mmap (see irstlm/src/*cpp and irstlm/src/*h)
- fixed a bug in querying IRST LMs with OOVs (LanguageModelIRST.cpp)
- some more checks on config file: if specified, existence of generation
and distortion files is checked (Parameter.cpp)
- 0 valued entries in binary phrase tables are loaded as 1.0e-38
(PhraseDictionaryTree.cpp)
git-svn-id: https://mosesdecoder.svn.sourceforge.net/svnroot/mosesdecoder/trunk@1189 1f5c12ca-751b-0410-a591-d2e778427230
error catching/fail more gracefully on tables/lm load error & consistent user output
consistent debugging output
cleaned up timing functions
cleaned up mose/moses-cmd api calls/interaction
split up loading of all data in StaticData into separate fns
got binary phrase table to work under WIN32 & passed regression !!
added more comments
deleted phrase table filtering code
deleted mysql support
change calls to ToString() which might affect decoding to a call to a non-debugging fn instead, eg GetString()
git-svn-id: https://mosesdecoder.svn.sourceforge.net/svnroot/mosesdecoder/trunk@988 1f5c12ca-751b-0410-a591-d2e778427230
Scores from Confusion Network input are output BEFORE any lexicon scores.
In nbest file, they are NOW labelled with "I:". Before they were "hidden" among other translation scores.
This change affect MERT, which will be changed ASAP.
git-svn-id: https://mosesdecoder.svn.sourceforge.net/svnroot/mosesdecoder/trunk@979 1f5c12ca-751b-0410-a591-d2e778427230
* more robust read functions
* correct treatment of epsilons
* code cleanup
- parameter: fixed check for binary phrase table
- staticData: do not read input phrases in case of binary phrase table
git-svn-id: https://mosesdecoder.svn.sourceforge.net/svnroot/mosesdecoder/trunk@260 1f5c12ca-751b-0410-a591-d2e778427230
-> new base class InputType is used throughout the decoder
instead of Sentence
Sentence and ConfusionNet derive from this class
-> Manager etc. do not know if the input is a sentence or a
confusion net (but could check if REALLY needed)
- two separate classes derived from TranslationOptionCollection:
one for text input and one for confusion net input
- score computation in PhraseDictionaryTree.cpp and some optimizations
git-svn-id: https://mosesdecoder.svn.sourceforge.net/svnroot/mosesdecoder/trunk@183 1f5c12ca-751b-0410-a591-d2e778427230
