src/base/CharacterSet.cc

   1 /*
   2  * Copyright (C) 1996-2017 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 #include "squid.h"
  10 #include "CharacterSet.h"
  11
  12 #include <algorithm>
  13 #include <iostream>
  14 #include <functional>
  15
  16 CharacterSet &
  17 CharacterSet::operator +=(const CharacterSet &src)
  18 {
  19     Storage::const_iterator s = src.chars_.begin();
  20     const Storage::const_iterator e = src.chars_.end();
  21     Storage::iterator d = chars_.begin();
  22     while (s != e) {
  23         if (*s)
  24             *d = 1;
  25         ++s;
  26         ++d;
  27     }
  28     return *this;
  29 }
  30
  31 CharacterSet &
  32 CharacterSet::operator -=(const CharacterSet &src)
  33 {
  34     Storage::const_iterator s = src.chars_.begin();
  35     const Storage::const_iterator e = src.chars_.end();
  36     Storage::iterator d = chars_.begin();
  37     while (s != e) {
  38         if (*s)
  39             *d = 0;
  40         ++s;
  41         ++d;
  42     }
  43     return *this;
  44 }
  45
  46 CharacterSet &
  47 CharacterSet::add(const unsigned char c)
  48 {
  49     chars_[static_cast<uint8_t>(c)] = 1;
  50     return *this;
  51 }
  52
  53 CharacterSet &
  54 CharacterSet::addRange(unsigned char low, unsigned char high)
  55 {
  56     //manual loop splitting is needed to cover case where high is 255
  57     // otherwise low will wrap, resulting in infinite loop
  58     while (low < high) {
  59         chars_[static_cast<uint8_t>(low)] = 1;
  60         ++low;
  61     }
  62     chars_[static_cast<uint8_t>(high)] = 1;
  63     return *this;
  64 }
  65
  66 CharacterSet
  67 CharacterSet::complement(const char *label) const
  68 {
  69     CharacterSet result((label ? label : "complement_of_some_other_set"), "");
  70     // negate each of our elements and add them to the result storage
  71     std::transform(chars_.begin(), chars_.end(), result.chars_.begin(),
  72                    std::logical_not<Storage::value_type>());
  73     return result;
  74 }
  75
  76 CharacterSet::CharacterSet(const char *label, const char * const c) :
  77     name(label ? label: "anonymous"),
  78     chars_(Storage(256,0))
  79 {
  80     const size_t clen = strlen(c);
  81     for (size_t i = 0; i < clen; ++i)
  82         add(c[i]);
  83 }
  84
  85 CharacterSet::CharacterSet(const char *label, unsigned char low, unsigned char high) :
  86     name(label ? label: "anonymous"),
  87     chars_(Storage(256,0))
  88 {
  89     addRange(low,high);
  90 }
  91
  92 CharacterSet::CharacterSet(const char *label, std::initializer_list<std::pair<uint8_t, uint8_t>> ranges) :
  93     name(label ? label: "anonymous"),
  94     chars_(Storage(256,0))
  95 {
  96     for (auto range: ranges)
  97         addRange(range.first, range.second);
  98 }
  99
 100 CharacterSet
 101 operator+ (CharacterSet lhs, const CharacterSet &rhs)
 102 {
 103     lhs += rhs;
 104     return lhs;
 105 }
 106
 107 CharacterSet
 108 operator- (CharacterSet lhs, const CharacterSet &rhs)
 109 {
 110     lhs -= rhs;
 111     return lhs;
 112 }
 113
 114 std::ostream&
 115 operator <<(std::ostream &s, const CharacterSet &c)
 116 {
 117     s << "CharacterSet(" << c.name << ')';
 118     return s;
 119 }
 120
 121 const CharacterSet
 122 // RFC 5234
 123 CharacterSet::ALPHA("ALPHA", "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"),
 124              CharacterSet::BIT("BIT","01"),
 125              CharacterSet::CR("CR","\r"),
 126 CharacterSet::CTL("CTL", {{0x01,0x1f},{0x7f,0x7f}}),
 127 CharacterSet::DIGIT("DIGIT","0123456789"),
 128 CharacterSet::DQUOTE("DQUOTE","\""),
 129 CharacterSet::HEXDIG("HEXDIG","0123456789aAbBcCdDeEfF"),
 130 CharacterSet::HTAB("HTAB","\t"),
 131 CharacterSet::LF("LF","\n"),
 132 CharacterSet::SP("SP"," "),
 133 CharacterSet::VCHAR("VCHAR", 0x21, 0x7e),
 134 // RFC 7230
 135 CharacterSet::WSP("WSP"," \t"),
 136 CharacterSet::CTEXT("ctext", {{0x09,0x09},{0x20,0x20},{0x2a,0x5b},{0x5d,0x7e},{0x80,0xff}}),
 137 CharacterSet::TCHAR("TCHAR","!#$%&'*+-.^_`|~0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ"),
 138 CharacterSet::SPECIAL("SPECIAL","()<>@,;:\\\"/[]?={}"),
 139 CharacterSet::QDTEXT("QDTEXT", {{0x09,0x09},{0x20,0x21},{0x23,0x5b},{0x5d,0x7e},{0x80,0xff}}),
 140 CharacterSet::OBSTEXT("OBSTEXT",0x80,0xff),
 141 // RFC 7232
 142 CharacterSet::ETAGC("ETAGC", {{0x21,0x21},{0x23,0x7e},{0x80,0xff}}),
 143 // RFC 7235
 144 CharacterSet::TOKEN68C("TOKEN68C","-._~+/0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ")
 145 ;
 146