50 most recent check-ins occurring on or after 2016-12-16 20:55:19.
More ↑
|
2017-02-22
| ||
| 22:45 | smaz-tests: add a test of multi-block verbatim sequence check-in: 3b0fbf12a9 user: nat tags: trunk | |
|
2017-02-21
| ||
| 21:11 | smaz_implementations-base_256: fix exception on 1-byte non-VLV blocks check-in: baef3cc8be user: nat tags: trunk | |
|
2017-02-20
| ||
| 20:16 | smaz_implementations-base_256: fix off-by-one error in Verbatim_Size check-in: 1183e8ca09 user: nat tags: trunk | |
|
2017-02-19
| ||
| 21:06 | smaz_tests: copy base-256 non-VLV tests from retired smaz test suite check-in: d389660384 user: nat tags: trunk | |
|
2017-02-18
| ||
| 22:14 | smaz_tests: rename base-256 sample string test for consistency check-in: beb52edfa5 user: nat tags: trunk | |
|
2017-02-17
| ||
| 21:33 | smaz: fix off-by-one error in Verbatim_Size check-in: 1f3ace98c0 user: nat tags: trunk | |
|
2017-02-16
| ||
| 20:56 | smaz-tests: add sample strings without variable-length verbatim check-in: f1bcc5d576 user: nat tags: trunk | |
|
2017-02-15
| ||
| 22:57 | coverage.sh: use gprbuild instead of gnatmake check-in: 89d897146c user: nat tags: trunk | |
|
2017-02-14
| ||
| 21:04 | smaz_generic-tools: fix the precondition of To_Dictionary check-in: 8458c28c3c user: nat tags: trunk | |
|
2017-02-13
| ||
| 21:12 | smaz_tests: add multi-block base-4096 verbatim, to reach full coverage check-in: 3110e5329d user: nat tags: trunk | |
|
2017-02-12
| ||
| 21:52 | smaz_tests: add base-4096 variable-length verbatim sample string test check-in: cf6b7c828f user: nat tags: trunk | |
|
2017-02-11
| ||
| 20:52 |
smaz_tests: fix verbatim parts of expected base-4096 streams
Why didn't just use standard base-64!? check-in: 0d15ccca30 user: nat tags: trunk | |
|
2017-02-10
| ||
| 20:58 | smaz_tests: update base-4096 expected streams for the new dictionaries check-in: 541596303a user: nat tags: trunk | |
|
2017-02-09
| ||
| 22:57 | smat_tests: add 256 digraphs to the base-4096 dictionaries check-in: 26569a2ae1 user: nat tags: trunk | |
|
2017-02-08
| ||
| 22:23 | smaz_tests: remove unneeded exception caused by byte-4096 hash function check-in: b2360c3935 user: nat tags: trunk | |
|
2017-02-07
| ||
| 20:48 | smaz_tests: remove 1-byte entries from 4096 dictionaries check-in: 61cc0370b5 user: nat tags: trunk | |
|
2017-02-06
| ||
| 22:09 | smaz_generic: add a description and an important warning check-in: 5793080d7d user: nat tags: trunk | |
|
2017-02-05
| ||
| 22:42 | smaz_tests: fix a mistake in a hand-made expected base-4096 stream check-in: 18dc02b3a0 user: nat tags: trunk | |
|
2017-02-04
| ||
| 22:25 | smaz_tests: fix endianness of the hand-encoded base-4096 streams check-in: 97980e3f18 user: nat tags: trunk | |
|
2017-02-03
| ||
| 21:08 | smaz_tests: fix base-4096 sample string test by actually calling it check-in: b4e3aec058 user: nat tags: trunk | |
|
2017-02-02
| ||
| 20:16 | smaz_tests: check base-4096 roundtrip of a few sample strings check-in: 46b1fe3862 user: nat tags: trunk | |
|
2017-02-01
| ||
| 21:37 | smaz_tests: add a base-4096 test harness check-in: 1d5eb501bc user: nat tags: trunk | |
|
2017-01-31
| ||
| 22:33 | tools/smaz: add support for base-4096 variant check-in: e6f252142c user: nat tags: trunk | |
|
2017-01-30
| ||
| 21:33 | smaz_4096: new instance of generic Smaz, base-64 and large dictionary check-in: c57a785ddb user: nat tags: trunk | |
|
2017-01-29
| ||
| 21:32 | tools/timekey: add CLI option to configure numer of subsecond digits check-in: b40ac98134 user: nat tags: trunk | |
|
2017-01-28
| ||
| 20:09 | tools/timekey: refactor the number of displayed subsecond digits check-in: f4f3b7b682 user: nat tags: trunk | |
|
2017-01-27
| ||
| 20:44 | tools/timekey: refactor command-line argument processing check-in: f9d3f1aa38 user: nat tags: trunk | |
|
2017-01-07
| ||
| 22:53 |
smaz_implementations-base_64: rewrite variable-length verbatim code
At some point in the development I changed the base-64 scheme, but failed to propagate it everywhere. Mostly I intend to use only one mode in base-64, and the other one is merely a place-holder, which explains how it got neglected. Now at least it works, even if it's useless. check-in: 210677c061 user: nat tags: trunk | |
|
2017-01-06
| ||
| 22:44 | smaz_tests: add a test for base-64 variable-length verbatim check-in: 3e04c004e7 user: nat tags: trunk | |
|
2017-01-05
| ||
| 22:56 | Add an ignore file check-in: 1877439212 user: nat tags: trunk | |
|
2017-01-04
| ||
| 23:07 | smaz_implementations-base_64_tools: fix exception on non-ASCII character check-in: a3ae81e9c1 user: nat tags: trunk | |
|
2017-01-03
| ||
| 20:41 | smaz_tests: add a check with non-ASCII characters check-in: 03f59a962d user: nat tags: trunk | |
|
2017-01-02
| ||
| 20:22 | smaz_tests: ASCIIfy the base-64 dictionary, to make it more portable check-in: 8f0534e6a4 user: nat tags: trunk | |
|
2017-01-01
| ||
| 21:43 | smaz_implementations-base_64: fix multi-block verbatim size computation check-in: c49fad3790 user: nat tags: trunk | |
|
2016-12-31
| ||
| 16:57 | smaz_tests: add a check to show a bug in long verbatim size computation check-in: 4f0ae7072f user: nat tags: trunk | |
|
2016-12-30
| ||
| 22:11 | smaz_generic-tools: fix removal of the first item of a dictionary check-in: dd10a5bea1 user: nat tags: trunk | |
|
2016-12-29
| ||
| 21:47 | smaz_implementations-base_64: fix index update in encoding 3n+2 vrbtim check-in: ebba07e9a7 user: nat tags: trunk | |
|
2016-12-28
| ||
| 22:29 |
smaz_implementations-base_64: fix reading of 3n+2 verbatim length
A sign error caused a direct constraint error, and using the wrong type caused overlong reads (now 4 is hardcoded just like it is in encoding code). check-in: 23d16059bd user: nat tags: trunk | |
|
2016-12-27
| ||
| 21:28 | smaz_tests: fix message of exceptions during decompression roundtrip check-in: 77d59bd0f4 user: nat tags: trunk | |
|
2016-12-26
| ||
| 20:00 | smaz_tests: add tests to fully cover non-variable-verbatim code check-in: c0e42e8031 user: nat tags: trunk | |
|
2016-12-25
| ||
| 21:39 | smaz_tests: add dictionary validity tests check-in: bc33d2a603 user: nat tags: trunk | |
|
2016-12-24
| ||
| 15:51 | smaz_tests: output base-64 stream directly, for more legibility check-in: 80ce774782 user: nat tags: trunk | |
|
2016-12-23
| ||
| 20:39 | smaz_tests: add a very basic test for base-64 Smaz check-in: a277512ddf user: nat tags: trunk | |
|
2016-12-22
| ||
| 20:47 | smaz_tests: update to make base-256-ness explicit check-in: 1a7b497cb0 user: nat tags: trunk | |
|
2016-12-21
| ||
| 20:24 | smaz_test_base_64_hash: sample generated base-64 smaz hash for tests check-in: d43e8f04d9 user: nat tags: trunk | |
|
2016-12-20
| ||
| 21:55 | tools/smaz: add support for base-64 variant check-in: 9de1423410 user: nat tags: trunk | |
|
2016-12-19
| ||
| 20:18 | smaz_64: new instance of generic smaz, outputs directly base-64 symbols check-in: 091a96ec6c user: nat tags: trunk | |
|
2016-12-18
| ||
| 21:16 | smaz_implementations-base_64_tools: new package for paddingless base-64 check-in: fd2ccb00b9 user: nat tags: trunk | |
|
2016-12-17
| ||
| 22:48 | tools/smaz: add a roundtrip check option to help debug new code check-in: f44feb5e3e user: nat tags: trunk | |
|
2016-12-16
| ||
| 20:55 |
smaz_generic: optimize compression
For some reason it seems even with -O3, calling Dict_Entry involves a string copy, which makes `memcpy` the larger time consumer of the copmpression algorithm. Inlining it manually improves performance a lot. check-in: 674fadc74b user: nat tags: trunk | |