-
Notifications
You must be signed in to change notification settings - Fork 5
/
Copy pathChangeLog
64 lines (57 loc) · 3.09 KB
/
ChangeLog
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
v3.1
- Added --translates, which outputs direct and indirect translations
- Added --orphan, which only outputs sentences that don’t belong to anyone
- Fixed misbehaviours --just-parse
- Added colored output, using --color
- Configure now has a --enable-python switch to add python bindings to the shared object
- Added --fuzzy N EXPR, which outputs the N sentences closest to EXPR.
- Numerous bug fixes.
- Changed behavior of --has-tag --in-list to treat names case-insensitively
v3.0
- Added support for lists.cv
- Fixed a bug in hdownloader.hpp
- Added python support
- Using boost.m4 in autoconf
- Added pkg-config for libcurl
- Added display-lang to output the language a sentence is written into
- Lowered autoconf requirements to 2.68
- Fixed unittests when DESTDIR was not /
v2.6
- Fixed a bug causing the parser to crash when a sentence of very high id was present in links.csv.
- The parser now works on windows
- Updated internal version of qlog
- Configure option --enable-download allows csv files to be downloaded via libcurl
- Fixed a crash when using display-first-translation
- ICU build options are now found through pkgconfig
- Internal symbols are no longer exported by the library
v2.5
- The code is now split into a library responsible for parsing the files into data structures,
and a binary using that library to filter the sentences in an appropriate way.
- Fixed a bug causing failure to return 0 instead of 1.
- Configure now has a --disable-debug to turn off debug.
- User option --lang can be supplied multiple times so that sentences from many languages can be kept.
- Unittests have been re-written to work with dash in addition to bash.
- Sentence parsing has been accelerated by estimating the number of sentences instead of counting the
number of lines in the file.
- Make has a new target called "check" to run unit tests.
- LDFLAGS are now passed to make.
v2.4
- Added --separator switch to change the way line numbers, sentence texts, and translations are separated.
- Added --has-id to look for a specific sentence
- Fixed a bug causing an invalid memory access at the end of the linkset container.
- The project now uses <boost/config.hpp> if it is found on the machine
- Added --display-first-translation to output a translation in addition to the sentence.
v2.3
- Fixed build when the user doesn’t have valgrind
- Fixed a bug causing a crash when links.csv and sentences.csv were inconsistent.
- mmap/munmap is no longer required (but are still used when available)
- boost::spirit is now used to parse sentences.csv. Thanks to liori.
- Fixed a lot of compilation problems reported by lool0 and sqrrl.
- Added unittests.
v2.2
- Automake is now used as a build system
- Added --is-translatable-in switch, which keeps sentences that can be translated in a given language.
- New switch --csv-path permits having the 3 main data files in a remote directory.
- Memory usage has been greatly reduced, from hundreds of MB to roughly 30 MB per data file.
v2.1
- First committed version. Tatoparser can filter sentences by language, regular expression and tags.