Text compare question

[brackca1]

Can you point me to any documenation on using Beyond Compare for 2 large text datasets. In our case there may be significant missing records or additional records(depending on how you look at it :) ) in one of the sets? What appears to be happening is the comparison works well until it hits some point where key match differences are significant. At that point even the records that match, are reported as mis-matches.

[Aaron]

Hello,

Thanks for the report. About how many rows are in your two files? Is your key just one column, or is it set as a combination of multiple columns? And about how large of a gap (number of rows) are you hitting?

[brackca1]

Aaron,

I'm sorry, after you asked this question I realized I gave wrong information. First it's not a text file compare, we are trying to compare very large PDF files. The documents contain text and in most cases will match record by record. However, under certain circumstances we have additional pages in one of the PDF files. When that occurs Beyond Compare throw's all other documents as mismatches. We're only expecting to see the additional pages as inserts with the other window being empty.

One point, it's 2 Large PDF documents we're comparing, each about 200MB.

Thanks for the help.

[Aaron]

The PDF files are loading in the BC3 Text Compare session, correct? By default, we should convert the PDF's to plain text and compare those plain text values.

If the alignment is a bit off, you may need to tweak the alignment settings in the Text Compare's Session menu -> Session Settings dialog -> Alignment tab. Perhaps enable "Never Align Mismatches" and increasing the slider would help. How large are the gaps (number of lines), and do your "records" match exactly?

Do you have smaller example files that you would be able to email in that also exhibit the issue?