1 /* Split source by line breaks, and calculate a simplistic checksum. */
3 * Copyright (c) 2020 Neels Hofmeyr <neels@hofmeyr.de>
5 * Permission to use, copy, modify, and distribute this software for any
6 * purpose with or without fee is hereby granted, provided that the above
7 * copyright notice and this permission notice appear in all copies.
9 * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
10 * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
11 * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
12 * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
13 * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
14 * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
15 * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
26 #include <arraylist.h>
27 #include <diff_main.h>
29 #include "diff_internal.h"
30 #include "diff_debug.h"
33 diff_atom_hash_update(unsigned int hash, unsigned char atom_byte)
35 return hash * 23 + atom_byte;
39 diff_data_atomize_text_lines_fd(struct diff_data *d)
42 const off_t end = pos + d->len;
43 unsigned int array_size_estimate = d->len / 50;
44 unsigned int pow2 = 1;
45 bool ignore_whitespace = (d->diff_flags & DIFF_FLAG_IGNORE_WHITESPACE);
46 bool embedded_nul = false;
48 while (array_size_estimate >>= 1)
51 ARRAYLIST_INIT(d->atoms, 1 << pow2);
53 if (fseek(d->root->f, 0L, SEEK_SET) == -1)
58 unsigned int hash = 0;
59 unsigned char buf[512];
61 struct diff_atom *atom;
64 while (eol == 0 && line_end < end) {
65 r = fread(buf, sizeof(char), sizeof(buf), d->root->f);
66 if (r == 0 && ferror(d->root->f))
69 while (eol == 0 && i < r) {
70 if (buf[i] != '\r' && buf[i] != '\n') {
71 if (!ignore_whitespace
72 || !isspace((unsigned char)buf[i]))
73 hash = diff_atom_hash_update(
84 /* When not at the end of data, the line ending char ('\r' or
85 * '\n') must follow */
88 /* If that was an '\r', also pull in any following '\n' */
89 if (line_end < end && eol == '\r') {
90 if (fseeko(d->root->f, line_end, SEEK_SET) == -1)
92 r = fread(buf, sizeof(char), sizeof(buf), d->root->f);
93 if (r == 0 && ferror(d->root->f))
95 if (r > 0 && buf[0] == '\n')
99 /* Record the found line as diff atom */
100 ARRAYLIST_ADD(atom, d->atoms);
104 *atom = (struct diff_atom){
107 .at = NULL, /* atom data is not memory-mapped */
108 .len = line_end - pos,
112 /* Starting point for next line: */
114 if (fseeko(d->root->f, pos, SEEK_SET) == -1)
118 /* File are considered binary if they contain embedded '\0' bytes. */
120 d->atomizer_flags |= DIFF_ATOMIZER_FOUND_BINARY_DATA;
126 diff_data_atomize_text_lines_mmap(struct diff_data *d)
128 const uint8_t *pos = d->data;
129 const uint8_t *end = pos + d->len;
130 bool ignore_whitespace = (d->diff_flags & DIFF_FLAG_IGNORE_WHITESPACE);
131 bool embedded_nul = false;
132 unsigned int array_size_estimate = d->len / 50;
133 unsigned int pow2 = 1;
134 while (array_size_estimate >>= 1)
137 ARRAYLIST_INIT(d->atoms, 1 << pow2);
140 const uint8_t *line_end = pos;
141 unsigned int hash = 0;
143 while (line_end < end && *line_end != '\r' && *line_end != '\n') {
144 if (!ignore_whitespace
145 || !isspace((unsigned char)*line_end))
146 hash = diff_atom_hash_update(hash, *line_end);
147 if (*line_end == '\0')
152 /* When not at the end of data, the line ending char ('\r' or
153 * '\n') must follow */
154 if (line_end < end && *line_end == '\r')
156 if (line_end < end && *line_end == '\n')
159 /* Record the found line as diff atom */
160 struct diff_atom *atom;
161 ARRAYLIST_ADD(atom, d->atoms);
165 *atom = (struct diff_atom){
167 .pos = (off_t)(pos - d->data),
169 .len = line_end - pos,
173 /* Starting point for next line: */
177 /* File are considered binary if they contain embedded '\0' bytes. */
179 d->atomizer_flags |= DIFF_ATOMIZER_FOUND_BINARY_DATA;
185 diff_data_atomize_text_lines(struct diff_data *d)
188 return diff_data_atomize_text_lines_fd(d);
190 return diff_data_atomize_text_lines_mmap(d);
194 diff_atomize_text_by_line(void *func_data, struct diff_data *d)
196 return diff_data_atomize_text_lines(d);