HISTORY runEmuMagic

06.11.17 : runEMUMagic 1.0 : first version
11.11.17 : runEMUMagic 1.3 :
           - added pipe type CHUNKPREP_G2P_MAUS_PHO2SYL for TextGrid/eaf
           input assuming that these file types contain a chunk segmentation
           (regardless the length of a chunk, i.e. it could happen
           that a label in a TextGrid/eaf file has more than 3000 words,
           although that is not very likely)
           - check TEXT input or the output of the ASR for word number: if
           larger than 3000 call pipe with CHUNKER
03.02.18 : runEMUMagic 1.4 : added default option PRESEG=true to maus.pipe call
07.09.19 : runEMUMagic 1.6 : added possible TEXT input CSV (chunk segmentation)
05.02.20 : runEMUMagic 2.0 : in earlier version 1.X this script created different emuDB structures dependent
           whether the spoken text contained more (with TRN tier) or less than 3000 words (without TRN tier).
           This was a problem because then we cannot run this script in batch over a large number
           of files and get them together in one emuDB.
           Starting with 2.0 the emuDB structure should always be the same
           bundle -> ORT+KAN+KAS -> MAU
                                 -> MAS -> MAU
                                 -> TRN
31.01.21 : runEMUMagic 2.1 : added dummy options for FormantAnalysis so that script accepts these from the backend server
16.02.21 : runEMUMagic 2.2 : added option --formantanalysisversion to pass on the version in the WebServices
08.03.21 : runEMUMagic 3.0 : extended input TEXT=*_annot.json : copy TEXT to OUT
           runEMUMagic 3.1 : added input TEXT=*.pdf
09.03.21 : runEMUMagic 3.2 : added input TEXT=*.doc,docx,odt,rtf
14.03.21 : runEMUMagic 3.3 : added G2P option imap to enable LANGUAGE=sampa 
