src/base/CharacterSet.h

   1 /*
   2  * Copyright (C) 1996-2018 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 <initializer_list>
  13 #include <iosfwd>
  14 #include <vector>
  15
  16 /// optimized set of C chars, with quick membership test and merge support
  17 class CharacterSet
  18 {
  19 public:
  20     typedef std::vector<uint8_t> Storage;
  21
  22     /// a character set with a given label and contents
  23     explicit CharacterSet(const char *label = "anonymous", const char * const chars = "");
  24
  25     /// define a character set with the given label ("anonymous" if nullptr)
  26     ///  containing characters defined in the supplied ranges
  27     /// \see addRange
  28     CharacterSet(const char *label, unsigned char low, unsigned char high);
  29
  30     /// define a character set with the given label ("anonymous" if nullptr)
  31     ///  containing characters defined in the supplied list of low-high ranges
  32     /// \see addRange
  33     CharacterSet(const char *label, std::initializer_list<std::pair<uint8_t,uint8_t>> ranges);
  34
  35     /// whether the set lacks any members
  36     bool isEmpty() const { return chars_.empty(); }
  37
  38     /// whether a given character exists in the set
  39     bool operator[](unsigned char c) const {return chars_[static_cast<uint8_t>(c)] != 0;}
  40
  41     /// add a given character to the character set
  42     CharacterSet & add(const unsigned char c);
  43
  44     /// add a list of character ranges, expressed as pairs [low,high], including both ends
  45     CharacterSet & addRange(unsigned char low, unsigned char high);
  46
  47     /// set addition: add to this set all characters that are also in rhs
  48     CharacterSet &operator +=(const CharacterSet &rhs);
  49
  50     /// set subtraction: remove all characters that are also in rhs
  51     CharacterSet &operator -=(const CharacterSet &rhs);
  52
  53     /// return a new CharacterSet containing characters not in this set
  54     ///  use the supplied label if provided, default is "complement_of_some_other_set"
  55     CharacterSet complement(const char *complementLabel = nullptr) const;
  56
  57     /// change name; handy in const declarations that use operators
  58     CharacterSet &rename(const char *label) { name = label; return *this; }
  59
  60     /// \note Ignores label
  61     bool operator == (const CharacterSet &cs) const { return chars_ == cs.chars_; }
  62     /// \note Ignores label
  63     bool operator != (const CharacterSet &cs) const { return !operator==(cs); }
  64
  65     /// prints all chars in arbitrary order, without any quoting/escaping
  66     void printChars(std::ostream &os) const;
  67
  68     /// optional set label for debugging (default: "anonymous")
  69     const char * name;
  70
  71     // common character sets, RFC 5234
  72     // A-Za-z
  73     static const CharacterSet ALPHA;
  74     // 0-1
  75     static const CharacterSet BIT;
  76     // carriage return
  77     static const CharacterSet CR;
  78     // controls
  79     static const CharacterSet CTL;
  80     // 0-9
  81     static const CharacterSet DIGIT;
  82     // double quote
  83     static const CharacterSet DQUOTE;
  84     // 0-9aAbBcCdDeEfF
  85     static const CharacterSet HEXDIG;
  86     // horizontal tab
  87     static const CharacterSet HTAB;
  88     // line feed
  89     static const CharacterSet LF;
  90     // white space
  91     static const CharacterSet SP;
  92     // visible (printable) characters
  93     static const CharacterSet VCHAR;
  94     // <space><tab>
  95     static const CharacterSet WSP;
  96
  97     // HTTP character sets, RFC 7230
  98     // ctext
  99     static const CharacterSet CTEXT;
 100     // XXX: maybe field-vchar = VCHAR / obs-text
 101     // any VCHAR except for SPECIAL
 102     static const CharacterSet TCHAR;
 103     // special VCHARs
 104     static const CharacterSet SPECIAL;
 105     // qdtext
 106     static const CharacterSet QDTEXT;
 107     // obs-text
 108     static const CharacterSet OBSTEXT;
 109
 110     // HTTP character sets, RFC 7232
 111     // etagc
 112     static const CharacterSet ETAGC;
 113
 114     // HTTP character sets, RFC 7235
 115     // token68 (internal charaters only, excludes '=' terminator)
 116     static const CharacterSet TOKEN68C;
 117
 118 private:
 119     /** index of characters in this set
 120      *
 121      * \note guaranteed to be always 256 slots big, as forced in the
 122      *  constructor. This assumption is relied upon in various methods
 123      */
 124     Storage chars_;
 125 };
 126
 127 /** CharacterSet addition
 128  *
 129  * \return a new CharacterSet containing all characters present both in lhs
 130  *  and rhs, labeled as lhs is
 131  */
 132 CharacterSet
 133 operator+ (CharacterSet lhs, const CharacterSet &rhs);
 134
 135 /** CharacterSet subtraction
 136  *
 137  * \return a new CharacterSet containing all characters present in lhs
 138  *  and not present in rhs, labeled as lhs is
 139  */
 140 CharacterSet
 141 operator- (CharacterSet lhs, const CharacterSet &rhs);
 142
 143 std::ostream&
 144 operator <<(std::ostream &, const CharacterSet &);
 145
 146 #endif /* _SQUID_SRC_PARSER_CHARACTERSET_H */
 147