I want to search scanned text (comprised of a few gigabytes of text).
The problem with this text (which is why I cannot make an index of words) is that it was OCRed, so there could be (and are) spaces in the middle of words (so, for example,
computer could be
c o m pute r).
On the other hand, to shrink the text (remove all spaces so the text of the page is
somethinglikethissentence and searching through
like "%this%" is slow.
Does anyone have any advice how to optimize the database?