A struct describing the options for a compressor or compression function.
These values are not stable and still subject to change!
The maximum number of checks to make in the hash table for matches.
Higher numbers mean slower, but better compression. Very high (say
will impact compression speed a lot. The maximum match length is 2^15, so values higher than
this won't make any difference, and will be truncated to 2^15 by the compression
Only lazy match if we have a length less than this value.
Higher values degrade compression slightly, but improve compression speed.
0: Never lazy match. (Same effect as setting
MatchingTypeto greedy, but may be slower).
1...257: Only check for a better match if the first match was shorter than this value.
258: Always lazy match.
As the maximum length of a match is
258, values higher than this will have
no further effect.
- Default value:
Whether to use lazy or greedy matching.
Lazy matching will provide better compression, at the expense of compression speed.
As a special case, if max_hash_checks is set to 0, and matching_type is set to lazy, compression using only run-length encoding (i.e maximum match distance of 1) is performed. (This may be changed in the future but is defined like this at the moment to avoid API breakage.
- Default value:
Force fixed/stored blocks (Not implemented yet).
- Default value:
pub fn high() -> CompressionOptions[src]
Returns compression settings roughly corresponding to the
HIGH(9) setting in miniz.
pub fn fast() -> CompressionOptions[src]
Returns a fast set of compression settings
Ideally this should roughly correspond to the
FAST(1) setting in miniz.
However, that setting makes miniz use a somewhat different algorithm,
so currently hte fast level in this library is slower and better compressing
than the corresponding level in miniz.
pub fn huffman_only() -> CompressionOptions[src]
Returns a set of compression settings that makes the compressor only compress using Huffman coding. (Ignoring any length/distance matching)
This will normally have the worst compression ratio (besides only using uncompressed data), but may be the fastest method in some cases.
pub fn rle() -> CompressionOptions[src]
Returns a set of compression settings that makes the compressor compress only using run-length encoding (i.e only looking for matches one byte back).
This is very fast, but tends to compress worse than looking for more matches using hash chains that the slower settings do. Works best on data that has runs of equivalent bytes, like binary or simple images, less good for text.
impl Clone for CompressionOptions[src]
impl Copy for CompressionOptions[src]
impl Debug for CompressionOptions[src]
impl Default for CompressionOptions[src]
fn default() -> CompressionOptions[src]
Returns the options describing the default compression level.
impl Eq for CompressionOptions[src]
impl From<Compression> for CompressionOptions[src]
impl Hash for CompressionOptions[src]
fn hash_slice<H>(data: &[Self], state: &mut H) where1.3.0[src]
fn eq(&self, other: &CompressionOptions) -> bool[src]
fn ne(&self, other: &CompressionOptions) -> bool[src]
impl StructuralEq for CompressionOptions[src]
Auto Trait Implementations
impl RefUnwindSafe for CompressionOptions
impl Send for CompressionOptions
impl Sync for CompressionOptions
impl Unpin for CompressionOptions
impl UnwindSafe for CompressionOptions
type Owned = T
The resulting type after obtaining ownership.
fn clone_into(&self, target: &mut T)[src]
type Error = Infallible
The type returned in the event of a conversion error.