This file is indexed.

/usr/include/jellyfish/int128.hpp is in libjellyfish-2.0-dev 2.1.4-1.

This file is owned by root:root, with mode 0o644.

The actual contents of the file can be viewed below.

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
/*  This file is part of Jellyfish.

    Jellyfish is free software: you can redistribute it and/or modify
    it under the terms of the GNU General Public License as published by
    the Free Software Foundation, either version 3 of the License, or
    (at your option) any later version.

    Jellyfish is distributed in the hope that it will be useful,
    but WITHOUT ANY WARRANTY; without even the implied warranty of
    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
    GNU General Public License for more details.

    You should have received a copy of the GNU General Public License
    along with Jellyfish.  If not, see <http://www.gnu.org/licenses/>.
*/

#ifndef _INT128_H_
#define _INT128_H_

#ifdef HAVE_CONFIG_H
#include <config.h>
#ifndef HAVE_INT128
#error "The type __int128 is not supported"
#endif
#endif

#include <unistd.h>
#include <stdint.h>
#include <iostream>
#include <sstream>
#include <limits>
#include <cstring>

// Output of __int128: this might be slow
namespace __int128_ns {
template<int base>
void __print_digits(std::ostream& os, unsigned __int128 x,
                    bool lower = true) {
  char buf[50];
  char* ptr = buf + sizeof(buf);
  do {
    int o  = x % base;
    if(o < 10) {
      *--ptr = '0' + o;
    } else {
      *--ptr = (lower ? 'a' : 'A') + o - 10;
    }
    x     /= base;
  } while (x > 0);
  os.write(ptr, buf + sizeof(buf) - ptr);
}

inline bool is_negative(unsigned __int128 x) { return false; }
inline bool is_negative(__int128 x) { return x < 0; }

template<typename T>
void __print_decimal(std::ostream& prefix, std::ostream& os, T x,
                     const std::ios::fmtflags& ff) {
  if((ff & std::ios::showpos) && x > 0)
    prefix << "+";
  if(x == 0) {
    os << "0";
    return;
  }
  if(is_negative(x)) {
    prefix << "-";
    x = -x;
  }
  __print_digits<10>(os, x);
}

void __print_bases(std::ostream& prefix, std::ostream& os,
                   unsigned __int128 x, 
                   const std::ios::fmtflags& ff);

template<typename T>
void __print_buf(std::ostream& prefix, std::ostream& os, T x,
                 const std::ios::fmtflags& ff) {
  if(ff & std::ios::dec)
    __print_decimal(prefix, os, x, ff);
  else
    __print_bases(prefix, os, (unsigned __int128)x, ff);
}

template<typename T>
void __print(std::ostream&os, T x) {
  const std::ios_base::fmtflags ff = os.flags();

  if(!(ff & std::ios::adjustfield))
    return __print_buf(os, os, x, ff);

  std::ostringstream prefix;
  std::ostringstream buf;
  __print_buf(prefix, buf, x, ff);
  ssize_t nb_padding = os.width() - (prefix.str().size() + buf.str().size());
  if(nb_padding <= 0) {
    os.write(prefix.str().c_str(), prefix.tellp());
    os.write(buf.str().c_str(), buf.tellp());
    return;
  }

  char padding[nb_padding];
  memset(padding, os.fill(), nb_padding);
  if(ff & std::ios::right)
    os.write(padding, nb_padding);
  os.write(prefix.str().c_str(), prefix.tellp());
  if(ff & std::ios::internal)
    os.write(padding, nb_padding);
  os.write(buf.str().c_str(), buf.tellp());
  if(ff & std::ios::left)
    os.write(padding, nb_padding);
}
}

inline
std::ostream& operator<<(std::ostream& os, __int128 x) {
  __int128_ns::__print(os, x);
  return os;
}

inline
std::ostream& operator<<(std::ostream& os, unsigned __int128 x) {
  __int128_ns::__print(os, x);
  return os;
}

#ifndef HAVE_NUMERIC_LIMITS128
namespace std {
template<>
class numeric_limits<__int128> {
public:
  static const bool is_specialized = true;
  static __int128 max() { return (unsigned __int128)-1 >> 1; }
  static __int128 min() { return  max() + 1; }
  static const int  digits     = 127;
  static const int  digits10   = 38;
#define NLS64 numeric_limits<int64_t>
  static const bool is_signed  = NLS64::is_signed;
  static const bool is_integer = NLS64::is_integer;
  static const bool is_exact   = NLS64::is_exact;
  static const int  radix      = NLS64::radix;
  static __int128 epsilon() { return NLS64::epsilon(); }
  static __int128 round_error() { return NLS64::round_error(); }
  static const int                min_exponent      = NLS64::min_exponent;
  static const int                min_exponent10    = NLS64::min_exponent10;
  static const int                max_exponent      = NLS64::max_exponent;
  static const int                max_exponent10    = NLS64::max_exponent10;
  static const bool               has_infinity      = NLS64::has_infinity;
  static const bool               has_quiet_NaN     = NLS64::has_quiet_NaN;
  static const bool               has_signaling_NaN = NLS64::has_signaling_NaN;
  static const float_denorm_style has_denorm        = NLS64::has_denorm;
  static const bool               has_denorm_loss   = NLS64::has_denorm_loss;
  static __int128 infinity() { return NLS64::infinity(); }
  static __int128 quiet_NaN() { return NLS64::quiet_NaN(); }
  static __int128 signaling_NaN() { return NLS64::signaling_NaN(); }
  static __int128 denorm_min() { return NLS64::denorm_min(); }
  static const bool              is_iec559       = NLS64::is_iec559;
  static const bool              is_bounded      = NLS64::is_bounded;
  static const bool              is_modulo       = NLS64::is_modulo;
  static const bool              traps           = NLS64::traps;
  static const bool              tinyness_before = NLS64::tinyness_before;
  static const float_round_style round_style     = NLS64::round_style;
};

template<>
class numeric_limits<unsigned __int128> {
public:
  static const bool is_specialized = true;
  static __int128 max() { return (unsigned __int128)-1; }
  static __int128 min() { return  0; }
  static const int  digits     = 128;
  static const int  digits10   = 39;
#define NLU64 numeric_limits<uint64_t>
  static const bool is_signed  = NLU64::is_signed;
  static const bool is_integer = NLU64::is_integer;
  static const bool is_exact   = NLU64::is_exact;
  static const int  radix      = NLU64::radix;
  static __int128 epsilon() { return NLU64::epsilon(); }
  static __int128 round_error() { return NLU64::round_error(); }
  static const int                min_exponent      = NLU64::min_exponent;
  static const int                min_exponent10    = NLU64::min_exponent10;
  static const int                max_exponent      = NLU64::max_exponent;
  static const int                max_exponent10    = NLU64::max_exponent10;
  static const bool               has_infinity      = NLU64::has_infinity;
  static const bool               has_quiet_NaN     = NLU64::has_quiet_NaN;
  static const bool               has_signaling_NaN = NLU64::has_signaling_NaN;
  static const float_denorm_style has_denorm        = NLU64::has_denorm;
  static const bool               has_denorm_loss   = NLU64::has_denorm_loss;
  static __int128 infinity() { return NLU64::infinity(); }
  static __int128 quiet_NaN() { return NLU64::quiet_NaN(); }
  static __int128 signaling_NaN() { return NLU64::signaling_NaN(); }
  static __int128 denorm_min() { return NLU64::denorm_min(); }
  static const bool              is_iec559       = NLU64::is_iec559;
  static const bool              is_bounded      = NLU64::is_bounded;
  static const bool              is_modulo       = NLU64::is_modulo;
  static const bool              traps           = NLU64::traps;
  static const bool              tinyness_before = NLU64::tinyness_before;
  static const float_round_style round_style     = NLU64::round_style;
};
} // namespace std
#endif /* HAVE_NUMERIC_LIMITS128 */

#endif /* _INT128_H_ */