1 /* readtokens.c -- Functions for reading tokens from an input stream.
2
3 Copyright (C) 1990-1991, 1999-2004, 2006, 2009-2021 Free Software
4 Foundation, Inc.
5
6 This program is free software: you can redistribute it and/or modify
7 it under the terms of the GNU General Public License as published by
8 the Free Software Foundation; either version 3 of the License, or
9 (at your option) any later version.
10
11 This program is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
14 GNU General Public License for more details.
15
16 You should have received a copy of the GNU General Public License
17 along with this program. If not, see <https://www.gnu.org/licenses/>.
18
19 Written by Jim Meyering. */
20
21 /* This almost supersedes xreadline stuff -- using delim="\n"
22 gives the same functionality, except that these functions
23 would never return empty lines. */
24
25 #include <config.h>
26
27 #include "readtokens.h"
28
29 #include <limits.h>
30 #include <stdio.h>
31 #include <stdlib.h>
32 #include <string.h>
33 #include <stdbool.h>
34
35 #include "xalloc.h"
36
37 #if USE_UNLOCKED_IO
38 # include "unlocked-io.h"
39 #endif
40
41 /* Initialize a tokenbuffer. */
42
43 void
init_tokenbuffer(token_buffer * tokenbuffer)44 init_tokenbuffer (token_buffer *tokenbuffer)
45 {
46 tokenbuffer->size = 0;
47 tokenbuffer->buffer = NULL;
48 }
49
50 typedef size_t word;
51 enum { bits_per_word = sizeof (word) * CHAR_BIT };
52
53 static bool
get_nth_bit(size_t n,word const * bitset)54 get_nth_bit (size_t n, word const *bitset)
55 {
56 return bitset[n / bits_per_word] >> n % bits_per_word & 1;
57 }
58
59 static void
set_nth_bit(size_t n,word * bitset)60 set_nth_bit (size_t n, word *bitset)
61 {
62 size_t one = 1;
63 bitset[n / bits_per_word] |= one << n % bits_per_word;
64 }
65
66 /* Read a token from STREAM into TOKENBUFFER.
67 A token is delimited by any of the N_DELIM bytes in DELIM.
68 Upon return, the token is in tokenbuffer->buffer and
69 has a trailing '\0' instead of any original delimiter.
70 The function value is the length of the token not including
71 the final '\0'. Upon EOF (i.e. on the call after the last
72 token is read) or error, return -1 without modifying tokenbuffer.
73 The EOF and error conditions may be distinguished in the caller
74 by testing ferror (STREAM).
75
76 This function works properly on lines containing NUL bytes
77 and on files that do not end with a delimiter. */
78
79 size_t
readtoken(FILE * stream,const char * delim,size_t n_delim,token_buffer * tokenbuffer)80 readtoken (FILE *stream,
81 const char *delim,
82 size_t n_delim,
83 token_buffer *tokenbuffer)
84 {
85 int c;
86 idx_t i;
87 word isdelim[(UCHAR_MAX + bits_per_word) / bits_per_word];
88
89 memset (isdelim, 0, sizeof isdelim);
90 for (i = 0; i < n_delim; i++)
91 {
92 unsigned char ch = delim[i];
93 set_nth_bit (ch, isdelim);
94 }
95
96 /* skip over any leading delimiters */
97 for (c = getc (stream); c >= 0 && get_nth_bit (c, isdelim); c = getc (stream))
98 {
99 /* empty */
100 }
101
102 char *p = tokenbuffer->buffer;
103 idx_t n = tokenbuffer->size;
104 i = 0;
105 for (;;)
106 {
107 if (c < 0 && i == 0)
108 return -1;
109
110 if (i == n)
111 p = xpalloc (p, &n, 1, -1, sizeof *p);
112
113 if (c < 0)
114 {
115 p[i] = 0;
116 break;
117 }
118 if (get_nth_bit (c, isdelim))
119 {
120 p[i] = 0;
121 break;
122 }
123 p[i++] = c;
124 c = getc (stream);
125 }
126
127 tokenbuffer->buffer = p;
128 tokenbuffer->size = n;
129 return i;
130 }
131
132 /* Build a NULL-terminated array of pointers to tokens
133 read from STREAM. Return the number of tokens read.
134 All storage is obtained through calls to xmalloc-like functions.
135
136 %%% Question: is it worth it to do a single
137 %%% realloc() of 'tokens' just before returning? */
138
139 size_t
readtokens(FILE * stream,size_t projected_n_tokens,const char * delim,size_t n_delim,char *** tokens_out,size_t ** token_lengths)140 readtokens (FILE *stream,
141 size_t projected_n_tokens,
142 const char *delim,
143 size_t n_delim,
144 char ***tokens_out,
145 size_t **token_lengths)
146 {
147 token_buffer tb, *token = &tb;
148 char **tokens;
149 size_t *lengths;
150 idx_t sz, n_tokens;
151
152 if (projected_n_tokens == 0)
153 projected_n_tokens = 64;
154 else
155 projected_n_tokens++; /* add one for trailing NULL pointer */
156
157 sz = projected_n_tokens;
158 tokens = xnmalloc (sz, sizeof *tokens);
159 lengths = xnmalloc (sz, sizeof *lengths);
160
161 n_tokens = 0;
162 init_tokenbuffer (token);
163 for (;;)
164 {
165 char *tmp;
166 size_t token_length = readtoken (stream, delim, n_delim, token);
167 if (n_tokens >= sz)
168 {
169 tokens = xpalloc (tokens, &sz, 1, -1, sizeof *tokens);
170 lengths = xreallocarray (lengths, sz, sizeof *lengths);
171 }
172
173 if (token_length == (size_t) -1)
174 {
175 /* don't increment n_tokens for NULL entry */
176 tokens[n_tokens] = NULL;
177 lengths[n_tokens] = 0;
178 break;
179 }
180 tmp = xnmalloc (token_length + 1, sizeof *tmp);
181 lengths[n_tokens] = token_length;
182 tokens[n_tokens] = memcpy (tmp, token->buffer, token_length + 1);
183 n_tokens++;
184 }
185
186 free (token->buffer);
187 *tokens_out = tokens;
188 if (token_lengths != NULL)
189 *token_lengths = lengths;
190 else
191 free (lengths);
192 return n_tokens;
193 }
194