Opened 9 years ago
Closed 9 years ago
#742 closed defect (fixed)
Import of large RDF file fails
| Reported by: | dstillman | Owned by: | simon |
|---|---|---|---|
| Priority: | major | Milestone: | 1.0 RC 4 |
| Component: | ingester | Version: | 1.0 |
| Keywords: | Cc: |
Description
[Exception... "Component returned failure code: 0x8000ffff (NS_ERROR_UNEXPECTED) [nsIRDFService.GetDataSourceBlocking]" nsresult: "0x8000ffff (NS_ERROR_UNEXPECTED)" location: "JS frame :: chrome://zotero/content/xpcom/translate.js :: anonymous :: line 1500" data: no] in opening IO for RDF
Problematic file attached.
Change History (2)
comment:1 Changed 9 years ago by simon
comment:2 Changed 9 years ago by simon
- Resolution set to fixed
- Status changed from new to closed
(In [1700]) - actually fix CrossRef translator (expects req_dat, not pid, for OpenURL 1.0)
- display translation failure if no items scraped during asynchronous translation, can be overridden by calling Zotero.done(true)
- always close scraping window following translation
- closes #742, Import of large RDF file fails. I've found a workaround for the export issue that causes this (actually a bug in Mozilla's RDF engine, which produces invalid XML under certain circumstances). However, old invalid RDF/XML documents still won't import.
Note: See
TracTickets for help on using
tickets.
Looks like it's a problem with Mozilla's RDF serializer (which should encode a Unicode entity), combined with an abstract containing a strange character (0x18 = "cancel" control character). We should probably just zap all ASCII chars <= 31 except for CR and LF, since these characters don't print anyway, and seem to screw up the XML.