src/base/CharacterSet.h

   1 /*
   2  * Copyright (C) 1996-2014 The Squid Software Foundation and contributors
   3  *
   4  * Squid software is distributed under GPLv2+ license and includes
   5  * contributions from numerous individuals and organizations.
   6  * Please see the COPYING and CONTRIBUTORS files for details.
   7  */
   8
   9 #ifndef _SQUID_SRC_PARSER_CHARACTERSET_H
  10 #define _SQUID_SRC_PARSER_CHARACTERSET_H
  11
  12 #include <vector>
  13
  14 /// optimized set of C chars, with quick membership test and merge support
  15 class CharacterSet
  16 {
  17 public:
  18     typedef std::vector<uint8_t> Storage;
  19
  20     /// define a character set with the given label ("anonymous" if NULL)
  21     ///  with specified initial contents
  22     CharacterSet(const char *label, const char * const initial);
  23
  24     /// define a character set with the given label ("anonymous" if NULL)
  25     ///  containing characters defined in the supplied ranges
  26     /// \see addRange
  27     CharacterSet(const char *label, unsigned char low, unsigned char high);
  28
  29     /// whether a given character exists in the set
  30     bool operator[](unsigned char c) const {return chars_[static_cast<uint8_t>(c)] != 0;}
  31
  32     /// add a given character to the character set
  33     CharacterSet & add(const unsigned char c);
  34
  35     /// add a list of character ranges, expressed as pairs [low,high], including both ends
  36     CharacterSet & addRange(unsigned char low, unsigned char high);
  37
  38     /// add all characters from the given CharacterSet to this one
  39     CharacterSet &operator +=(const CharacterSet &src);
  40
  41     /// return a new CharacterSet containing the union of two sets
  42     CharacterSet operator +(const CharacterSet &src) const;
  43
  44     /// return a new CharacterSet containing characters not in this set
  45     CharacterSet complement(const char *complementLabel = NULL) const;
  46
  47     /// change name; handy in const declarations that use operators
  48     CharacterSet &rename(const char *label) { name = label; return *this; }
  49
  50     /// optional set label for debugging (default: "anonymous")
  51     const char * name;
  52
  53     // common character sets, RFC 5234
  54     // A-Za-z
  55     static const CharacterSet ALPHA;
  56     // 0-1
  57     static const CharacterSet BIT;
  58     // carriage return
  59     static const CharacterSet CR;
  60     // controls
  61 #if __cplusplus == 201103L
  62     // ready but disabled as needs C++11 constructor
  63     //static const CharacterSet CTL;
  64 #endif
  65     // 0-9
  66     static const CharacterSet DIGIT;
  67     // double quote
  68     static const CharacterSet DQUOTE;
  69     // 0-9aAbBcCdDeEfF
  70     static const CharacterSet HEXDIG;
  71     // horizontal tab
  72     static const CharacterSet HTAB;
  73     // line feed
  74     static const CharacterSet LF;
  75     // white space
  76     static const CharacterSet SP;
  77     // visible (printable) characters
  78     static const CharacterSet VCHAR;
  79     // <space><tab>
  80     static const CharacterSet WSP;
  81
  82     // HTTP character sets, RFC 7230
  83     // ctext
  84 #if __cplusplus == 201103L
  85     // ready but disabled as needs C++11 constructor
  86     //static const CharacterSet CTEXT;
  87 #endif
  88     // XXX: maybe field-vchar = VCHAR / obs-text
  89     // any VCHAR except for SPECIAL
  90     static const CharacterSet TCHAR;
  91     // special VCHARs
  92     static const CharacterSet SPECIAL;
  93     // qdtext
  94 #if __cplusplus == 201103L
  95     // ready but disabled as needs C++11 constructor
  96     //static const CharacterSet QDTEXT;
  97 #endif
  98     // obs-text
  99     static const CharacterSet OBSTEXT;
 100
 101     // HTTP character sets, RFC 7232
 102     // etagc
 103 #if __cplusplus == 201103L
 104     // ready but disabled as needs C++11 constructor
 105     //static const CharacterSet ETAGC;
 106 #endif
 107
 108     // HTTP character sets, RFC 7235
 109     // token68 (internal charaters only, excludes '=' terminator)
 110     static const CharacterSet TOKEN68C;
 111
 112 private:
 113     /** index of characters in this set
 114      *
 115      * \note guaranteed to be always 256 slots big, as forced in the
 116      *  constructor. This assumption is relied upon in operator[], add,
 117      *  operator+=
 118      */
 119     Storage chars_;
 120 };
 121
 122 #endif /* _SQUID_SRC_PARSER_CHARACTERSET_H */
 123