| 1 | /* |
| 2 | * Copyright (c) Meta Platforms, Inc. and affiliates. |
| 3 | * All rights reserved. |
| 4 | * |
| 5 | * This source code is licensed under both the BSD-style license (found in the |
| 6 | * LICENSE file in the root directory of this source tree) and the GPLv2 (found |
| 7 | * in the COPYING file in the root directory of this source tree). |
| 8 | * You may select, at your option, one of the above-listed licenses. |
| 9 | */ |
| 10 | |
| 11 | #ifndef ZSTD_LDM_H |
| 12 | #define ZSTD_LDM_H |
| 13 | |
| 14 | #if defined (__cplusplus) |
| 15 | extern "C" { |
| 16 | #endif |
| 17 | |
| 18 | #include "zstd_compress_internal.h" /* ldmParams_t, U32 */ |
| 19 | #include "../zstd.h" /* ZSTD_CCtx, size_t */ |
| 20 | |
| 21 | /*-************************************* |
| 22 | * Long distance matching |
| 23 | ***************************************/ |
| 24 | |
| 25 | #define ZSTD_LDM_DEFAULT_WINDOW_LOG ZSTD_WINDOWLOG_LIMIT_DEFAULT |
| 26 | |
| 27 | void ZSTD_ldm_fillHashTable( |
| 28 | ldmState_t* state, const BYTE* ip, |
| 29 | const BYTE* iend, ldmParams_t const* params); |
| 30 | |
| 31 | /** |
| 32 | * ZSTD_ldm_generateSequences(): |
| 33 | * |
| 34 | * Generates the sequences using the long distance match finder. |
| 35 | * Generates long range matching sequences in `sequences`, which parse a prefix |
| 36 | * of the source. `sequences` must be large enough to store every sequence, |
| 37 | * which can be checked with `ZSTD_ldm_getMaxNbSeq()`. |
| 38 | * @returns 0 or an error code. |
| 39 | * |
| 40 | * NOTE: The user must have called ZSTD_window_update() for all of the input |
| 41 | * they have, even if they pass it to ZSTD_ldm_generateSequences() in chunks. |
| 42 | * NOTE: This function returns an error if it runs out of space to store |
| 43 | * sequences. |
| 44 | */ |
| 45 | size_t ZSTD_ldm_generateSequences( |
| 46 | ldmState_t* ldms, rawSeqStore_t* sequences, |
| 47 | ldmParams_t const* params, void const* src, size_t srcSize); |
| 48 | |
| 49 | /** |
| 50 | * ZSTD_ldm_blockCompress(): |
| 51 | * |
| 52 | * Compresses a block using the predefined sequences, along with a secondary |
| 53 | * block compressor. The literals section of every sequence is passed to the |
| 54 | * secondary block compressor, and those sequences are interspersed with the |
| 55 | * predefined sequences. Returns the length of the last literals. |
| 56 | * Updates `rawSeqStore.pos` to indicate how many sequences have been consumed. |
| 57 | * `rawSeqStore.seq` may also be updated to split the last sequence between two |
| 58 | * blocks. |
| 59 | * @return The length of the last literals. |
| 60 | * |
| 61 | * NOTE: The source must be at most the maximum block size, but the predefined |
| 62 | * sequences can be any size, and may be longer than the block. In the case that |
| 63 | * they are longer than the block, the last sequences may need to be split into |
| 64 | * two. We handle that case correctly, and update `rawSeqStore` appropriately. |
| 65 | * NOTE: This function does not return any errors. |
| 66 | */ |
| 67 | size_t ZSTD_ldm_blockCompress(rawSeqStore_t* rawSeqStore, |
| 68 | ZSTD_matchState_t* ms, seqStore_t* seqStore, U32 rep[ZSTD_REP_NUM], |
| 69 | ZSTD_paramSwitch_e useRowMatchFinder, |
| 70 | void const* src, size_t srcSize); |
| 71 | |
| 72 | /** |
| 73 | * ZSTD_ldm_skipSequences(): |
| 74 | * |
| 75 | * Skip past `srcSize` bytes worth of sequences in `rawSeqStore`. |
| 76 | * Avoids emitting matches less than `minMatch` bytes. |
| 77 | * Must be called for data that is not passed to ZSTD_ldm_blockCompress(). |
| 78 | */ |
| 79 | void ZSTD_ldm_skipSequences(rawSeqStore_t* rawSeqStore, size_t srcSize, |
| 80 | U32 const minMatch); |
| 81 | |
| 82 | /* ZSTD_ldm_skipRawSeqStoreBytes(): |
| 83 | * Moves forward in rawSeqStore by nbBytes, updating fields 'pos' and 'posInSequence'. |
| 84 | * Not to be used in conjunction with ZSTD_ldm_skipSequences(). |
| 85 | * Must be called for data with is not passed to ZSTD_ldm_blockCompress(). |
| 86 | */ |
| 87 | void ZSTD_ldm_skipRawSeqStoreBytes(rawSeqStore_t* rawSeqStore, size_t nbBytes); |
| 88 | |
| 89 | /** ZSTD_ldm_getTableSize() : |
| 90 | * Estimate the space needed for long distance matching tables or 0 if LDM is |
| 91 | * disabled. |
| 92 | */ |
| 93 | size_t ZSTD_ldm_getTableSize(ldmParams_t params); |
| 94 | |
| 95 | /** ZSTD_ldm_getSeqSpace() : |
| 96 | * Return an upper bound on the number of sequences that can be produced by |
| 97 | * the long distance matcher, or 0 if LDM is disabled. |
| 98 | */ |
| 99 | size_t ZSTD_ldm_getMaxNbSeq(ldmParams_t params, size_t maxChunkSize); |
| 100 | |
| 101 | /** ZSTD_ldm_adjustParameters() : |
| 102 | * If the params->hashRateLog is not set, set it to its default value based on |
| 103 | * windowLog and params->hashLog. |
| 104 | * |
| 105 | * Ensures that params->bucketSizeLog is <= params->hashLog (setting it to |
| 106 | * params->hashLog if it is not). |
| 107 | * |
| 108 | * Ensures that the minMatchLength >= targetLength during optimal parsing. |
| 109 | */ |
| 110 | void ZSTD_ldm_adjustParameters(ldmParams_t* params, |
| 111 | ZSTD_compressionParameters const* cParams); |
| 112 | |
| 113 | #if defined (__cplusplus) |
| 114 | } |
| 115 | #endif |
| 116 | |
| 117 | #endif /* ZSTD_FAST_H */ |