>I created an index from a database wherein a field contains the path
>and filename of a PDF file (another fields contain text data), thus
>every row contains different filename all in PDF format.
>
>There is nothing wrong with indexing and searching, I only notice that
>when a searching using with certain conditions (i.e. <fieldname
>contains(something)>) and try to click the result PDF filename it does
>generate a correct file and correct search but it misses the correct
>highlighted hits. I expect the highlighted hits in PDF to point on the
>keyword stated in the search criteria but it hihglights different
>words that are not even in the criteria.

Usually hit highlighting problems mean that FileConverter was set up
with incomplete data about the input document, although other
possible causes are (1) changes to the file since it was indexed, (2)
changes to the dtSearch Engine version, or (3) providing inconsistent
input, especially when indexing using the data source API (the input
to FileConverter has to exactly match what you returned in your
DataSource implementation).

Building the index with caching of original files enabled, and using
the dtsConvertGetFromCache flag in FileConverter, can greatly
simplify inconsistency problems because FileConverter can get your
original data from the cache.

For other troubleshooting steps, see:
http://support.dtsearch.com/faq/dts0208.htm

David Thede
dtSearch Corp.