112 Commits

Author SHA1 Message Date
Vikas Arora
781c01f421 Simple Huffman code changes.
No empty trees are codified with the simple Huffman code. The simple Huffman
code is simplified to be either a 1-bit code or 8-bit code for symbols.

Change-Id: I3e2813027b5a643862729339303d80197c497aff
2012-05-07 14:25:21 -07:00
Urvang Joshi
a2849bc502 Lossless decoder: remove an unneeded param in ReadHuffmanCodeLengths().
Change-Id: I279452fdf38b680737d5ba6e868a219281bc8962
2012-05-07 14:25:20 -07:00
Pascal Massimino
8415ddf3be further simplification for the meta-Huffman coding
* don't transmit the number of Huffman tree group explicitly
* move color-cache information before the meta-Huffman block

* also add a check that color_cache_bits is in [1..11] range, as per spec.

Change-Id: I81d7711068653b509cdbc1151d93e229c4254580
2012-05-07 14:25:07 -07:00
Vikas Arora
e38b40a996 Factorize code for clearing HtreeGroup.
Change-Id: I29de6dab7383b8cf071eec155e01340d5fdadee5
2012-05-07 14:24:53 -07:00
James Zern
5c5be8ba69 VP8[L]GetInfo: check input pointers
validate data before using & width/height before assigning.

Change-Id: I0872e80fcbfea295d7c633b0d4cb7809e1d1883b
2012-05-02 16:05:02 -07:00
Pascal Massimino
23be6edfd4 few cosmetics
Change-Id: I85b44d811ecc7c551b75e74e64115e74645836b6
2012-04-24 11:53:10 -07:00
Vikas Arora
231ec1fb6d Removing the indirection of meta-huffman tables.
Now, the indexing refers directly to 5 huffman codes  that must be encoded separately.

Change-Id: I4deeb04de1997e6d20c376046d2053ec7ee918de
2012-04-23 08:36:14 -07:00
James Zern
9b3d1f3a1b decode.h: use size_t consistently
replaces mixed use of int/uint32_t for buffer sizes
further changes the API/ABI.

Change-Id: I91d70fd82ee3e1ac34b884b8ead9a114a9b1015a
2012-04-19 14:31:31 -07:00
Pascal Massimino
fac0f12e1b rename BitReader to VP8LBitReader
Change-Id: I192b76422e131a94fb58c2c4a5520a5dba807126
2012-04-13 01:56:31 -07:00
Urvang Joshi
4af1858a10 Fix a crash due to max symbol in a tree >= alphabet size
(cherry picked from commit 15253f91567ce7a2a4a3de8d70e54ba4b0641af3)

Change-Id: I4fbeaea4c712dae3bef078aef9d302b44bc39ffb
2012-04-10 23:07:13 -07:00
Pascal Massimino
6f01b830e2 split the VP8 and VP8L decoding properly
* each with their own decoder instances.
* Refactor the incremental buffer-update code a lot.
* remove br_offset_ for VP8LDecoder along the way
* make VP8GetHeaders() be used only for VP8, not VP8L bitstream
* remove VP8LInitDecoder()
* rename VP8LBitReaderResize() to VP8LBitReaderSetBuffer()
(cherry picked from commit 5529a2e6d47212a721ca4ab003215f97bd88ebb4)

Change-Id: I58f0b8abe1ef31c8b0e1a6175d2d86b863793ead
2012-04-10 23:06:58 -07:00
James Zern
19f6398ed7 add dec/vp8l{i.h,.c} from experimental
Pulled from 5529a2e^.
The history of this and related files is a bit entangled so rather
trying to split the changes and introduce some noise in master's history
we'll start with a fresh snapshot.
The file progression is still available in the experimental branch.

Change-Id: I490a49cb0084d556a29d7c5257c986ea695ce60c
2012-04-10 23:05:14 -07:00