reftable/block: create public interface for reading blocks
While users of the reftable library wouldn't generally require access to individual blocks in a reftable table, there are valid usecases where one may require low-level access to them. One such upcoming usecase in the Git codebase is to implement consistency checks for the reftable library where we want to verify each block individually. Create a public interface for reading blocks. The interface isn't yet complete and lacks e.g. a way to read individual records from a block. Such missing functionality will be backfilled in subsequent commits. Note that this change also requires us to expose `reftable_buf`, which is used by the `reftable_block_first_key()` function. Signed-off-by: Patrick Steinhardt <ps@pks.im> Signed-off-by: Junio C Hamano <gitster@pobox.com>
Patrick Steinhardt committed
Apr 7, 2025 at 15:16 UTC
655e18d6b4f845090b0ba4761105b32726893ecb
4 files changed
+78
-54
reftable/basics.h
-7
@@ -18,13 +18,6 @@
18
19
#define REFTABLE_UNUSED __attribute__((__unused__))
20
21
-struct reftable_buf {
22
- size_t alloc;
23
- size_t len;
24
- char *buf;
25
-};
26
-#define REFTABLE_BUF_INIT { 0 }
27
-
21
/*
22
* Initialize the buffer such that it is ready for use. This is equivalent to
23
* using REFTABLE_BUF_INIT for stack-allocated variables.
reftable/block.h
+1
-47
@@ -11,6 +11,7 @@
11
12
#include "basics.h"
13
#include "record.h"
14
+#include "reftable-block.h"
15
#include "reftable-blocksource.h"
16
17
/*
@@ -62,53 +63,6 @@ int block_writer_finish(struct block_writer *w);
63
/* clears out internally allocated block_writer members. */
64
void block_writer_release(struct block_writer *bw);
65
65
-/*
66
- * A block part of a reftable. Contains records as well as some metadata
67
- * describing them.
68
- */
69
-struct reftable_block {
70
- /* offset of the block header; nonzero for the first block in a
71
- * reftable. */
72
- uint32_t header_off;
73
-
74
- /* the memory block */
75
- struct reftable_block_data block_data;
76
- uint32_t hash_size;
77
-
78
- /* Uncompressed data for log entries. */
79
- struct z_stream_s *zstream;
80
- unsigned char *uncompressed_data;
81
- size_t uncompressed_cap;
82
-
83
- /*
84
- * Restart point data. Restart points are located after the block's
85
- * record data.
86
- */
87
- uint16_t restart_count;
88
- uint32_t restart_off;
89
-
90
- /* size of the data in the file. For log blocks, this is the compressed
91
- * size. */
92
- uint32_t full_block_size;
93
- uint8_t block_type;
94
-};
95
-
96
-/*
97
- * Initialize a reftable block from the given block source.
98
- */
99
-int reftable_block_init(struct reftable_block *b,
100
- struct reftable_block_source *source,
101
- uint32_t offset, uint32_t header_size,
102
- uint32_t table_block_size, uint32_t hash_size);
103
-
104
-void reftable_block_release(struct reftable_block *b);
105
-
106
-/* Returns the block type (eg. 'r' for refs) */
107
-uint8_t reftable_block_type(const struct reftable_block *b);
108
-
109
-/* Decodes the first key in the block */
110
-int reftable_block_first_key(const struct reftable_block *b, struct reftable_buf *key);
111
-
66
/* Iterate over entries in a block */
67
struct block_iter {
68
/* offset within the block of the next entry to read. */
reftable/reftable-basics.h
+8
@@ -11,6 +11,14 @@
11
12
#include <stddef.h>
13
14
+/* A buffer that contains arbitrary byte slices. */
15
+struct reftable_buf {
16
+ size_t alloc;
17
+ size_t len;
18
+ char *buf;
19
+};
20
+#define REFTABLE_BUF_INIT { 0 }
21
+
22
/*
23
* Hash functions understood by the reftable library. Note that the values are
24
* arbitrary and somewhat random such that we can easily detect cases where the
reftable/reftable-block.h
new
+69
@@ -0,0 +1,69 @@
1
+/*
2
+ * Copyright 2020 Google LLC
3
+ *
4
+ * Use of this source code is governed by a BSD-style
5
+ * license that can be found in the LICENSE file or at
6
+ * https://developers.google.com/open-source/licenses/bsd
7
+ */
8
+
9
+#ifndef REFTABLE_BLOCK_H
10
+#define REFTABLE_BLOCK_H
11
+
12
+#include <stdint.h>
13
+
14
+#include "reftable-basics.h"
15
+#include "reftable-blocksource.h"
16
+
17
+struct z_stream_s;
18
+
19
+/*
20
+ * A block part of a reftable. Contains records as well as some metadata
21
+ * describing them.
22
+ */
23
+struct reftable_block {
24
+ /*
25
+ * Offset of the block header; nonzero for the first block in a
26
+ * reftable.
27
+ */
28
+ uint32_t header_off;
29
+
30
+ /* The memory block. */
31
+ struct reftable_block_data block_data;
32
+ uint32_t hash_size;
33
+
34
+ /* Uncompressed data for log entries. */
35
+ struct z_stream_s *zstream;
36
+ unsigned char *uncompressed_data;
37
+ size_t uncompressed_cap;
38
+
39
+ /*
40
+ * Restart point data. Restart points are located after the block's
41
+ * record data.
42
+ */
43
+ uint16_t restart_count;
44
+ uint32_t restart_off;
45
+
46
+ /*
47
+ * Size of the data in the file. For log blocks, this is the compressed
48
+ * size.
49
+ */
50
+ uint32_t full_block_size;
51
+ uint8_t block_type;
52
+};
53
+
54
+/* Initialize a reftable block from the given block source. */
55
+int reftable_block_init(struct reftable_block *b,
56
+ struct reftable_block_source *source,
57
+ uint32_t offset, uint32_t header_size,
58
+ uint32_t table_block_size, uint32_t hash_size);
59
+
60
+/* Release resources allocated by the block. */
61
+void reftable_block_release(struct reftable_block *b);
62
+
63
+/* Returns the block type (eg. 'r' for refs). */
64
+uint8_t reftable_block_type(const struct reftable_block *b);
65
+
66
+/* Decodes the first key in the block. */
67
+int reftable_block_first_key(const struct reftable_block *b, struct reftable_buf *key);
68
+
69
+#endif /* REFTABLE_BLOCK_H */