xlsx with new Lines (CR+LF) : Parsing error #392
Description
Hi, I'm trying to open an xslx file.
I get this error : Current state not START_ELEMENT, END_ELEMENT or ENTITY_REFERENCE
This error lies in OPCPackage.extractFormat() :
In the while loop, the instruction if ("numFmt".equals(reader.getLocalName())) {
creates en error because the current token type of the reader is CHARACTERS, and then getLocalName() create an exception.
It is because my xlsx file has new lines (CRL+LF) in the styles.xml and so, when we are in the loop that comes after getting the <cellXFS>, insideCellXfs is true, but we can't call getLocalName() since there are CHARACTERS (the new line).
I hope I'm clear in spite of my bad english.
To get this error, take a valid xlsx file, and put CR+LF after each tag in the styles.xml
Excel doesn't put CRLF in the xml it creates, but my xlsx file comes from another tools which put theses CRLF. I think that xlsx format doesn't forbid this, so fastexcel should consider this possibility.
Another side effect of adding CRLF after each tag, is when fastexcel reads sharedStrings.xml :
instead of keeping the text in <t> tags, it retrieves all the text between <si> tags, so including all the CRLF. As a consequence, the strings returned are not the good one (but it doesn't crash the reading, compare to the styles.xml problem explained above).