When parsing E-BUGS-107.idf.txt, according to the log, MAGETABParser.parse() method returns investigation prior to parsing 100% of the SDRF file. I am using version 1.0.7.2.
E-BUGS-107.idf.txt...
Apr 28, 2011 5:40:20 PM uk.ac.ebi.arrayexpress2.magetab.parser.MAGETABParser parse
INFO: Parsing MAGE-TAB from file:/Users/hesfalling/data/transmit/experiments/BUGS/E-BUGS-107/E-BUGS-107.idf.txt
Apr 28, 2011 5:40:20 PM uk.ac.ebi.arrayexpress2.magetab.parser.IDFParser parse
INFO: Starting IDF parsing...
Apr 28, 2011 5:40:20 PM uk.ac.ebi.arrayexpress2.magetab.parser.IDFParser parse
INFO: IDF parsing and syntactic validation finished
Apr 28, 2011 5:40:20 PM uk.ac.ebi.arrayexpress2.magetab.parser.SDRFParser parse
INFO: Starting SDRF parsing...
Apr 28, 2011 5:40:20 PM uk.ac.ebi.arrayexpress2.magetab.parser.SDRFParser parse
INFO: SDRF parsing and syntactic validation finished
Apr 28, 2011 5:40:20 PM uk.ac.ebi.arrayexpress2.magetab.parser.MAGETABParser$3 run
INFO: Parsed 98% of E-BUGS-107.
-- my application code log starts here
Hi - the parser did finish, in fact - the "INFO: IDF parsing and syntactic validation finished" and the SDRF equivalent confirms this. The 98% refers to the amount of the file that was successfully included in the parse - so the fact that it is not 100% suggests that either some lines in the IDF or columns in the SDRF contained illegal headers, or that some parts could not be parsed - sometimes you'll see this with a weird protocol layout or else if there are cycles introduced in the columns where the datafiles are described (e.g. due to multiple derived datafiles appearing to derive from one another).
Could you check if either of these things appear to be the case in the file you're trying to load, or else forward me the relevant file and I'll try it for you and let you know?