/* -*- Mode: C; tab-width: 8; indent-tabs-mode: nil; c-basic-offset: 4; coding: utf-8 -*- * * moolang-rules.c * * Copyright (C) 2004-2005 by Yevgen Muntyan * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or * (at your option) any later version. * * See COPYING file that comes with this distribution. */ #define MOOEDIT_COMPILATION #include "mooedit/moolang-rules.h" #include "mooedit/moolang-aux.h" typedef MooRuleMatchFlags MatchFlags; #define MATCH_START_ONLY MOO_RULE_MATCH_START_ONLY #define MooRuleString MooRuleAsciiString #define MooRuleChar MooRuleAsciiChar #define MooRule2Char MooRuleAscii2Char #define MooRuleAnyChar MooRuleAsciiAnyChar typedef MooRule* (*MatchFunc) (MooRule *self, MooRuleMatchData *data, MooRuleMatchResult *result, MooRuleMatchFlags flags); typedef void (*DestroyFunc) (MooRule *self); static MooRule *rule_new (MooRuleFlags flags, const char *style, MatchFunc match_func, DestroyFunc destroy_func); static void child_rules_match (MooRuleArray *array, MatchData *data, MatchResult *result); static MooRule *rules_match_real (MooRuleArray *array, MatchData *data, MatchResult *result, MatchFlags flags); void moo_match_data_init (MatchData *data, int line_number, const GtkTextIter *line_start, const GtkTextIter *line_end) { GtkTextBuffer *buffer; g_assert (data != NULL); g_assert (line_start && gtk_text_iter_starts_line (line_start)); g_assert (line_number == gtk_text_iter_get_line (line_start)); g_assert (!line_end || gtk_text_iter_ends_line (line_end)); data->line_start = *line_start; data->line_number = line_number; if (line_end) { data->line_end = *line_end; } else { data->line_end = *line_start; if (!gtk_text_iter_ends_line (&data->line_end)) gtk_text_iter_forward_to_line_end (&data->line_end); } buffer = gtk_text_iter_get_buffer (line_start); data->line_string = gtk_text_buffer_get_slice (buffer, line_start, &data->line_end, TRUE); data->line_string_len = strlen (data->line_string); data->start_iter = *line_start; data->start = data->line_string; data->start_offset = 0; } void moo_match_data_set_start (MatchData *data, const GtkTextIter *start_iter, char *start, int start_offset) { g_assert (data != NULL); g_assert (start != NULL); g_assert (start_offset >= 0); data->start = start; data->start_offset = start_offset; if (start_iter) { data->start_iter = *start_iter; } else { data->start_iter = data->line_start; gtk_text_iter_forward_chars (&data->start_iter, start_offset); } } void moo_match_data_destroy (MatchData *data) { if (data->line_string) g_free (data->line_string); } static MooRule* rules_match_real (MooRuleArray *array, MatchData *data, MatchResult *result, MatchFlags flags) { guint i; MooRule *matched = NULL; MatchResult tmp; g_assert (array != NULL); if (!array->len) return NULL; if (flags & MATCH_START_ONLY) { data->limit = data->start; data->limit_offset = 0; } else { data->limit = NULL; data->limit_offset = SIZE_NOT_SET; } for (i = 0; i < array->len; ++i) { MooRule *rule = array->data[i]; MooRule *matched_here = NULL; if (!(flags & MATCH_START_ONLY)) { /* TODO: first-non-blank */ if ((rule->flags & MOO_RULE_MATCH_FIRST_CHAR) && data->start != data->line_string) continue; } if ((rule->flags & MOO_RULE_MATCH_FIRST_LINE) && data->line_number != 0) continue; matched_here = rule->match (rule, data, &tmp, flags); if (matched_here) { if (!(flags & MATCH_START_ONLY)) { if (!matched || tmp.match_start < result->match_start) { matched = matched_here ? matched_here : rule; *result = tmp; data->limit = tmp.match_start; data->limit_offset = tmp.match_offset; } if (tmp.match_start == data->start) break; } else { matched = matched_here ? matched_here : rule; *result = tmp; break; } } } if (matched && matched->child_rules) child_rules_match (matched->child_rules, data, result); return matched; } static void child_rules_match (MooRuleArray *array, MatchData *data, MatchResult *result) { MatchResult tmp; MooRule *matched; char *saved_start; g_assert (array != NULL); g_assert (result->match_start != NULL); g_assert (result->match_end != NULL); g_assert (result->match_start <= result->match_end); tmp = *result; saved_start = data->start; data->start = result->match_end; matched = rules_match_real (array, data, &tmp, MATCH_START_ONLY); if (matched) { g_return_if_fail (tmp.match_start == result->match_end); result->match_end = tmp.match_end; if (tmp.match_len >= 0) { if (result->match_len >= 0) result->match_len += tmp.match_len; } else { result->match_len = -1; } } data->start = saved_start; } MooRule* moo_rule_array_match (MooRuleArray *array, MatchData *data, MatchResult *result) { return rules_match_real (array, data, result, 0); } static MooRule* rule_new (MooRuleFlags flags, const char *style, MatchFunc match_func, DestroyFunc destroy_func) { MooRule *rule; g_return_val_if_fail (match_func != NULL, NULL); rule = g_new0 (MooRule, 1); rule->match = match_func; rule->destroy = destroy_func; rule->flags = flags; rule->style = g_strdup (style); return rule; } void moo_rule_free (MooRule *rule) { guint i; if (!rule) return; if (rule->destroy) rule->destroy (rule); if (rule->child_rules) { for (i = 0; i < rule->child_rules->len; ++i) moo_rule_free (rule->child_rules->data[i]); g_ptr_array_free ((GPtrArray*) rule->child_rules, TRUE); } g_free (rule->style); g_free (rule); } void moo_rule_add_child_rule (MooRule *rule, MooRule *child_rule) { g_return_if_fail (rule != NULL && child_rule != NULL); if (!rule->child_rules) rule->child_rules = (MooRuleArray*) g_ptr_array_new (); g_ptr_array_add ((GPtrArray*) rule->child_rules, child_rule); } void moo_rule_set_end_stay (MooRule *rule) { g_return_if_fail (rule != NULL); rule->exit.type = MOO_CONTEXT_STAY; rule->exit.num = 0; } void moo_rule_set_end_pop (MooRule *rule, guint num) { g_return_if_fail (rule != NULL && num != 0); rule->exit.type = MOO_CONTEXT_POP; rule->exit.num = num; } void moo_rule_set_end_switch (MooRule *rule, MooContext *target) { g_return_if_fail (rule != NULL && target != 0); rule->exit.type = MOO_CONTEXT_SWITCH; rule->exit.ctx = target; } /*************************************************************************/ /* String match */ static MooRule* rule_string_match (MooRule *rule, MatchData *data, MatchResult *result, MatchFlags flags) { /* TODO: limit */ result->match_start = NULL; if (rule->str.caseless) { if (flags & MATCH_START_ONLY) { if (!g_ascii_strncasecmp (data->start, rule->str.string, rule->str.length)) result->match_start = data->start; } else { result->match_start = ascii_casestrstr (data->start, rule->str.string); } } else { if (flags & MATCH_START_ONLY) { if (!strncmp (data->start, rule->str.string, rule->str.length)) result->match_start = data->start; } else { result->match_start = strstr (data->start, rule->str.string); } } if (!result->match_start) return NULL; result->match_end = result->match_start + rule->str.length; result->match_len = rule->str.length; result->match_offset = -1; return rule; } static void rule_string_destroy (MooRule *rule) { g_free (rule->str.string); } MooRule* moo_rule_string_new (const char *string, MooRuleFlags flags, const char *style) { MooRule *rule; guint length; g_return_val_if_fail (string && string[0], NULL); g_return_val_if_fail (g_utf8_validate (string, -1, NULL), NULL); g_return_val_if_fail (string_is_ascii (string), NULL); length = strlen (string); g_return_val_if_fail (length != 0, NULL); rule = rule_new (flags, style, rule_string_match, rule_string_destroy); g_return_val_if_fail (rule != NULL, NULL); rule->str.caseless = (flags & MOO_RULE_MATCH_CASELESS) ? TRUE : FALSE; if (rule->str.caseless) rule->str.string = g_ascii_strdown (string, -1); else rule->str.string = g_strdup (string); rule->str.length = length; return rule; } /*************************************************************************/ /* Regex match */ static MooRule* rule_regex_match (MooRule *rule, MatchData *data, MatchResult *result, MatchFlags flags) { /* TODO: limit */ /* XXX line start and stuff */ int n_matches, start_pos, end_pos; EggRegexMatchFlags regex_flags = 0; egg_regex_clear (rule->regex.regex); if (flags & MATCH_START_ONLY) regex_flags |= EGG_REGEX_MATCH_ANCHORED; n_matches = egg_regex_match_extended (rule->regex.regex, data->line_string, data->line_string_len, data->start - data->line_string, regex_flags); if (n_matches <= 0) return NULL; egg_regex_fetch_pos (rule->regex.regex, data->line_string, 0, &start_pos, &end_pos); result->match_start = data->line_string + start_pos; result->match_end = data->line_string + end_pos; result->match_len = -1; result->match_offset = -1; return rule; } static void rule_regex_destroy (MooRule *rule) { egg_regex_free (rule->regex.regex); } MooRule* moo_rule_regex_new (const char *pattern, gboolean non_empty, EggRegexCompileFlags regex_compile_options, EggRegexMatchFlags regex_match_options, MooRuleFlags flags, const char *style) { MooRule *rule; EggRegex *regex; GError *error = NULL; g_return_val_if_fail (pattern && pattern[0], NULL); if (flags & MOO_RULE_MATCH_CASELESS) regex_compile_options |= EGG_REGEX_CASELESS; if (non_empty) regex_match_options |= EGG_REGEX_MATCH_NOTEMPTY; regex = egg_regex_new (pattern, regex_compile_options, regex_match_options, &error); if (error) { g_warning ("could not compile pattern '%s': %s", pattern, error->message); g_error_free (error); egg_regex_unref (regex); return NULL; } egg_regex_optimize (regex, &error); if (error) { g_warning ("egg_regex_optimize() failed: %s", error->message); g_error_free (error); } if (pattern[0] == '^') flags |= MOO_RULE_MATCH_FIRST_CHAR; rule = rule_new (flags, style, rule_regex_match, rule_regex_destroy); if (!rule) { egg_regex_free (regex); return NULL; } rule->regex.regex = regex; return rule; } /*************************************************************************/ /* Char match */ static MooRule* rule_char_match (MooRule *rule, MatchData *data, MatchResult *result, MatchFlags flags) { result->match_start = NULL; if (flags & MATCH_START_ONLY) { if (rule->_char.caseless) { if (data->start[0] == rule->_char.ch) result->match_start = data->start; } else { if (g_ascii_tolower (data->start[0]) == rule->_char.ch) result->match_start = data->start; } } else { if (rule->_char.caseless) result->match_start = ascii_lower_strchr (data->start, rule->_char.ch); else result->match_start = strchr (data->start, rule->_char.ch); } if (!result->match_start) return NULL; result->match_end = result->match_start + 1; result->match_len = 1; result->match_offset = -1; return rule; } static MooRule* rule_2char_match (MooRule *rule, MatchData *data, MatchResult *result, MatchFlags flags) { result->match_start = NULL; if (flags & MATCH_START_ONLY) { if (data->start[0] == rule->_2char.str[0] && data->start[1] == rule->_2char.str[1]) result->match_start = data->start; } else { result->match_start = strstr (data->start, rule->_2char.str); } if (!result->match_start) return NULL; result->match_end = result->match_start + 2; result->match_len = 2; result->match_offset = -1; return rule; } MooRule* moo_rule_char_new (char ch, MooRuleFlags flags, const char *style) { MooRule *rule; g_return_val_if_fail (ch && CHAR_IS_ASCII (ch), NULL); rule = rule_new (flags, style, rule_char_match, NULL); g_return_val_if_fail (rule != NULL, NULL); if (flags & MOO_RULE_MATCH_CASELESS) { rule->_char.ch = g_ascii_tolower (ch); rule->_char.caseless = TRUE; } else { rule->_char.ch = ch; } return rule; } MooRule* moo_rule_2char_new (char ch1, char ch2, MooRuleFlags flags, const char *style) { MooRule *rule; g_return_val_if_fail (ch1 && CHAR_IS_ASCII (ch1), NULL); g_return_val_if_fail (ch2 && CHAR_IS_ASCII (ch2), NULL); rule = rule_new (flags, style, rule_2char_match, NULL); g_return_val_if_fail (rule != NULL, NULL); if (flags & MOO_RULE_MATCH_CASELESS) { ch1 = g_ascii_tolower (ch1); ch2 = g_ascii_tolower (ch2); } rule->_2char.str[0] = ch1; rule->_2char.str[1] = ch2; rule->_2char.str[2] = 0; return rule; } /*************************************************************************/ /* AnyChar match */ static MooRule* rule_any_char_match (MooRule *rule, MatchData *data, MatchResult *result, MatchFlags flags) { guint i; result->match_start = NULL; if (flags & MATCH_START_ONLY) { for (i = 0; i < rule->anychar.n_chars; ++i) { if (data->start[0] == rule->anychar.chars[i]) { result->match_start = data->start; break; } } } else { for (i = 0; i < rule->anychar.n_chars; ++i) { if (!result->match_start) { result->match_start = strchr (data->start, rule->anychar.chars[i]); } else if (result->match_start == data->start + 1) { if (data->start[0] == rule->anychar.chars[i]) { result->match_start = data->start; break; } } else { char *tmp = strchr (data->start, rule->anychar.chars[i]); if (tmp < result->match_start) result->match_start = tmp; } if (result->match_start == data->start) break; } } if (!result->match_start) return NULL; result->match_end = result->match_start + 1; result->match_len = 1; result->match_offset = -1; return rule; } static void rule_any_char_destroy (MooRule *rule) { g_free (rule->anychar.chars); } MooRule* moo_rule_any_char_new (const char *string, MooRuleFlags flags, const char *style) { MooRule *rule; guint i, len; g_return_val_if_fail (string && string[0], NULL); len = strlen (string); for (i = 0; i < len; ++i) g_return_val_if_fail (CHAR_IS_ASCII (string[i]), NULL); rule = rule_new (flags, style, rule_any_char_match, rule_any_char_destroy); g_return_val_if_fail (rule != NULL, NULL); rule->anychar.n_chars = len; rule->anychar.chars = g_strdup (string); return rule; } /*************************************************************************/ /* Keywords */ MooRule* moo_rule_keywords_new (GSList *words, MooRuleFlags flags, const char *style) { GSList *l; GString *pattern; MooRule *rule; g_return_val_if_fail (words != NULL, NULL); for (l = words; l != NULL; l = l->next) { char *word = l->data; g_return_val_if_fail (word && word[0], NULL); g_return_val_if_fail (g_utf8_validate (word, -1, NULL), NULL); } pattern = g_string_new ("\\b("); for (l = words; l != NULL; l = l->next) { if (l != words) g_string_append_c (pattern, '|'); g_string_append (pattern, l->data); } g_string_append (pattern, ")\\b"); rule = moo_rule_regex_new (pattern->str, TRUE, 0, 0, flags, style); g_string_free (pattern, TRUE); return rule; } /*************************************************************************/ /* IncludeRules */ static MooRule* rule_include_match (MooRule *rule, MatchData *data, MatchResult *result, MatchFlags flags) { return rules_match_real (rule->incl.ctx->rules, data, result, flags); } MooRule* moo_rule_include_new (MooContext *ctx) { MooRule *rule; g_return_val_if_fail (ctx != NULL, NULL); rule = rule_new (0, NULL, rule_include_match, NULL); g_return_val_if_fail (rule != NULL, NULL); rule->incl.ctx = ctx; return rule; } /*************************************************************************/ /* Special sequences */ #if 0 #define ISDIGIT(c__) (c__ >= '0' && c__ <= '9') static MooRule* rule_int_match (MooRule *rule, MatchData *data, MatchResult *result, MatchFlags flags) { if (flags & MATCH_START_ONLY) { if (ISDIGIT(data->start[0])) { guint i; for (i = 1; ISDIGIT(data->start[i]); ++i) ; result->match_start = data->start; result->match_end = result->match_start + i; result->match_len = i; result->match_offset = -1; return rule; } else { return NULL; } } else { guint i; for (i = 0; data->start[i] && !ISDIGIT(data->start[i]); ++i) ; if (!data->start[i]) return NULL; result->match_start = data->start + i; for ( ; ISDIGIT(data->start[i]); ++i) ; result->match_end = result->match_start + i; result->match_len = result->match_end - result->match_start; result->match_offset = -1; return rule; } } #endif #define PATTERN_INT "[0-9]*" #define PATTERN_FLOAT "[0-9]*\\.[0-9]*" #define PATTERN_OCTAL "0[0-7]+" #define PATTERN_HEX "0x[0-9A-Fa-f]+" #define PATTERN_ESC_CHAR "\\\\([abefnrtv\"'?\\\\]|0[0-7]*|x[0-9A-Fa-f])" #define PATTERN_C_CHAR "'" PATTERN_ESC_CHAR "'" #define PATTERN_IDENTIFIER "[a-zA-Z_][a-zA-Z0-9_]*" #define PATTERN_WHITESPACE "\\s+" MooRule* moo_rule_int_new (MooRuleFlags flags, const char *style) { return moo_rule_regex_new (PATTERN_INT, TRUE, 0, 0, flags, style); } MooRule* moo_rule_float_new (MooRuleFlags flags, const char *style) { return moo_rule_regex_new (PATTERN_FLOAT, TRUE, 0, 0, flags, style); } MooRule* moo_rule_octal_new (MooRuleFlags flags, const char *style) { return moo_rule_regex_new (PATTERN_OCTAL, TRUE, 0, 0, flags, style); } MooRule* moo_rule_hex_new (MooRuleFlags flags, const char *style) { return moo_rule_regex_new (PATTERN_HEX, TRUE, 0, 0, flags, style); } MooRule* moo_rule_escaped_char_new (MooRuleFlags flags, const char *style) { return moo_rule_regex_new (PATTERN_ESC_CHAR, TRUE, 0, 0, flags, style); } MooRule* moo_rule_c_char_new (MooRuleFlags flags, const char *style) { return moo_rule_regex_new (PATTERN_C_CHAR, TRUE, 0, 0, flags, style); } MooRule* moo_rule_whitespace_new (MooRuleFlags flags, const char *style) { return moo_rule_regex_new (PATTERN_WHITESPACE, TRUE, 0, 0, flags, style); } MooRule* moo_rule_identifier_new (MooRuleFlags flags, const char *style) { return moo_rule_regex_new (PATTERN_IDENTIFIER, TRUE, 0, 0, flags, style); } MooRule* moo_rule_line_continue_new (MooRuleFlags flags, const char *style) { MooRule *rule = moo_rule_regex_new ("\\\\$", TRUE, 0, 0, flags, style); g_return_val_if_fail (rule != NULL, NULL); rule->include_eol = TRUE; return rule; }