Hi Peter, I opened this issue:
https://issues.apache.org/jira/browse/UIMA-6131 Cheers, Mario > On 26 Sep 2019, at 16:15 , Peter Klügl <[email protected]> wrote: > > Hi, > > > there is no reason. > > > I think there are two, I remember. One in the test utils, which can be > replaced, and one in the generated code of the jflex lexer, which I do > not know if it can be replaced. Maybe a newer version of jflex avoids > this? Or, it could be catched and wrapped in an exception... > > > Do you want to open an issue for this? > > > Best, > > > Peter > > Am 25.09.2019 um 20:32 schrieb Mario Juric: >> Hi Peter, >> >> Just one more thing that came to my mind. Is there a particular reason for >> throwing a java.lang.Error instead of an exception? >> >> Normally that is something only thrown by the JVM when it’s really >> impossible to continue the process, e.g. out of memory, linkage errors or >> fatal VM failures. It is normally not meant to be caught so our UIMA runtime >> environment exits because of this, although it’s not a big issue when we run >> the process as a service since it is then restarted automatically. I just >> thought it’s maybe a bit drastic behaviour when only the document in >> question needs to fail. >> >> Cheers, >> Mario >> >> >> >> >> >> >> >> >> >> >> >> >> >>> On 23 Sep 2019, at 09:48 , Mario Juric <[email protected]> wrote: >>> >>> Thanks Peter, >>> >>> I will await your confirmation of the fix, but I guess we will then stick >>> with 2.6.1 until the next Ruta release :) >>> >>> Cheers, >>> Mario >>> >>> >>> >>> >>> >>> >>> >>> >>> >>>> On 20 Sep 2019, at 18:09 , Peter Klügl <[email protected] >>>> <mailto:[email protected]> <mailto:[email protected] >>>> <mailto:[email protected]>>> wrote: >>>> >>>> Hi Mario, >>>> >>>> >>>> I did not have the chance to have a look at your example yet... >>>> >>>> >>>> Most likely, this problem is already fixed in the current trunk, but I >>>> was not able to find the time for a new release. In 2.7.0, there was a >>>> small modification in the lexer rules for the seeding, which had >>>> unfortunately some unintended side effects in the generated code >>>> especially with unusual unicode characters. I'll try to verify that with >>>> your example the next days. >>>> >>>> >>>> Best, >>>> >>>> >>>> Peter >>>> >>>> Am 19.09.2019 um 12:35 schrieb Mario Juric: >>>>> Hi Peter, >>>>> >>>>> After upgrading to Ruta 2.7.0 a while ago we started getting some >>>>> errors from the SeedLexer, which we didn’t have before. It appears >>>>> related to odd unicode characters that we haven’t cleaned properly >>>>> upstream, but it is consumed by the previous version 2.6.1 where our >>>>> pipeline completes without error. I attached a small sample program >>>>> with a dummy ruta script to reproduce it. >>>>> >>>>> Which version has the correct behaviour in such cases? 2.7.0 or 2.6.1? >>>>> >>>>> >>>>> Cheers, >>>>> Mario >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>>> >>>> -- >>>> Dr. Peter Klügl >>>> R&D Text Mining/Machine Learning >>>> >>>> Averbis GmbH >>>> Salzstr. 15 >>>> 79098 Freiburg >>>> Germany >>>> >>>> Fon: +49 761 708 394 0 >>>> Fax: +49 761 708 394 10 >>>> Email: [email protected] <mailto:[email protected]> >>>> <mailto:[email protected] <mailto:[email protected]>> >>>> Web: https://averbis.com <https://averbis.com/> >>>> >>>> Headquarters: Freiburg im Breisgau >>>> Register Court: Amtsgericht Freiburg im Breisgau, HRB 701080 >>>> Managing Directors: Dr. med. Philipp Daumke, Dr. Kornél Markó >>>> >> > -- > Dr. Peter Klügl > R&D Text Mining/Machine Learning > > Averbis GmbH > Salzstr. 15 > 79098 Freiburg > Germany > > Fon: +49 761 708 394 0 > Fax: +49 761 708 394 10 > Email: [email protected] <mailto:[email protected]> > Web: https://averbis.com <https://averbis.com/> > > Headquarters: Freiburg im Breisgau > Register Court: Amtsgericht Freiburg im Breisgau, HRB 701080 > Managing Directors: Dr. med. Philipp Daumke, Dr. Kornél Markó
