libhl/source/regex.c

660 lines
18 KiB
C
Raw Normal View History

2023-08-24 06:24:46 -04:00
/* regex.c
* Copyright 2023 Anon Anonson, Ognjen 'xolatile' Milan Robovic, Emil Williams
* SPDX Identifier: GPL-3.0-only / NO WARRANTY / NO GUARANTEE */
2023-08-23 21:16:55 -04:00
#include "regex.h"
2023-08-23 19:59:59 -04:00
#include <assert.h>
#include <string.h>
2023-08-28 09:42:42 -04:00
#include <limits.h>
2023-08-28 15:58:20 -04:00
#include <stdlib.h>
2023-08-19 18:49:10 -04:00
2023-08-24 09:19:35 -04:00
// ------------------
2023-08-24 09:00:44 -04:00
// ### Char tests ###
2023-08-24 09:19:35 -04:00
// ------------------
2023-08-23 22:11:46 -04:00
static bool is_quantifier(const char c) {
for (const char * s = "+*?="; *s != '\00'; s++) {
2023-08-23 19:59:59 -04:00
if (*s == c) {
2023-08-19 18:49:10 -04:00
return true;
}
}
return false;
}
2023-08-23 22:11:46 -04:00
bool is_magic(const char c) {
if (is_quantifier(c)) {
return true;
}
2023-08-29 10:23:10 -04:00
for (const char * s = "\\[].^"; *s != '\00'; s++) {
2023-08-23 22:11:46 -04:00
if (*s == c) {
return true;
}
}
return false;
}
2023-08-24 09:19:35 -04:00
// ----------------------
2023-08-24 09:00:44 -04:00
// ### Internal Types ###
2023-08-24 09:19:35 -04:00
// ----------------------
2023-08-24 09:00:44 -04:00
typedef struct {
int in;
char input;
int to;
int width;
2023-08-24 09:00:44 -04:00
} delta_t;
typedef struct {
int in;
int to;
2023-08-29 10:23:10 -04:00
int width;
2023-08-24 09:00:44 -04:00
} offshoot_t;
2023-08-24 15:08:31 -04:00
typedef struct {
2023-08-31 18:15:22 -04:00
// XXX:
// These should share a mask
// Not even sure why they are pointers to begin with
2023-08-24 15:08:31 -04:00
bool * do_catch;
bool * is_negative;
2023-08-31 18:15:22 -04:00
bool is_at_the_beginning;
bool do_skip;
2023-08-29 10:23:10 -04:00
// these might be obsolite but im leaving them for now
bool * do_loop_hook;
bool * do_follow_hook;
bool * do_loop_shoot;
bool * do_follow_shoot;
// ---
2023-08-24 15:08:31 -04:00
int * state;
int * width;
char * whitelist;
2023-08-28 09:42:42 -04:00
char * blacklist;
2023-08-24 15:08:31 -04:00
regex_t * regex;
} compiler_state;
2023-08-24 09:00:44 -04:00
2023-08-24 09:19:35 -04:00
// ----------------------------------
2023-08-24 09:00:44 -04:00
// ### Regex creation/destruction ###
2023-08-24 09:19:35 -04:00
// ----------------------------------
2023-08-31 18:15:22 -04:00
#define HALT_AND_CATCH_FIRE INT_MIN
static void HOOK_ALL( int from,
const char * const str,
int to,
compiler_state * cs) {
int hook_to = (to == HALT_AND_CATCH_FIRE) ? HALT_AND_CATCH_FIRE : ((*cs->state) + to);
for (const char * s = str; *s != '\0'; s++) {
delta_t * delta = malloc(sizeof(delta_t));
delta->in = *cs->state + from;
delta->input = *s;
delta->to = hook_to;
delta->width = *cs->width;
vector_push(&cs->regex->delta_table,
&delta);
}
}
static void ABSOLUTE_OFFSHOOT(int from,
int to,
int width,
compiler_state * cs) {
offshoot_t * offshoot = malloc(sizeof(offshoot_t));
offshoot->in = from;
offshoot->to = to;
offshoot->width = width;
vector_push(&cs->regex->catch_table,
&offshoot);
}
static void OFFSHOOT(int from,
int to,
int width,
compiler_state * cs) {
ABSOLUTE_OFFSHOOT(*cs->state + from, *cs->state + to, width, cs);
}
2023-08-28 09:42:42 -04:00
static int escape_1_to_1(const char c, compiler_state * cs) {
char * target_list = (*cs->is_negative) ? cs->blacklist : cs->whitelist;
2023-08-24 11:21:21 -04:00
switch (c) {
2023-08-23 19:59:59 -04:00
case 't': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "\t");
2023-08-23 19:59:59 -04:00
} return 1;
case 'n': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "\n");
2023-08-23 19:59:59 -04:00
} return 1;
case 'r': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "\r");
2023-08-23 19:59:59 -04:00
} return 1;
case 'b': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "\b");
2023-08-23 19:59:59 -04:00
} return 1;
case '[': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "[");
2023-08-23 19:59:59 -04:00
} return 1;
case ']': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "]");
2023-08-23 19:59:59 -04:00
} return 1;
case '.': {
2023-08-28 09:42:42 -04:00
strcat(target_list, ".");
2023-08-23 19:59:59 -04:00
} return 1;
2023-08-29 10:23:10 -04:00
case '^': {
strcat(target_list, "^");
} return 1;
case '=': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "=");
} return 1;
2023-08-23 19:59:59 -04:00
case '?': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "?");
2023-08-23 19:59:59 -04:00
} return 1;
case '+': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "+");
2023-08-23 19:59:59 -04:00
} return 1;
case '*': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "*");
2023-08-23 19:59:59 -04:00
} return 1;
case '\\': {
2023-08-28 09:42:42 -04:00
strcat(target_list, "\\");
2023-08-23 19:59:59 -04:00
} return 1;
}
return 0;
2023-08-19 18:49:10 -04:00
}
2023-08-28 09:42:42 -04:00
static int escape_1_to_N(const char c, compiler_state * cs) {
char * target_list = (*cs->is_negative) ? cs->blacklist : cs->whitelist;
2023-08-23 19:59:59 -04:00
switch(c) {
case 'i': {
2023-08-24 11:21:21 -04:00
const char identifier_chars[] = "@0123456789_"
"\300\301\302\303\304"
2023-08-25 12:51:50 -04:00
"\305\306\307\310\311"
"\312\313\314\315\316"
"\317\320\321\322\323"
"\324\325\326\327\330"
"\331\332\333\334\335"
"\336\337";
2023-08-28 09:42:42 -04:00
strcpy(target_list, identifier_chars);
2023-08-23 19:59:59 -04:00
return sizeof(identifier_chars)-1;
};
case 'I': {
2023-08-24 11:21:21 -04:00
const char identifier_chars[] = "@_"
"\300\301\302\303\304"
2023-08-25 12:51:50 -04:00
"\305\306\307\310\311"
"\312\313\314\315\316"
"\317\320\321\322\323"
"\324\325\326\327\330"
"\331\332\333\334\335"
"\336\337";
2023-08-28 09:42:42 -04:00
strcpy(target_list, identifier_chars);
2023-08-23 19:59:59 -04:00
return sizeof(identifier_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'k': {
2023-08-24 11:21:21 -04:00
const char keyword_chars[] = "@0123456789_"
"\300\301\302\303\304"
2023-08-25 12:51:50 -04:00
"\305\306\307\310\311"
"\312\313\314\315\316"
"\317\320\321\322\323"
"\324\325\326\327\330"
"\331\332\333\334\335"
"\336\337";
2023-08-28 09:42:42 -04:00
strcpy(target_list, keyword_chars);
2023-08-23 19:59:59 -04:00
return sizeof(keyword_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'K': {
2023-08-24 11:21:21 -04:00
const char keyword_chars[] = "@_"
2023-08-25 12:51:50 -04:00
"\300\301\302\303\304"
"\305\306\307\310\311"
"\312\313\314\315\316"
"\317\320\321\322\323"
"\324\325\326\327\330"
"\331\332\333\334\335"
"\336\337";
2023-08-28 09:42:42 -04:00
strcpy(target_list, keyword_chars);
2023-08-23 19:59:59 -04:00
return sizeof(keyword_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'f': {
const char filename_chars[] = "@0123456789/.-_+,#$%~=";
2023-08-28 09:42:42 -04:00
strcpy(target_list, filename_chars);
2023-08-23 21:16:55 -04:00
return sizeof(filename_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'F': {
const char filename_chars[] = "@/.-_+,#$%~=";
2023-08-28 09:42:42 -04:00
strcpy(target_list, filename_chars);
2023-08-23 21:16:55 -04:00
return sizeof(filename_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'p': {
2023-08-24 11:21:21 -04:00
const char printable_chars[] = "@"
2023-08-25 12:51:50 -04:00
"\241\242\243\244\245"
"\246\247\250\251\252"
"\253\254\255\256\257"
"\260\261\262\263\264"
"\265\266\267\270\271"
"\272\273\274\275\276"
"\277"
"\300\301\302\303\304"
"\305\306\307\310\311"
"\312\313\314\315\316"
"\317\320\321\322\323"
"\324\325\326\327\330"
"\331\332\333\334\335"
"\336\337";
2023-08-28 09:42:42 -04:00
strcpy(target_list, printable_chars);
2023-08-23 19:59:59 -04:00
return sizeof(printable_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'P': {
2023-08-24 11:21:21 -04:00
const char printable_chars[] = "@"
2023-08-25 12:51:50 -04:00
"\241\242\243\244\245"
"\246\247\250\251\252"
"\253\254\255\256\257"
"\260\261\262\263\264"
"\265\266\267\270\271"
"\272\273\274\275\276"
"\277"
"\300\301\302\303\304"
"\305\306\307\310\311"
"\312\313\314\315\316"
"\317\320\321\322\323"
"\324\325\326\327\330"
"\331\332\333\334\335"
"\336\337";
2023-08-28 09:42:42 -04:00
strcpy(target_list, printable_chars);
2023-08-23 19:59:59 -04:00
return sizeof(printable_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 's': {
const char whitespace_chars[] = " \t\v\n";
2023-08-28 09:42:42 -04:00
strcpy(target_list, whitespace_chars);
2023-08-23 19:59:59 -04:00
return sizeof(whitespace_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'd': {
const char digit_chars[] = "0123456789";
2023-08-28 09:42:42 -04:00
strcpy(target_list, digit_chars);
2023-08-23 19:59:59 -04:00
return sizeof(digit_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'x': {
2023-08-24 11:21:21 -04:00
const char hex_chars[] = "0123456789"
2023-08-25 12:51:50 -04:00
"abcdef"
"ABCDEF";
2023-08-28 09:42:42 -04:00
strcpy(target_list, hex_chars);
2023-08-23 19:59:59 -04:00
return sizeof(hex_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'o': {
const char oct_chars[] = "01234567";
2023-08-28 09:42:42 -04:00
strcpy(target_list, oct_chars);
2023-08-23 19:59:59 -04:00
return sizeof(oct_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'w': {
2023-08-24 11:21:21 -04:00
const char word_chars[] = "0123456789"
2023-08-25 12:51:50 -04:00
"abcdefghijklmnopqrstuwxyz"
"ABCDEFGHIJKLMNOPQRSTUWXYZ"
"_";
2023-08-28 09:42:42 -04:00
strcpy(target_list, word_chars);
2023-08-23 19:59:59 -04:00
return sizeof(word_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'h': {
2023-08-24 11:21:21 -04:00
const char very_word_chars[] = "abcdefghijklmnopqrstuwxyz"
2023-08-25 12:51:50 -04:00
"ABCDEFGHIJKLMNOPQRSTUWXYZ"
"_";
2023-08-28 09:42:42 -04:00
strcpy(target_list, very_word_chars);
2023-08-23 19:59:59 -04:00
return sizeof(very_word_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'a': {
2023-08-24 11:21:21 -04:00
const char alpha_chars[] = "abcdefghijklmnopqrstuwxyz"
2023-08-25 12:51:50 -04:00
"ABCDEFGHIJKLMNOPQRSTUWXYZ";
2023-08-28 09:42:42 -04:00
strcpy(target_list, alpha_chars);
2023-08-23 19:59:59 -04:00
return sizeof(alpha_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'l': {
const char lower_alpha_chars[] = "abcdefghijklmnopqrstuwxyz";
2023-08-28 09:42:42 -04:00
strcpy(target_list, lower_alpha_chars);
2023-08-23 19:59:59 -04:00
return sizeof(lower_alpha_chars)-1;
2023-08-19 18:49:10 -04:00
};
2023-08-23 19:59:59 -04:00
case 'u': {
const char upper_alpha_chars[] = "ABCDEFGHIJKLMNOPQRSTUWXYZ";
2023-08-28 09:42:42 -04:00
strcpy(target_list, upper_alpha_chars);
2023-08-23 19:59:59 -04:00
return sizeof(upper_alpha_chars)-1;
2023-08-19 18:49:10 -04:00
};
}
2023-08-23 19:59:59 -04:00
return 0;
2023-08-19 18:49:10 -04:00
}
2023-08-28 09:43:25 -04:00
static int escape_to_negative(const char c,
compiler_state * cs) {
switch (c) {
case 'D': {
const char digit_chars[] = "0123456789";
strcpy(cs->blacklist, digit_chars);
*cs->is_negative = true;
return sizeof(digit_chars)-1;
};
}
return 0;
}
2023-08-31 18:15:22 -04:00
static int escape_hologram(const char c, compiler_state * cs) {
switch (c) {
case '<': {
if (cs->is_at_the_beginning) {
ABSOLUTE_OFFSHOOT(0, 2, 0, cs);
cs->do_skip = true;
}
const char very_word_chars[] = "abcdefghijklmnopqrstuwxyz"
"ABCDEFGHIJKLMNOPQRSTUWXYZ"
"_";
*cs->is_negative = true; // effectless currently; should be used to trigger the following lines in the main compile loop
strcat(cs->blacklist, very_word_chars);
HOOK_ALL(0, cs->blacklist, HALT_AND_CATCH_FIRE, cs);
OFFSHOOT(0, 0, 1, cs);
return sizeof(very_word_chars)-1;
};
case '>': {
const char very_word_chars[] = "abcdefghijklmnopqrstuwxyz"
"ABCDEFGHIJKLMNOPQRSTUWXYZ"
"_";
*cs->is_negative = true;
strcat(cs->blacklist, very_word_chars);
return 1;
}
}
return 0;
}
2023-08-28 09:43:25 -04:00
static int compile_dot(compiler_state * cs) {
*cs->do_catch = true;
return true;
}
static int compile_escape(const char c,
compiler_state * cs) {
2023-08-31 18:15:22 -04:00
return escape_1_to_1(c, cs)
|| escape_1_to_N(c, cs)
2023-08-28 09:43:25 -04:00
|| escape_to_negative(c, cs)
2023-08-31 18:15:22 -04:00
|| escape_hologram(c, cs)
2023-08-28 09:43:25 -04:00
;
}
static int compile_range(const char * const range,
compiler_state * cs) {
assert((range[0] == '[') && "Not a range.");
2023-08-19 18:49:10 -04:00
2023-08-23 19:59:59 -04:00
const char * s;
if (range[1] == '^') {
2023-08-28 09:43:25 -04:00
*cs->is_negative = true;
s = range + 2;
} else {
s = range + 1;
}
2023-08-31 18:15:22 -04:00
char * target_list = (*cs->is_negative) ? cs->blacklist : cs->whitelist;
for (; *s != ']'; s++) {
assert((*s != '\0') && "Unclosed range.");
2023-08-23 19:59:59 -04:00
char c = *s;
2023-08-28 09:42:42 -04:00
if (c == '\\') {
s += 1;
assert(compile_escape(*s, cs) && "Unknown escape.");
2023-08-23 19:59:59 -04:00
} else if (*(s+1) == '-') {
char end = *(s+2);
assert((c < end) && "Endless range.");
2023-08-23 19:59:59 -04:00
for (char cc = c; cc < end+1; cc++) {
2023-08-28 09:42:42 -04:00
strncat(target_list, &cc, 1);
strncat(target_list, "\0", 1);
2023-08-19 18:49:10 -04:00
}
2023-08-23 19:59:59 -04:00
s += 2;
} else {
2023-08-28 09:42:42 -04:00
strncat(target_list, &c, 1);
2023-08-23 19:59:59 -04:00
}
}
2023-08-19 18:49:10 -04:00
2023-08-23 19:59:59 -04:00
return ((s - range) + 1);
}
2023-08-19 18:49:10 -04:00
2023-08-28 18:19:40 -04:00
void filter_blacklist(const char * whitelist,
const char * blacklist,
char * filtered) {
for (; *blacklist != '\0'; blacklist++) {
for(; *whitelist != '\0'; whitelist++) {
if (*blacklist == *whitelist) {
2023-08-28 09:42:42 -04:00
goto long_continue;
}
}
2023-08-28 18:19:40 -04:00
strncat(filtered, blacklist, 1);
2023-08-31 18:15:22 -04:00
long_continue:
;
2023-08-24 15:08:31 -04:00
}
}
2023-08-23 19:59:59 -04:00
regex_t * regex_compile(const char * const pattern) {
2023-08-23 21:16:55 -04:00
regex_t * regex = (regex_t *)malloc(sizeof(regex_t));
2023-08-23 19:59:59 -04:00
regex->str = strdup(pattern);
vector_init(&regex->delta_table, sizeof(delta_t*), 0UL);
vector_init(&regex->catch_table, sizeof(offshoot_t*), 0UL);
2023-08-19 18:49:10 -04:00
2023-08-29 10:23:10 -04:00
int state = 2;
2023-08-19 18:49:10 -04:00
2023-08-31 18:15:22 -04:00
// this is plain retarded
2023-08-23 19:59:59 -04:00
bool do_catch;
bool is_negative;
2023-08-29 10:23:10 -04:00
bool do_loop_hook;
bool do_follow_hook;
bool do_loop_shoot;
bool do_follow_shoot;
int width;
2023-08-24 15:08:31 -04:00
char whitelist[64];
2023-08-28 09:42:42 -04:00
char blacklist[64];
2023-08-24 15:08:31 -04:00
compiler_state cs = {
2023-08-31 18:15:22 -04:00
.do_catch = &do_catch,
.is_negative = &is_negative,
.is_at_the_beginning = true,
.do_skip = false,
.state = &state,
.width = &width,
.whitelist = whitelist,
.blacklist = blacklist,
.regex = regex,
2023-08-24 15:08:31 -04:00
};
2023-08-23 19:59:59 -04:00
for (const char * s = pattern; *s != '\00';) {
2023-08-28 09:42:42 -04:00
// Reset the compiler
2023-08-23 19:59:59 -04:00
assert(!is_quantifier(*pattern) && "Pattern starts with quantifier.");
2023-08-29 10:23:10 -04:00
whitelist[0] = '\0';
blacklist[0] = '\0';
do_catch = false;
is_negative = false;
2023-08-31 18:15:22 -04:00
cs.do_skip = false;
/**/
2023-08-29 10:23:10 -04:00
do_loop_hook = false;
do_follow_hook = false;
do_loop_shoot = false;
do_follow_shoot = false;
2023-08-31 18:15:22 -04:00
/**/
width = 1;
2023-08-28 09:42:42 -04:00
// Translate char
2023-08-23 19:59:59 -04:00
switch (*s) {
2023-08-29 10:23:10 -04:00
case '^': {
2023-08-31 18:15:22 -04:00
if (cs.is_at_the_beginning) {
2023-08-29 10:23:10 -04:00
ABSOLUTE_OFFSHOOT(0, 2, 0, &cs);
ABSOLUTE_OFFSHOOT(1, HALT_AND_CATCH_FIRE, 0, &cs);
}
whitelist[0] = '\n';
whitelist[1] = '\0';
HOOK_ALL(0, whitelist, 0, &cs);
if (s != pattern) {
state += 1;
}
2023-08-31 18:15:22 -04:00
cs.do_skip = true;
2023-08-29 10:23:10 -04:00
} break;
2023-08-23 19:59:59 -04:00
case '.': {
2023-08-28 09:42:42 -04:00
compile_dot(&cs);
2023-08-23 19:59:59 -04:00
} break;
case '\\': {
2023-08-28 09:42:42 -04:00
s += 1;
assert(compile_escape(*s, &cs) && "Unknown escape.");
2023-08-23 19:59:59 -04:00
} break;
case '[': {
2023-08-28 09:42:42 -04:00
s += compile_range(s, &cs) - 1;
2023-08-23 19:59:59 -04:00
} break;
default: {
2023-08-29 10:23:10 -04:00
whitelist[0] = *s;
whitelist[1] = '\0';
2023-08-23 19:59:59 -04:00
} break;
}
2023-08-31 18:15:22 -04:00
2023-08-28 09:42:42 -04:00
s += 1;
2023-08-19 18:49:10 -04:00
2023-08-31 18:15:22 -04:00
if (cs.do_skip) {
goto long_continue;
}
2023-08-28 09:42:42 -04:00
// Compile with quantifier
2023-08-23 19:59:59 -04:00
switch (*s) {
case '=':
2023-08-23 19:59:59 -04:00
case '?': {
2023-08-29 10:23:10 -04:00
do_loop_hook = true;
2023-08-24 15:08:31 -04:00
HOOK_ALL(0, whitelist, +1, &cs);
2023-08-28 09:42:42 -04:00
if (do_catch || is_negative) {
2023-08-29 10:23:10 -04:00
OFFSHOOT(0, +1, 1, &cs);
2023-08-28 09:42:42 -04:00
}
s += 1;
2023-08-23 19:59:59 -04:00
} break;
case '*': {
2023-08-24 15:08:31 -04:00
HOOK_ALL(0, whitelist, 0, &cs);
2023-08-28 09:42:42 -04:00
if (do_catch) {
2023-08-29 10:23:10 -04:00
OFFSHOOT(0, +1, 1, &cs);
2023-08-28 09:42:42 -04:00
} else if (is_negative) {
2023-08-29 10:23:10 -04:00
OFFSHOOT(0, 0, 1, &cs);
2023-08-28 09:42:42 -04:00
}
s += 1;
2023-08-23 19:59:59 -04:00
} break;
case '+': {
2023-08-24 15:08:31 -04:00
HOOK_ALL(0, whitelist, +1, &cs);
2023-08-28 09:42:42 -04:00
if (do_catch || is_negative) {
2023-08-29 10:23:10 -04:00
OFFSHOOT(0, +1, 1, &cs);
2023-08-28 09:42:42 -04:00
}
2023-08-23 19:59:59 -04:00
state += 1;
2023-08-24 15:08:31 -04:00
HOOK_ALL(0, whitelist, 0, &cs);
2023-08-28 09:42:42 -04:00
if (do_catch || is_negative) {
2023-08-29 10:23:10 -04:00
OFFSHOOT(0, 0, 1, &cs);
2023-08-28 09:42:42 -04:00
}
s += 1;
2023-08-23 19:59:59 -04:00
} break;
default: { // Literal
2023-08-24 15:08:31 -04:00
HOOK_ALL(0, whitelist, +1, &cs);
2023-08-28 09:42:42 -04:00
if (do_catch || is_negative) {
2023-08-29 10:23:10 -04:00
OFFSHOOT(0, +1, 1, &cs);
2023-08-28 09:42:42 -04:00
}
2023-08-23 19:59:59 -04:00
state += 1;
} break;
2023-08-19 18:49:10 -04:00
}
2023-08-28 09:42:42 -04:00
// Compile blacklist
if (*blacklist) {
char filtered_blacklist[64];
filtered_blacklist[0] = '\0';
filter_blacklist(whitelist, blacklist, filtered_blacklist);
HOOK_ALL(0, filtered_blacklist, HALT_AND_CATCH_FIRE, &cs);
}
2023-08-31 18:15:22 -04:00
2023-08-29 10:23:10 -04:00
long_continue:
2023-08-31 18:15:22 -04:00
cs.is_at_the_beginning = false;
2023-08-29 13:09:55 -04:00
long_continue:;
2023-08-23 19:59:59 -04:00
}
2023-08-19 18:49:10 -04:00
2023-08-23 19:59:59 -04:00
regex->accepting_state = state;
2023-08-23 21:16:55 -04:00
return regex;
2023-08-23 19:59:59 -04:00
}
int regex_free(regex_t * const regex) {
free(regex->str);
2023-08-23 21:16:55 -04:00
vector_free(&regex->delta_table);
vector_free(&regex->catch_table);
free(regex);
2023-08-23 19:59:59 -04:00
return 0;
}
2023-08-24 09:19:35 -04:00
// -----------------
2023-08-24 09:00:44 -04:00
// ### Searching ###
2023-08-24 09:19:35 -04:00
// -----------------
2023-08-29 10:23:10 -04:00
static int catch_(const regex_t * const regex,
2023-08-25 19:16:05 -04:00
int * const state) {
for (size_t i = 0; i < regex->catch_table.element_count; i++){
const offshoot_t * const offshoot = *(offshoot_t**)vector_get(&regex->catch_table, i);
if (offshoot->in == *state) {
*state = offshoot->to;
2023-08-29 10:23:10 -04:00
return offshoot->width;
2023-08-25 19:16:05 -04:00
}
}
2023-08-29 10:23:10 -04:00
return HALT_AND_CATCH_FIRE;
2023-08-25 19:16:05 -04:00
}
2023-08-29 10:23:10 -04:00
static int regex_assert(const regex_t * const regex,
const char * const string,
const int string_offset,
int state,
int width) { // XXX: im pretty sure this is actually redundant and the width should be calculated from string - s
for (const char * s = (string + string_offset); *s != '\00';) {
2023-08-31 18:15:22 -04:00
// XXX: this should be a jump search for the instate and then a linear
2023-08-23 19:59:59 -04:00
// delta
2023-08-31 18:15:22 -04:00
//int left = 0;
//int right = regex->delta_table.element_count - 1;
//int i;
//while(left <= right) }
for (size_t i = 0; i < regex->delta_table.element_count; i++) {
2023-08-31 18:15:22 -04:00
//i = (left + right) / 2;
2023-08-25 15:06:26 -04:00
const delta_t * const delta = *(delta_t**)vector_get(&regex->delta_table, i);
2023-08-23 21:16:55 -04:00
if ((delta->in == state)
2023-08-23 19:59:59 -04:00
&& (delta->input == *s)) {
2023-08-29 10:23:10 -04:00
int r = regex_assert(regex, string, (s - string) + delta->width, delta->to, width + 1);
2023-08-26 16:39:07 -04:00
if(r){
return r;
2023-08-23 19:59:59 -04:00
}
}
}
2023-08-29 10:23:10 -04:00
const int catch_width = catch_(regex, &state);
if ((catch_width != HALT_AND_CATCH_FIRE)
&& (state != HALT_AND_CATCH_FIRE)) {
s += catch_width;
2023-08-23 19:59:59 -04:00
continue;
}
2023-08-31 18:15:22 -04:00
// XXX: the extra catch might not be necessary if we were to compile to a simpler form
catch_(regex, &state);
2023-08-26 16:39:07 -04:00
return (state == regex->accepting_state) ? width : false;
2023-08-23 19:59:59 -04:00
}
2023-08-25 19:16:05 -04:00
return false;
2023-08-23 19:59:59 -04:00
}
2023-08-29 10:23:10 -04:00
int regex_match( regex_t * regex,
const char * const string,
const bool is_start_of_string,
const int string_offset) { // XXX: remove this useless piece of shit of a parameter nigger
2023-08-23 19:59:59 -04:00
if (regex == NULL) {
return false;
}
if (string == NULL) {
return true;
2023-08-21 14:07:39 -04:00
}
2023-08-23 19:59:59 -04:00
2023-08-29 10:23:10 -04:00
const int initial_state = (int)(!is_start_of_string);
return regex_assert(regex, string, string_offset, initial_state, 0);
2023-08-19 18:49:10 -04:00
}
2023-08-25 19:16:05 -04:00
bool regex_search( regex_t * regex,
const char * const string) {
2023-08-29 10:23:10 -04:00
return (bool)regex_match(regex, string, true, 0);
2023-08-25 19:16:05 -04:00
}