Home

83 check-ins by user spaskalev

2015-09-22
09:14
imgui design notes additions Leaf check-in: 1e328bcd88 user: spaskalev tags: trunk
06:54
added design notes for imgui check-in: 4d1ed64397 user: spaskalev tags: trunk
2015-09-09
22:22
initial imgui api ideas check-in: bc1a46b4d9 user: spaskalev tags: trunk
2015-04-06
19:20
sync tbd check-in: 87e5dc43ff user: spaskalev tags: trunk
2015-03-19
20:51
Added parse.Digit and a test check-in: 61937abf4c user: spaskalev tags: trunk
20:40
Added a test for parse.Defer check-in: b37c997693 user: spaskalev tags: trunk
20:36
Added a string matching parser. check-in: 893c36d683 user: spaskalev tags: trunk
20:17
A simple parser combinator library check-in: 061baeefcb user: spaskalev tags: trunk
2015-01-26
23:02
full dep traversal in the collecting handler check-in: bca8ee3e12 user: spaskalev tags: trunk
21:32
Added missing depends interation in the collecting handler check-in: 3861ae4c57 user: spaskalev tags: trunk
20:40
Added tbd.slide as well check-in: 96db071674 user: spaskalev tags: trunk
20:30
Added tbd from gophergala/tbd :) check-in: 61b0c5b57a user: spaskalev tags: trunk
2015-01-14
20:46
Use for range in mtf's encoder for the position table. check-in: 3ea011cf07 user: spaskalev tags: trunk
2015-01-11
08:44
Switched encoding/fibonacci.Encoder back to using a 2-bytes array as a preallocated write buffer. It seems that it otherwise escapes everytime and calls a causes a lot of allocations and subsequent gc work. check-in: 90877fa64f user: spaskalev tags: trunk
2015-01-10
21:28
encoding/mtf optimizations check-in: 5c9c0a2164 user: spaskalev tags: trunk
2015-01-04
16:45
Use a lookup table for fibonacci's encoding of bytes. check-in: 9ded78a659 user: spaskalev tags: trunk
12:52
commands/mtf now uses fibonacci representation when encoding check-in: 6bd6e6d5c7 user: spaskalev tags: trunk
12:24
encoding/mtf.Encoder now returns an io.Reader. This allows to encoding in place without allocating buffers. check-in: 470d7e947b user: spaskalev tags: trunk
2015-01-02
17:55
update the copyright notice for 2015 check-in: 783d0b7f51 user: spaskalev tags: trunk
14:35
Minor optimizations to fibonacci.Encoder check-in: 98661a7373 user: spaskalev tags: trunk
2015-01-01
15:19
[fibonacci] renamed Writer->Encoder, Reader->Decoder. [mtf] moved package mtf to encoding\mtf check-in: cb736e7ad3 user: spaskalev tags: trunk
15:12
Use a common fibonacci number source for each encoder and decoder. Dropped the two-bytes buffer from the encoder struct. check-in: f1a8d5baa9 user: spaskalev tags: trunk
14:11
Initial implementation and test of fibonacci.Reader(io.Reader) io.Reader that decodes bytes encoded from fibonacci.Writer(..). CC at 97.5% check-in: 4b9f9bff39 user: spaskalev tags: trunk
06:34
Fixed Decode's length return and added a test for it. Reduce encoder's buffer to 2 bytes. CC at 98.3% check-in: ea97951fcd user: spaskalev tags: trunk
05:30
Added a proper test for fibonacci.Writer that writes 0-255, converts the resulting bits to a string and compares against the result of Numbers.Code for each value. check-in: c7c8d6445f user: spaskalev tags: trunk
04:58
Added fibonacci.Writer(io.Writer) io.Writer that encodes bytes. check-in: 93fcb281a1 user: spaskalev tags: trunk
2014-12-30
22:59
Dropped the Nth method and return a populated slice by fibonacci.New(size). Changed all access to direct indexing. CC at 100% check-in: ffb139e305 user: spaskalev tags: trunk
15:01
Minor opmitization to fibonacci's decoding. Changed plaindiff to show line numbers starting from 1. check-in: 9f5054e305 user: spaskalev tags: trunk
14:03
Added encoding/fibonacci (cc: 100%) check-in: 7a1684ea05 user: spaskalev tags: trunk
2014-12-28
18:06
[mtf] Removed :to indices from copy(..) destinations. check-in: c07658474d user: spaskalev tags: trunk
17:19
Integrate ioutil.ReadByte from the bpe branch into trunk. check-in: 00c4e0e448 user: spaskalev tags: trunk
2014-12-26
21:35
Added 0dev.org/types, providing aliases that implement sort.Interface for [u]int{8|16|32|64} check-in: 129d90b4a8 user: spaskalev tags: trunk
16:02
Initial implementation of commands/mtf (based on commands/pdc) check-in: e72b637df4 user: spaskalev tags: trunk
15:50
Initial implementation of package mtf - a move-to-front transform. check-in: ad5c25d9ab user: spaskalev tags: trunk
2014-12-25
08:48
Allocate the predictor's decompression buffer per Read call. check-in: f06f8cd058 user: spaskalev tags: trunk
01:25
Removed pdc's output buffer when decompressing as io.Copy uses a sufficiently-large buffer internally. check-in: 2b1ed8e45e user: spaskalev tags: trunk
00:55
Predictor's compressor and decompressor structures now implement io.Writer/io.Reader in order to deal away with function indirection but they do not follow the required semantics. Those are provided by the SizedWriter/SizedReader wrappers returned by the constructor functions. check-in: 4dfcff962c user: spaskalev tags: trunk
00:43
Extracted predictor's compressor and decompressor code into separate structs that embed Sized{Writer,Reader} check-in: 50507bd510 user: spaskalev tags: trunk
00:26
Extracted SizedWriter to a sizedWriter struct with a Write() method. check-in: 46da7a6ae9 user: spaskalev tags: trunk
2014-12-24
23:58
0dev.org/ioutil/SizedReader is now as fast as bufio.Reader if not faster for a buffer of 4096 bytes. Switched pdc to use it for decompression buffering and removed profiling code from the former check-in: 2cec92909f user: spaskalev tags: trunk
23:45
Extracted SizedReader to a sizedReader type with a Read() method. Closures seem to be slower in Go 1.4 and there is no real need for SizedReader to be a closure. check-in: b703c38e0b user: spaskalev tags: trunk
23:33
Made SizedReader faster by keeping explicit buffer indices. check-in: 701ac713de user: spaskalev tags: trunk
22:32
Use 0dev.org/ioutil.SizedWriter as an output buffer for pdc in both compress and decompress modes. check-in: 1717bfae3b user: spaskalev tags: trunk
21:40
Fixed SizedWriter behavior so that it follows io.Writer's Write(...) contract. Added more tests for 100% CC on the ioutil package. Predictor's compressor now uses SizedWriter and no longer has to do any internal buffering. check-in: e1778aba98 user: spaskalev tags: trunk
19:07
Implemented ioutil/SizedWriter. CC at 100%. check-in: c28a763d5e user: spaskalev tags: trunk
08:24
Added an explicit copyright notice in the code. check-in: ffd1ab7b0c user: spaskalev tags: trunk
2014-12-23
20:53
Renamed BlockReader to SizedReader, modified ioutil tests for 100% code coverage check-in: 11d1c50cd5 user: spaskalev tags: trunk
19:18
Fixing ioutil tests to compile :) check-in: b0ff11dfcd user: spaskalev tags: trunk
18:52
Renamed MinReader to BlockReader. The later is now used by predictor's decompressor to simplify the code and deal away with the need for internal buffering. check-in: 38f8e62c81 user: spaskalev tags: trunk
18:39
Additional fixes and code simplification for MinReader check-in: 70896f73e9 user: spaskalev tags: trunk
14:27
minReader fixes check-in: 4195e7817d user: spaskalev tags: trunk
10:38
Added MinReader to ioutils, CC at 100% check-in: 47b221d5b4 user: spaskalev tags: trunk
09:32
Removed the compressor alias type from predictor, use ioutil.WriterFunc check-in: c62581c2a6 user: spaskalev tags: trunk
08:15
Removed the decompressor alias type from predictor, use ioutil.ReaderFunc check-in: 2b049247ed user: spaskalev tags: trunk
07:52
Added package ioutil with io.Writer and io.Reader function wrappers check-in: 2bcd5307ea user: spaskalev tags: trunk
2014-12-22
19:52
Added documentation for the decompressor check-in: 89bfe97384 user: spaskalev tags: trunk
19:28
Extracted the predictor's hash function as a method of the context struct. Minor changes to the decompressor's variables. check-in: 9dfd3cb1a2 user: spaskalev tags: trunk
17:15
Calculate the decompressed block length outside of the predictor loop check-in: 27ecac81d3 user: spaskalev tags: trunk
16:41
Integrate the decompressor2 branch into trunk now that it is faster. check-in: 6d10a1d28f user: spaskalev tags: trunk
16:35
Decompressor - try to fill as much as possible in the output buffer in a single pass. Closed-Leaf check-in: 42ba1f458d user: spaskalev tags: decompressor2
15:34
Fixed a nasty variable shadowing bug :) check-in: e9b80a705b user: spaskalev tags: decompressor2
14:23
Read available bytes and the end of the buffer and decompress in a singe pass check-in: ae0940d072 user: spaskalev tags: decompressor2
2014-12-21
23:26
Closing the decompressor2 branch as this implementation is slower than the naive one. check-in: 52e14c83da user: spaskalev tags: decompressor2
23:24
The decompressor now tries to fill in the whole output buffer in a single call. check-in: 1f73d90f65 user: spaskalev tags: trunk
22:52
Removed TODOs, renamed readCount->rc, wrapped->reader check-in: 630530df49 user: spaskalev tags: trunk
22:12
Check in the new decompressor implementation in a separate branch check-in: bd1368b81f user: spaskalev tags: decompressor2
19:38
Added debug/pprof to ease basic cpu profiling check-in: 1a4bdf36e2 user: spaskalev tags: trunk
17:23
Fixed a rare case of losing data from the decompressor's internal result buffer. check-in: 7b74fd57f8 user: spaskalev tags: trunk
01:59
Added a function that reverses the bits in a byte. Coverage: 100.0% of statements. check-in: 2be2ff6bf7 user: spaskalev tags: trunk
2014-12-20
17:44
Buffer the input on decompressing, not the decompressor itself. This takes pdc -d < linux-3.18.1.tar.pdc > linux-3.18.1.tar down to 11 seconds from 13 minutes :) check-in: 4f0d26907d user: spaskalev tags: trunk
13:04
Decompressor might loose part of the underlying buffer array by reslicing, fixed by copy check-in: d516e7425d user: spaskalev tags: trunk
11:52
[predictor] Removed the buffer from the context struct, allocate the input slice buffer on creation with make. check-in: 723ffeb1fd user: spaskalev tags: trunk
2014-12-19
21:54
Implemented commands/pdc using predictor. Made predictor's Compressor(...) return an io.Writer. check-in: c9f3a59cb6 user: spaskalev tags: trunk
2014-12-16
23:29
Removed err variable from compressor check-in: 0f4bc650d1 user: spaskalev tags: trunk
22:56
Reworked the compressor's buffering code, switched to table testing, add step testing for the compressor. Code coverage at 88.3% check-in: 1847f77062 user: spaskalev tags: trunk
16:04
Removed goto from predictor's compressor, added more tests that invoke both compress/decompress check-in: 10013ae789 user: spaskalev tags: trunk
04:03
Fixed issues with both compressor and decompressor, added more tests check-in: b838653282 user: spaskalev tags: trunk
01:55
Added hamming weight lookup table for bytes in package bits. Added a PoC predictor decompressor implementation check-in: 60ca5b4b7b user: spaskalev tags: trunk
2014-12-15
21:22
Added diff.D struct; Added an RFC1978 Predictor compressor implementation and a test check-in: 67794bdf14 user: spaskalev tags: trunk
2014-12-14
20:41
Buffer stdout, use os.Args instead of flag check-in: 074a26f76b user: spaskalev tags: trunk
12:11
Skip diagonals shorter than the current match check-in: e01b36769d user: spaskalev tags: trunk
01:03
[bits] fixed package documentation [bits] added tests for the bool-backed vector [diff] added a real test, commented out the visual debug aids [plaindiff] added comments, removed commented code, added a helper []line getter [math] removed math package as it is currently unused check-in: d5adabf68e user: spaskalev tags: trunk
00:16
Initial implementation of a diff program and required libraries check-in: be5950faa4 user: spaskalev tags: trunk