2 This file is part of GNUnet
3 Copyright (C) 2012, 2013 GNUnet e.V.
5 GNUnet is free software; you can redistribute it and/or modify
6 it under the terms of the GNU General Public License as published
7 by the Free Software Foundation; either version 3, or (at your
8 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 General Public License for more details.
15 You should have received a copy of the GNU General Public License
16 along with GNUnet; see the file COPYING. If not, write to the
17 Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
18 Boston, MA 02110-1301, USA.
21 * @file regex/regex_internal_lib.h
22 * @brief library to parse regular expressions into dfa
23 * @author Maximilian Szengel
26 #ifndef REGEX_INTERNAL_LIB_H
27 #define REGEX_INTERNAL_LIB_H
29 #include "gnunet_util_lib.h"
30 #include "gnunet_dht_service.h"
31 #include "gnunet_statistics_service.h"
32 #include "regex_block_lib.h"
37 #if 0 /* keep Emacsens' auto-indent happy */
44 * Automaton (NFA/DFA) representation.
46 struct REGEX_INTERNAL_Automaton;
50 * Construct DFA for the given 'regex' of length 'len'.
52 * Path compression means, that for example a DFA o -> a -> b -> c -> o will be
53 * compressed to o -> abc -> o. Note that this parameter influences the
54 * non-determinism of states of the resulting NFA in the DHT (number of outgoing
55 * edges with the same label). For example for an application that stores IPv4
56 * addresses as bitstrings it could make sense to limit the path compression to
59 * @param regex regular expression string.
60 * @param len length of the regular expression.
61 * @param max_path_len limit the path compression length to the
62 * given value. If set to 1, no path compression is applied. Set to 0 for
63 * maximal possible path compression (generally not desireable).
64 * @return DFA, needs to be freed using #REGEX_INTERNAL_automaton_destroy().
66 struct REGEX_INTERNAL_Automaton *
67 REGEX_INTERNAL_construct_dfa (const char *regex,
69 unsigned int max_path_len);
73 * Free the memory allocated by constructing the REGEX_INTERNAL_Automaton.
76 * @param a automaton to be destroyed.
79 REGEX_INTERNAL_automaton_destroy (struct REGEX_INTERNAL_Automaton *a);
83 * Evaluates the given 'string' against the given compiled regex.
86 * @param string string to check.
88 * @return 0 if string matches, non 0 otherwise.
91 REGEX_INTERNAL_eval (struct REGEX_INTERNAL_Automaton *a,
96 * Get the first key for the given @a input_string. This hashes
97 * the first x bits of the @a input_string.
99 * @param input_string string.
100 * @param string_len length of the @a input_string.
101 * @param key pointer to where to write the hash code.
102 * @return number of bits of @a input_string that have been consumed
103 * to construct the key
106 REGEX_INTERNAL_get_first_key (const char *input_string,
108 struct GNUNET_HashCode * key);
112 * Iterator callback function.
114 * @param cls closure.
115 * @param key hash for current state.
116 * @param proof proof for current state
117 * @param accepting #GNUNET_YES if this is an accepting state, #GNUNET_NO if not.
118 * @param num_edges number of edges leaving current state.
119 * @param edges edges leaving current state.
122 (*REGEX_INTERNAL_KeyIterator)(void *cls,
123 const struct GNUNET_HashCode *key,
126 unsigned int num_edges,
127 const struct REGEX_BLOCK_Edge *edges);
131 * Iterate over all edges starting from start state of automaton 'a'. Calling
132 * iterator for each edge.
134 * @param a automaton.
135 * @param iterator iterator called for each edge.
136 * @param iterator_cls closure.
139 REGEX_INTERNAL_iterate_all_edges (struct REGEX_INTERNAL_Automaton *a,
140 REGEX_INTERNAL_KeyIterator iterator,
145 * Iterate over all edges of automaton 'a' that are reachable from a state with
146 * a proof of at least #GNUNET_REGEX_INITIAL_BYTES characters.
148 * Call the iterator for each such edge.
150 * @param a automaton.
151 * @param iterator iterator called for each reachable edge.
152 * @param iterator_cls closure.
155 REGEX_INTERNAL_iterate_reachable_edges (struct REGEX_INTERNAL_Automaton *a,
156 REGEX_INTERNAL_KeyIterator iterator,
162 * Handle to store cached data about a regex announce.
164 struct REGEX_INTERNAL_Announcement;
167 * Handle to store data about a regex search.
169 struct REGEX_INTERNAL_Search;
173 * Announce a regular expression: put all states of the automaton in the DHT.
174 * Does not free resources, must call #REGEX_INTERNAL_announce_cancel() for that.
176 * @param dht An existing and valid DHT service handle. CANNOT be NULL.
177 * @param priv our private key, must remain valid until the announcement is cancelled
178 * @param regex Regular expression to announce.
179 * @param compression How many characters per edge can we squeeze?
180 * @param stats Optional statistics handle to report usage. Can be NULL.
181 * @return Handle to reuse o free cached resources.
182 * Must be freed by calling #REGEX_INTERNAL_announce_cancel().
184 struct REGEX_INTERNAL_Announcement *
185 REGEX_INTERNAL_announce (struct GNUNET_DHT_Handle *dht,
186 const struct GNUNET_CRYPTO_EddsaPrivateKey *priv,
188 uint16_t compression,
189 struct GNUNET_STATISTICS_Handle *stats);
193 * Announce again a regular expression previously announced.
194 * Does use caching to speed up process.
196 * @param h Handle returned by a previous #REGEX_INTERNAL_announce() call.
199 REGEX_INTERNAL_reannounce (struct REGEX_INTERNAL_Announcement *h);
203 * Clear all cached data used by a regex announce.
204 * Does not close DHT connection.
206 * @param h Handle returned by a previous #REGEX_INTERNAL_announce() call.
209 REGEX_INTERNAL_announce_cancel (struct REGEX_INTERNAL_Announcement *h);
213 * Search callback function.
215 * @param cls Closure provided in #REGEX_INTERNAL_search().
216 * @param id Peer providing a regex that matches the string.
217 * @param get_path Path of the get request.
218 * @param get_path_length Length of @a get_path.
219 * @param put_path Path of the put request.
220 * @param put_path_length Length of the @a put_path.
223 (*REGEX_INTERNAL_Found)(void *cls,
224 const struct GNUNET_PeerIdentity *id,
225 const struct GNUNET_PeerIdentity *get_path,
226 unsigned int get_path_length,
227 const struct GNUNET_PeerIdentity *put_path,
228 unsigned int put_path_length);
232 * Search for a peer offering a regex matching certain string in the DHT.
233 * The search runs until #REGEX_INTERNAL_search_cancel() is called, even if results
236 * @param dht An existing and valid DHT service handle.
237 * @param string String to match against the regexes in the DHT.
238 * @param callback Callback for found peers.
239 * @param callback_cls Closure for @c callback.
240 * @param stats Optional statistics handle to report usage. Can be NULL.
241 * @return Handle to stop search and free resources.
242 * Must be freed by calling #REGEX_INTERNAL_search_cancel().
244 struct REGEX_INTERNAL_Search *
245 REGEX_INTERNAL_search (struct GNUNET_DHT_Handle *dht,
247 REGEX_INTERNAL_Found callback,
249 struct GNUNET_STATISTICS_Handle *stats);
252 * Stop search and free all data used by a #REGEX_INTERNAL_search() call.
253 * Does not close DHT connection.
255 * @param h Handle returned by a previous #REGEX_INTERNAL_search() call.
258 REGEX_INTERNAL_search_cancel (struct REGEX_INTERNAL_Search *h);
261 #if 0 /* keep Emacsens' auto-indent happy */
268 /* end of regex_internal_lib.h */