I'm trying to index XML files containing a corpus annotated with a positional tagset. I'm using the index tools and "index wizzard" to do this. Everything works correctly until the indexing finishes (it ends with error 7). Quick examination of etc/corpuslog.xml shows huge number of errors like this:
<loc sysid="file1.xml" line="5" col="1" />
<msg>Task stack too deep, switched off.</msg>
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I want to use all the 19 annotation attributes (+ the base/lemma). When I cut down them and use only 3 of them, the indexing process completes correctly. Is the number of attributes somehow connected with the error message? How can I resolve the problem with task stack, whatever it is?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm having the same problem-if I don't process any of the morphological tagging, the indexer works fine, but if I include any of it (even just one attribute) the indexer throws up thousands of "Task stack too deep, switched off." errors. I can then open the corpus with Xaira, but it crashes if I try to use any of the morphological tagging.
I think the problem is the number of attributes you're trying to index. You need to tell the indexer not to create indexes for all of them. By default, an index is created for every attribute found, which is clearly pointless in cases like (for example) xml:id
and just wastes stack space during indexing.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Indexing xml:id is indeed pointless, but all other attributes probably are needed. If the indexer cannot deal with 4 attributes removing xml:id won't help.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I'm trying to index XML files containing a corpus annotated with a positional tagset. I'm using the index tools and "index wizzard" to do this. Everything works correctly until the indexing finishes (it ends with error 7). Quick examination of etc/corpuslog.xml shows huge number of errors like this:
<loc sysid="file1.xml" line="5" col="1" />
<msg>Task stack too deep, switched off.</msg>
_
The above error appears several times for each line containing a token delimiting XML tag, for any file.
The line of XML file looks like this:
_
I want to use all the 19 annotation attributes (+ the base/lemma). When I cut down them and use only 3 of them, the indexing process completes correctly. Is the number of attributes somehow connected with the error message? How can I resolve the problem with task stack, whatever it is?
I'm having the same problem-if I don't process any of the morphological tagging, the indexer works fine, but if I include any of it (even just one attribute) the indexer throws up thousands of "Task stack too deep, switched off." errors. I can then open the corpus with Xaira, but it crashes if I try to use any of the morphological tagging.
The xml looks like this:
Anyone have any ideas about this problem?
I think the problem is the number of attributes you're trying to index. You need to tell the indexer not to create indexes for all of them. By default, an index is created for every attribute found, which is clearly pointless in cases like (for example) xml:id
and just wastes stack space during indexing.
Indexing xml:id is indeed pointless, but all other attributes probably are needed. If the indexer cannot deal with 4 attributes removing xml:id won't help.
I've just tried removing xml:id from the XML files-and now the indexer just crashes. Perhaps unrelated? Very mysterious.
But when you say tell the indexer not to create indexes for all of them, how would I do that? Just by not creating a lemma scheme for some of them?
Thanks! :)