R.Wieser wrote:
Paul,

as long as it has "search providers" for such formats.

Due to that description overlapping with "internet search providers" it
seems to be hard to find anything about it. Do you perhaps have any other
name for me (from the registry perhaps) ? Not that I currently would want
to create one myself mind you, but just so I can have a look at them (read:
curiosity, plain and simple).


They chose to call it an iFilter here. Acrobat Reader is known
to provide a file doing things like that.

http://www.documentsnap.com/how-to-f...dows-7-64-bit/

In some cases, the COM interface is used instead.
Say an email tool stores messages in a custom database
format. The search indexer "doesn't like binary" and
won't munch on that directly. However, an email tool
can have a COM interface, "pulling one .mbx at a time",
and this affords a way to pull text messages one
at a time, which the indexer can use. In the outlook
articles, there is a registry setting to enable
indexing email, which could be using such a method.

The technique in the modern OSes may be called "federated search",
but that does not appear to be a very good search term either.
The Google results are easily diluted by the need to "sell socks".

A search provider, is a mechanism for converting a proprietary
format, into a format the indexer can use. I guess "filter" is
as good a term as any.

Around Acrobat Reader three or four or so, the Reader actually
had an inverted search index capability and could index all the
PDFs and put them into the one index file for search. Like the
windows.edb on Windows Search, this can be slow when the database
gets large, and the updater tries to "merge" an index into
the main index. It's very time consuming, and doesn't scale
particularly well. My Acrobat attempts kept breaking so I
gave up on using the indexing feature. And this means, since
Adobe built this for their own usage, eons ago, designing
an iFilter is old hat for them.

Paul