[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 >= 35) || PCRE2_MAJOR >= 11
#define GIT_PCRE2_VERSION_10_35_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 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;
	int max_count;
	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, \
	.max_count = -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, \
}

struct config_context;
int grep_config(const char *var, const char *value,
		const struct config_context *ctx, void *data);
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
14b9a044 MK	10	#if (PCRE2_MAJOR >= 10 && PCRE2_MINOR >= 35) \|\| PCRE2_MAJOR >= 11
	11	#define GIT_PCRE2_VERSION_10_35_OR_HIGHER
	12	#endif
b76bf27f ÆAB	13	#if (PCRE2_MAJOR >= 10 && PCRE2_MINOR >= 34) \|\| PCRE2_MAJOR >= 11
	14	#define GIT_PCRE2_VERSION_10_34_OR_HIGHER
	15	#endif
94da9193 ÆAB	16	#else
	17	typedef int pcre2_code;
	18	typedef int pcre2_match_data;
	19	typedef int pcre2_compile_context;
cbe81e65	20	typedef int pcre2_general_context;
94da9193	21	#endif
95ca1f98 ÆAB	22	#ifndef PCRE2_MATCH_INVALID_UTF
	23	/* PCRE2_MATCH_* dummy also with !USE_LIBPCRE2, for test-pcre2-config.c */
	24	#define PCRE2_MATCH_INVALID_UTF 0
	25	#endif
0579f91d	26	#include "thread-utils.h"
94ad9d9e	27	#include "userdiff.h"
83b5d2f5	28
38bbc2ea NTND	29	struct repository;
38bbc2ea NTND	30
83b5d2f5 JH	31	enum grep_pat_token {
83b5d2f5 JH	32	GREP_PATTERN,
480c1ca6 JH	33	GREP_PATTERN_HEAD,
480c1ca6 JH	34	GREP_PATTERN_BODY,
83b5d2f5 JH	35	GREP_AND,
	36	GREP_OPEN_PAREN,
	37	GREP_CLOSE_PAREN,
	38	GREP_NOT,
4b05548f	39	GREP_OR
83b5d2f5 JH	40	};
83b5d2f5 JH	41
480c1ca6 JH	42	enum grep_context {
480c1ca6 JH	43	GREP_CONTEXT_HEAD,
4b05548f	44	GREP_CONTEXT_BODY
480c1ca6 JH	45	};
480c1ca6 JH	46
a4d7d2c6	47	enum grep_header_field {
3ce3ffb8 AP	48	GREP_HEADER_FIELD_MIN = 0,
3ce3ffb8 AP	49	GREP_HEADER_AUTHOR = GREP_HEADER_FIELD_MIN,
ad4813b3	50	GREP_HEADER_COMMITTER,
72fd13f7	51	GREP_HEADER_REFLOG,
ad4813b3 NTND	52
	53	/* Must be at the end of the enum */
	54	GREP_HEADER_FIELD_MAX
a4d7d2c6 JH	55	};
a4d7d2c6 JH	56
fa151dc5 NTND	57	enum grep_color {
	58	GREP_COLOR_CONTEXT,
	59	GREP_COLOR_FILENAME,
	60	GREP_COLOR_FUNCTION,
	61	GREP_COLOR_LINENO,
d036d667	62	GREP_COLOR_COLUMNNO,
fa151dc5 NTND	63	GREP_COLOR_MATCH_CONTEXT,
	64	GREP_COLOR_MATCH_SELECTED,
	65	GREP_COLOR_SELECTED,
	66	GREP_COLOR_SEP,
	67	NR_GREP_COLORS
	68	};
	69
83b5d2f5 JH	70	struct grep_pat {
	71	struct grep_pat *next;
	72	const char *origin;
	73	int no;
	74	enum grep_pat_token token;
526a858a	75	char *pattern;
ed40a095	76	size_t patternlen;
a4d7d2c6	77	enum grep_header_field field;
83b5d2f5	78	regex_t regexp;
94da9193 ÆAB	79	pcre2_code *pcre2_pattern;
	80	pcre2_match_data *pcre2_match_data;
	81	pcre2_compile_context *pcre2_compile_context;
cbe81e65	82	pcre2_general_context *pcre2_general_context;
10da030a	83	const uint8_t *pcre2_tables;
94da9193	84	uint32_t pcre2_jit_on;
c822255c	85	unsigned fixed:1;
09872f64	86	unsigned is_fixed:1;
5183bf67	87	unsigned ignore_case:1;
d7eb527d	88	unsigned word_regexp:1;
83b5d2f5 JH	89	};
	90
	91	enum grep_expr_node {
	92	GREP_NODE_ATOM,
	93	GREP_NODE_NOT,
	94	GREP_NODE_AND,
5aaeb733	95	GREP_NODE_TRUE,
4b05548f	96	GREP_NODE_OR
83b5d2f5 JH	97	};
83b5d2f5 JH	98
84befcd0 S	99	enum grep_pattern_type {
	100	GREP_PATTERN_TYPE_UNSPECIFIED = 0,
	101	GREP_PATTERN_TYPE_BRE,
	102	GREP_PATTERN_TYPE_ERE,
	103	GREP_PATTERN_TYPE_FIXED,
	104	GREP_PATTERN_TYPE_PCRE
	105	};
	106
83b5d2f5 JH	107	struct grep_expr {
83b5d2f5 JH	108	enum grep_expr_node node;
0ab7befa	109	unsigned hit;
83b5d2f5 JH	110	union {
	111	struct grep_pat *atom;
	112	struct grep_expr *unary;
	113	struct {
	114	struct grep_expr *left;
	115	struct grep_expr *right;
	116	} binary;
	117	} u;
	118	};
	119
	120	struct grep_opt {
	121	struct grep_pat *pattern_list;
	122	struct grep_pat **pattern_tail;
80235ba7 JH	123	struct grep_pat *header_list;
80235ba7 JH	124	struct grep_pat **header_tail;
83b5d2f5	125	struct grep_expr *pattern_expression;
0693806b JT	126
	127	/*
	128	* NEEDSWORK: See if we can remove this field, because the repository
	129	* should probably be per-source. That is, grep.c functions using this
	130	* field should probably start using "repo" in "struct grep_source"
	131	* instead.
	132	*
	133	* This is potentially the cause of at least one bug - "git grep"
45bde58e MT	134	* using the textconv attributes from the superproject on the
	135	* submodules. See the failing "git grep --textconv" tests in
	136	* t7814-grep-recurse-submodules.sh for more information.
0693806b	137	*/
38bbc2ea	138	struct repository *repo;
0693806b	139
3e230fa1	140	int linenum;
017c0fcf	141	int columnnum;
3e230fa1	142	int invert;
5183bf67	143	int ignore_case;
3e230fa1 RS	144	int status_only;
	145	int name_only;
	146	int unmatch_name_only;
	147	int count;
	148	int word_regexp;
3e230fa1	149	int all_match;
794c0002 RS	150	int no_body_match;
794c0002 RS	151	int body_hit;
83b5d2f5 JH	152	#define GREP_BINARY_DEFAULT 0
	153	#define GREP_BINARY_NOMATCH 1
	154	#define GREP_BINARY_TEXT 2
3e230fa1	155	int binary;
335ec3bf	156	int allow_textconv;
baa6378f	157	int use_reflog_filter;
3e230fa1 RS	158	int relative;
	159	int pathname;
	160	int null_following_name;
9d8db06e	161	int only_matching;
7e8f59d5	162	int color;
a91f453f	163	int max_depth;
2944e4e6	164	int funcname;
ba8ea749	165	int funcbody;
84befcd0	166	int extended_regexp_option;
321ee436	167	enum grep_pattern_type pattern_type_option;
44570188	168	int ignore_locale;
fa151dc5	169	char colors[NR_GREP_COLORS][COLOR_MAXLEN];
83b5d2f5 JH	170	unsigned pre_context;
83b5d2f5 JH	171	unsigned post_context;
5dd06d38	172	unsigned last_shown;
046802d0	173	int show_hunk_mark;
a8f0e764	174	int file_break;
1d84f72e	175	int heading;
68437ede	176	int max_count;
60ecac98	177	void *priv;
5b594f45 FK	178
	179	void (output)(struct grep_opt opt, const void *data, size_t size);
	180	void *output_priv;
83b5d2f5 JH	181	};
83b5d2f5 JH	182
72365bb4 ÆAB	183	#define GREP_OPT_INIT { \
	184	.relative = 1, \
	185	.pathname = 1, \
	186	.max_depth = -1, \
68437ede	187	.max_count = -1, \
72365bb4 ÆAB	188	.pattern_type_option = GREP_PATTERN_TYPE_UNSPECIFIED, \
	189	.colors = { \
	190	[GREP_COLOR_CONTEXT] = "", \
	191	[GREP_COLOR_FILENAME] = GIT_COLOR_MAGENTA, \
	192	[GREP_COLOR_FUNCTION] = "", \
	193	[GREP_COLOR_LINENO] = GIT_COLOR_GREEN, \
	194	[GREP_COLOR_COLUMNNO] = GIT_COLOR_GREEN, \
	195	[GREP_COLOR_MATCH_CONTEXT] = GIT_COLOR_BOLD_RED, \
	196	[GREP_COLOR_MATCH_SELECTED] = GIT_COLOR_BOLD_RED, \
	197	[GREP_COLOR_SELECTED] = "", \
	198	[GREP_COLOR_SEP] = GIT_COLOR_CYAN, \
	199	}, \
	200	.only_matching = 0, \
	201	.color = -1, \
	202	.output = std_output, \
	203	}
	204
a4e7e317 GC	205	struct config_context;
	206	int grep_config(const char var, const char value,
	207	const struct config_context ctx, void data);
9725c8dd	208	void grep_init(struct grep_opt , struct repository repo);
7687a054	209
55454427 DL	210	void append_grep_pat(struct grep_opt opt, const char pat, size_t patlen, const char *origin, int no, enum grep_pat_token t);
	211	void append_grep_pattern(struct grep_opt opt, const char pat, const char *origin, int no, enum grep_pat_token t);
	212	void append_header_grep_pattern(struct grep_opt , enum grep_header_field, const char );
	213	void compile_grep_patterns(struct grep_opt *opt);
	214	void free_grep_patterns(struct grep_opt *opt);
1e668716	215	int grep_buffer(struct grep_opt opt, const char buf, unsigned long size);
83b5d2f5	216
3f566c4e HM	217	/* The field parameter is only used to filter header patterns
	218	* (where appropriate). If filtering isn't desirable
	219	* GREP_HEADER_FIELD_MAX should be supplied.
	220	*/
	221	int grep_next_match(struct grep_opt *opt,
	222	const char bol, const char eol,
	223	enum grep_context ctx, regmatch_t *pmatch,
	224	enum grep_header_field field, int eflags);
	225
e1327023 JK	226	struct grep_source {
	227	char *name;
	228
	229	enum grep_source_type {
1c41c82b	230	GREP_SOURCE_OID,
e1327023 JK	231	GREP_SOURCE_FILE,
	232	GREP_SOURCE_BUF,
	233	} type;
	234	void *identifier;
0693806b	235	struct repository repo; / if GREP_SOURCE_OID */
e1327023	236
1e668716	237	const char *buf;
e1327023	238	unsigned long size;
94ad9d9e	239
55c61688	240	char path; / for attribute lookups */
94ad9d9e	241	struct userdiff_driver *driver;
e1327023 JK	242	};
e1327023 JK	243
50d92b5f JT	244	void grep_source_init_file(struct grep_source gs, const char name,
	245	const char *path);
	246	void grep_source_init_oid(struct grep_source gs, const char name,
0693806b JT	247	const char path, const struct object_id oid,
0693806b JT	248	struct repository *repo);
e1327023 JK	249	void grep_source_clear_data(struct grep_source *gs);
e1327023 JK	250	void grep_source_clear(struct grep_source *gs);
acd00ea0 NTND	251	void grep_source_load_driver(struct grep_source *gs,
acd00ea0 NTND	252	struct index_state *istate);
07a7d656	253
e1327023 JK	254
	255	int grep_source(struct grep_opt opt, struct grep_source gs);
	256
55454427	257	struct grep_opt grep_opt_dup(const struct grep_opt opt);
5b594f45	258
0579f91d TR	259	/*
	260	* Mutex used around access to the attributes machinery if
	261	* opt->use_threads. Must be initialized/destroyed by callers!
	262	*/
78db6ea9	263	extern int grep_use_locks;
0579f91d	264	extern pthread_mutex_t grep_attr_mutex;
b3aeb285	265
83b5d2f5	266	#endif