCRE2_VERSION_10_34_OR_HIGHER
#endif
#else
typedef int pcre2_code;
typedef int pcre2_match_data;
typedef int pcre2_compile_context;
typedef int pcre2_general_context;
#endif
#ifndef PCRE2_MATCH_INVALID_UTF
/* PCRE2_MATCH_* dummy also with !USE_LIBPCRE2, for test-pcre2-config.c */
#define PCRE2_MATCH_INVALID_UTF 0
#endif
#include "thread-utils.h"
#include "userdiff.h"

struct repository;

enum grep_pat_token {
	GREP_PATTERN,
	GREP_PATTERN_HEAD,
	GREP_PATTERN_BODY,
	GREP_AND,
	GREP_OPEN_PAREN,
	GREP_CLOSE_PAREN,
	GREP_NOT,
	GREP_OR
};

enum grep_context {
	GREP_CONTEXT_HEAD,
	GREP_CONTEXT_BODY
};

enum grep_header_field {
	GREP_HEADER_FIELD_MIN = 0,
	GREP_HEADER_AUTHOR = GREP_HEADER_FIELD_MIN,
	GREP_HEADER_COMMITTER,
	GREP_HEADER_REFLOG,

	/* Must be at the end of the enum */
	GREP_HEADER_FIELD_MAX
};

enum grep_color {
	GREP_COLOR_CONTEXT,
	GREP_COLOR_FILENAME,
	GREP_COLOR_FUNCTION,
	GREP_COLOR_LINENO,
	GREP_COLOR_COLUMNNO,
	GREP_COLOR_MATCH_CONTEXT,
	GREP_COLOR_MATCH_SELECTED,
	GREP_COLOR_SELECTED,
	GREP_COLOR_SEP,
	NR_GREP_COLORS
};

struct grep_pat {
	struct grep_pat *next;
	const char *origin;
	int no;
	enum grep_pat_token token;
	char *pattern;
	size_t patternlen;
	enum grep_header_field field;
	regex_t regexp;
	pcre2_code *pcre2_pattern;
	pcre2_match_data *pcre2_match_data;
	pcre2_compile_context *pcre2_compile_context;
	pcre2_general_context *pcre2_general_context;
	const uint8_t *pcre2_tables;
	uint32_t pcre2_jit_on;
	unsigned fixed:1;
	unsigned is_fixed:1;
	unsigned ignore_case:1;
	unsigned word_regexp:1;
};

enum grep_expr_node {
	GREP_NODE_ATOM,
	GREP_NODE_NOT,
	GREP_NODE_AND,
	GREP_NODE_TRUE,
	GREP_NODE_OR
};

enum grep_pattern_type {
	GREP_PATTERN_TYPE_UNSPECIFIED = 0,
	GREP_PATTERN_TYPE_BRE,
	GREP_PATTERN_TYPE_ERE,
	GREP_PATTERN_TYPE_FIXED,
	GREP_PATTERN_TYPE_PCRE
};

struct grep_expr {
	enum grep_expr_node node;
	unsigned hit;
	union {
		struct grep_pat *atom;
		struct grep_expr *unary;
		struct {
			struct grep_expr *left;
			struct grep_expr *right;
		} binary;
	} u;
};

struct grep_opt {
	struct grep_pat *pattern_list;
	struct grep_pat **pattern_tail;
	struct grep_pat *header_list;
	struct grep_pat **header_tail;
	struct grep_expr *pattern_expression;

	/*
	 * NEEDSWORK: See if we can remove this field, because the repository
	 * should probably be per-source. That is, grep.c functions using this
	 * field should probably start using "repo" in "struct grep_source"
	 * instead.
	 *
	 * This is potentially the cause of at least one bug - "git grep"
	 * using the textconv attributes from the superproject on the
	 * submodules. See the failing "git grep --textconv" tests in
	 * t7814-grep-recurse-submodules.sh for more information.
	 */
	struct repository *repo;

	int linenum;
	int columnnum;
	int invert;
	int ignore_case;
	int status_only;
	int name_only;
	int unmatch_name_only;
	int count;
	int word_regexp;
	int all_match;
	int no_body_match;
	int body_hit;
#define GREP_BINARY_DEFAULT	0
#define GREP_BINARY_NOMATCH	1
#define GREP_BINARY_TEXT	2
	int binary;
	int allow_textconv;
	int extended;
	int use_reflog_filter;
	int relative;
	int pathname;
	int null_following_name;
	int only_matching;
	int color;
	int max_depth;
	int funcname;
	int funcbody;
	int extended_regexp_option;
	enum grep_pattern_type pattern_type_option;
	int ignore_locale;
	char colors[NR_GREP_COLORS][COLOR_MAXLEN];
	unsigned pre_context;
	unsigned post_context;
	unsigned last_shown;
	int show_hunk_mark;
	int file_break;
	int heading;
	void *priv;

	void (*output)(struct grep_opt *opt, const void *data, size_t size);
	void *output_priv;
};

#define GREP_OPT_INIT { \
	.relative = 1, \
	.pathname = 1, \
	.max_depth = -1, \
	.pattern_type_option = GREP_PATTERN_TYPE_UNSPECIFIED, \
	.colors = { \
		[GREP_COLOR_CONTEXT] = "", \
		[GREP_COLOR_FILENAME] = GIT_COLOR_MAGENTA, \
		[GREP_COLOR_FUNCTION] = "", \
		[GREP_COLOR_LINENO] = GIT_COLOR_GREEN, \
		[GREP_COLOR_COLUMNNO] = GIT_COLOR_GREEN, \
		[GREP_COLOR_MATCH_CONTEXT] = GIT_COLOR_BOLD_RED, \
		[GREP_COLOR_MATCH_SELECTED] = GIT_COLOR_BOLD_RED, \
		[GREP_COLOR_SELECTED] = "", \
		[GREP_COLOR_SEP] = GIT_COLOR_CYAN, \
	}, \
	.only_matching = 0, \
	.color = -1, \
	.output = std_output, \
}

int grep_config(const char *var, const char *value, void *);
void grep_init(struct grep_opt *, struct repository *repo);

void append_grep_pat(struct grep_opt *opt, const char *pat, size_t patlen, const char *origin, int no, enum grep_pat_token t);
void append_grep_pattern(struct grep_opt *opt, const char *pat, const char *origin, int no, enum grep_pat_token t);
void append_header_grep_pattern(struct grep_opt *, enum grep_header_field, const char *);
void compile_grep_patterns(struct grep_opt *opt);
void free_grep_patterns(struct grep_opt *opt);
int grep_buffer(struct grep_opt *opt, const char *buf, unsigned long size);

/* The field parameter is only used to filter header patterns
 * (where appropriate). If filtering isn't desirable
 * GREP_HEADER_FIELD_MAX should be supplied.
 */
int grep_next_match(struct grep_opt *opt,
		    const char *bol, const char *eol,
		    enum grep_context ctx, regmatch_t *pmatch,
		    enum grep_header_field field, int eflags);

struct grep_source {
	char *name;

	enum grep_source_type {
		GREP_SOURCE_OID,
		GREP_SOURCE_FILE,
		GREP_SOURCE_BUF,
	} type;
	void *identifier;
	struct repository *repo; /* if GREP_SOURCE_OID */

	const char *buf;
	unsigned long size;

	char *path; /* for attribute lookups */
	struct userdiff_driver *driver;
};

void grep_source_init_file(struct grep_source *gs, const char *name,
			   const char *path);
void grep_source_init_oid(struct grep_source *gs, const char *name,
			  const char *path, const struct object_id *oid,
			  struct repository *repo);
void grep_source_clear_data(struct grep_source *gs);
void grep_source_clear(struct grep_source *gs);
void grep_source_load_driver(struct grep_source *gs,
			     struct index_state *istate);


int grep_source(struct grep_opt *opt, struct grep_source *gs);

struct grep_opt *grep_opt_dup(const struct grep_opt *opt);

/*
 * Mutex used around access to the attributes machinery if
 * opt->use_threads.  Must be initialized/destroyed by callers!
 */
extern int grep_use_locks;
extern pthread_mutex_t grep_attr_mutex;

#endif
Commit	Line	Data
83b5d2f5 JH	1	#ifndef GREP_H
83b5d2f5 JH	2	#define GREP_H
7e8f59d5	3	#include "color.h"
94da9193 ÆAB	4	#ifdef USE_LIBPCRE2
	5	#define PCRE2_CODE_UNIT_WIDTH 8
	6	#include <pcre2.h>
797c3599 ÆAB	7	#if (PCRE2_MAJOR >= 10 && PCRE2_MINOR >= 36) \|\| PCRE2_MAJOR >= 11
	8	#define GIT_PCRE2_VERSION_10_36_OR_HIGHER
	9	#endif
b76bf27f ÆAB	10	#if (PCRE2_MAJOR >= 10 && PCRE2_MINOR >= 34) \|\| PCRE2_MAJOR >= 11
	11	#define GIT_PCRE2_VERSION_10_34_OR_HIGHER
	12	#endif
94da9193 ÆAB	13	#else
	14	typedef int pcre2_code;
	15	typedef int pcre2_match_data;
	16	typedef int pcre2_compile_context;
cbe81e65	17	typedef int pcre2_general_context;
94da9193	18	#endif
95ca1f98 ÆAB	19	#ifndef PCRE2_MATCH_INVALID_UTF
	20	/* PCRE2_MATCH_* dummy also with !USE_LIBPCRE2, for test-pcre2-config.c */
	21	#define PCRE2_MATCH_INVALID_UTF 0
	22	#endif
0579f91d	23	#include "thread-utils.h"
94ad9d9e	24	#include "userdiff.h"
83b5d2f5	25
38bbc2ea NTND	26	struct repository;
38bbc2ea NTND	27
83b5d2f5 JH	28	enum grep_pat_token {
83b5d2f5 JH	29	GREP_PATTERN,
480c1ca6 JH	30	GREP_PATTERN_HEAD,
480c1ca6 JH	31	GREP_PATTERN_BODY,
83b5d2f5 JH	32	GREP_AND,
	33	GREP_OPEN_PAREN,
	34	GREP_CLOSE_PAREN,
	35	GREP_NOT,
4b05548f	36	GREP_OR
83b5d2f5 JH	37	};
83b5d2f5 JH	38
480c1ca6 JH	39	enum grep_context {
480c1ca6 JH	40	GREP_CONTEXT_HEAD,
4b05548f	41	GREP_CONTEXT_BODY
480c1ca6 JH	42	};
480c1ca6 JH	43
a4d7d2c6	44	enum grep_header_field {
3ce3ffb8 AP	45	GREP_HEADER_FIELD_MIN = 0,
3ce3ffb8 AP	46	GREP_HEADER_AUTHOR = GREP_HEADER_FIELD_MIN,
ad4813b3	47	GREP_HEADER_COMMITTER,
72fd13f7	48	GREP_HEADER_REFLOG,
ad4813b3 NTND	49
	50	/* Must be at the end of the enum */
	51	GREP_HEADER_FIELD_MAX
a4d7d2c6 JH	52	};
a4d7d2c6 JH	53
fa151dc5 NTND	54	enum grep_color {
	55	GREP_COLOR_CONTEXT,
	56	GREP_COLOR_FILENAME,
	57	GREP_COLOR_FUNCTION,
	58	GREP_COLOR_LINENO,
d036d667	59	GREP_COLOR_COLUMNNO,
fa151dc5 NTND	60	GREP_COLOR_MATCH_CONTEXT,
	61	GREP_COLOR_MATCH_SELECTED,
	62	GREP_COLOR_SELECTED,
	63	GREP_COLOR_SEP,
	64	NR_GREP_COLORS
	65	};
	66
83b5d2f5 JH	67	struct grep_pat {
	68	struct grep_pat *next;
	69	const char *origin;
	70	int no;
	71	enum grep_pat_token token;
526a858a	72	char *pattern;
ed40a095	73	size_t patternlen;
a4d7d2c6	74	enum grep_header_field field;
83b5d2f5	75	regex_t regexp;
94da9193 ÆAB	76	pcre2_code *pcre2_pattern;
	77	pcre2_match_data *pcre2_match_data;
	78	pcre2_compile_context *pcre2_compile_context;
cbe81e65	79	pcre2_general_context *pcre2_general_context;
10da030a	80	const uint8_t *pcre2_tables;
94da9193	81	uint32_t pcre2_jit_on;
c822255c	82	unsigned fixed:1;
09872f64	83	unsigned is_fixed:1;
5183bf67	84	unsigned ignore_case:1;
d7eb527d	85	unsigned word_regexp:1;
83b5d2f5 JH	86	};
	87
	88	enum grep_expr_node {
	89	GREP_NODE_ATOM,
	90	GREP_NODE_NOT,
	91	GREP_NODE_AND,
5aaeb733	92	GREP_NODE_TRUE,
4b05548f	93	GREP_NODE_OR
83b5d2f5 JH	94	};
83b5d2f5 JH	95
84befcd0 S	96	enum grep_pattern_type {
	97	GREP_PATTERN_TYPE_UNSPECIFIED = 0,
	98	GREP_PATTERN_TYPE_BRE,
	99	GREP_PATTERN_TYPE_ERE,
	100	GREP_PATTERN_TYPE_FIXED,
	101	GREP_PATTERN_TYPE_PCRE
	102	};
	103
83b5d2f5 JH	104	struct grep_expr {
83b5d2f5 JH	105	enum grep_expr_node node;
0ab7befa	106	unsigned hit;
83b5d2f5 JH	107	union {
	108	struct grep_pat *atom;
	109	struct grep_expr *unary;
	110	struct {
	111	struct grep_expr *left;
	112	struct grep_expr *right;
	113	} binary;
	114	} u;
	115	};
	116
	117	struct grep_opt {
	118	struct grep_pat *pattern_list;
	119	struct grep_pat **pattern_tail;
80235ba7 JH	120	struct grep_pat *header_list;
80235ba7 JH	121	struct grep_pat **header_tail;
83b5d2f5	122	struct grep_expr *pattern_expression;
0693806b JT	123
	124	/*
	125	* NEEDSWORK: See if we can remove this field, because the repository
	126	* should probably be per-source. That is, grep.c functions using this
	127	* field should probably start using "repo" in "struct grep_source"
	128	* instead.
	129	*
	130	* This is potentially the cause of at least one bug - "git grep"
45bde58e MT	131	* using the textconv attributes from the superproject on the
	132	* submodules. See the failing "git grep --textconv" tests in
	133	* t7814-grep-recurse-submodules.sh for more information.
0693806b	134	*/
38bbc2ea	135	struct repository *repo;
0693806b	136
3e230fa1	137	int linenum;
017c0fcf	138	int columnnum;
3e230fa1	139	int invert;
5183bf67	140	int ignore_case;
3e230fa1 RS	141	int status_only;
	142	int name_only;
	143	int unmatch_name_only;
	144	int count;
	145	int word_regexp;
3e230fa1	146	int all_match;
794c0002 RS	147	int no_body_match;
794c0002 RS	148	int body_hit;
83b5d2f5 JH	149	#define GREP_BINARY_DEFAULT 0
	150	#define GREP_BINARY_NOMATCH 1
	151	#define GREP_BINARY_TEXT 2
3e230fa1	152	int binary;
335ec3bf	153	int allow_textconv;
3e230fa1	154	int extended;
baa6378f	155	int use_reflog_filter;
3e230fa1 RS	156	int relative;
	157	int pathname;
	158	int null_following_name;
9d8db06e	159	int only_matching;
7e8f59d5	160	int color;
a91f453f	161	int max_depth;
2944e4e6	162	int funcname;
ba8ea749	163	int funcbody;
84befcd0	164	int extended_regexp_option;
321ee436	165	enum grep_pattern_type pattern_type_option;
44570188	166	int ignore_locale;
fa151dc5	167	char colors[NR_GREP_COLORS][COLOR_MAXLEN];
83b5d2f5 JH	168	unsigned pre_context;
83b5d2f5 JH	169	unsigned post_context;
5dd06d38	170	unsigned last_shown;
046802d0	171	int show_hunk_mark;
a8f0e764	172	int file_break;
1d84f72e	173	int heading;
60ecac98	174	void *priv;
5b594f45 FK	175
	176	void (output)(struct grep_opt opt, const void *data, size_t size);
	177	void *output_priv;
83b5d2f5 JH	178	};
83b5d2f5 JH	179
72365bb4 ÆAB	180	#define GREP_OPT_INIT { \
	181	.relative = 1, \
	182	.pathname = 1, \
	183	.max_depth = -1, \
	184	.pattern_type_option = GREP_PATTERN_TYPE_UNSPECIFIED, \
	185	.colors = { \
	186	[GREP_COLOR_CONTEXT] = "", \
	187	[GREP_COLOR_FILENAME] = GIT_COLOR_MAGENTA, \
	188	[GREP_COLOR_FUNCTION] = "", \
	189	[GREP_COLOR_LINENO] = GIT_COLOR_GREEN, \
	190	[GREP_COLOR_COLUMNNO] = GIT_COLOR_GREEN, \
	191	[GREP_COLOR_MATCH_CONTEXT] = GIT_COLOR_BOLD_RED, \
	192	[GREP_COLOR_MATCH_SELECTED] = GIT_COLOR_BOLD_RED, \
	193	[GREP_COLOR_SELECTED] = "", \
	194	[GREP_COLOR_SEP] = GIT_COLOR_CYAN, \
	195	}, \
	196	.only_matching = 0, \
	197	.color = -1, \
	198	.output = std_output, \
	199	}
	200
55454427	201	int grep_config(const char var, const char value, void *);
9725c8dd	202	void grep_init(struct grep_opt , struct repository repo);
7687a054	203
55454427 DL	204	void append_grep_pat(struct grep_opt opt, const char pat, size_t patlen, const char *origin, int no, enum grep_pat_token t);
	205	void append_grep_pattern(struct grep_opt opt, const char pat, const char *origin, int no, enum grep_pat_token t);
	206	void append_header_grep_pattern(struct grep_opt , enum grep_header_field, const char );
	207	void compile_grep_patterns(struct grep_opt *opt);
	208	void free_grep_patterns(struct grep_opt *opt);
1e668716	209	int grep_buffer(struct grep_opt opt, const char buf, unsigned long size);
83b5d2f5	210
3f566c4e HM	211	/* The field parameter is only used to filter header patterns
	212	* (where appropriate). If filtering isn't desirable
	213	* GREP_HEADER_FIELD_MAX should be supplied.
	214	*/
	215	int grep_next_match(struct grep_opt *opt,
	216	const char bol, const char eol,
	217	enum grep_context ctx, regmatch_t *pmatch,
	218	enum grep_header_field field, int eflags);
	219
e1327023 JK	220	struct grep_source {
	221	char *name;
	222
	223	enum grep_source_type {
1c41c82b	224	GREP_SOURCE_OID,
e1327023 JK	225	GREP_SOURCE_FILE,
	226	GREP_SOURCE_BUF,
	227	} type;
	228	void *identifier;
0693806b	229	struct repository repo; / if GREP_SOURCE_OID */
e1327023	230
1e668716	231	const char *buf;
e1327023	232	unsigned long size;
94ad9d9e	233
55c61688	234	char path; / for attribute lookups */
94ad9d9e	235	struct userdiff_driver *driver;
e1327023 JK	236	};
e1327023 JK	237
50d92b5f JT	238	void grep_source_init_file(struct grep_source gs, const char name,
	239	const char *path);
	240	void grep_source_init_oid(struct grep_source gs, const char name,
0693806b JT	241	const char path, const struct object_id oid,
0693806b JT	242	struct repository *repo);
e1327023 JK	243	void grep_source_clear_data(struct grep_source *gs);
e1327023 JK	244	void grep_source_clear(struct grep_source *gs);
acd00ea0 NTND	245	void grep_source_load_driver(struct grep_source *gs,
acd00ea0 NTND	246	struct index_state *istate);
07a7d656	247
e1327023 JK	248
	249	int grep_source(struct grep_opt opt, struct grep_source gs);
	250
55454427	251	struct grep_opt grep_opt_dup(const struct grep_opt opt);
5b594f45	252
0579f91d TR	253	/*
	254	* Mutex used around access to the attributes machinery if
	255	* opt->use_threads. Must be initialized/destroyed by callers!
	256	*/
78db6ea9	257	extern int grep_use_locks;
0579f91d	258	extern pthread_mutex_t grep_attr_mutex;
b3aeb285	259
83b5d2f5	260	#endif