Closed RoxvilleRooster closed 4 years ago
I boosted the ANT_OPTS to -Xmx8192m but that had no affect on the result of the pdf5-ml transformation.
I have no experience to got this error in ahf:getClearAtts (dita2fo_attribute.xsl Line#: 87) until now. Is it possible to send me the test data you are using?
Please try -Xss
parameter. You can increase stack size by this parameter.
Setting the -Xss
made no real difference.
It seems that the looping behavior, where the transformation looked for the dlentry element's title content, would run endlessly. After 30 mins, my process timed out (that was my setting in Jenkins) - the builds did not end before that.
But, after inspecting the two DITA files mentioned in the console log, I manually changed the DITA .
Here are the log entries that directed me:
[xslt] [getXrefTitle 035W] Xref/@href destination dlentry has no title content. dlentry id=obligor file=file:/C:/algodocs/svn/projects/algo_aggregation/awa/FRTBAggregation_input_data_portfolio.dita
[xslt] [getXrefTitle 035W] Xref/@href destination dlentry has no title content. dlentry id=underlyings file=file:/C:/algodocs/svn/projects/algo_aggregation/awa/FRTBAggregation_input_data_context_data.dita
In this case, the DITA was valid but I'd suggest that our usage was suboptimal. I really do not think that it is a great idea to have an xref to a dlentry but it is valid and we were doing this. The xref pointed to an id of a dlentry element like this:
<xref href="FRTBAggregation_input_data_context_data.dita#input_files_context_data/instruments"/>
Once I added link titles in xref elements, the pdf5.ml transformation progressed.
<xref href="FRTBAggregation_input_data_context_data.dita#input_files_context_data/instruments">Instruments.csv</xref>
Makita-san, I'd suggest that this case could be a known issue or, better yet, a defect that you might address in the future.
I'd suggest that this case could be a known issue or, better yet, a defect that you might address in the future.
Recently another user requested me to enable xref links to dlentry
. In this case the build process ended normally. Based on your suggestion, I will investigate this problem as soon as possible.
I have done several test cases for referencing dlentry from xref. But I couldn't reproduce the looping. Could you kindly offer the relevant contents referenced by "FRTBAggregation_input_data_context_data.dita#input_files_context_data/instruments"?
Since the DITA has been changed, I cannot reproduce the looping either. Let's close this and, if I see this looping again, we can re-open this and I will attach files. I can create an xref to a dlentry and that works fine in the pd5.ml transform.
Hello, Makita-san,
I am testing the pdf5-ml transformation. I have encountered this fatal error.
This same map builds successfully when I build using other transform: "DITA Map PDF - based on HTML5 & CSS" in Oxygen Author 21.1. I needed to bump up the Java heap size to -Xmx4096m to get a result but the PDF rendered successfully. In this case the CSS Processor was Oxygen PDF Chemistry. I was not testing the use of the Antenna House CCS Processor in this case. I was really just looking to see if I could generate a PDF of this same boo with another OT PDF transform.
So I expect this indicates a defect in pdf5.ml. I have tried boosting the Java heap space to -Xmx4086m but to no avail yet.
consoleText.txt