lib/getndelim2.c

   1 /* getndelim2 - Read a line from a stream, stopping at one of 2 delimiters,
   2    with bounded memory allocation.
   3
   4    Copyright (C) 1993, 1996-1998, 2000, 2003-2004, 2006, 2008-2019 Free
   5    Software Foundation, Inc.
   6
   7    This program is free software: you can redistribute it and/or modify
   8    it under the terms of the GNU General Public License as published by
   9    the Free Software Foundation; either version 3 of the License, or
  10    (at your option) any later version.
  11
  12    This program is distributed in the hope that it will be useful,
  13    but WITHOUT ANY WARRANTY; without even the implied warranty of
  14    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  15    GNU General Public License for more details.
  16
  17    You should have received a copy of the GNU General Public License
  18    along with this program.  If not, see <https://www.gnu.org/licenses/>.  */
  19
  20 /* Originally written by Jan Brittenson, bson@gnu.ai.mit.edu.  */
  21
  22 #include <config.h>
  23
  24 #include "getndelim2.h"
  25
  26 #include <stdbool.h>
  27 #include <stddef.h>
  28 #include <stdlib.h>
  29 #include <string.h>
  30
  31 #if USE_UNLOCKED_IO
  32 # include "unlocked-io.h"
  33 #endif
  34 #if !HAVE_FLOCKFILE
  35 # undef flockfile
  36 # define flockfile(x) ((void) 0)
  37 #endif
  38 #if !HAVE_FUNLOCKFILE
  39 # undef funlockfile
  40 # define funlockfile(x) ((void) 0)
  41 #endif
  42
  43 #include <limits.h>
  44 #include <stdint.h>
  45
  46 #include "freadptr.h"
  47 #include "freadseek.h"
  48 #include "memchr2.h"
  49
  50 #ifndef SSIZE_MAX
  51 # define SSIZE_MAX ((ssize_t) (SIZE_MAX / 2))
  52 #endif
  53
  54 /* Use this to suppress gcc's "...may be used before initialized" warnings. */
  55 #if defined GCC_LINT || defined lint
  56 # define IF_LINT(Code) Code
  57 #else
  58 # define IF_LINT(Code) /* empty */
  59 #endif
  60
  61 /* The maximum value that getndelim2 can return without suffering from
  62    overflow problems, either internally (because of pointer
  63    subtraction overflow) or due to the API (because of ssize_t).  */
  64 #define GETNDELIM2_MAXIMUM (PTRDIFF_MAX < SSIZE_MAX ? PTRDIFF_MAX : SSIZE_MAX)
  65
  66 /* Try to add at least this many bytes when extending the buffer.
  67    MIN_CHUNK must be no greater than GETNDELIM2_MAXIMUM.  */
  68 #define MIN_CHUNK 64
  69
  70 ssize_t
  71 getndelim2 (char **lineptr, size_t *linesize, size_t offset, size_t nmax,
  72             int delim1, int delim2, FILE *stream)
  73 {
  74   size_t nbytes_avail;          /* Allocated but unused bytes in *LINEPTR.  */
  75   char *read_pos;               /* Where we're reading into *LINEPTR. */
  76   ssize_t bytes_stored = -1;
  77   char *ptr = *lineptr;
  78   size_t size = *linesize;
  79   bool found_delimiter;
  80
  81   if (!ptr)
  82     {
  83       size = nmax < MIN_CHUNK ? nmax : MIN_CHUNK;
  84       ptr = malloc (size);
  85       if (!ptr)
  86         return -1;
  87     }
  88
  89   if (size < offset)
  90     goto done;
  91
  92   nbytes_avail = size - offset;
  93   read_pos = ptr + offset;
  94
  95   if (nbytes_avail == 0 && nmax <= size)
  96     goto done;
  97
  98   /* Normalize delimiters, since memchr2 doesn't handle EOF.  */
  99   if (delim1 == EOF)
 100     delim1 = delim2;
 101   else if (delim2 == EOF)
 102     delim2 = delim1;
 103
 104   flockfile (stream);
 105
 106   found_delimiter = false;
 107   do
 108     {
 109       /* Here always ptr + size == read_pos + nbytes_avail.
 110          Also nbytes_avail > 0 || size < nmax.  */
 111
 112       int c IF_LINT (= 0);
 113       const char *buffer;
 114       size_t buffer_len;
 115
 116       buffer = freadptr (stream, &buffer_len);
 117       if (buffer)
 118         {
 119           if (delim1 != EOF)
 120             {
 121               const char *end = memchr2 (buffer, delim1, delim2, buffer_len);
 122               if (end)
 123                 {
 124                   buffer_len = end - buffer + 1;
 125                   found_delimiter = true;
 126                 }
 127             }
 128         }
 129       else
 130         {
 131           c = getc (stream);
 132           if (c == EOF)
 133             {
 134               /* Return partial line, if any.  */
 135               if (read_pos == ptr)
 136                 goto unlock_done;
 137               else
 138                 break;
 139             }
 140           if (c == delim1 || c == delim2)
 141             found_delimiter = true;
 142           buffer_len = 1;
 143         }
 144
 145       /* We always want at least one byte left in the buffer, since we
 146          always (unless we get an error while reading the first byte)
 147          NUL-terminate the line buffer.  */
 148
 149       if (nbytes_avail < buffer_len + 1 && size < nmax)
 150         {
 151           /* Grow size proportionally, not linearly, to avoid O(n^2)
 152              running time.  */
 153           size_t newsize = size < MIN_CHUNK ? size + MIN_CHUNK : 2 * size;
 154           char *newptr;
 155
 156           /* Increase newsize so that it becomes
 157              >= (read_pos - ptr) + buffer_len.  */
 158           if (newsize - (read_pos - ptr) < buffer_len + 1)
 159             newsize = (read_pos - ptr) + buffer_len + 1;
 160           /* Respect nmax.  This handles possible integer overflow.  */
 161           if (! (size < newsize && newsize <= nmax))
 162             newsize = nmax;
 163
 164           if (GETNDELIM2_MAXIMUM < newsize - offset)
 165             {
 166               size_t newsizemax = offset + GETNDELIM2_MAXIMUM + 1;
 167               if (size == newsizemax)
 168                 goto unlock_done;
 169               newsize = newsizemax;
 170             }
 171
 172           nbytes_avail = newsize - (read_pos - ptr);
 173           newptr = realloc (ptr, newsize);
 174           if (!newptr)
 175             goto unlock_done;
 176           ptr = newptr;
 177           size = newsize;
 178           read_pos = size - nbytes_avail + ptr;
 179         }
 180
 181       /* Here, if size < nmax, nbytes_avail >= buffer_len + 1.
 182          If size == nmax, nbytes_avail > 0.  */
 183
 184       if (1 < nbytes_avail)
 185         {
 186           size_t copy_len = nbytes_avail - 1;
 187           if (buffer_len < copy_len)
 188             copy_len = buffer_len;
 189           if (buffer)
 190             memcpy (read_pos, buffer, copy_len);
 191           else
 192             *read_pos = c;
 193           read_pos += copy_len;
 194           nbytes_avail -= copy_len;
 195         }
 196
 197       /* Here still nbytes_avail > 0.  */
 198
 199       if (buffer && freadseek (stream, buffer_len))
 200         goto unlock_done;
 201     }
 202   while (!found_delimiter);
 203
 204   /* Done - NUL terminate and return the number of bytes read.
 205      At this point we know that nbytes_avail >= 1.  */
 206   *read_pos = '\0';
 207
 208   bytes_stored = read_pos - (ptr + offset);
 209
 210  unlock_done:
 211   funlockfile (stream);
 212
 213  done:
 214   *lineptr = ptr;
 215   *linesize = size;
 216   return bytes_stored ? bytes_stored : -1;
 217 }