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
|
Version 2.1.8
=============
* partitioning of the database table
- this allows much faster import of new corpora, even if there are already
big corpora in the database
- some performance improvement since the database tables to load are smaller
- performance of smaller corpora is not affected by bigger corpora
* the "new" partitur visualizer is now able to display overlapping spans so the
old version is obsolete and can be replaced
* highlight the matches belonging to different parts of the query in different
colors, especially helpful if you query something like cat="S" & tok & #1 > #2
* cache for resolver entries for faster display of the results
* text exporter which is exporting the the token annotations
* general cleanup in the code
* fix bug that didn't allow "+" to be used in exporters
* don't throw an exception if you search for a metadata value that does not
exist in the corpus (just deliver 0 results)
* more escaping of user input in SQL
* updated to the latest version (3.2.1) of the ExtJS JavaScript-Framework
(bug-fix release)
* fixed some size issues in the GUI
Version 2.1.7
=============
* Negation of word forms (e.g. in the pcc2 demo corpus: tok!=”in”), of attribute
values (pos!=”NN”) and of edge labels
(cat=”S” & cat=”PP” & #1 >[func!=”MO”] #2).
* Configurable namespaces and display names for visualizers, allowing e.g.
multiple tree visualizations for multiple parses etc. (see attached user
guide).
* Preliminary support for parallel corpora: parallel corpora are now importable
and alignments on all levels can be searched for; there is as yet no
visualization for the alignment edges – results with multiple languages are
arranged under each other in the KWIC view with hit elements highlighted
(see ANNIS site for a picture).
* Pointing relations can now carry both types and labels, allowing for both
annotated dependency trees (e.g. relations like #1 ->dep[func="DET"] #2 are
searchable, though there is no special visualization for them yet) and labeled
alignments (e.g. for “fuzzy” vs. “good” alignment etc.).
* A new basic KWIC exporter in the export tab (just the matched tokens with
context, in plain text)
* Acceleration and parallelization of certain queries (as a side effect, a first
page of hits may now be retrieved before the complete match count is
calculated)
* Improvements in the behavior of resizing, highlighting and scrollbars in
visualizers
* Various bug fixes
|