1 /* Generic infrastructure to implement various diff algorithms. */
2 /*
3  * Copyright (c) 2020 Neels Hofmeyr <neels@hofmeyr.de>
4  *
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.
8  *
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.
16  */
17 
18 #ifndef MAX
19 #define MAX(A,B) ((A)>(B)?(A):(B))
20 #endif
21 #ifndef MIN
22 #define MIN(A,B) ((A)<(B)?(A):(B))
23 #endif
24 
25 static inline bool
26 diff_range_empty(const struct diff_range *r)
27 {
28 	return r->start == r->end;
29 }
30 
31 static inline bool
32 diff_ranges_touch(const struct diff_range *a, const struct diff_range *b)
33 {
34 	return (a->end >= b->start) && (a->start <= b->end);
35 }
36 
37 static inline void
38 diff_ranges_merge(struct diff_range *a, const struct diff_range *b)
39 {
40 	*a = (struct diff_range){
41 		.start = MIN(a->start, b->start),
42 		.end = MAX(a->end, b->end),
43 	};
44 }
45 
46 static inline int
47 diff_range_len(const struct diff_range *r)
48 {
49 	if (!r)
50 		return 0;
51 	return r->end - r->start;
52 }
53 
54 /* List of all possible return codes of a diff invocation. */
55 #define DIFF_RC_USE_DIFF_ALGO_FALLBACK	-1
56 #define DIFF_RC_OK			0
57 /* Any positive return values are errno values from sys/errno.h */
58 
59 /* Indicate whether two given diff atoms match. */
60 int
61 diff_atom_same(bool *same,
62 	       const struct diff_atom *left,
63 	       const struct diff_atom *right);
64 
65 /* A diff chunk represents a set of atoms on the left and/or a set of atoms on
66  * the right.
67  *
68  * If solved == false:
69  * The diff algorithm has divided the source file, and this is a chunk that the
70  * inner_algo should run on next.
71  * The lines on the left should be diffed against the lines on the right.
72  * (If there are no left lines or no right lines, it implies solved == true,
73  * because there is nothing to diff.)
74  *
75  * If solved == true:
76  * If there are only left atoms, it is a chunk removing atoms from the left ("a
77  * minus chunk").
78  * If there are only right atoms, it is a chunk adding atoms from the right ("a
79  * plus chunk").
80  * If there are both left and right lines, it is a chunk of equal content on
81  * both sides, and left_count == right_count:
82  *
83  * - foo  }
84  * - bar  }-- diff_chunk{ left_start = &left.atoms.head[0], left_count = 3,
85  * - baz  }               right_start = NULL, right_count = 0 }
86  *   moo    }
87  *   goo    }-- diff_chunk{ left_start = &left.atoms.head[3], left_count = 3,
88  *   zoo    }              right_start = &right.atoms.head[0], right_count = 3 }
89  *  +loo      }
90  *  +roo      }-- diff_chunk{ left_start = NULL, left_count = 0,
91  *  +too      }            right_start = &right.atoms.head[3], right_count = 3 }
92  *
93  */
94 struct diff_chunk {
95 	bool solved;
96 	struct diff_atom *left_start;
97 	unsigned int left_count;
98 	struct diff_atom *right_start;
99 	unsigned int right_count;
100 };
101 
102 #define DIFF_RESULT_ALLOC_BLOCKSIZE 128
103 
104 enum diff_chunk_type {
105 	CHUNK_EMPTY,
106 	CHUNK_PLUS,
107 	CHUNK_MINUS,
108 	CHUNK_SAME,
109 	CHUNK_ERROR,
110 };
111 
112 static inline enum diff_chunk_type
113 diff_chunk_type(const struct diff_chunk *chunk)
114 {
115 	if (!chunk->left_count && !chunk->right_count)
116 		return CHUNK_EMPTY;
117 	if (!chunk->solved)
118 		return CHUNK_ERROR;
119 	if (!chunk->right_count)
120 		return CHUNK_MINUS;
121 	if (!chunk->left_count)
122 		return CHUNK_PLUS;
123 	if (chunk->left_count != chunk->right_count)
124 		return CHUNK_ERROR;
125 	return CHUNK_SAME;
126 }
127 
128 struct diff_chunk_context;
129 
130 bool
131 diff_chunk_context_empty(const struct diff_chunk_context *cc);
132 
133 bool
134 diff_chunk_contexts_touch(const struct diff_chunk_context *cc,
135 			  const struct diff_chunk_context *other);
136 
137 void
138 diff_chunk_contexts_merge(struct diff_chunk_context *cc,
139 			  const struct diff_chunk_context *other);
140 
141 struct diff_state {
142 	/* The final result passed to the original diff caller. */
143 	struct diff_result *result;
144 
145 	/* The root diff_data is in result->left,right, these are (possibly)
146 	 * subsections of the root data. */
147 	struct diff_data left;
148 	struct diff_data right;
149 
150 	unsigned int recursion_depth_left;
151 
152 	/* Remaining chunks from one diff algorithm pass, if any solved == false
153 	 * chunks came up. */
154 	diff_chunk_arraylist_t temp_result;
155 
156  	/* State buffer used by Myers algorithm. */
157 	int *kd_buf;
158 	size_t kd_buf_size; /* in units of sizeof(int), not bytes */
159 };
160 
161 struct diff_chunk *diff_state_add_chunk(struct diff_state *state, bool solved,
162 					struct diff_atom *left_start,
163 					unsigned int left_count,
164 					struct diff_atom *right_start,
165 					unsigned int right_count);
166 
167 struct diff_output_info;
168 
169 int diff_output_lines(struct diff_output_info *output_info, FILE *dest,
170 		       const char *prefix, struct diff_atom *start_atom,
171 		       unsigned int count);
172 
173 int diff_output_trailing_newline_msg(struct diff_output_info *outinfo,
174 				     FILE *dest,
175 				     const struct diff_chunk *c);
176 #define DIFF_FUNCTION_CONTEXT_SIZE	55
177 int diff_output_match_function_prototype(char *prototype, size_t prototype_size,
178 					 int *last_prototype_idx,
179 					 const struct diff_result *result,
180 					 const struct diff_chunk_context *cc);
181 
182 struct diff_output_info *diff_output_info_alloc(void);
183 
184 void
185 diff_data_init_subsection(struct diff_data *d, struct diff_data *parent,
186 			  struct diff_atom *from_atom, unsigned int atoms_count);