1 /*
2  * Copyright (c) 2018 Stefan Sperling <stsp@openbsd.org>
3  *
4  * Permission to use, copy, modify, and distribute this software for any
5  * purpose with or without fee is hereby granted, provided that the above
6  * copyright notice and this permission notice appear in all copies.
7  *
8  * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
9  * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
10  * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
11  * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
12  * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
13  * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
14  * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
15  */
16 
17 /* An open pack file. */
18 struct got_pack {
19 	char *path_packfile;
20 	int fd;
21 	uint8_t *map;
22 	size_t filesize;
23 };
24 
25 const struct got_error *got_pack_close(struct got_pack *);
26 
27 /* See Documentation/technical/pack-format.txt in Git. */
28 
29 struct got_packidx_trailer {
30 	u_int8_t	packfile_sha1[SHA1_DIGEST_LENGTH];
31 	u_int8_t	packidx_sha1[SHA1_DIGEST_LENGTH];
32 } __attribute__((__packed__));
33 
34 struct got_packidx_object_id {
35 	u_int8_t sha1[SHA1_DIGEST_LENGTH];
36 } __attribute__((__packed__));
37 
38 /* Ignore pack index version 1 which is no longer written by Git. */
39 #define GOT_PACKIDX_VERSION 2
40 
41 struct got_packidx_v2_hdr {
42 	uint32_t	*magic;		/* big endian */
43 #define GOT_PACKIDX_V2_MAGIC 0xff744f63	/* "\377t0c" */
44 	uint32_t	*version;
45 
46 	/* 
47 	 * Each entry N in the fanout table contains the number of objects in
48 	 * the packfile whose SHA1 begins with a byte less than or equal to N.
49 	 * The last entry (index 255) contains the number of objects in the
50 	 * pack file whose first SHA1 byte is <= 0xff, and thus records the
51 	 * total number of objects in the pack file. All pointer variables
52 	 * below point to tables with a corresponding number of entries.
53 	 */
54 	uint32_t	*fanout_table;	/* values are big endian */
55 #define GOT_PACKIDX_V2_FANOUT_TABLE_ITEMS (0xff + 1)
56 
57 	/* Sorted SHA1 checksums for each object in the pack file. */
58 	struct got_packidx_object_id *sorted_ids;
59 
60 	/* CRC32 of the packed representation of each object. */
61 	uint32_t	*crc32;
62 
63 	/* Offset into the pack file for each object. */
64 	uint32_t	*offsets;		/* values are big endian */
65 #define GOT_PACKIDX_OFFSET_VAL_MASK		0x7fffffff
66 #define GOT_PACKIDX_OFFSET_VAL_IS_LARGE_IDX	0x80000000
67 
68 	/* Large offsets table is empty for pack files < 2 GB. */
69 	uint64_t	*large_offsets;		/* values are big endian */
70 
71 	struct got_packidx_trailer *trailer;
72 };
73 
74 /* An open pack index file. */
75 struct got_packidx {
76 	char *path_packidx; /* actual on-disk path */
77 	int fd;
78 	uint8_t *map;
79 	size_t len;
80 	struct got_packidx_v2_hdr hdr; /* convenient pointers into map */
81 };
82 
83 struct got_packfile_hdr {
84 	uint32_t	signature;
85 #define GOT_PACKFILE_SIGNATURE	0x5041434b	/* 'P' 'A' 'C' 'K' */
86 	uint32_t	version;	/* big endian */
87 #define GOT_PACKFILE_VERSION 2
88 	uint32_t	nobjects;	/* big endian */
89 };
90 
91 struct got_packfile_obj_hdr {
92 	/* 
93 	 * The object size field uses a variable length encoding:
94 	 * size0...sizeN form a 4+7+7+...+7 bit integer, where size0 is the
95 	 * least significant part and sizeN is the most significant part.
96 	 * If the MSB of a size byte is set, an additional size byte follows.
97 	 * Of the 7 remaining bits of size0, the first 3 bits indicate the
98 	 * object's type, and the remaining 4 bits contribute to the size.
99 	 */
100 	uint8_t *size;		/* variable length */
101 #define GOT_PACK_OBJ_SIZE_MORE		0x80
102 #define GOT_PACK_OBJ_SIZE0_TYPE_MASK	0x70 /* See struct got_object->type */
103 #define GOT_PACK_OBJ_SIZE0_TYPE_MASK_SHIFT	4
104 #define GOT_PACK_OBJ_SIZE0_VAL_MASK	0x0f
105 #define GOT_PACK_OBJ_SIZE_VAL_MASK	0x7f
106 };
107 
108 /* If object is not a DELTA type. */
109 struct got_packfile_object_data {
110 	uint8_t *data;	/* compressed */
111 };
112 
113 /* If object is of type	GOT_OBJ_TYPE_REF_DELTA. */
114 struct got_packfile_object_data_ref_delta {
115 	uint8_t sha1[SHA1_DIGEST_LENGTH];
116 	uint8_t *delta_data;		/* compressed */
117 };
118 
119 /* If object is of type GOT_OBJ_TYPE_OFFSET_DELTA. */
120 struct got_packfile_object_data_offset_delta {
121 	/* 
122 	 * This offset is interpreted as a negative offset from
123 	 * the got_packfile_obj_hdr corresponding to this object.
124 	 * The size provided in the header specifies the amount
125 	 * of compressed delta data that follows.
126 	 *
127 	 * This field uses a variable length encoding of N bytes,
128 	 * where the MSB is always set except for the last byte.
129 	 * The value is encoded as a series of N 7 bit integers,
130 	 * which are concatenated, and if N > 1 the value 2^7 +
131 	 * 2^14 + ... + 2^(7 * (n-1)) is added to the result.
132 	 */
133 	uint8_t *offset;	/* variable length */
134 #define GOT_PACK_OBJ_DELTA_OFF_MORE		0x80
135 #define GOT_PACK_OBJ_DELTA_OFF_VAL_MASK		0x7f
136 	uint8_t *delta_data;		/* compressed */
137 };
138 
139 struct got_packfile_obj_data {
140 	union {
141 		struct got_packfile_object_data data;
142 		struct got_packfile_object_data_ref_delta ref_delta;
143 		struct got_packfile_object_data_offset_delta offset_delta;
144 	} __attribute__((__packed__));
145 } __attribute__((__packed__));
146 
147 const struct got_error *got_packidx_open(struct got_packidx **,
148     const char *, int);
149 const struct got_error* got_packidx_close(struct got_packidx *);
150 
151 const struct got_error *got_packfile_open_object(struct got_object **,
152     struct got_object_id *, struct got_repository *);
153 const struct got_error *got_packfile_extract_object(FILE **,
154     struct got_object *, struct got_repository *);
155 const struct got_error *got_packfile_extract_object_to_mem(uint8_t **, size_t *,
156     struct got_object *, struct got_repository *);