2 * InspIRCd -- Internet Relay Chat Daemon
4 * Copyright (C) 2009 Daniel De Graaf <danieldg@inspircd.org>
5 * Copyright (C) 2005-2009 Craig Edwards <craigedwards@brainbox.cc>
6 * Copyright (C) 2007-2008 Robin Burchell <robin+git@viroteck.net>
7 * Copyright (C) 2007 Dennis Friis <peavey@inspircd.org>
9 * This file is part of InspIRCd. InspIRCd is free software: you can
10 * redistribute it and/or modify it under the terms of the GNU General Public
11 * License as published by the Free Software Foundation, version 2.
13 * This program is distributed in the hope that it will be useful, but WITHOUT
14 * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
15 * FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
18 * You should have received a copy of the GNU General Public License
19 * along with this program. If not, see <http://www.gnu.org/licenses/>.
25 /******************************************************
27 * The hash functions of InspIRCd are the centrepoint
28 * of the entire system. If these functions are
29 * inefficient or wasteful, the whole program suffers
30 * as a result. A lot of C programmers in the ircd
31 * scene spend a lot of time debating (arguing) about
32 * the best way to write hash functions to hash irc
33 * nicknames, channels etc.
34 * We are lucky as C++ developers as unordered_map does
35 * a lot of this for us. It does intellegent memory
36 * requests, bucketing, search functions, insertion
37 * and deletion etc. All we have to do is write some
38 * overloaded comparison and hash value operators which
39 * cause it to act in an irc-like way. The features we
40 * add to the standard hash_map are:
42 * Case insensitivity: The hash_map will be case
45 * Scandanavian Comparisons: The characters [, ], \ will
46 * be considered the lowercase of {, } and |.
48 ******************************************************/
52 * A case insensitive mapping of characters from upper case to lower case for
53 * the ASCII character set.
55 unsigned const char ascii_case_insensitive_map[256] = {
56 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, // 0-9
57 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, // 10-19
58 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, // 20-29
59 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, // 30-39
60 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, // 40-49
61 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, // 50-59
62 60, 61, 62, 63, 64, 97, 98, 99, 100, 101, // 60-69
63 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, // 70-79
64 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, // 80-89
65 122, 91, 92, 93, 94, 95, 96, 97, 98, 99, // 90-99
66 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, // 100-109
67 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, // 110-119
68 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, // 120-129
69 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, // 130-139
70 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, // 140-149
71 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, // 150-159
72 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, // 160-169
73 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, // 170-179
74 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, // 180-189
75 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, // 190-199
76 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, // 200-209
77 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, // 210-219
78 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, // 220-229
79 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, // 230-249
80 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, // 240-249
81 250, 251, 252, 253, 254, 255, // 250-255
87 * A case insensitive mapping of characters from upper case to lower case for
88 * the character set of RFC 1459. This is identical to ASCII with the small
89 * exception of {}| being considered to be the lower case equivalents of the
90 * characters []\ respectively.
92 unsigned const char rfc_case_insensitive_map[256] = {
93 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, // 0-9
94 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, // 10-19
95 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, // 20-29
96 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, // 30-39
97 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, // 40-49
98 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, // 50-59
99 60, 61, 62, 63, 64, 97, 98, 99, 100, 101, // 60-69
100 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, // 70-79
101 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, // 80-89
102 122, 123, 124, 125, 94, 95, 96, 97, 98, 99, // 90-99
103 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, // 100-109
104 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, // 110-119
105 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, // 120-129
106 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, // 130-139
107 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, // 140-149
108 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, // 150-159
109 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, // 160-169
110 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, // 170-179
111 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, // 180-189
112 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, // 190-199
113 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, // 200-209
114 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, // 210-219
115 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, // 220-229
116 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, // 230-239
117 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, // 240-249
118 250, 251, 252, 253, 254, 255, // 250-255
121 bool irc::equals(const std::string& s1, const std::string& s2)
123 const unsigned char* n1 = (const unsigned char*)s1.c_str();
124 const unsigned char* n2 = (const unsigned char*)s2.c_str();
125 for (; *n1 && *n2; n1++, n2++)
126 if (national_case_insensitive_map[*n1] != national_case_insensitive_map[*n2])
128 return (national_case_insensitive_map[*n1] == national_case_insensitive_map[*n2]);
131 size_t irc::find(const std::string& haystack, const std::string& needle)
133 // The haystack can't contain the needle if it is smaller than it.
134 if (needle.length() > haystack.length())
135 return std::string::npos;
137 // The inner loop checks the characters between haystack_last and the end of the haystack.
138 size_t haystack_last = haystack.length() - needle.length();
139 for (size_t hpos = 0; hpos <= haystack_last; ++hpos)
141 // Check for the needle at the current haystack position.
143 for (size_t npos = 0; npos < needle.length(); ++npos)
145 if (national_case_insensitive_map[(unsigned char)needle[npos]] != national_case_insensitive_map[(unsigned char)haystack[hpos + npos]])
147 // Uh-oh, characters at the current haystack position don't match.
153 // The entire needle was found in the haystack!
158 // We didn't find anything.
159 return std::string::npos;
163 bool irc::insensitive_swo::operator()(const std::string& a, const std::string& b) const
165 const unsigned char* charmap = national_case_insensitive_map;
166 std::string::size_type asize = a.size();
167 std::string::size_type bsize = b.size();
168 std::string::size_type maxsize = std::min(asize, bsize);
170 for (std::string::size_type i = 0; i < maxsize; i++)
172 unsigned char A = charmap[(unsigned char)a[i]];
173 unsigned char B = charmap[(unsigned char)b[i]];
179 return (asize < bsize);
182 size_t irc::insensitive::operator()(const std::string &s) const
184 /* XXX: NO DATA COPIES! :)
185 * The hash function here is practically
186 * a copy of the one in STL's hash_fun.h,
187 * only with *x replaced with national_case_insensitive_map[*x].
188 * This avoids a copy to use hash<const char*>
191 for (std::string::const_iterator x = s.begin(); x != s.end(); ++x) /* ++x not x++, as its faster */
192 t = 5 * t + national_case_insensitive_map[(unsigned char)*x];
196 irc::tokenstream::tokenstream(const std::string& msg, size_t start)
197 : message(msg, start)
202 bool irc::tokenstream::GetMiddle(std::string& token)
204 // If we are past the end of the string we can't do anything.
205 if (position >= message.length())
211 // If we can't find another separator this is the last token in the message.
212 size_t separator = message.find(' ', position);
213 if (separator == std::string::npos)
215 token.assign(message, position, std::string::npos);
216 position = message.length();
220 token.assign(message, position, separator - position);
221 position = message.find_first_not_of(' ', separator);
225 bool irc::tokenstream::GetTrailing(std::string& token)
227 // If we are past the end of the string we can't do anything.
228 if (position >= message.length())
234 // If this is true then we have a <trailing> token!
235 if (message[position] == ':')
237 token.assign(message, position + 1, std::string::npos);
238 position = message.length();
242 // There is no <trailing> token so it must be a <middle> token.
243 return GetMiddle(token);
246 irc::sepstream::sepstream(const std::string& source, char separator, bool allowempty)
247 : tokens(source), sep(separator), pos(0), allow_empty(allowempty)
251 bool irc::sepstream::GetToken(std::string &token)
253 if (this->StreamEnd())
259 if (!this->allow_empty)
261 this->pos = this->tokens.find_first_not_of(this->sep, this->pos);
262 if (this->pos == std::string::npos)
264 this->pos = this->tokens.length() + 1;
270 size_t p = this->tokens.find(this->sep, this->pos);
271 if (p == std::string::npos)
272 p = this->tokens.length();
274 token.assign(tokens, this->pos, p - this->pos);
280 const std::string irc::sepstream::GetRemaining()
282 return !this->StreamEnd() ? this->tokens.substr(this->pos) : "";
285 bool irc::sepstream::StreamEnd()
287 return this->pos > this->tokens.length();
290 irc::portparser::portparser(const std::string &source, bool allow_overlapped)
291 : sep(source), in_range(0), range_begin(0), range_end(0), overlapped(allow_overlapped)
295 bool irc::portparser::Overlaps(long val)
300 return (!overlap_set.insert(val).second);
303 long irc::portparser::GetToken()
308 if (in_range <= range_end)
310 if (!Overlaps(in_range))
316 while (((Overlaps(in_range)) && (in_range <= range_end)))
319 if (in_range <= range_end)
333 while (Overlaps(atoi(x.c_str())))
335 if (!sep.GetToken(x))
339 std::string::size_type dash = x.rfind('-');
340 if (dash != std::string::npos)
342 std::string sbegin(x, 0, dash);
343 range_begin = atoi(sbegin.c_str());
344 range_end = atoi(x.c_str()+dash+1);
346 if ((range_begin > 0) && (range_end > 0) && (range_begin < 65536) && (range_end < 65536) && (range_begin < range_end))
348 in_range = range_begin;
353 /* Assume its just the one port */
354 return atoi(sbegin.c_str());
359 return atoi(x.c_str());