Debian Codegraph @

.cargo/vendor/png/CHANGES.md

raw
1 ## 0.18.0
2
3 ### API Breaking Changes
4
5 * Removed deprecated `Info::encode` and `Encoder::set_srgb` methods.
6 * Improved the compression settings API for encoding.
7 * `Decoder` now requires a reader that implements `Seek` and `BufRead` traits.
8 * Bump bitflags dependency to 2.0.
9 * `StreamingDecoder::update` now takes a structured `UnfilterBuf` argument
10 instead of a direct reference to a vector. This allows in-place
11 decompression. There is a public constructor for `UnfilterBuf`.
12 * The methods `Decoder::output_buffer_size` and `output_line_size` now return
13 `Option<usize>` to reflect that these calculations no longer overflow on some
14 targets where the required buffers can not be represented in the address
15 space. They return the mathematically correct size where possible.
16 * The `Decoded` enum returned from `StreamingDecoder::update` was simplified to
17 no longer contains any chunk payload data. Instead, it now contains only
18 chunk events where every chunk that was started will eventually be ended by
19 `ChunkComplete`, `BadAncillaryChunk` or `SkippedAncillaryChunk`.
20 * Ancillary chunks, i.e. those not critical to decoder interpretation of the
21 file, which fail to parse are now terminated with a `BadAncillaryChunk` event
22 but no longer returned a `DecodingError`. This includes text chunks as well
23 as many metadata chunks (except for `fcTL` that we deem crucial to the
24 parser's understanding of the image sequence in an APNG despite being
25 technically ancillary).
26
27 ### Additions
28
29 * Added `Reader::read_row` method.
30 * Add support for parsing eXIf chunk.
31 * Treat most auxiliary chunk errors as benign.
32 * Added `splat_interlaced_row`, which implements an alternative method for
33 merging Adam7 interlaced lines into the output buffer that is more suitable
34 for the presentation of progressive states of the buffer.
35 * Added `Adam7Variant` documenting the various methods for applying interlaced
36 rows and to prepare an API to progressively read frames through `Decoder`.
37
38 ### Changes
39
40 * The decoding of Adam7 interlaced data is now much faster.
41 * The `acTL` chunk is now ignored when it is invalid, instead of producing
42 errors while reading or decoding the following APNG chunks.
43 * The requirement of the `fcTL` chunk for the default image to match the IHDR's
44 indicate image size is now enforced.
45 * More minor format errors in auxiliary chunks are now ignored by the decoder,
46 instead disregarding the malformed chunk.
47 * Adam7 Interlacing on 32-bit targets now handles some cases correctly that
48 previously wrote some bytes to the wrong pixel indices due to overflows.
49
50 ## 0.17.16
51
52 * Make gAMA and cHRM fallback optional for sRGB ([#547])
53 * Pass through nightly feature to crc32fast crate to get SIMD crc32 on Aarch64 ([#545])
54 * Fix bug in iCCP chunk encoding ([#458])
55 * Deprecate Info::encode ([#550])
56
57 [#545]: https://github.com/image-rs/image-png/pull/545
58 [#458]: https://github.com/image-rs/image-png/pull/548
59 [#550]: https://github.com/image-rs/image-png/pull/550
60 [#547]: https://github.com/image-rs/image-png/pull/547
61
62 ## 0.17.15
63
64 ### Added
65
66 * Add a public API to advance to the next frame in APNG decoder ([#518])
67 * Add APIs to write ICC and EXIF chunks ([#526])
68 * Add support for parsing the sBIT chunk ([#524])
69 * Add support for parsing the bKGD chunk ([#538])
70 * Add support for parsing the cICP chunk ([#529])
71 * Add support for parsing mDCV and cLLI chunks ([#528], ([#543]))
72
73 ### Changed
74
75 * Improve performance of Paeth filter during decoding ([#512], [#539])
76
77 ### Fixed
78
79 * Avoid an infinite loop when retrying after a fatal error using `StreamingDecoder` ([#520])
80 * Fixed chunk order in encoded files ([#526])
81
82 [#495]: https://github.com/image-rs/image-png/pull/495
83 [#496]: https://github.com/image-rs/image-png/pull/496
84 [#512]: https://github.com/image-rs/image-png/pull/512
85 [#518]: https://github.com/image-rs/image-png/pull/518
86 [#520]: https://github.com/image-rs/image-png/pull/520
87 [#524]: https://github.com/image-rs/image-png/pull/524
88 [#526]: https://github.com/image-rs/image-png/pull/526
89 [#528]: https://github.com/image-rs/image-png/pull/528
90 [#529]: https://github.com/image-rs/image-png/pull/529
91 [#538]: https://github.com/image-rs/image-png/pull/538
92 [#539]: https://github.com/image-rs/image-png/pull/539
93 [#543]: https://github.com/image-rs/image-png/pull/543
94
95 ## 0.17.14
96
97 * Updated to miniz_oxide 0.8.0.
98 * Added public API to consume interlaced rows one by one ([#495])
99 * Improved support for resuming decoding after an `UnexpectedEof`, which lets you start parsing a file before it's fully received over the network ([#496])
100 * Fixed some broken links in documentation, improved some documentation comments
101
102 [#495]: https://github.com/image-rs/image-png/pull/495
103 [#496]: https://github.com/image-rs/image-png/pull/496
104
105
106 ## 0.17.13
107
108 * Fix `Send` bound on `Reader`.
109
110 ## 0.17.12
111
112 * Reject zero-sized frames.
113 * Optimized decoding of paletted images.
114 * Removed remaining uses of miniz_oxide for decoding.
115 * Correct lifetime used for `Info` struct.
116 * Fix build issue with `-Z minimal-versions`.
117
118 ## 0.17.11
119
120 * Ignore subsequent iCCP chunks to match libpng behavior.
121 * Added an option to ignore ancillary chunks with invalid CRC.
122 * Added `new_with_info` constructor for encoder.
123 * Removed hard-coded memory limits.
124 * No longer allow zero sized images.
125 * Added `Reader::finish` to read all the auxiliary chunks that comes after the
126 image.
127
128 ## 0.17.10
129
130 * Added Transformations::ALPHA
131 * Enable encoding pixel dimensions
132
133 ## 0.17.9
134
135 * Fixed a bug in ICC profile decompression.
136 * Improved unfilter performance.
137
138 ## 0.17.8
139
140 * Increased MSRV to 1.57.0.
141 * Substantially optimized encoding and decoding:
142 - Autovectorize filtering and unfiltering.
143 - Make the "fast" compression preset use fdeflate.
144 - Switch decompression to always use fdeflate.
145 - Updated to miniz_oxide 0.7.
146 - Added an option to ignore checksums.
147 * Added corpus-bench example which measures the compression ratio and time to
148 re-encode and subsequently decode a corpus of images.
149 * More fuzz testing.
150
151 ## 0.17.7
152
153 * Fixed handling broken tRNS chunk.
154 * Updated to miniz_oxide 0.6.
155
156 ## 0.17.6
157
158 * Added `Decoder::read_header_info` to query the information contained in the
159 PNG header.
160 * Switched to using the flate2 crate for encoding.
161
162 ## 0.17.5
163
164 * Fixed a regression, introduced by chunk validation, that made the decoder
165 sensitive to the order of `gAMA`, `cHRM`, and `sRGB` chunks.
166
167 ## 0.17.4
168
169 * Added `{Decoder,StreamDecoder}::set_ignore_text_chunk` to disable decoding of
170 ancillary text chunks during the decoding process (chunks decoded by default).
171 * Added duplicate chunk checks. The decoder now enforces that standard chunks
172 such as palette, gamma, … occur at most once as specified.
173 * Added `#[forbid(unsafe_code)]` again. This may come at a minor performance
174 cost when decoding ASCII text for now.
175 * Fixed a bug where decoding of large chunks (>32kB) failed to produce the
176 correct result, or fail the image decoding. As new chunk types are decoded
177 this introduced regressions relative to previous versions.
178
179 ## 0.17.3
180
181 * Fixed a bug where `Writer::finish` would not drop the underlying writer. This
182 would fail to flush and leak memory when using a buffered file writers.
183 * Calling `Writer::finish` will now eagerly flush the underlying writer,
184 returning any error that this operation may result in.
185 * Errors in inflate are now diagnosed with more details.
186 * The color and depth combination is now checked in stream decoder.
187
188 ## 0.17.2
189
190 * Added support for encoding and decoding tEXt/zTXt/iTXt chunks.
191 * Added `Encoder::validate_sequence` to enable validation of the written frame
192 sequence, that is, if the number of written images is consistent with the
193 animation state.
194 * Validation is now off by default. The basis of the new validation had been
195 introduced in 0.17 but this fixes some cases where this validation was too
196 aggressive compared to previous versions.
197 * Added `Writer::finish` to fully check the write of the end of an image
198 instead of silently ignoring potential errors in `Drop`.
199 * The `Writer::write_chunk` method now validates that the computed chunk length
200 does not overflow the limit set by PNG.
201 * Fix an issue where the library would panic or even abort the process when
202 `flush` or `write` of an underlying writer panicked, or in some other uses of
203 `StreamWriter`.
204
205 ## 0.17.1
206
207 * Fix panic in adaptive filter method `sum_buffer`
208
209 ## 0.17.0
210
211 * Increased MSRV to 1.46.0
212 * Rework output info usage
213 * Implement APNG encoding
214 * Improve ergonomics of encoder set_palette and set_trns methods
215 * Make Info struct non-exhaustive
216 * Make encoder a core feature
217 * Default Transformations to Identity
218 * Add Adaptive filtering method for encoding
219 * Fix SCREAM_CASE on ColorType variants
220 * Forbid unsafe code
221
222 ## 0.16.7
223
224 * Added `Encoder::set_trns` to register a transparency table to be written.
225
226 ## 0.16.6
227
228 * Fixed silent integer overflows in buffer size calculation, resulting in
229 panics from assertions and out-of-bounds accesses when actually decoding.
230 This improves the stability of 32-bit and 16-bit targets and make decoding
231 run as stable as on 64-bit.
232 * Reject invalid color/depth combinations. Some would lead to mismatched output
233 buffer size and panics during decoding.
234 * Add `Clone` impl for `Info` struct.
235
236 ## 0.16.5
237
238 * Decoding of APNG subframes is now officially supported and specified. Note
239 that dispose ops and positioning in the image need to be done by the caller.
240 * Added encoding of indexed data.
241 * Switched to `miniz_oxide` for decompressing image data, with 30%-50% speedup
242 in common cases and up to 200% in special ones.
243 * Fix accepting images only with consecutive IDAT chunks, rules out data loss.
244
245 ## 0.16.4
246
247 * The fdAT frames are no longer inspected when the main image is read. This
248 would previously be the case for non-interlaced images. This would lead to
249 incorrect failure and, e.g. an error of the form `"invalid filter method"`.
250 * Fix always validating the last IDAT-chunks checksum, was sometimes ignored.
251 * Prevent encoding color/bit-depth combinations forbidden by the specification.
252 * The fixes for APNG/fdAT enable further implementation. The _next_ release is
253 expected to officially support APNG.
254
255 ## 0.16.3
256
257 * Fix encoding with filtering methods Up, Avg, Paeth
258 * Optimize decoding throughput by up to +30%
259
260 ## 0.16.2
261
262 * Added method constructing an owned stream encoder.
263
264 ## 0.16.1
265
266 * Addressed files bloating the packed crate
267
268 ## 0.16.0
269
270 * Fix a bug compressing images with deflate
271 * Address use of deprecated error interfaces
272
273 ## 0.15.3
274
275 * Fix panic while trying to encode empty images. Such images are no longer
276 accepted and error when calling `write_header` before any data has been
277 written. The specification does not permit empty images.
278
279 ## 0.15.2
280
281 * Fix `EXPAND` transformation to leave bit depths above 8 unchanged
282
283 ## 0.15.1
284
285 * Fix encoding writing invalid chunks. Images written can be corrected: see
286 https://github.com/image-rs/image/issues/1074 for a recovery.
287 * Fix a panic in bit unpacking with checked arithmetic (e.g. in debug builds)
288 * Added better fuzzer integration
289 * Update `term`, `rand` dev-dependency
290 * Note: The `show` example program requires a newer compiler than 1.34.2 on
291 some targets due to depending on `glium`. This is not considered a breaking
292 bug.
293
294 ## 0.15
295
296 Begin of changelog