allowing doc process toolkit to parse other forms of documents #3
Conversation
There was a problem hiding this comment.
Wouldn't a simpler check be to try and extract the text from a document and see if that returns anything?
There was a problem hiding this comment.
I actually run, 2 checks for text. The first is a quick scan to make sure that text even exists in the document. The second occurs after Tika extracts the document text to make sure that the text extraction worked. I figured it would save some Tika processing time.
|
General suggestion: it looks like you've got lots of "if it's this type of document, do X, if it's that, do Y". That's really what standard interfaces are for -- consider having a PDFTextExtractor, an XLSTextExtractor, etc. so that those type checks are made implicit |
|
That's a good idea @cmc333333 and would likely lead to a cleaner design of this. |
|
@cmc333333 @khandelwal I've added the suggested changes to PR #4 |
allowing doc process toolkit to parse other forms of documents
No description provided.