Note: This was originally posted by an inactive account. Content was preserved by moving under an admin account.
Originally posted by: sdbThe XML Data node seems to operate in a wierd exponential time and throws an EOF kind of error when the input file exceeds circa 512KB.
I am using it to read the structure of spreadsheets (merged cells and the like), something one cannot do directly with the excel node.
Save one tab from an Excel spreadsheet as an XML Spreadsheet 2003 format, and the file is about 450kB and takes around 30 secs to process - fine.
Save two or three tabs in the same way as XML sheets in one file, and the file is 944KB or more and the node really slows down exponentially with file size - as long as an hour.
Why is this node so slow???
But then it throws and EOF kind of error. I say kind because it always refers to the last line and last column in the error message, eg:
Error occurred parsing XML data:
Error occurred while processing tokens to be sent to the node outputs.
At Line 17,891, Column 12
When it hits an error it also sometimes does not flush all read token data, so the outputs are short-changed.
I have turned off reference ID generation, but it makes no difference to speed.
What is this node doing beyond reading a strictly hierarchical file linearly once?
Any tips for speed up would be gratefully received.
Using Input raw and reading from the data field makes no difference.
LAL version 5.1.35
Looks like a memory leak?
Anyone seen this and know workarounds?
I doubt there is anything wrong with the XML.