1 /* Commandline diff utility to test diff implementations. */
2 /*
3  * Copyright (c) 2018 Martin Pieuchot
4  * Copyright (c) 2020 Neels Hofmeyr <neels@hofmeyr.de>
5  *
6  * Permission to use, copy, modify, and distribute this software for any
7  * purpose with or without fee is hereby granted, provided that the above
8  * copyright notice and this permission notice appear in all copies.
9  *
10  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
11  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
12  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
13  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
14  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
15  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
16  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
17  */
18 
19 #include <sys/mman.h>
20 #include <sys/stat.h>
21 #include <sys/types.h>
22 
23 #include <err.h>
24 #include <fcntl.h>
25 #include <inttypes.h>
26 #include <stdio.h>
27 #include <stdlib.h>
28 #include <stdbool.h>
29 #include <string.h>
30 #include <unistd.h>
31 
32 #include <arraylist.h>
33 #include <diff_main.h>
34 #include <diff_output.h>
35 
36 enum diffreg_algo {
37 	DIFFREG_ALGO_MYERS_THEN_MYERS_DIVIDE = 0,
38 	DIFFREG_ALGO_MYERS_THEN_PATIENCE = 1,
39 	DIFFREG_ALGO_PATIENCE = 2,
40 	DIFFREG_ALGO_NONE = 3,
41 };
42 
43 __dead void	 usage(void);
44 int		 diffreg(char *, char *, enum diffreg_algo, bool, bool,
45 			 int, bool);
46 FILE *		 openfile(const char *, char **, struct stat *);
47 
48 __dead void
49 usage(void)
50 {
51 	fprintf(stderr,
52 		"usage: %s [-pPQTwe] [-U n] file1 file2\n"
53 		"\n"
54 		"  -p   Show function prototypes in hunk headers\n"
55 		"  -P   Use Patience Diff (slower but often nicer)\n"
56 		"  -Q   Use forward-Myers for small files, otherwise Patience\n"
57 		"  -T   Trivial algo: detect similar start and end only\n"
58 		"  -w   Ignore Whitespace\n"
59 		"  -U n Number of Context Lines\n"
60 		"  -e   Produce ed script output\n"
61 		, getprogname());
62 	exit(1);
63 }
64 
65 int
66 main(int argc, char *argv[])
67 {
68 	int ch, rc;
69 	bool ignore_whitespace = false;
70 	bool show_function_prototypes = false;
71 	bool edscript = false;
72 	int context_lines = 3;
73 	enum diffreg_algo algo = DIFFREG_ALGO_MYERS_THEN_MYERS_DIVIDE;
74 
75 	while ((ch = getopt(argc, argv, "pPQTwU:e")) != -1) {
76 		switch (ch) {
77 		case 'p':
78 			show_function_prototypes = true;
79 			break;
80 		case 'P':
81 			algo = DIFFREG_ALGO_PATIENCE;
82 			break;
83 		case 'Q':
84 			algo = DIFFREG_ALGO_MYERS_THEN_PATIENCE;
85 			break;
86 		case 'T':
87 			algo = DIFFREG_ALGO_NONE;
88 			break;
89 		case 'w':
90 			ignore_whitespace = true;
91 			break;
92 		case 'U':
93 			context_lines = atoi(optarg);
94 			break;
95 		case 'e':
96 			edscript = true;
97 			break;
98 		default:
99 			usage();
100 		}
101 	}
102 
103 	argc -= optind;
104 	argv += optind;
105 
106 	if (argc != 2)
107 		usage();
108 
109 	rc = diffreg(argv[0], argv[1], algo, ignore_whitespace,
110 	    show_function_prototypes, context_lines, edscript);
111 	if (rc != DIFF_RC_OK) {
112 		fprintf(stderr, "diff: %s\n", strerror(rc));
113 		return 1;
114 	}
115 	return 0;
116 }
117 
118 const struct diff_algo_config myers_then_patience;
119 const struct diff_algo_config myers_then_myers_divide;
120 const struct diff_algo_config patience;
121 const struct diff_algo_config myers_divide;
122 
123 const struct diff_algo_config myers_then_patience = (struct diff_algo_config){
124 	.impl = diff_algo_myers,
125 	.permitted_state_size = 1024 * 1024 * sizeof(int),
126 	.fallback_algo = &patience,
127 };
128 
129 const struct diff_algo_config myers_then_myers_divide =
130 	(struct diff_algo_config){
131 	.impl = diff_algo_myers,
132 	.permitted_state_size = 1024 * 1024 * sizeof(int),
133 	.fallback_algo = &myers_divide,
134 };
135 
136 const struct diff_algo_config patience = (struct diff_algo_config){
137 	.impl = diff_algo_patience,
138 	/* After subdivision, do Patience again: */
139 	.inner_algo = &patience,
140 	/* If subdivision failed, do Myers Divide et Impera: */
141 	.fallback_algo = &myers_then_myers_divide,
142 };
143 
144 const struct diff_algo_config myers_divide = (struct diff_algo_config){
145 	.impl = diff_algo_myers_divide,
146 	/* When division succeeded, start from the top: */
147 	.inner_algo = &myers_then_myers_divide,
148 	/* (fallback_algo = NULL implies diff_algo_none). */
149 };
150 
151 const struct diff_algo_config no_algo = (struct diff_algo_config){
152 	.impl = diff_algo_none,
153 };
154 
155 /* If the state for a forward-Myers is small enough, use Myers, otherwise first
156  * do a Myers-divide. */
157 const struct diff_config diff_config_myers_then_myers_divide = {
158 	.atomize_func = diff_atomize_text_by_line,
159 	.algo = &myers_then_myers_divide,
160 };
161 
162 /* If the state for a forward-Myers is small enough, use Myers, otherwise first
163  * do a Patience. */
164 const struct diff_config diff_config_myers_then_patience = {
165 	.atomize_func = diff_atomize_text_by_line,
166 	.algo = &myers_then_patience,
167 };
168 
169 /* Directly force Patience as a first divider of the source file. */
170 const struct diff_config diff_config_patience = {
171 	.atomize_func = diff_atomize_text_by_line,
172 	.algo = &patience,
173 };
174 
175 /* Directly force Patience as a first divider of the source file. */
176 const struct diff_config diff_config_no_algo = {
177 	.atomize_func = diff_atomize_text_by_line,
178 };
179 
180 int
181 diffreg(char *file1, char *file2, enum diffreg_algo algo, bool ignore_whitespace,
182     bool show_function_prototypes, int context_lines, bool edscript)
183 {
184 	char *str1, *str2;
185 	FILE *f1, *f2;
186 	struct stat st1, st2;
187 	struct diff_input_info info = {
188 		.left_path = file1,
189 		.right_path = file2,
190 	};
191 	struct diff_result *result;
192 	int rc;
193 	const struct diff_config *cfg;
194 	int diff_flags = 0;
195 	
196 	switch (algo) {
197 	default:
198 	case DIFFREG_ALGO_MYERS_THEN_MYERS_DIVIDE:
199 		cfg = &diff_config_myers_then_myers_divide;
200 		break;
201 	case DIFFREG_ALGO_MYERS_THEN_PATIENCE:
202 		cfg = &diff_config_myers_then_patience;
203 		break;
204 	case DIFFREG_ALGO_PATIENCE:
205 		cfg = &diff_config_patience;
206 		break;
207 	case DIFFREG_ALGO_NONE:
208 		cfg = &diff_config_no_algo;
209 		break;
210 	}
211 
212 	f1 = openfile(file1, &str1, &st1);
213 	f2 = openfile(file2, &str2, &st2);
214 
215 	if (ignore_whitespace)
216 		diff_flags |= DIFF_FLAG_IGNORE_WHITESPACE;
217 	if (show_function_prototypes)
218 		diff_flags |= DIFF_FLAG_SHOW_PROTOTYPES;
219 
220 	result = diff_main(cfg, f1, str1, st1.st_size, f2, str2, st2.st_size,
221 	    diff_flags);
222 #if 0
223 	rc = diff_output_plain(stdout, &info, result);
224 #else
225 	if (edscript)
226 		rc = diff_output_edscript(NULL, stdout, &info, result);
227 	else {
228 		rc = diff_output_unidiff(NULL, stdout, &info, result,
229 		    context_lines);
230 	}
231 #endif
232 	diff_result_free(result);
233 
234 	if (str1)
235 		munmap(str1, st1.st_size);
236 	if (str2)
237 		munmap(str2, st2.st_size);
238 	fclose(f1);
239 	fclose(f2);
240 
241 	return rc;
242 }
243 
244 FILE *
245 openfile(const char *path, char **p, struct stat *st)
246 {
247 	FILE *f = NULL;
248 
249 	f = fopen(path, "r");
250 	if (f == NULL)
251 		err(2, "%s", path);
252 
253 	if (fstat(fileno(f), st) == -1)
254 		err(2, "%s", path);
255 
256 #ifndef DIFF_NO_MMAP
257 	*p = mmap(NULL, st->st_size, PROT_READ, MAP_PRIVATE, fileno(f), 0);
258 	if (*p == MAP_FAILED)
259 #endif
260 		*p = NULL; /* fall back on file I/O */
261 
262 	return f;
263 }