ddidderr/zstd-rs
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

update new api presentation in zstd.h and manual

Browse Source
This commit is contained in:
Yann Collet
2017-06-19 12:56:25 -07:00
parent c7dcf0f379
commit bf99150be3
3 changed files with 431 additions and 435 deletions
Download Patch File Download Diff File Expand all files Collapse all files
doc/zstd_manual.html
+197 -198
View File
@@ -516,204 +516,6 @@ size_t ZSTD_estimateDDictSize(size_t dictSize, unsigned byReference);
</b><p> Same as ZSTD_compress_usingCDict(), with fine-tune control over frame parameters
</p></pre><BR>
<h3>New experimental advanced parameters API</h3><pre></pre><b><pre></pre></b><BR>
<pre><b>typedef enum {
</b>/* compression parameters */<b>
ZSTD_p_compressionLevel=100, </b>/* Update all compression parameters according to pre-defined cLevel table<b>
* Default level is 3.
* Special: value 0 means "do not change cLevel". */
ZSTD_p_windowLog, </b>/* Maximum allowed back-reference distance, expressed as power of 2.<b>
* Must be clamped between ZSTD_WINDOWLOG_MIN and ZSTD_WINDOWLOG_MAX.
* default value : set through compressionLevel.
* Special: value 0 means "do not change windowLog". */
ZSTD_p_hashLog, </b>/* Size of the probe table, as a power of 2.<b>
* Resulting table size is (1 << (hashLog+2)).
* Must be clamped between ZSTD_HASHLOG_MIN and ZSTD_HASHLOG_MAX.
* Larger tables improve compression ratio of strategies <= dFast,
* and improve speed of strategies > dFast.
* Special: value 0 means "do not change hashLog". */
ZSTD_p_chainLog, </b>/* Size of the full-search table, as a power of 2.<b>
* Resulting table size is (1 << (chainLog+2)).
* Larger tables result in better and slower compression.
* This parameter is useless when using "fast" strategy.
* Special: value 0 means "do not change chainLog". */
ZSTD_p_searchLog, </b>/* Number of search attempts, as a power of 2.<b>
* More attempts result in better and slower compression.
* This parameter is useless when using "fast" and "dFast" strategies.
* Special: value 0 means "do not change searchLog". */
ZSTD_p_minMatch, </b>/* Minimum size of searched matches (note : repCode matches can be smaller).<b>
* Larger values make faster compression and decompression, but decrease ratio.
* Must be clamped between ZSTD_SEARCHLENGTH_MIN and ZSTD_SEARCHLENGTH_MAX.
* Note that currently, for all strategies < btopt, effective minimum is 4.
* Note that currently, for all strategies > fast, effective maximum is 6.
* Special: value 0 means "do not change minMatchLength". */
ZSTD_p_targetLength, </b>/* Only useful for strategies >= btopt.<b>
* Length of Match considered "good enough" to stop search.
* Larger values make compression stronger and slower.
* Special: value 0 means "do not change targetLength". */
ZSTD_p_compressionStrategy, </b>/* See ZSTD_strategy enum definition.<b>
* Cast selected strategy as unsigned for ZSTD_CCtx_setParameter() compatibility.
* The higher the value of selected strategy, the more complex it is,
* resulting in stronger and slower compression.
* Special: value 0 means "do not change strategy". */
#if 0
ZSTD_p_windowSize, </b>/* Maximum allowed back-reference distance.<b>
* Can be set to a more precise value than windowLog.
* Will be transparently reduced to closest possible inferior value
* (see Zstandard compression format) */
</b>/* Not ready yet ! */<b>
#endif
</b>/* frame parameters */<b>
ZSTD_p_contentSizeFlag=200, </b>/* Content size is written into frame header _whenever known_ (default:1) */<b>
ZSTD_p_checksumFlag, </b>/* A 32-bits checksum of content is written at end of frame (default:0) */<b>
ZSTD_p_dictIDFlag, </b>/* When applicable, dictID of dictionary is provided in frame header (default:1) */<b>
</b>/* dictionary parameters */<b>
ZSTD_p_refDictContent=300, </b>/* Content of dictionary content will be referenced, instead of copied (default:0).<b>
* This avoids duplicating dictionary content.
* But it also requires that dictionary buffer outlives its user (CDict) */
</b>/* Not ready yet ! */<b>
ZSTD_p_rawContentDict, </b>/* load dictionary in "content-only" mode (no header analysis) (default:0) */<b>
</b>/* question : should there be an option to load dictionary only in zstd format, rejecting others with an error code ? */<b>
</b>/* multi-threading parameters */<b>
ZSTD_p_nbThreads=400, </b>/* Select how many threads a compression job can spawn (default:1)<b>
* More threads improve speed, but also increase memory usage.
* Can only receive a value > 1 if ZSTD_MULTITHREAD is enabled.
* Special: value 0 means "do not change nbThreads" */
ZSTD_p_jobSize, </b>/* Size of a compression job. Each compression job is completed in parallel.<b>
* 0 means default, which is dynamically determined based on compression parameters.
* Job size must be a minimum of overlapSize, or 1 KB, whichever is largest
* The minimum size is automatically and transparently enforced */
ZSTD_p_overlapSizeLog, </b>/* Size of previous input reloaded at the beginning of each job.<b>
* 0 => no overlap, 6(default) => use 1/8th of windowSize, >=9 => use full windowSize */
</b>/* advanced parameters - may not remain available after API update */<b>
ZSTD_p_forceMaxWindow=1100, </b>/* Force back-references to remain < windowSize,<b>
* even when referencing into Dictionary content
* default : 0 when using a CDict, 1 when using a Prefix */
} ZSTD_cParameter;
</b></pre><BR>
<pre><b>size_t ZSTD_CCtx_setParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, unsigned value);
</b><p> Set one compression parameter, selected by enum ZSTD_cParameter.
Note : when `value` is an enum, cast it to unsigned for proper type checking.
@result : 0, or an error code (which can be tested with ZSTD_isError()).
</p></pre><BR>
<pre><b>size_t ZSTD_CCtx_setPledgedSrcSize(ZSTD_CCtx* cctx, unsigned long long pledgedSrcSize);
</b><p> Total input data size to be compressed into a single frame.
This value will be controlled at the end, and result in error if not respected.
@result : 0, or an error code (which can be tested with ZSTD_isError()).
Note 1 : 0 means zero, empty.
In order to mean "unknown content size", pass constant ZSTD_CONTENTSIZE_UNKNOWN.
Note that ZSTD_CONTENTSIZE_UNKNOWN is default value for all new compression jobs.
Note 2 : If all data is provided and consumed in a single round,
this value is overriden by srcSize instead.
</p></pre><BR>
<pre><b>size_t ZSTD_CCtx_loadDictionary(ZSTD_CCtx* cctx, const void* dict, size_t dictSize);
</b><p> Create an internal CDict from dict buffer.
Decompression will have to use same buffer.
@result : 0, or an error code (which can be tested with ZSTD_isError()).
Special : Adding a NULL (or 0-size) dictionary invalidates any previous prefix,
meaning "return to no-dictionary mode".
Note 1 : Dictionary content will be copied internally,
except if ZSTD_p_refDictContent is set.
Note 2 : Loading a dictionary involves building tables, which are dependent on compression parameters.
For this reason, compression parameters cannot be changed anymore after loading a prefix.
It's also a CPU-heavy operation, with non-negligible impact on latency.
Note 3 : Dictionary will be used for all future compression jobs.
To return to "no-dictionary" situation, load a NULL dictionary
</p></pre><BR>
<pre><b>size_t ZSTD_CCtx_refPrefix(ZSTD_CCtx* cctx, const void* prefix, size_t prefixSize); </b>/* Not ready yet ! */<b>
</b><p> Reference a prefix (content-only dictionary) to bootstrap next compression job.
Decompression will have to use same prefix.
@result : 0, or an error code (which can be tested with ZSTD_isError()).
Special : Adding a NULL (or 0-size) dictionary invalidates any previous prefix, meaning "return to no-dictionary mode".
Note 1 : Prefix buffer is referenced. It must outlive compression job.
Note 3 : Prefix is only used once. Tables are discarded at end of compression job.
If there is a need to use same prefix multiple times, consider embedding it into a ZSTD_CDict.
Note 2 : Referencing a prefix involves building tables, which are dependent on compression parameters.
For this reason, compression parameters cannot be changed anymore after loading a prefix.
It's also a CPU-heavy operation, with non-negligible impact on latency.
</p></pre><BR>
<pre><b>ZSTD_CDict* ZSTD_CDict_createEmpty(void); </b>/* Not ready yet ! */<b>
size_t ZSTD_CDict_setParameter(ZSTD_CDict* cdict, ZSTD_cParameter param, unsigned value); </b>/* Not ready yet ! */<b>
size_t ZSTD_CDict_loadDictionary(ZSTD_CDict* cdict, const void* dict, size_t dictSize); </b>/* Not ready yet ! */<b>
</b><p> Create a CDict object which is still mutable after creation.
It's the only one case allowing usage of ZSTD_CDict_setParameter().
Once all compression parameters are selected,
it's possible to load the target dictionary, using ZSTD_CDict_loadDictionary().
Dictionary content will be copied internally (except if ZSTD_p_refDictContent is set).
After loading the dictionary, no more change is possible.
The only remaining operation is to free CDict object.
Note : An unfinished CDict behaves the same as a NULL CDict if referenced into a CCtx.
</p></pre><BR>
<pre><b>size_t ZSTD_CCtx_refCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict);
</b><p> Add a prepared dictionary to cctx, to be used for next compression jobs.
Note that compression parameters will be enforced from within CDict.
Currently, they supercede any compression parameter previously set within CCtx.
The dictionary will remain valid for future compression jobs using same cctx.
@result : 0, or an error code (which can be tested with ZSTD_isError()).
Special : adding a NULL CDict means "return to no-dictionary mode".
Note 1 : Currently, only one dictionary can be managed.
Adding a new dictionary effectively "discards" any previous one.
Note 2 : CDict is just referenced, its lifetime must outlive CCtx.
</p></pre><BR>
<pre><b>typedef enum {
ZSTD_e_continue=0, </b>/* collect more data, encoder transparently decides when to output result, for optimal conditions */<b>
ZSTD_e_flush, </b>/* flush any data provided so far - frame will continue, future data can still reference previous data for better compression */<b>
ZSTD_e_end </b>/* flush any remaining data and ends current frame. Any future compression starts a new frame. */<b>
} ZSTD_EndDirective;
</b></pre><BR>
<pre><b>size_t ZSTD_compress_generic (ZSTD_CCtx* cctx,
ZSTD_outBuffer* output,
ZSTD_inBuffer* input,
ZSTD_EndDirective endOp);
</b><p> Behave about the same as ZSTD_compressStream. To note :
- Compression parameters are pushed into CCtx before starting compression, using ZSTD_setCCtxParameter()
- Compression parameters cannot be changed once compression is started.
- *dstPos must be <= dstCapacity, *srcPos must be <= srcSize
- *dspPos and *srcPos will be updated. They are guaranteed to remain below their respective limit.
- @return provides the amount of data ready to flush within internal buffers
or an error code, which can be tested using ZSTD_isError().
if @return != 0, flush is not fully completed, and must be called again to empty internal buffers.
- after a ZSTD_e_end directive, if internal buffer is not fully flushed,
only ZSTD_e_end and ZSTD_e_flush operations are allowed.
It is necessary to fully flush internal buffers
before changing compression parameters or start a new compression job.
</p></pre><BR>
<pre><b>void ZSTD_CCtx_reset(ZSTD_CCtx* cctx); </b>/* Not ready yet ! */<b>
</b><p> Return a CCtx to clean state.
Useful after an error, or to interrupt an ongoing compression job and start a new one.
Any internal data not yet flushed is cancelled.
Dictionary (if any) is dropped, next compression starts with srcSize==unknown by default.
It's possible to modify compression parameters after a reset.
</p></pre><BR>
<pre><b>size_t ZSTD_compress_generic_simpleArgs (
ZSTD_CCtx* cctx,
void* dst, size_t dstCapacity, size_t* dstPos,
const void* src, size_t srcSize, size_t* srcPos,
ZSTD_EndDirective endOp);
</b><p> Same as ZSTD_compress_generic(),
but using only simple integral types as arguments.
Argument list is less expressive than ZSTD_{in,out}Buffer,
but can be helpful for binders to dynamic languages
which have troubles handling structures containing memory pointers.
</p></pre><BR>
<a name="Chapter15"></a><h2>Advanced decompression functions</h2><pre></pre>
<pre><b>unsigned ZSTD_isFrame(const void* buffer, size_t size);
@@ -931,6 +733,203 @@ void ZSTD_copyDCtx(ZSTD_DCtx* dctx, const ZSTD_DCtx* preparedDCtx);
</pre></b><BR>
<pre><b>typedef enum { ZSTDnit_frameHeader, ZSTDnit_blockHeader, ZSTDnit_block, ZSTDnit_lastBlock, ZSTDnit_checksum, ZSTDnit_skippableFrame } ZSTD_nextInputType_e;
</b></pre><BR>
<h3>New advanced API (experimental, and compression only)</h3><pre></pre><b><pre></pre></b><BR>
<pre><b>typedef enum {
</b>/* compression parameters */<b>
ZSTD_p_compressionLevel=100, </b>/* Update all compression parameters according to pre-defined cLevel table<b>
* Default level is ZSTD_CLEVEL_DEFAULT==3.
* Special: value 0 means "do not change cLevel". */
ZSTD_p_windowLog, </b>/* Maximum allowed back-reference distance, expressed as power of 2.<b>
* Must be clamped between ZSTD_WINDOWLOG_MIN and ZSTD_WINDOWLOG_MAX.
* default value : set through compressionLevel.
* Special: value 0 means "do not change windowLog". */
ZSTD_p_hashLog, </b>/* Size of the probe table, as a power of 2.<b>
* Resulting table size is (1 << (hashLog+2)).
* Must be clamped between ZSTD_HASHLOG_MIN and ZSTD_HASHLOG_MAX.
* Larger tables improve compression ratio of strategies <= dFast,
* and improve speed of strategies > dFast.
* Special: value 0 means "do not change hashLog". */
ZSTD_p_chainLog, </b>/* Size of the full-search table, as a power of 2.<b>
* Resulting table size is (1 << (chainLog+2)).
* Larger tables result in better and slower compression.
* This parameter is useless when using "fast" strategy.
* Special: value 0 means "do not change chainLog". */
ZSTD_p_searchLog, </b>/* Number of search attempts, as a power of 2.<b>
* More attempts result in better and slower compression.
* This parameter is useless when using "fast" and "dFast" strategies.
* Special: value 0 means "do not change searchLog". */
ZSTD_p_minMatch, </b>/* Minimum size of searched matches (note : repCode matches can be smaller).<b>
* Larger values make faster compression and decompression, but decrease ratio.
* Must be clamped between ZSTD_SEARCHLENGTH_MIN and ZSTD_SEARCHLENGTH_MAX.
* Note that currently, for all strategies < btopt, effective minimum is 4.
* Note that currently, for all strategies > fast, effective maximum is 6.
* Special: value 0 means "do not change minMatchLength". */
ZSTD_p_targetLength, </b>/* Only useful for strategies >= btopt.<b>
* Length of Match considered "good enough" to stop search.
* Larger values make compression stronger and slower.
* Special: value 0 means "do not change targetLength". */
ZSTD_p_compressionStrategy, </b>/* See ZSTD_strategy enum definition.<b>
* Cast selected strategy as unsigned for ZSTD_CCtx_setParameter() compatibility.
* The higher the value of selected strategy, the more complex it is,
* resulting in stronger and slower compression.
* Special: value 0 means "do not change strategy". */
#if 0
ZSTD_p_windowSize, </b>/* Maximum allowed back-reference distance.<b>
* Can be set to a more precise value than windowLog.
* Will be transparently reduced to closest possible inferior value
* (see Zstandard compression format) */
</b>/* Not ready yet ! */<b>
#endif
</b>/* frame parameters */<b>
ZSTD_p_contentSizeFlag=200, </b>/* Content size is written into frame header _whenever known_ (default:1) */<b>
ZSTD_p_checksumFlag, </b>/* A 32-bits checksum of content is written at end of frame (default:0) */<b>
ZSTD_p_dictIDFlag, </b>/* When applicable, dictID of dictionary is provided in frame header (default:1) */<b>
</b>/* dictionary parameters */<b>
ZSTD_p_refDictContent=300, </b>/* Content of dictionary content will be referenced, instead of copied (default:0).<b>
* This avoids duplicating dictionary content.
* But it also requires that dictionary buffer outlives its user (CDict) */
</b>/* Not ready yet ! */<b>
ZSTD_p_rawContentDict, </b>/* load dictionary in "content-only" mode (no header analysis) (default:0) */<b>
</b>/* question : should there be an option to load dictionary only in zstd format, rejecting others with an error code ? */<b>
</b>/* multi-threading parameters */<b>
ZSTD_p_nbThreads=400, </b>/* Select how many threads a compression job can spawn (default:1)<b>
* More threads improve speed, but also increase memory usage.
* Can only receive a value > 1 if ZSTD_MULTITHREAD is enabled.
* Special: value 0 means "do not change nbThreads" */
ZSTD_p_jobSize, </b>/* Size of a compression job. Each compression job is completed in parallel.<b>
* 0 means default, which is dynamically determined based on compression parameters.
* Job size must be a minimum of overlapSize, or 1 KB, whichever is largest
* The minimum size is automatically and transparently enforced */
ZSTD_p_overlapSizeLog, </b>/* Size of previous input reloaded at the beginning of each job.<b>
* 0 => no overlap, 6(default) => use 1/8th of windowSize, >=9 => use full windowSize */
</b>/* advanced parameters - may not remain available after API update */<b>
ZSTD_p_forceMaxWindow=1100, </b>/* Force back-references to remain < windowSize,<b>
* even when referencing into Dictionary content
* default : 0 when using a CDict, 1 when using a Prefix */
} ZSTD_cParameter;
</b></pre><BR>
<pre><b>size_t ZSTD_CCtx_setParameter(ZSTD_CCtx* cctx, ZSTD_cParameter param, unsigned value);
</b><p> Set one compression parameter, selected by enum ZSTD_cParameter.
Note : when `value` is an enum, cast it to unsigned for proper type checking.
@result : 0, or an error code (which can be tested with ZSTD_isError()).
</p></pre><BR>
<pre><b>size_t ZSTD_CCtx_setPledgedSrcSize(ZSTD_CCtx* cctx, unsigned long long pledgedSrcSize);
</b><p> Total input data size to be compressed as a single frame.
This value will be controlled at the end, and result in error if not respected.
@result : 0, or an error code (which can be tested with ZSTD_isError()).
Note 1 : 0 means zero, empty.
In order to mean "unknown content size", pass constant ZSTD_CONTENTSIZE_UNKNOWN.
Note that ZSTD_CONTENTSIZE_UNKNOWN is default value for new compression jobs.
Note 2 : If all data is provided and consumed in a single round,
this value is overriden by srcSize instead.
</p></pre><BR>
<pre><b>size_t ZSTD_CCtx_loadDictionary(ZSTD_CCtx* cctx, const void* dict, size_t dictSize);
</b><p> Create an internal CDict from dict buffer.
Decompression will have to use same buffer.
@result : 0, or an error code (which can be tested with ZSTD_isError()).
Special : Adding a NULL (or 0-size) dictionary invalidates any previous prefix,
meaning "return to no-dictionary mode".
Note 1 : Dictionary content will be copied internally,
except if ZSTD_p_refDictContent is set.
Note 2 : Loading a dictionary involves building tables, which are dependent on compression parameters.
For this reason, compression parameters cannot be changed anymore after loading a prefix.
It's also a CPU-heavy operation, with non-negligible impact on latency.
Note 3 : Dictionary will be used for all future compression jobs.
To return to "no-dictionary" situation, load a NULL dictionary
</p></pre><BR>
<pre><b>size_t ZSTD_CCtx_refCDict(ZSTD_CCtx* cctx, const ZSTD_CDict* cdict);
</b><p> Ref a prepared dictionary, to be used for all next compression jobs.
Note that compression parameters are enforced from within CDict,
and supercede any compression parameter previously set within CCtx.
The dictionary will remain valid for future compression jobs using same CCtx.
@result : 0, or an error code (which can be tested with ZSTD_isError()).
Special : adding a NULL CDict means "return to no-dictionary mode".
Note 1 : Currently, only one dictionary can be managed.
Adding a new dictionary effectively "discards" any previous one.
Note 2 : CDict is just referenced, its lifetime must outlive CCtx.
</p></pre><BR>
<pre><b>size_t ZSTD_CCtx_refPrefix(ZSTD_CCtx* cctx, const void* prefix, size_t prefixSize); </b>/* Not ready yet ! */<b>
</b><p> Reference a prefix (content-only dictionary) to bootstrap next compression job.
Decompression will have to use same prefix.
Prefix is only used once. Tables are discarded at end of compression job.
If there is a need to use same prefix multiple times, consider embedding it into a ZSTD_CDict.
@result : 0, or an error code (which can be tested with ZSTD_isError()).
Special : Adding a NULL (or 0-size) dictionary invalidates any previous prefix, meaning "return to no-dictionary mode".
Note 1 : Prefix buffer is referenced. It must outlive compression job.
Note 2 : Referencing a prefix involves building tables, which are dependent on compression parameters.
It's a CPU-heavy operation, with non-negligible impact on latency.
</p></pre><BR>
<pre><b>typedef enum {
ZSTD_e_continue=0, </b>/* collect more data, encoder transparently decides when to output result, for optimal conditions */<b>
ZSTD_e_flush, </b>/* flush any data provided so far - frame will continue, future data can still reference previous data for better compression */<b>
ZSTD_e_end </b>/* flush any remaining data and ends current frame. Any future compression starts a new frame. */<b>
} ZSTD_EndDirective;
</b></pre><BR>
<pre><b>size_t ZSTD_compress_generic (ZSTD_CCtx* cctx,
ZSTD_outBuffer* output,
ZSTD_inBuffer* input,
ZSTD_EndDirective endOp);
</b><p> Behave about the same as ZSTD_compressStream. To note :
- Compression parameters are pushed into CCtx before starting compression, using ZSTD_setCCtxParameter()
- Compression parameters cannot be changed once compression is started.
- *dstPos must be <= dstCapacity, *srcPos must be <= srcSize
- *dspPos and *srcPos will be updated. They are guaranteed to remain below their respective limit.
- @return provides the minimum amount of data still to flush from internal buffers
or an error code, which can be tested using ZSTD_isError().
if @return != 0, flush is not fully completed, and must be called again to empty internal buffers.
- after a ZSTD_e_end directive, if internal buffer is not fully flushed,
only ZSTD_e_end and ZSTD_e_flush operations are allowed.
It is necessary to fully flush internal buffers
before starting a new compression job, or changing compression parameters.
</p></pre><BR>
<pre><b>void ZSTD_CCtx_reset(ZSTD_CCtx* cctx); </b>/* Not ready yet ! */<b>
</b><p> Return a CCtx to clean state.
Useful after an error, or to interrupt an ongoing compression job and start a new one.
Any internal data not yet flushed is cancelled.
Dictionary (if any) is dropped.
It's possible to modify compression parameters after a reset.
</p></pre><BR>
<pre><b>size_t ZSTD_compress_generic_simpleArgs (
ZSTD_CCtx* cctx,
void* dst, size_t dstCapacity, size_t* dstPos,
const void* src, size_t srcSize, size_t* srcPos,
ZSTD_EndDirective endOp);
</b><p> Same as ZSTD_compress_generic(),
but using only integral types as arguments.
Argument list is larger and less expressive than ZSTD_{in,out}Buffer,
but can be helpful for binders from dynamic languages
which have troubles handling structures containing memory pointers.
</p></pre><BR>
<pre><b>ZSTD_CDict* ZSTD_CDict_createEmpty(void); </b>/* Not ready yet ! */<b>
size_t ZSTD_CDict_setParameter(ZSTD_CDict* cdict, ZSTD_cParameter param, unsigned value); </b>/* Not ready yet ! */<b>
size_t ZSTD_CDict_loadDictionary(ZSTD_CDict* cdict, const void* dict, size_t dictSize); </b>/* Not ready yet ! */<b>
</b><p> Create a CDict object which is still mutable after creation.
It's the only one case allowing usage of ZSTD_CDict_setParameter().
Once all compression parameters are selected,
it's possible to load the target dictionary, using ZSTD_CDict_loadDictionary().
Dictionary content will be copied internally (except if ZSTD_p_refDictContent is set).
After loading the dictionary, no more change is possible.
The only remaining operation is to free CDict object.
Note : An unfinished CDict behaves the same as a NULL CDict if referenced into a CCtx.
</p></pre><BR>
<a name="Chapter20"></a><h2>Block functions</h2><pre>
Block functions produce and decode raw zstd blocks, without frame metadata.
Frame metadata cost is typically ~18 bytes, which can be non-negligible for very small blocks (< 100 bytes).