danvics/oxipng
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
861 commits 4 branches 85 tags 55 MiB
Commit graph

25 commits

Author SHA1 Message Date
andrews05
bbde68ddce
Refactor optimize_raw (#670) 
Code always tends to get messy over time. I've found the `optimize_raw`
function increasingly harder to read, particularly after the addition of
fast mode, so I've taken some time to refactor and simplify it.

One change of note here is the main compression trials now use the
Evaluator. This means verbose output is a little different which is
shown below.

There is no change to performance or output size.

`-vvo2`: master
```
Processing: tests/files/rgba_8_should_be_palette_4.png
    500x400 pixels, PNG format
    8-bit RGB + Alpha, non-interlaced
    IDAT size = 2757 bytes
    File size = 18109 bytes
Eval: 4-bit Indexed (5 colors)      None       1837 bytes
Eval: 8-bit Indexed (5 colors)      None       1988 bytes
Eval: 4-bit Indexed (5 colors)      Bigrams   >1837 bytes
Eval: 8-bit Indexed (5 colors)      Bigrams   >1837 bytes
Transformed image to 4-bit Indexed (5 colors), non-interlaced
Evaluating: 2 filters
Eval: 4-bit Indexed (5 colors)      Sub       >1810 bytes
Eval: 4-bit Indexed (5 colors)      Entropy   >1810 bytes
Trying: None
    zc = 11  f = None      1583 bytes
Found better combination:
    zc = 11  f = None      1583 bytes
    IDAT size = 1583 bytes (1174 bytes decrease)
    file size = 16962 bytes (1147 bytes = 6.33% decrease)
16962 bytes (6.33% smaller): Running in pretend mode, no output
```

`-vvo2`: PR
```
Processing: tests/files/rgba_8_should_be_palette_4.png
    500x400 pixels, PNG format
    8-bit RGB + Alpha, non-interlaced
    IDAT size = 2757 bytes
    File size = 18109 bytes
Eval: 4-bit Indexed (5 colors)      None       1837 bytes
Eval: 8-bit Indexed (5 colors)      None       1988 bytes
Eval: 4-bit Indexed (5 colors)      Bigrams   >1837 bytes
Eval: 8-bit Indexed (5 colors)      Bigrams   >1837 bytes
Transformed image to 4-bit Indexed (5 colors), non-interlaced
Evaluating 2 filters
Eval: 4-bit Indexed (5 colors)      Sub       >1810 bytes
Eval: 4-bit Indexed (5 colors)      Entropy   >1810 bytes
Trying filter None with zc = 11
1610 bytes
Found better result:
    zc = 11, f = None
    IDAT size = 1583 bytes (1174 bytes decrease)
    file size = 16962 bytes (1147 bytes = 6.33% decrease)
16962 bytes (6.33% smaller): Running in pretend mode, no output
```

`-vvZo5`: master
```
Processing: tests/files/rgba_8_should_be_palette_4.png
    500x400 pixels, PNG format
    8-bit RGB + Alpha, non-interlaced
    IDAT size = 2757 bytes
    File size = 18109 bytes
Eval: 8-bit Indexed (battiato sort) None       1821 bytes
Eval: 4-bit Indexed (5 colors)      None       1657 bytes
Eval: 8-bit Indexed (mzeng sort)    None       1821 bytes
Eval: 8-bit Indexed (5 colors)      None       1821 bytes
Eval: 8-bit Indexed (battiato sort) Bigrams   >1821 bytes
Eval: 4-bit Indexed (5 colors)      Bigrams   >1657 bytes
Eval: 8-bit Indexed (mzeng sort)    Bigrams   >1657 bytes
Eval: 8-bit Indexed (5 colors)      Bigrams   >1657 bytes
Transformed image to 4-bit Indexed (5 colors), non-interlaced
Trying: 8 filters
    zc = zopfli  f = Brute     1562 bytes
    zc = zopfli  f = Sub      >1562 bytes
    zc = zopfli  f = Bigrams  >1562 bytes
    zc = zopfli  f = None      1407 bytes
    zc = zopfli  f = Up       >1407 bytes
    zc = zopfli  f = MinSum   >1407 bytes
    zc = zopfli  f = BigEnt   >1407 bytes
    zc = zopfli  f = Entropy  >1407 bytes
Found better combination:
    zc = zopfli  f = None      1407 bytes
    IDAT size = 1407 bytes (1350 bytes decrease)
    file size = 16786 bytes (1323 bytes = 7.31% decrease)
16786 bytes (7.31% smaller): Running in pretend mode, no output
```

`-vvZo5`: PR
```
Processing: tests/files/rgba_8_should_be_palette_4.png
    500x400 pixels, PNG format
    8-bit RGB + Alpha, non-interlaced
    IDAT size = 2757 bytes
    File size = 18109 bytes
Eval: 8-bit Indexed (battiato sort) None       1821 bytes
Eval: 4-bit Indexed (5 colors)      None       1657 bytes
Eval: 8-bit Indexed (mzeng sort)    None       1821 bytes
Eval: 8-bit Indexed (5 colors)      None       1821 bytes
Eval: 8-bit Indexed (battiato sort) Bigrams   >1657 bytes
Eval: 4-bit Indexed (5 colors)      Bigrams   >1657 bytes
Eval: 8-bit Indexed (mzeng sort)    Bigrams   >1657 bytes
Eval: 8-bit Indexed (5 colors)      Bigrams   >1657 bytes
Transformed image to 4-bit Indexed (5 colors), non-interlaced
Trying 8 filters with zopfli, zi = 15
Eval: 4-bit Indexed (5 colors)      Brute      1589 bytes
Eval: 4-bit Indexed (5 colors)      Bigrams    1641 bytes
Eval: 4-bit Indexed (5 colors)      Sub        1711 bytes
Eval: 4-bit Indexed (5 colors)      None       1434 bytes
Eval: 4-bit Indexed (5 colors)      Up         1764 bytes
Eval: 4-bit Indexed (5 colors)      MinSum     1760 bytes
Eval: 4-bit Indexed (5 colors)      BigEnt     1742 bytes
Eval: 4-bit Indexed (5 colors)      Entropy    1748 bytes
Found better result:
    zopfli, zi = 15, f = None
    IDAT size = 1407 bytes (1350 bytes decrease)
    file size = 16786 bytes (1323 bytes = 7.31% decrease)
16786 bytes (7.31% smaller): Running in pretend mode, no output
```
 2025-01-29 20:53:11 +01:00
Kornel
a72bc945fc Clippy 2025-01-02 09:15:24 +13:00
Alejandro González
0608ab9cab
Fix some new nightly Clippy lints and run rustfmt with prettier config 2024-02-20 19:12:35 +01:00
Josh Holmer
31e796c4e8 Revert "Provide BufferedZopfliDeflater and allow user to pass in a custom Deflater (#530)" 
This reverts commit 2a59419bdf.
 2023-07-09 13:40:20 -04:00
Josh Holmer
2a59419bdf
Provide BufferedZopfliDeflater and allow user to pass in a custom Deflater (#530) 
* Add .whitesource configuration file

* Experimental: allow Zopfli to use any size BufWriter

* Allow user to specify the output buffer size as well

* Allow user to specify maximum block splits

* Reformat and fix warnings

* Use deflater on iCCP chunk as well

* Bug fix: need to implement Zlib format

* Make functions const when possible

* Switch to using zopfli::Options in prep for https://github.com/zopfli-rs/zopfli/pull/21

* Switch to using zopfli::Options in prep for https://github.com/zopfli-rs/zopfli/pull/21

* Cargo fmt

* Fix compilation

* Fix tests

* Fix more lints

* Fix more lints

* Fix compilation more

---------

Co-authored-by: mend-bolt-for-github[bot] <42819689+mend-bolt-for-github[bot]@users.noreply.github.com>
Co-authored-by: Chris Hennick <hennickc@amazon.com>
Co-authored-by: Chris Hennick <4961925+Pr0methean@users.noreply.github.com>
 2023-07-07 22:35:01 -04:00
Andrew
c71fd44454 Prevent conversion from grayscale when disabled 2023-06-18 01:02:11 -04:00
andrews05
ea5f1884be
Refactor aux chunk handling (#505) 2023-05-21 15:34:23 -04:00
andrews05
d8b7ebaf47
Further small reduction improvements (#504) 2023-05-16 20:53:05 -04:00
andrews05
9a500941d8
Raw API (#482) 2023-05-15 08:37:49 -04:00
andrews05
2f622fc7bd
Minor reduction improvements (#502) 
* Fix bKGD conversion from gray to palette

* Allow grayscale reduction from 16 to 4 or less

* Refactor reduction evaluation sequence

* Separate palette into new file
 2023-05-07 17:46:58 -04:00
andrews05
a26d225d81
More flexible verbosity, improve verbose test (#501) 
* More flexible verbosity, improve verbose test

* Tweak reporting format

* Add evaluation reporting at trace level
 2023-05-06 10:45:31 -04:00
andrews05
c97572be3e
ColorType Refactor (#500) 
* Refactor ColorType
Move transparency and palette data into the ColorType

* Fixup tests

* Make more use of helper functions

* Change BitDepth to u8 representation with TryFrom

* Fix clippy lints

* Don't use unstable language features

* Restore documentation on transparency/palette
 2023-04-27 19:49:54 -04:00
andrews05
131a3c6af7
🚀 Revamp alpha optimisation (#475) 2022-12-10 21:14:50 -05:00
Josh Holmer
17f5cc36ca Fix clippy lints 2022-07-05 10:05:06 -04:00
Kornel
14867c7abc
Clippy lints (#401) 2021-05-19 10:04:31 -04:00
Ingvar Stepanyan
63e36ed43f
Simplify comparer (#208) 
* Skip baselines earlier in the evaluator

We know that we'll throw them away anyway, so there is no point in even sending them to Sender and comparing them with others once we used them for the best candidate size.

* Simplify comparison logic
 2020-04-14 17:09:52 -04:00
Josh Holmer
d86bc80f43 Prepare for next release 2019-02-01 17:27:03 -05:00
Kornel
9af874d21b Make reductions return a new uncompressed image (#158) 
* Fix verbose message

* No cloning when restoring original data

* Make reductions return a new uncompressed image

Partially fixes #145

* Async reduction evaluator

* Assert

* Faster bit depth check

* Also try 4-bit depth for small-depth images

* Skip test when using miniz

* Ensure palette is trimmed after depth reduction

Fixes #159

* Fudge factor for reductions to prefer better reductions even if gzip estimation says otherwise
 2019-01-27 03:22:49 -05:00
Josh Holmer
f862a0df24 Apply clippy fixes 2018-07-28 16:51:54 -04:00
Josh Holmer
348c7dbc56 Extract a function into a submodule 2018-01-30 22:21:31 -05:00
Josh Holmer
f41ef551d3 Enable option for multiple alpha transformations 2017-07-26 00:37:55 -04:00
Josh Holmer
413f1aac7c Automatically change transparent pixels to black 2017-07-24 08:47:27 -04:00
Josh Holmer
5f4ddbb8f4 Upgrade dependencies 2017-06-14 18:28:07 -04:00
Josh Holmer
64edc9110e Start using quickcheck and add inline hints 2016-07-30 23:14:36 -04:00
Josh Holmer
7f9ad263ec Refactor code into modules 2016-07-14 11:10:10 -04:00