History of anonymizer

...
04.03.19 : version 0.6 : revised alpha, ready for test server
06.03.19 : version 0.8 : ANNOTONLY=true causes the annotation file to be output into a ZIP
           (to be compatible with ANNOTONLY=false); bug fix for WAVE file input
06.03.19 : version 0.9 : column separator in output BPF is now conform to input BPF
09.03.19 : version 0.10 : TRN tier is anonymized token by token; that way the original transcript
           (as e.g. provided by CHUNKPREP in TRN) is retained except for the tokens that match aTerms
12.03.19 : version 0.11 : fixed bug that occurred when no aTerm was found to be anonymized
20.03.19 : version 0.12 : aTerms that contain at least 2 capital chars are expanded to vrsions with 
           dashes (as g2p.pl would do in word normalisation mode, e.g. 'GmbH' -> 'G-m-b-H'), and 
           aTerms that contain at least 2 capital chars and at least one dash are expanded to a 
           version without dashes ('B-B-C' -> 'BBC'); that way chances are higher that all aTerms 
           the user requests via ATERMS are anonymized, even if they differ in BPF tiers (e.g. 
           'B-B-C' in ORT , 'BBC' in TRN/TRO). 
           version 0.13 : self-reference name in emuR output was wrong - fixed,
           call of par2TextGrid was out-dated - fixed
08.04.19 : version 0.14 : bug fix AWORD="*DELETED*" caused an undocumented ERROR
15.04.19 : version 0.16 : improved robustness of ATERMS list: an ERROR is issued, if the list
           contains meta-characters that are not searchable (e.g. '(').
19.04.19 : version 0.17 : DOS-style mismatch between BPF and ATERMS leads to no aTerm matches ->
           fixed by filtering '\r' from both inputs before matching; improved fade in/out between
           speech parts and masking signals (max. fade time is 0.02sec)
24.04.19 : version 0.18 : filter input BPF and ATERMS list by initial BOM and CR+LF.
26.09.19 : version 1.0 : added ASIGNAL=silence (= total silence with fade in/out)
02.10.19 : version 2.0 : integrated annotConv for output conversion
