1 /* +------------------------------------+
\r * | Inspire Internet Relay Chat Daemon |
\r * +------------------------------------+
\r *
\r * InspIRCd: (C) 2002-2007 InspIRCd Development Team
\r * See: http://www.inspircd.org/wiki/index.php/Credits
\r *
\r * This program is free but copyrighted software; see
\r * the file COPYING for details.
\r *
\r * ---------------------------------------------------
\r */
\r\r#ifndef _HASHCOMP_H_
\r#define _HASHCOMP_H_
\r\r#include "inspircd_config.h"
\r#include "socket.h"
\r#include "hash_map.h"
\r\r/*******************************************************
\r * This file contains classes and templates that deal
\r * with the comparison and hashing of 'irc strings'.
\r * An 'irc string' is a string which compares in a
\r * case insensitive manner, and as per RFC 1459 will
\r * treat [ identical to {, ] identical to }, and \
\r * as identical to |.
\r *
\r * Our hashing functions are designed to accept
\r * std::string and compare/hash them as type irc::string
\r * by converting them internally. This makes them
\r * backwards compatible with other code which is not
\r * aware of irc::string.
\r *******************************************************/
\r\r/** Required namespaces and symbols */
\rusing namespace std;
\r\r/** aton() */
\rusing irc::sockets::insp_aton;
\r\r/** nota() */
\rusing irc::sockets::insp_ntoa;
\r\r#ifndef LOWERMAP
\r#define LOWERMAP
\r/** A mapping of uppercase to lowercase, including scandinavian
\r * 'oddities' as specified by RFC1459, e.g. { -> [, and | -> \
\r */
\runsigned const char lowermap[256] = { 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, /* 0-19 */
\r 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, /* 20-39 */
\r 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, /* 40-59 */
\r 60, 61, 62, 63, 64, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, /* 60-79 */
\r 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 94, 95, 96, 97, 98, 99, /* 80-99 */
\r 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, /* 100-119 */
\r 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, /* 120-139 */
\r 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, /* 140-159 */
\r 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, /* 160-179 */
\r 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, /* 180-199 */
\r 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, /* 200-219 */
\r 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, /* 220-239 */
\r 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 255 /* 240-255 */
\r};
\r#endif
\r\r/** The irc namespace contains a number of helper classes.
\r */
\rnamespace irc
\r{
\r\r /** This class returns true if two strings match.
\r * Case sensitivity is ignored, and the RFC 'character set'
\r * is adhered to
\r */
\r struct StrHashComp
\r {
\r /** The operator () does the actual comparison in hash_map
\r */
\r bool operator()(const std::string& s1, const std::string& s2) const;
\r };
\r\r /** The irc_char_traits class is used for RFC-style comparison of strings.
\r * This class is used to implement irc::string, a case-insensitive, RFC-
\r * comparing string class.
\r */
\r struct irc_char_traits : std::char_traits<char> {
\r\r /** Check if two chars match.
\r * @param c1st First character
\r * @param c2nd Second character
\r * @return true if the characters are equal
\r */
\r static bool eq(char c1st, char c2nd);
\r\r /** Check if two chars do NOT match.
\r * @param c1st First character
\r * @param c2nd Second character
\r * @return true if the characters are unequal
\r */
\r static bool ne(char c1st, char c2nd);
\r\r /** Check if one char is less than another.
\r * @param c1st First character
\r * @param c2nd Second character
\r * @return true if c1st is less than c2nd
\r */
\r static bool lt(char c1st, char c2nd);
\r\r /** Compare two strings of size n.
\r * @param str1 First string
\r * @param str2 Second string
\r * @param n Length to compare to
\r * @return similar to strcmp, zero for equal, less than zero for str1
\r * being less and greater than zero for str1 being greater than str2.
\r */
\r static CoreExport int compare(const char* str1, const char* str2, size_t n);
\r\r /** Find a char within a string up to position n.
\r * @param s1 String to find in
\r * @param n Position to search up to
\r * @param c Character to search for
\r * @return Pointer to the first occurance of c in s1
\r */
\r static CoreExport const char* find(const char* s1, int n, char c);
\r };
\r\r /** Compose a hex string from raw data.
\r * @param raw The raw data to compose hex from
\r * @pram rawsz The size of the raw data buffer
\r * @return The hex string.
\r */
\r CoreExport std::string hex(const unsigned char *raw, size_t rawsz);
\r\r /** This typedef declares irc::string based upon irc_char_traits.
\r */
\r typedef basic_string<char, irc_char_traits, allocator<char> > string;
\r\r /** irc::stringjoiner joins string lists into a string, using
\r * the given seperator string.
\r * This class can join a vector of std::string, a deque of
\r * std::string, or a const char** array, using overloaded
\r * constructors.
\r */
\r class CoreExport stringjoiner
\r {
\r private:
\r /** Output string
\r */
\r std::string joined;
\r public:
\r /** Join elements of a vector, between (and including) begin and end
\r * @param seperator The string to seperate values with
\r * @param sequence One or more items to seperate
\r * @param begin The starting element in the sequence to be joined
\r * @param end The ending element in the sequence to be joined
\r */
\r stringjoiner(const std::string &seperator, const std::vector<std::string> &sequence, int begin, int end);
\r /** Join elements of a deque, between (and including) begin and end
\r * @param seperator The string to seperate values with
\r * @param sequence One or more items to seperate
\r * @param begin The starting element in the sequence to be joined
\r * @param end The ending element in the sequence to be joined
\r */
\r stringjoiner(const std::string &seperator, const std::deque<std::string> &sequence, int begin, int end);
\r /** Join elements of an array of char arrays, between (and including) begin and end
\r * @param seperator The string to seperate values with
\r * @param sequence One or more items to seperate
\r * @param begin The starting element in the sequence to be joined
\r * @param end The ending element in the sequence to be joined
\r */
\r stringjoiner(const std::string &seperator, const char** sequence, int begin, int end);
\r\r /** Get the joined sequence
\r * @return A reference to the joined string
\r */
\r std::string& GetJoined();
\r };
\r\r /** irc::modestacker stacks mode sequences into a list.
\r * It can then reproduce this list, clamped to a maximum of MAXMODES
\r * values per line.
\r */
\r class CoreExport modestacker
\r {
\r private:
\r /** The mode sequence and its parameters
\r */
\r std::deque<std::string> sequence;
\r /** True if the mode sequence is initially adding
\r * characters, false if it is initially removing
\r * them
\r */
\r bool adding;
\r public:
\r /** Construct a new modestacker.
\r * @param add True if the stack is adding modes,
\r * false if it is removing them
\r */
\r modestacker(bool add);
\r /** Push a modeletter and its parameter onto the stack.
\r * No checking is performed as to if this mode actually
\r * requires a parameter. If you stack invalid mode
\r * sequences, they will be tidied if and when they are
\r * passed to a mode parser.
\r * @param modeletter The mode letter to insert
\r * @param parameter The parameter for the mode
\r */
\r void Push(char modeletter, const std::string ¶meter);
\r /** Push a modeletter without parameter onto the stack.
\r * No checking is performed as to if this mode actually
\r * requires a parameter. If you stack invalid mode
\r * sequences, they will be tidied if and when they are
\r * passed to a mode parser.
\r * @param modeletter The mode letter to insert
\r */
\r void Push(char modeletter);
\r /** Push a '+' symbol onto the stack.
\r */
\r void PushPlus();
\r /** Push a '-' symbol onto the stack.
\r */
\r void PushMinus();
\r /** Return zero or more elements which form the
\r * mode line. This will be clamped to a max of
\r * MAXMODES items (MAXMODES-1 mode parameters and
\r * one mode sequence string), and max_line_size
\r * characters. As specified below, this function
\r * should be called in a loop until it returns zero,
\r * indicating there are no more modes to return.
\r * @param result The deque to populate. This will
\r * be cleared before it is used.
\r * @param max_line_size The maximum size of the line
\r * to build, in characters, seperate to MAXMODES.
\r * @return The number of elements in the deque.
\r * The function should be called repeatedly until it
\r * returns 0, in case there are multiple lines of
\r * mode changes to be obtained.
\r */
\r int GetStackedLine(std::deque<std::string> &result, int max_line_size = 360);
\r };
\r\r /** irc::tokenstream reads a string formatted as per RFC1459 and RFC2812.
\r * It will split the string into 'tokens' each containing one parameter
\r * from the string.
\r * For instance, if it is instantiated with the string:
\r * "PRIVMSG #test :foo bar baz qux"
\r * then each successive call to tokenstream::GetToken() will return
\r * "PRIVMSG", "#test", "foo bar baz qux", "".
\r * Note that if the whole string starts with a colon this is not taken
\r * to mean the string is all one parameter, and the first item in the
\r * list will be ":item". This is to allow for parsing 'source' fields
\r * from data.
\r */
\r class CoreExport tokenstream
\r {
\r private:
\r /** Original string
\r */
\r std::string tokens;
\r /** Last position of a seperator token
\r */
\r std::string::iterator last_starting_position;
\r /** Current string position
\r */
\r std::string::iterator n;
\r /** True if the last value was an ending value
\r */
\r bool last_pushed;
\r public:
\r /** Create a tokenstream and fill it with the provided data
\r */
\r tokenstream(const std::string &source);
\r ~tokenstream();
\r\r /** Fetch the next token from the stream as a std::string
\r * @param token The next token available, or an empty string if none remain
\r * @return True if tokens are left to be read, false if the last token was just retrieved.
\r */
\r bool GetToken(std::string &token);
\r\r /** Fetch the next token from the stream as an irc::string
\r * @param token The next token available, or an empty string if none remain
\r * @return True if tokens are left to be read, false if the last token was just retrieved.
\r */
\r bool GetToken(irc::string &token);
\r\r /** Fetch the next token from the stream as an integer
\r * @param token The next token available, or undefined if none remain
\r * @return True if tokens are left to be read, false if the last token was just retrieved.
\r */
\r bool GetToken(int &token);
\r\r /** Fetch the next token from the stream as a long integer
\r * @param token The next token available, or undefined if none remain
\r * @return True if tokens are left to be read, false if the last token was just retrieved.
\r */
\r bool GetToken(long &token);
\r };
\r\r /** irc::sepstream allows for splitting token seperated lists.
\r * Each successive call to sepstream::GetToken() returns
\r * the next token, until none remain, at which point the method returns
\r * an empty string.
\r */
\r class CoreExport sepstream : public classbase
\r {
\r private:
\r /** Original string.
\r */
\r std::string tokens;
\r /** Last position of a seperator token
\r */
\r std::string::iterator last_starting_position;
\r /** Current string position
\r */
\r std::string::iterator n;
\r /** Seperator value
\r */
\r char sep;
\r public:
\r /** Create a sepstream and fill it with the provided data
\r */
\r sepstream(const std::string &source, char seperator);
\r virtual ~sepstream();
\r\r /** Fetch the next token from the stream
\r * @return The next token is returned, or an empty string if none remain
\r */
\r virtual const std::string GetToken();
\r \r /** Fetch the entire remaining stream, without tokenizing
\r * @return The remaining part of the stream
\r */
\r virtual const std::string GetRemaining();
\r \r /** Returns true if the end of the stream has been reached
\r * @return True if the end of the stream has been reached, otherwise false
\r */
\r virtual bool StreamEnd();
\r };
\r\r /** A derived form of sepstream, which seperates on commas
\r */
\r class CoreExport commasepstream : public sepstream
\r {
\r public:
\r /** Initialize with comma seperator
\r */
\r commasepstream(const std::string &source) : sepstream(source, ',')
\r {
\r }
\r };
\r\r /** A derived form of sepstream, which seperates on spaces
\r */
\r class CoreExport spacesepstream : public sepstream
\r {
\r public:
\r /** Initialize with space seperator
\r */
\r spacesepstream(const std::string &source) : sepstream(source, ' ')
\r {
\r }
\r };
\r\r /** The portparser class seperates out a port range into integers.
\r * A port range may be specified in the input string in the form
\r * "6660,6661,6662-6669,7020". The end of the stream is indicated by
\r * a return value of 0 from portparser::GetToken(). If you attempt
\r * to specify an illegal range (e.g. one where start >= end, or
\r * start or end < 0) then GetToken() will return the first element
\r * of the pair of numbers.
\r */
\r class CoreExport portparser : public classbase
\r {
\r private:
\r /** Used to split on commas
\r */
\r commasepstream* sep;
\r /** Current position in a range of ports
\r */
\r long in_range;
\r /** Starting port in a range of ports
\r */
\r long range_begin;
\r /** Ending port in a range of ports
\r */
\r long range_end;
\r /** Allow overlapped port ranges
\r */
\r bool overlapped;
\r /** Used to determine overlapping of ports
\r * without O(n) algorithm being used
\r */
\r std::map<long, bool> overlap_set;
\r /** Returns true if val overlaps an existing range
\r */
\r bool Overlaps(long val);
\r public:
\r /** Create a portparser and fill it with the provided data
\r * @param source The source text to parse from
\r * @param allow_overlapped Allow overlapped ranges
\r */
\r portparser(const std::string &source, bool allow_overlapped = true);
\r /** Frees the internal commasepstream object
\r */
\r ~portparser();
\r /** Fetch the next token from the stream
\r * @return The next port number is returned, or 0 if none remain
\r */
\r long GetToken();
\r };
\r\r /** Used to hold a bitfield definition in dynamicbitmask.
\r * You must be allocated one of these by dynamicbitmask::Allocate(),
\r * you should not fill the values yourself!
\r */
\r typedef std::pair<size_t, unsigned char> bitfield;
\r\r /** The irc::dynamicbitmask class is used to maintain a bitmap of
\r * boolean values, which can grow to any reasonable size no matter
\r * how many bitfields are in it.
\r *
\r * It starts off at 32 bits in size, large enough to hold 32 boolean
\r * values, with a memory allocation of 8 bytes. If you allocate more
\r * than 32 bits, the class will grow the bitmap by 8 bytes at a time
\r * for each set of 8 bitfields you allocate with the Allocate()
\r * method.
\r *
\r * This method is designed so that multiple modules can be allocated
\r * bit values in a bitmap dynamically, rather than having to define
\r * costs in a fixed size unsigned integer and having the possibility
\r * of collisions of values in different third party modules.
\r *
\r * IMPORTANT NOTE:
\r *
\r * To use this class, you must derive from it.
\r * This is because each derived instance has its own freebits array
\r * which can determine what bitfields are allocated on a TYPE BY TYPE
\r * basis, e.g. an irc::dynamicbitmask type for userrecs, and one for
\r * chanrecs, etc. You should inheret it in a very simple way as follows.
\r * The base class will resize and maintain freebits as required, you are
\r * just required to make the pointer static and specific to this class
\r * type.
\r *
\r * \code
\r * class mydbitmask : public irc::dynamicbitmask
\r * {
\r * private:
\r *
\r * static unsigned char* freebits;
\r *
\r * public:
\r *
\r * mydbitmask() : irc::dynamicbitmask()
\r * {
\r * freebits = new unsigned char[this->bits_size];
\r * memset(freebits, 0, this->bits_size);
\r * }
\r *
\r * ~mydbitmask()
\r * {
\r * delete[] freebits;
\r * }
\r *
\r * unsigned char* GetFreeBits()
\r * {
\r * return freebits;
\r * }
\r *
\r * void SetFreeBits(unsigned char* freebt)
\r * {
\r * freebits = freebt;
\r * }
\r * };
\r * \endcode
\r */
\r class CoreExport dynamicbitmask : public classbase
\r {
\r private:
\r /** Data bits. We start with four of these,
\r * and we grow the bitfield as we allocate
\r * more than 32 entries with Allocate().
\r */
\r unsigned char* bits;
\r protected:
\r /** Current set size (size of freebits and bits).
\r * Both freebits and bits will ALWAYS be the
\r * same length.
\r */
\r unsigned char bits_size;
\r public:
\r /** Allocate the initial memory for bits and
\r * freebits and zero the memory.
\r */
\r dynamicbitmask();
\r\r /** Free the memory used by bits and freebits
\r */
\r virtual ~dynamicbitmask();
\r\r /** Allocate an irc::bitfield.
\r * @return An irc::bitfield which can be used
\r * with Get, Deallocate and Toggle methods.
\r * @throw Can throw std::bad_alloc if there is
\r * no ram left to grow the bitmask.
\r */
\r bitfield Allocate();
\r\r /** Deallocate an irc::bitfield.
\r * @param An irc::bitfield to deallocate.
\r * @return True if the bitfield could be
\r * deallocated, false if it could not.
\r */
\r bool Deallocate(bitfield &pos);
\r\r /** Toggle the value of a bitfield.
\r * @param pos A bitfield to allocate, previously
\r * allocated by dyamicbitmask::Allocate().
\r * @param state The state to set the field to.
\r */
\r void Toggle(bitfield &pos, bool state);
\r\r /** Get the value of a bitfield.
\r * @param pos A bitfield to retrieve, previously
\r * allocated by dyamicbitmask::Allocate().
\r * @return The value of the bitfield.
\r * @throw Will throw ModuleException if the bitfield
\r * you provide is outside of the range
\r * 0 >= bitfield.first < size_bits.
\r */
\r bool Get(bitfield &pos);
\r\r /** Return the size in bytes allocated to the bits
\r * array.
\r * Note that the actual allocation is twice this,
\r * as there are an equal number of bytes allocated
\r * for the freebits array.
\r */
\r unsigned char GetSize();
\r\r /** Get free bits mask
\r */
\r virtual unsigned char* GetFreeBits() { return NULL; }
\r\r /** Set free bits mask
\r */
\r virtual void SetFreeBits(unsigned char* freebits) { }
\r };
\r\r /** Turn _ characters in a string into spaces
\r * @param n String to translate
\r * @return The new value with _ translated to space.
\r */
\r CoreExport const char* Spacify(const char* n);
\r}
\r\r/* Define operators for using >> and << with irc::string to an ostream on an istream. */
\r/* This was endless fun. No. Really. */
\r/* It was also the first core change Ommeh made, if anyone cares */
\r\r/** Operator << for irc::string
\r */
\rinline std::ostream& operator<<(std::ostream &os, const irc::string &str) { return os << str.c_str(); }
\r\r/** Operator >> for irc::string
\r */
\rinline std::istream& operator>>(std::istream &is, irc::string &str)
\r{
\r std::string tmp;
\r is >> tmp;
\r str = tmp.c_str();
\r return is;
\r}
\r\r/* Define operators for + and == with irc::string to std::string for easy assignment
\r * and comparison
\r *
\r * Operator +
\r */
\rinline std::string operator+ (std::string& leftval, irc::string& rightval)
\r{
\r return leftval + std::string(rightval.c_str());
\r}
\r\r/* Define operators for + and == with irc::string to std::string for easy assignment
\r * and comparison
\r *
\r * Operator +
\r */
\rinline irc::string operator+ (irc::string& leftval, std::string& rightval)
\r{
\r return leftval + irc::string(rightval.c_str());
\r}
\r\r/* Define operators for + and == with irc::string to std::string for easy assignment
\r * and comparison
\r *
\r * Operator ==
\r */
\rinline bool operator== (const std::string& leftval, const irc::string& rightval)
\r{
\r return (leftval.c_str() == rightval);
\r}
\r\r/* Define operators for + and == with irc::string to std::string for easy assignment
\r * and comparison
\r *
\r * Operator ==
\r */
\rinline bool operator== (const irc::string& leftval, const std::string& rightval)
\r{
\r return (leftval == rightval.c_str());
\r}
\r\r/** Assign an irc::string to a std::string.
\r */
\rinline std::string assign(const irc::string &other) { return other.c_str(); }
\r\r/** Assign a std::string to an irc::string.
\r */
\rinline irc::string assign(const std::string &other) { return other.c_str(); }
\r\r/** Trim the leading and trailing spaces from a std::string.
\r */
\rinline std::string& trim(std::string &str)
\r{
\r std::string::size_type start = str.find_first_not_of(" ");
\r std::string::size_type end = str.find_last_not_of(" ");
\r if (start == std::string::npos || end == std::string::npos)
\r str = "";
\r else
\r str = str.substr(start, end-start+1);
\r\r return str;
\r}
\r\r/** Hashing stuff is totally different on vc++'s hash_map implementation, so to save a buttload of
\r * #ifdefs we'll just do it all at once
\r */
\rnamespace nspace
\r{
\r /** Hashing function to hash irc::string
\r */
\r#ifdef WINDOWS
\r template<> class CoreExport hash_compare<irc::string, std::less<irc::string> >
\r {
\r public:
\r enum { bucket_size = 4, min_buckets = 8 }; /* Got these numbers from the CRT source, if anyone wants to change them feel free. */
\r\r /** Compare two irc::string values for hashing in hash_map
\r */
\r bool operator()(const irc::string & s1, const irc::string & s2) const
\r {
\r if(s1.length() != s2.length()) return true;
\r return (irc::irc_char_traits::compare(s1.c_str(), s2.c_str(), s1.length()) < 0);
\r }
\r \r /** Hash an irc::string value for hash_map
\r */
\r size_t operator()(const irc::string & s) const;
\r };
\r\r template<> class CoreExport hash_compare<std::string, std::less<std::string> >
\r {
\r public:
\r enum { bucket_size = 4, min_buckets = 8 }; /* Again, from the CRT source */
\r\r /** Compare two std::string values for hashing in hash_map
\r */
\r bool operator()(const std::string & s1, const std::string & s2) const
\r {
\r if(s1.length() != s2.length()) return true;
\r return (irc::irc_char_traits::compare(s1.c_str(), s2.c_str(), s1.length()) < 0);
\r }
\r \r /** Hash a std::string using RFC1459 case sensitivity rules
\r * @param s A string to hash
\r * @return The hash value
\r */
\r size_t operator()(const std::string & s) const;
\r };
\r#else
\r template<> struct hash<irc::string>
\r {
\r /** Hash an irc::string using RFC1459 case sensitivity rules
\r * @param s A string to hash
\r * @return The hash value
\r */
\r size_t operator()(const irc::string &s) const;
\r };
\r\r template<> struct hash<std::string>
\r {
\r /** Hash a std::string using RFC1459 case sensitivity rules
\r * @param s A string to hash
\r * @return The hash value
\r */
\r size_t operator()(const string &s) const;
\r };
\r#endif
\r\r /** Convert a string to lower case respecting RFC1459
\r * @param n A string to lowercase
\r */
\r void strlower(char *n);
\r}
\r\r#endif
\r\r