2 This file is part of GNUnet
3 Copyright (C) 2012 GNUnet e.V.
5 GNUnet is free software: you can redistribute it and/or modify it
6 under the terms of the GNU Affero General Public License as published
7 by the Free Software Foundation, either version 3 of the License,
8 or (at your option) any later version.
10 GNUnet is distributed in the hope that it will be useful, but
11 WITHOUT ANY WARRANTY; without even the implied warranty of
12 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 Affero General Public License for more details.
15 You should have received a copy of the GNU Affero General Public License
16 along with this program. If not, see <http://www.gnu.org/licenses/>.
18 SPDX-License-Identifier: AGPL3.0-or-later
21 * @file src/regex/regex_internal.h
22 * @brief common internal definitions for regex library.
23 * @author Maximilian Szengel
25 #ifndef REGEX_INTERNAL_H
26 #define REGEX_INTERNAL_H
28 #include "regex_internal_lib.h"
33 #if 0 /* keep Emacsens' auto-indent happy */
39 * char array of literals that are allowed inside a regex (apart from the
42 #define ALLOWED_LITERALS \
43 "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"
47 * Transition between two states. Transitions are stored at the states from
48 * which they origin ('from_state'). Each state can have 0-n transitions.
49 * If label is NULL, this is considered to be an epsilon transition.
51 struct REGEX_INTERNAL_Transition
54 * This is a linked list.
56 struct REGEX_INTERNAL_Transition *prev;
59 * This is a linked list.
61 struct REGEX_INTERNAL_Transition *next;
64 * Unique id of this transition.
69 * Label for this transition. This is basically the edge label for the graph.
74 * State to which this transition leads.
76 struct REGEX_INTERNAL_State *to_state;
79 * State from which this transition origins.
81 struct REGEX_INTERNAL_State *from_state;
86 * A state. Can be used in DFA and NFA automatons.
88 struct REGEX_INTERNAL_State;
94 struct REGEX_INTERNAL_StateSet
99 struct REGEX_INTERNAL_State **states;
102 * Number of entries in *use* in the 'states' array.
107 * Length of the 'states' array.
114 * A state. Can be used in DFA and NFA automatons.
116 struct REGEX_INTERNAL_State
119 * This is a linked list to keep states in an automaton.
121 struct REGEX_INTERNAL_State *prev;
124 * This is a linked list to keep states in an automaton.
126 struct REGEX_INTERNAL_State *next;
129 * This is a multi DLL for StateSet_MDLL.
131 struct REGEX_INTERNAL_State *prev_SS;
134 * This is a multi DLL for StateSet_MDLL.
136 struct REGEX_INTERNAL_State *next_SS;
139 * This is a multi DLL for StateSet_MDLL Stack.
141 struct REGEX_INTERNAL_State *prev_ST;
144 * This is a multi DLL for StateSet_MDLL Stack.
146 struct REGEX_INTERNAL_State *next_ST;
154 * Unique state id that is used for traversing the automaton. It is guaranteed
155 * to be > 0 and < state_count.
157 unsigned int traversal_id;
160 * If this is an accepting state or not.
165 * Marking of the state. This is used for marking all visited states when
166 * traversing all states of an automaton and for cases where the state id
167 * cannot be used (dfa minimization).
172 * Marking the state as contained. This is used for checking, if the state is
173 * contained in a set in constant time.
178 * Marking the state as part of an SCC (Strongly Connected Component). All
179 * states with the same scc_id are part of the same SCC. scc_id is 0, if state
180 * is not a part of any SCC.
185 * Used for SCC detection.
190 * Used for SCC detection.
195 * Human readable name of the state. Used for debugging and graph
203 struct GNUNET_HashCode hash;
206 * Linear state ID accquired by depth-first-search. This ID should be used for
207 * storing information about the state in an array, because the 'id' of the
208 * state is not guaranteed to be linear. The 'dfs_id' is guaranteed to be > 0
209 * and < 'state_count'.
214 * Proof for this state.
219 * Number of transitions from this state to other states.
221 unsigned int transition_count;
224 * DLL of transitions.
226 struct REGEX_INTERNAL_Transition *transitions_head;
229 * DLL of transitions.
231 struct REGEX_INTERNAL_Transition *transitions_tail;
234 * Number of incoming transitions. Used for compressing DFA paths.
236 unsigned int incoming_transition_count;
239 * Set of states on which this state is based on. Used when creating a DFA out
240 * of several NFA states.
242 struct REGEX_INTERNAL_StateSet nfa_set;
247 * Type of an automaton.
249 enum REGEX_INTERNAL_AutomatonType
257 * Automaton representation.
259 struct REGEX_INTERNAL_Automaton
262 * Linked list of NFAs used for partial NFA creation.
264 struct REGEX_INTERNAL_Automaton *prev;
267 * Linked list of NFAs used for partial NFA creation.
269 struct REGEX_INTERNAL_Automaton *next;
272 * First state of the automaton. This is mainly used for constructing an NFA,
273 * where each NFA itself consists of one or more NFAs linked together.
275 struct REGEX_INTERNAL_State *start;
278 * End state of the partial NFA. This is undefined for DFAs
280 struct REGEX_INTERNAL_State *end;
283 * Number of states in the automaton.
285 unsigned int state_count;
290 struct REGEX_INTERNAL_State *states_head;
295 struct REGEX_INTERNAL_State *states_tail;
298 * Type of the automaton.
300 enum REGEX_INTERNAL_AutomatonType type;
308 * Canonical regex (result of RX->NFA->DFA->RX)
310 char *canonical_regex;
313 * GNUNET_YES, if multi strides have been added to the Automaton.
320 * Construct an NFA by parsing the regex string of length 'len'.
322 * @param regex regular expression string.
323 * @param len length of the string.
325 * @return NFA, needs to be freed using REGEX_INTERNAL_automaton_destroy.
327 struct REGEX_INTERNAL_Automaton *
328 REGEX_INTERNAL_construct_nfa (const char *regex, const size_t len);
332 * Function that get's passed to automaton traversal and is called before each
333 * next traversal from state 's' using transition 't' to check if traversal
334 * should proceed. Return GNUNET_NO to stop traversal or GNUNET_YES to continue.
336 * @param cls closure for the check.
337 * @param s current state in the traversal.
338 * @param t current transition from state 's' that will be used for the next
341 * @return GNUNET_YES to proceed traversal, GNUNET_NO to stop.
343 typedef int (*REGEX_INTERNAL_traverse_check) (void *cls,
344 struct REGEX_INTERNAL_State *s,
345 struct REGEX_INTERNAL_Transition *
350 * Function that is called with each state, when traversing an automaton.
352 * @param cls closure.
353 * @param count current count of the state, from 0 to a->state_count -1.
356 typedef void (*REGEX_INTERNAL_traverse_action) (void *cls,
357 const unsigned int count,
358 struct REGEX_INTERNAL_State *s);
362 * Traverses the given automaton using depth-first-search (DFS) from it's start
363 * state, visiting all reachable states and calling 'action' on each one of
366 * @param a automaton to be traversed.
367 * @param start start state, pass a->start or NULL to traverse the whole automaton.
368 * @param check function that is checked before advancing on each transition
370 * @param check_cls closure for check.
371 * @param action action to be performed on each state.
372 * @param action_cls closure for action
375 REGEX_INTERNAL_automaton_traverse (const struct REGEX_INTERNAL_Automaton *a,
376 struct REGEX_INTERNAL_State *start,
377 REGEX_INTERNAL_traverse_check check,
379 REGEX_INTERNAL_traverse_action action,
383 * Get the canonical regex of the given automaton.
384 * When constructing the automaton a proof is computed for each state,
385 * consisting of the regular expression leading to this state. A complete
386 * regex for the automaton can be computed by combining these proofs.
387 * As of now this function is only useful for testing.
389 * @param a automaton for which the canonical regex should be returned.
391 * @return canonical regex string.
394 REGEX_INTERNAL_get_canonical_regex (struct REGEX_INTERNAL_Automaton *a);
398 * Get the number of transitions that are contained in the given automaton.
400 * @param a automaton for which the number of transitions should be returned.
402 * @return number of transitions in the given automaton.
405 REGEX_INTERNAL_get_transition_count (struct REGEX_INTERNAL_Automaton *a);
409 * Context that contains an id counter for states and transitions as well as a
410 * DLL of automatons used as a stack for NFA construction.
412 struct REGEX_INTERNAL_Context
417 unsigned int state_id;
420 * Unique transition id.
422 unsigned int transition_id;
425 * DLL of REGEX_INTERNAL_Automaton's used as a stack.
427 struct REGEX_INTERNAL_Automaton *stack_head;
430 * DLL of REGEX_INTERNAL_Automaton's used as a stack.
432 struct REGEX_INTERNAL_Automaton *stack_tail;
437 * Adds multi-strided transitions to the given 'dfa'.
439 * @param regex_ctx regex context needed to add transitions to the automaton.
440 * @param dfa DFA to which the multi strided transitions should be added.
441 * @param stride_len length of the strides.
444 REGEX_INTERNAL_dfa_add_multi_strides (struct REGEX_INTERNAL_Context *regex_ctx,
445 struct REGEX_INTERNAL_Automaton *dfa,
446 const unsigned int stride_len);
449 #if 0 /* keep Emacsens' auto-indent happy */