I’ve set up a lexicon and configured the tier types and analyzers in the way that seems correct as far as I can tell, but I can’t get the analyzer to actually parse and tag the whole transcript. The Whitespace Analyzer properly splits the intonation units into words on the Word tier, but the Parse Analyzer and Gloss Analyzer won’t work properly. Sometimes it will walk through one or two lines of the transcript, pausing to ask me to choose between parsing options and glossing options, but it always stops after that and never goes through the entire transcript. In addition to the Parse Analyzer + Gloss Analyzer approach, I also tried setting up the Lexicon Analyzer, but that won’t run at all. It gives me a message saying “No analyzer configuration is selected as active for interlinearization.”
Also, I’ve made a version of the intonation unit tier with hyphens inserted to mark all morpheme boundaries, thinking that would help the parser identify the boundaries correctly. However, it doesn’t seem to recognize them, even though I’ve set the hyphen as the Affix Character Marker when configuring the analyzer.
I’ve tried to reproduce the problem, so far without success. If I understand correctly once the automatic, sequential processing stops, it is not possible to re-start it from there? And doesn’t manual, word-by-word parsing (via the context menu or CTRL+I with the input annotation active) work either in that situation? I’m wondering if there are maybe clear, interlinearization related error messages in the log (View->View Log...) in that case (no need to post the complete log here, in case of doubt you can send the entire log to me by mail: han.sloetjes AT mpi.nl)?
Some general remarks: the sequential processing stops when e.g. the Esc key is pressed (I’m not assuming that is the case here). In principal, if it is possible to configure the Parse and Gloss analyzer separately, it should also be possible to configure the Lexicon analyzer. The affix marker character only tells the analyzer how entries in the lexicon should be interpreted, these markers are not expected to be found in the input for the parser.
I’m happy to help to get the Lexicon analyzer configured, maybe some screenshots of the settings for the separate parse and gloss analyzers and of the lexicon analyzer would help (per email).
