Tokenized Text Aligner Activity
Aligns tokens in two versions of a text with differing tokenization.
Status: Alpha
Brought to you by:
rainsfordtm
Activity for Tokenized Text Aligner
-
T. M. Rainsford committed [210f15] on Code
Modified error message for fixed aligner if token IDs don't match
-
Added mismatch option to inject from multiple B_IDs
-
Updated TTA to generate line IDS
-
Debug lib/io.py
-
Updated so that XML-W is added
-
Added XML-W as file format
-
Updated readme
-
Added support for .tsv files
-
Tidied untracked files
-
Update to io.py to deal with empty lines.
-
Quick bug fix in lib.io
-
Added an exact inject algorithm - one to one correspondances only.
-
Removed old v0.1.1 files
-
Disabled all but the generic injector
-
Updated readme.txt
-
Added sample files
-
Added Latin 1 encoding as an option.
-
Tweak/bug fix
-
Minor modifications (debugs, new error messages) made during testing on SRCMF-UD.
-
Added support for CSV with headers, initial testing completed.
-
Test phase of TTA script: succesfully injected into CONLL.
-
Universal launch script completed, as yet untested.
-
Forgotten modifications.
-
Began work on universal launch script - all formats, inject and align
-
Continued work on integrated launch script.
-
Added writer for CONLL-U format.
-
Completion of CONLL-U input module, start development of inject module.
-
Inject for CONLL-U implemented.
-
Work on new branch with IO module begun for CONLL upgrade
-
Added script to inject based on existing alignment table.
-
Added script to inject based on existing alignment table.
-
aligner library updated, removed BST specific m...
-
Threshold fixed, added readme.
-
Added "newline=''" to correct newlines in CSV o...
-
Corrected aligner reference, added __init__.py,...
-
Initial commit: aligncsv and aligner
1
