1 /* +------------------------------------+
2 * | Inspire Internet Relay Chat Daemon |
3 * +------------------------------------+
5 * InspIRCd: (C) 2002-2009 InspIRCd Development Team
6 * See: http://wiki.inspircd.org/Credits
8 * This program is free but copyrighted software; see
9 * the file COPYING for details.
11 * ---------------------------------------------------
25 /*******************************************************
26 * This file contains classes and templates that deal
27 * with the comparison and hashing of 'irc strings'.
28 * An 'irc string' is a string which compares in a
29 * case insensitive manner, and as per RFC 1459 will
30 * treat [ identical to {, ] identical to }, and \
33 * Our hashing functions are designed to accept
34 * std::string and compare/hash them as type irc::string
35 * by converting them internally. This makes them
36 * backwards compatible with other code which is not
37 * aware of irc::string.
38 *******************************************************/
40 /** Seperate from the other casemap tables so that code *can* still exclusively rely on RFC casemapping
43 * This is provided as a pointer so that modules can change it to their custom mapping tables,
44 * e.g. for national character support.
46 CoreExport extern unsigned const char *national_case_insensitive_map;
48 /** A mapping of uppercase to lowercase, including scandinavian
49 * 'oddities' as specified by RFC1459, e.g. { -> [, and | -> \
51 CoreExport extern unsigned const char rfc_case_insensitive_map[256];
53 /** Case insensitive map, ASCII rules.
57 CoreExport extern unsigned const char ascii_case_insensitive_map[256];
59 /** Case sensitive (identity) map.
61 CoreExport extern unsigned const char rfc_case_sensitive_map[256];
63 template<typename T> const T& SearchAndReplace(T& text, const T& pattern, const T& replace)
66 if ((!pattern.empty()) && (!text.empty()))
68 for (std::string::size_type n = 0; n != text.length(); ++n)
70 if (text.length() >= pattern.length() && text.substr(n, pattern.length()) == pattern)
72 /* Found the pattern in the text, replace it, and advance */
73 replacement.append(replace);
74 n = n + pattern.length() - 1;
78 replacement += text[n];
86 /** The irc namespace contains a number of helper classes.
91 /** This class returns true if two strings match.
92 * Case sensitivity is ignored, and the RFC 'character set'
97 /** The operator () does the actual comparison in hash_map
99 bool operator()(const std::string& s1, const std::string& s2) const;
102 /** The irc_char_traits class is used for RFC-style comparison of strings.
103 * This class is used to implement irc::string, a case-insensitive, RFC-
104 * comparing string class.
106 struct irc_char_traits : std::char_traits<char> {
108 /** Check if two chars match.
109 * @param c1st First character
110 * @param c2nd Second character
111 * @return true if the characters are equal
113 static bool eq(char c1st, char c2nd);
115 /** Check if two chars do NOT match.
116 * @param c1st First character
117 * @param c2nd Second character
118 * @return true if the characters are unequal
120 static bool ne(char c1st, char c2nd);
122 /** Check if one char is less than another.
123 * @param c1st First character
124 * @param c2nd Second character
125 * @return true if c1st is less than c2nd
127 static bool lt(char c1st, char c2nd);
129 /** Compare two strings of size n.
130 * @param str1 First string
131 * @param str2 Second string
132 * @param n Length to compare to
133 * @return similar to strcmp, zero for equal, less than zero for str1
134 * being less and greater than zero for str1 being greater than str2.
136 static CoreExport int compare(const char* str1, const char* str2, size_t n);
138 /** Find a char within a string up to position n.
139 * @param s1 String to find in
140 * @param n Position to search up to
141 * @param c Character to search for
142 * @return Pointer to the first occurance of c in s1
144 static CoreExport const char* find(const char* s1, int n, char c);
147 /** Compose a hex string from raw data.
148 * @param raw The raw data to compose hex from
149 * @pram rawsz The size of the raw data buffer
150 * @return The hex string.
152 CoreExport std::string hex(const unsigned char *raw, size_t rawsz);
154 /** This typedef declares irc::string based upon irc_char_traits.
156 typedef std::basic_string<char, irc_char_traits, std::allocator<char> > string;
158 /** irc::stringjoiner joins string lists into a string, using
159 * the given seperator string.
160 * This class can join a vector of std::string, a deque of
161 * std::string, or a const char* const* array, using overloaded
164 class CoreExport stringjoiner
174 /** Join elements of a vector, between (and including) begin and end
175 * @param seperator The string to seperate values with
176 * @param sequence One or more items to seperate
177 * @param begin The starting element in the sequence to be joined
178 * @param end The ending element in the sequence to be joined
180 stringjoiner(const std::string &seperator, const std::vector<std::string> &sequence, int begin, int end);
182 /** Join elements of a deque, between (and including) begin and end
183 * @param seperator The string to seperate values with
184 * @param sequence One or more items to seperate
185 * @param begin The starting element in the sequence to be joined
186 * @param end The ending element in the sequence to be joined
188 stringjoiner(const std::string &seperator, const std::deque<std::string> &sequence, int begin, int end);
190 /** Join elements of an array of char arrays, between (and including) begin and end
191 * @param seperator The string to seperate values with
192 * @param sequence One or more items to seperate
193 * @param begin The starting element in the sequence to be joined
194 * @param end The ending element in the sequence to be joined
196 stringjoiner(const std::string &seperator, const char* const* sequence, int begin, int end);
198 /** Get the joined sequence
199 * @return A reference to the joined string
201 std::string& GetJoined();
204 /** irc::modestacker stacks mode sequences into a list.
205 * It can then reproduce this list, clamped to a maximum of MAXMODES
208 class CoreExport modestacker
211 /** The mode sequence and its parameters
213 std::deque<std::string> sequence;
215 /** True if the mode sequence is initially adding
216 * characters, false if it is initially removing
222 /** Construct a new modestacker.
223 * @param add True if the stack is adding modes,
224 * false if it is removing them
226 modestacker(bool add);
228 /** Push a modeletter and its parameter onto the stack.
229 * No checking is performed as to if this mode actually
230 * requires a parameter. If you stack invalid mode
231 * sequences, they will be tidied if and when they are
232 * passed to a mode parser.
233 * @param modeletter The mode letter to insert
234 * @param parameter The parameter for the mode
236 void Push(char modeletter, const std::string ¶meter);
238 /** Push a modeletter without parameter onto the stack.
239 * No checking is performed as to if this mode actually
240 * requires a parameter. If you stack invalid mode
241 * sequences, they will be tidied if and when they are
242 * passed to a mode parser.
243 * @param modeletter The mode letter to insert
245 void Push(char modeletter);
247 /** Push a '+' symbol onto the stack.
251 /** Push a '-' symbol onto the stack.
255 /** Return zero or more elements which form the
256 * mode line. This will be clamped to a max of
257 * MAXMODES items (MAXMODES-1 mode parameters and
258 * one mode sequence string), and max_line_size
259 * characters. As specified below, this function
260 * should be called in a loop until it returns zero,
261 * indicating there are no more modes to return.
262 * @param result The vector to populate. This will not
263 * be cleared before it is used.
264 * @param max_line_size The maximum size of the line
265 * to build, in characters, seperate to MAXMODES.
266 * @return The number of elements in the deque.
267 * The function should be called repeatedly until it
268 * returns 0, in case there are multiple lines of
269 * mode changes to be obtained.
271 int GetStackedLine(std::vector<std::string> &result, int max_line_size = 360);
273 /** deprecated compatability interface - TODO remove */
274 int GetStackedLine(std::deque<std::string> &result, int max_line_size = 360) {
275 std::vector<std::string> r;
276 int n = GetStackedLine(r, max_line_size);
278 result.insert(result.end(), r.begin(), r.end());
283 /** irc::tokenstream reads a string formatted as per RFC1459 and RFC2812.
284 * It will split the string into 'tokens' each containing one parameter
286 * For instance, if it is instantiated with the string:
287 * "PRIVMSG #test :foo bar baz qux"
288 * then each successive call to tokenstream::GetToken() will return
289 * "PRIVMSG", "#test", "foo bar baz qux", "".
290 * Note that if the whole string starts with a colon this is not taken
291 * to mean the string is all one parameter, and the first item in the
292 * list will be ":item". This is to allow for parsing 'source' fields
295 class CoreExport tokenstream
303 /** Last position of a seperator token
305 std::string::iterator last_starting_position;
307 /** Current string position
309 std::string::iterator n;
311 /** True if the last value was an ending value
316 /** Create a tokenstream and fill it with the provided data
318 tokenstream(const std::string &source);
324 /** Fetch the next token from the stream as a std::string
325 * @param token The next token available, or an empty string if none remain
326 * @return True if tokens are left to be read, false if the last token was just retrieved.
328 bool GetToken(std::string &token);
330 /** Fetch the next token from the stream as an irc::string
331 * @param token The next token available, or an empty string if none remain
332 * @return True if tokens are left to be read, false if the last token was just retrieved.
334 bool GetToken(irc::string &token);
336 /** Fetch the next token from the stream as an integer
337 * @param token The next token available, or undefined if none remain
338 * @return True if tokens are left to be read, false if the last token was just retrieved.
340 bool GetToken(int &token);
342 /** Fetch the next token from the stream as a long integer
343 * @param token The next token available, or undefined if none remain
344 * @return True if tokens are left to be read, false if the last token was just retrieved.
346 bool GetToken(long &token);
349 /** irc::sepstream allows for splitting token seperated lists.
350 * Each successive call to sepstream::GetToken() returns
351 * the next token, until none remain, at which point the method returns
354 class CoreExport sepstream
360 /** Last position of a seperator token
362 std::string::iterator last_starting_position;
363 /** Current string position
365 std::string::iterator n;
370 /** Create a sepstream and fill it with the provided data
372 sepstream(const std::string &source, char seperator);
376 virtual ~sepstream();
378 /** Fetch the next token from the stream
379 * @param token The next token from the stream is placed here
380 * @return True if tokens still remain, false if there are none left
382 virtual bool GetToken(std::string &token);
384 /** Fetch the entire remaining stream, without tokenizing
385 * @return The remaining part of the stream
387 virtual const std::string GetRemaining();
389 /** Returns true if the end of the stream has been reached
390 * @return True if the end of the stream has been reached, otherwise false
392 virtual bool StreamEnd();
395 /** A derived form of sepstream, which seperates on commas
397 class CoreExport commasepstream : public sepstream
400 /** Initialize with comma seperator
402 commasepstream(const std::string &source) : sepstream(source, ',')
407 /** A derived form of sepstream, which seperates on spaces
409 class CoreExport spacesepstream : public sepstream
412 /** Initialize with space seperator
414 spacesepstream(const std::string &source) : sepstream(source, ' ')
419 /** The portparser class seperates out a port range into integers.
420 * A port range may be specified in the input string in the form
421 * "6660,6661,6662-6669,7020". The end of the stream is indicated by
422 * a return value of 0 from portparser::GetToken(). If you attempt
423 * to specify an illegal range (e.g. one where start >= end, or
424 * start or end < 0) then GetToken() will return the first element
425 * of the pair of numbers.
427 class CoreExport portparser
431 /** Used to split on commas
435 /** Current position in a range of ports
439 /** Starting port in a range of ports
443 /** Ending port in a range of ports
447 /** Allow overlapped port ranges
451 /** Used to determine overlapping of ports
452 * without O(n) algorithm being used
454 std::map<long, bool> overlap_set;
456 /** Returns true if val overlaps an existing range
458 bool Overlaps(long val);
461 /** Create a portparser and fill it with the provided data
462 * @param source The source text to parse from
463 * @param allow_overlapped Allow overlapped ranges
465 portparser(const std::string &source, bool allow_overlapped = true);
467 /** Frees the internal commasepstream object
471 /** Fetch the next token from the stream
472 * @return The next port number is returned, or 0 if none remain
477 /** Turn _ characters in a string into spaces
478 * @param n String to translate
479 * @return The new value with _ translated to space.
481 CoreExport const char* Spacify(const char* n);
484 /* Define operators for using >> and << with irc::string to an ostream on an istream. */
485 /* This was endless fun. No. Really. */
486 /* It was also the first core change Ommeh made, if anyone cares */
488 /** Operator << for irc::string
490 inline std::ostream& operator<<(std::ostream &os, const irc::string &str) { return os << str.c_str(); }
492 /** Operator >> for irc::string
494 inline std::istream& operator>>(std::istream &is, irc::string &str)
502 /* Define operators for + and == with irc::string to std::string for easy assignment
507 inline std::string operator+ (std::string& leftval, irc::string& rightval)
509 return leftval + std::string(rightval.c_str());
512 /* Define operators for + and == with irc::string to std::string for easy assignment
517 inline irc::string operator+ (irc::string& leftval, std::string& rightval)
519 return leftval + irc::string(rightval.c_str());
522 /* Define operators for + and == with irc::string to std::string for easy assignment
527 inline bool operator== (const std::string& leftval, const irc::string& rightval)
529 return (leftval.c_str() == rightval);
532 /* Define operators for + and == with irc::string to std::string for easy assignment
537 inline bool operator== (const irc::string& leftval, const std::string& rightval)
539 return (leftval == rightval.c_str());
542 /* Define operators != for irc::string to std::string for easy comparison
544 inline bool operator!= (const irc::string& leftval, const std::string& rightval)
546 return !(leftval == rightval.c_str());
549 /* Define operators != for std::string to irc::string for easy comparison
551 inline bool operator!= (const std::string& leftval, const irc::string& rightval)
553 return !(leftval.c_str() == rightval);
556 /** Assign an irc::string to a std::string.
558 inline std::string assign(const irc::string &other) { return other.c_str(); }
560 /** Assign a std::string to an irc::string.
562 inline irc::string assign(const std::string &other) { return other.c_str(); }
564 /** Trim the leading and trailing spaces from a std::string.
566 inline std::string& trim(std::string &str)
568 std::string::size_type start = str.find_first_not_of(" ");
569 std::string::size_type end = str.find_last_not_of(" ");
570 if (start == std::string::npos || end == std::string::npos)
573 str = str.substr(start, end-start+1);
578 /** Hashing stuff is totally different on vc++'s hash_map implementation, so to save a buttload of
579 * #ifdefs we'll just do it all at once. Except, of course, with TR1, when it's the same as GCC.
581 BEGIN_HASHMAP_NAMESPACE
583 /** Hashing function to hash irc::string
585 #if defined(WINDOWS) && !defined(HAS_TR1_UNORDERED)
586 template<> class CoreExport hash_compare<irc::string, std::less<irc::string> >
589 enum { bucket_size = 4, min_buckets = 8 }; /* Got these numbers from the CRT source, if anyone wants to change them feel free. */
591 /** Compare two irc::string values for hashing in hash_map
593 bool operator()(const irc::string & s1, const irc::string & s2) const
595 if(s1.length() != s2.length()) return true;
596 return (irc::irc_char_traits::compare(s1.c_str(), s2.c_str(), (size_t)s1.length()) < 0);
599 /** Hash an irc::string value for hash_map
601 size_t operator()(const irc::string & s) const;
604 template<> class CoreExport hash_compare<std::string, std::less<std::string> >
607 enum { bucket_size = 4, min_buckets = 8 }; /* Again, from the CRT source */
609 /** Compare two std::string values for hashing in hash_map
611 bool operator()(const std::string & s1, const std::string & s2) const
613 if(s1.length() != s2.length()) return true;
614 return (irc::irc_char_traits::compare(s1.c_str(), s2.c_str(), (size_t)s1.length()) < 0);
617 /** Hash a std::string using RFC1459 case sensitivity rules
618 * @param s A string to hash
619 * @return The hash value
621 size_t operator()(const std::string & s) const;
625 template<> struct hash<irc::string>
627 /** Hash an irc::string using RFC1459 case sensitivity rules
628 * @param s A string to hash
629 * @return The hash value
631 size_t CoreExport operator()(const irc::string &s) const;
634 /* XXX FIXME: Implement a hash function overriding std::string's that works with TR1! */
636 #ifdef HASHMAP_DEPRECATED
639 CoreExport template<> struct hash<std::string>
642 size_t CoreExport operator()(const std::string &s) const;
647 /** Convert a string to lower case respecting RFC1459
648 * @param n A string to lowercase
650 void strlower(char *n);
652 END_HASHMAP_NAMESPACE