1 /* readtokens.c -- Functions for reading tokens from an input stream.
3 Copyright (C) 1990-1991, 1999-2004, 2006, 2009-2018 Free Software
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.
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.
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/>.
19 Written by Jim Meyering. */
21 /* This almost supersedes xreadline stuff -- using delim="\n"
22 gives the same functionality, except that these functions
23 would never return empty lines. */
27 #include "readtokens.h"
38 # include "unlocked-io.h"
41 /* Initialize a tokenbuffer. */
44 init_tokenbuffer (token_buffer
*tokenbuffer
)
46 tokenbuffer
->size
= 0;
47 tokenbuffer
->buffer
= NULL
;
51 enum { bits_per_word
= sizeof (word
) * CHAR_BIT
};
54 get_nth_bit (size_t n
, word
const *bitset
)
56 return bitset
[n
/ bits_per_word
] >> n
% bits_per_word
& 1;
60 set_nth_bit (size_t n
, word
*bitset
)
63 bitset
[n
/ bits_per_word
] |= one
<< n
% bits_per_word
;
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).
76 This function works properly on lines containing NUL bytes
77 and on files that do not end with a delimiter. */
80 readtoken (FILE *stream
,
83 token_buffer
*tokenbuffer
)
88 word isdelim
[(UCHAR_MAX
+ bits_per_word
) / bits_per_word
];
90 memset (isdelim
, 0, sizeof isdelim
);
91 for (i
= 0; i
< n_delim
; i
++)
93 unsigned char ch
= delim
[i
];
94 set_nth_bit (ch
, isdelim
);
97 /* skip over any leading delimiters */
98 for (c
= getc (stream
); c
>= 0 && get_nth_bit (c
, isdelim
); c
= getc (stream
))
103 p
= tokenbuffer
->buffer
;
104 n
= tokenbuffer
->size
;
112 p
= x2nrealloc (p
, &n
, sizeof *p
);
119 if (get_nth_bit (c
, isdelim
))
128 tokenbuffer
->buffer
= p
;
129 tokenbuffer
->size
= n
;
133 /* Build a NULL-terminated array of pointers to tokens
134 read from STREAM. Return the number of tokens read.
135 All storage is obtained through calls to xmalloc-like functions.
137 %%% Question: is it worth it to do a single
138 %%% realloc() of 'tokens' just before returning? */
141 readtokens (FILE *stream
,
142 size_t projected_n_tokens
,
146 size_t **token_lengths
)
148 token_buffer tb
, *token
= &tb
;
154 if (projected_n_tokens
== 0)
155 projected_n_tokens
= 64;
157 projected_n_tokens
++; /* add one for trailing NULL pointer */
159 sz
= projected_n_tokens
;
160 tokens
= xnmalloc (sz
, sizeof *tokens
);
161 lengths
= xnmalloc (sz
, sizeof *lengths
);
164 init_tokenbuffer (token
);
168 size_t token_length
= readtoken (stream
, delim
, n_delim
, token
);
171 tokens
= x2nrealloc (tokens
, &sz
, sizeof *tokens
);
172 lengths
= xnrealloc (lengths
, sz
, sizeof *lengths
);
175 if (token_length
== (size_t) -1)
177 /* don't increment n_tokens for NULL entry */
178 tokens
[n_tokens
] = NULL
;
179 lengths
[n_tokens
] = 0;
182 tmp
= xnmalloc (token_length
+ 1, sizeof *tmp
);
183 lengths
[n_tokens
] = token_length
;
184 tokens
[n_tokens
] = memcpy (tmp
, token
->buffer
, token_length
+ 1);
188 free (token
->buffer
);
189 *tokens_out
= tokens
;
190 if (token_lengths
!= NULL
)
191 *token_lengths
= lengths
;