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 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/>.
19 * @file regex/regex_internal_lib.h
20 * @brief library to parse regular expressions into dfa
21 * @author Maximilian Szengel
24 #ifndef REGEX_INTERNAL_LIB_H
25 #define REGEX_INTERNAL_LIB_H
27 #include "gnunet_util_lib.h"
28 #include "gnunet_dht_service.h"
29 #include "gnunet_statistics_service.h"
30 #include "regex_block_lib.h"
35 #if 0 /* keep Emacsens' auto-indent happy */
42 * Automaton (NFA/DFA) representation.
44 struct REGEX_INTERNAL_Automaton;
48 * Construct DFA for the given 'regex' of length 'len'.
50 * Path compression means, that for example a DFA o -> a -> b -> c -> o will be
51 * compressed to o -> abc -> o. Note that this parameter influences the
52 * non-determinism of states of the resulting NFA in the DHT (number of outgoing
53 * edges with the same label). For example for an application that stores IPv4
54 * addresses as bitstrings it could make sense to limit the path compression to
57 * @param regex regular expression string.
58 * @param len length of the regular expression.
59 * @param max_path_len limit the path compression length to the
60 * given value. If set to 1, no path compression is applied. Set to 0 for
61 * maximal possible path compression (generally not desireable).
62 * @return DFA, needs to be freed using #REGEX_INTERNAL_automaton_destroy().
64 struct REGEX_INTERNAL_Automaton *
65 REGEX_INTERNAL_construct_dfa (const char *regex,
67 unsigned int max_path_len);
71 * Free the memory allocated by constructing the REGEX_INTERNAL_Automaton.
74 * @param a automaton to be destroyed.
77 REGEX_INTERNAL_automaton_destroy (struct REGEX_INTERNAL_Automaton *a);
81 * Evaluates the given 'string' against the given compiled regex.
84 * @param string string to check.
86 * @return 0 if string matches, non 0 otherwise.
89 REGEX_INTERNAL_eval (struct REGEX_INTERNAL_Automaton *a,
94 * Get the first key for the given @a input_string. This hashes
95 * the first x bits of the @a input_string.
97 * @param input_string string.
98 * @param string_len length of the @a input_string.
99 * @param key pointer to where to write the hash code.
100 * @return number of bits of @a input_string that have been consumed
101 * to construct the key
104 REGEX_INTERNAL_get_first_key (const char *input_string,
106 struct GNUNET_HashCode * key);
110 * Iterator callback function.
112 * @param cls closure.
113 * @param key hash for current state.
114 * @param proof proof for current state
115 * @param accepting #GNUNET_YES if this is an accepting state, #GNUNET_NO if not.
116 * @param num_edges number of edges leaving current state.
117 * @param edges edges leaving current state.
120 (*REGEX_INTERNAL_KeyIterator)(void *cls,
121 const struct GNUNET_HashCode *key,
124 unsigned int num_edges,
125 const struct REGEX_BLOCK_Edge *edges);
129 * Iterate over all edges starting from start state of automaton 'a'. Calling
130 * iterator for each edge.
132 * @param a automaton.
133 * @param iterator iterator called for each edge.
134 * @param iterator_cls closure.
137 REGEX_INTERNAL_iterate_all_edges (struct REGEX_INTERNAL_Automaton *a,
138 REGEX_INTERNAL_KeyIterator iterator,
143 * Iterate over all edges of automaton 'a' that are reachable from a state with
144 * a proof of at least #GNUNET_REGEX_INITIAL_BYTES characters.
146 * Call the iterator for each such edge.
148 * @param a automaton.
149 * @param iterator iterator called for each reachable edge.
150 * @param iterator_cls closure.
153 REGEX_INTERNAL_iterate_reachable_edges (struct REGEX_INTERNAL_Automaton *a,
154 REGEX_INTERNAL_KeyIterator iterator,
160 * Handle to store cached data about a regex announce.
162 struct REGEX_INTERNAL_Announcement;
165 * Handle to store data about a regex search.
167 struct REGEX_INTERNAL_Search;
171 * Announce a regular expression: put all states of the automaton in the DHT.
172 * Does not free resources, must call #REGEX_INTERNAL_announce_cancel() for that.
174 * @param dht An existing and valid DHT service handle. CANNOT be NULL.
175 * @param priv our private key, must remain valid until the announcement is cancelled
176 * @param regex Regular expression to announce.
177 * @param compression How many characters per edge can we squeeze?
178 * @param stats Optional statistics handle to report usage. Can be NULL.
179 * @return Handle to reuse o free cached resources.
180 * Must be freed by calling #REGEX_INTERNAL_announce_cancel().
182 struct REGEX_INTERNAL_Announcement *
183 REGEX_INTERNAL_announce (struct GNUNET_DHT_Handle *dht,
184 const struct GNUNET_CRYPTO_EddsaPrivateKey *priv,
186 uint16_t compression,
187 struct GNUNET_STATISTICS_Handle *stats);
191 * Announce again a regular expression previously announced.
192 * Does use caching to speed up process.
194 * @param h Handle returned by a previous #REGEX_INTERNAL_announce() call.
197 REGEX_INTERNAL_reannounce (struct REGEX_INTERNAL_Announcement *h);
201 * Clear all cached data used by a regex announce.
202 * Does not close DHT connection.
204 * @param h Handle returned by a previous #REGEX_INTERNAL_announce() call.
207 REGEX_INTERNAL_announce_cancel (struct REGEX_INTERNAL_Announcement *h);
211 * Search callback function.
213 * @param cls Closure provided in #REGEX_INTERNAL_search().
214 * @param id Peer providing a regex that matches the string.
215 * @param get_path Path of the get request.
216 * @param get_path_length Length of @a get_path.
217 * @param put_path Path of the put request.
218 * @param put_path_length Length of the @a put_path.
221 (*REGEX_INTERNAL_Found)(void *cls,
222 const struct GNUNET_PeerIdentity *id,
223 const struct GNUNET_PeerIdentity *get_path,
224 unsigned int get_path_length,
225 const struct GNUNET_PeerIdentity *put_path,
226 unsigned int put_path_length);
230 * Search for a peer offering a regex matching certain string in the DHT.
231 * The search runs until #REGEX_INTERNAL_search_cancel() is called, even if results
234 * @param dht An existing and valid DHT service handle.
235 * @param string String to match against the regexes in the DHT.
236 * @param callback Callback for found peers.
237 * @param callback_cls Closure for @c callback.
238 * @param stats Optional statistics handle to report usage. Can be NULL.
239 * @return Handle to stop search and free resources.
240 * Must be freed by calling #REGEX_INTERNAL_search_cancel().
242 struct REGEX_INTERNAL_Search *
243 REGEX_INTERNAL_search (struct GNUNET_DHT_Handle *dht,
245 REGEX_INTERNAL_Found callback,
247 struct GNUNET_STATISTICS_Handle *stats);
250 * Stop search and free all data used by a #REGEX_INTERNAL_search() call.
251 * Does not close DHT connection.
253 * @param h Handle returned by a previous #REGEX_INTERNAL_search() call.
256 REGEX_INTERNAL_search_cancel (struct REGEX_INTERNAL_Search *h);
259 #if 0 /* keep Emacsens' auto-indent happy */
266 /* end of regex_internal_lib.h */