"invalid UTF-8 character" in nfo:tableOfContents from PDFs
If a PDF has actions of the type "goto dest", we will try to include the destination name in the TOC. However, despite PopplerDest.named_dest being a char*
, it seems to contain not-quite-strings, and definitely not in a readable format.
The result looks something like:
nfo:tableOfContents "Social networks and non-market valuations �� Introduction �� Preferences �� Solving the model �� Networks and individual valuation �� Networks and aggregate valuation �� Project choice and opinion leadership �� Conclusion �� Acknowledgments �� References �� " ;
Which will trip tracker-store
as SPARQL is expected to be UTF-8.
The solution seems to be to ignore those named destinations.