renamed record filter module sources
Omit CVS Id. Update copyright year.
Changed type of index_type char/int to string. The index_type is part of the RecWord structure which is used by all record filters. It used to be a int char (e.g. 'w'). It is now a string so that arbitrary index types can be specified . For example there might be a Greek Title (e.g. word:greek or word_greek).
Fixes for snippet/scan where we do "reextract". The setStoreData handle was assumed to be present. The alvis and dom filter now check for it. For the reextract case it's NULL and we do not do splitting.
Removed snippet code from alvis filter
Another and hopefully, last, YAZ OID DB update
Update for YAZ 3s new OID system
WRBUF updates.
Added always the XML parsing flag XML_PARSE_NONET to any XML_PARSE_XINCLUDE to avoid spoofing Zebra to fetch megabyte from an external xincluded url. pretty normal safety thing to do, we just did forget before.
Fixed minor memory leak
Fixed bug #884: Entity declarations in input are lost at retrieval time.
2007.
Using atoizint for staticrank
Use atoll/atoi for ascii to zint conversion in alvis filter.
Fixed bug #767: Zebra indexing stylesheet fails to work. The problem was that the root node was not of type element and split_level = 0 (deal with whole record / no split). We make split level 0 a special case (again) and use extract_full for this.
Reformat a bit.
For alvis filter make XML pass-through if stylesheet is attribute is omitted (not specified).
Fixed bug #730: Bad XML for update with alvis filter does not fail as it should. The Libxml2 function xmlReadIO does not report errors so we use xml reader interface always for ALVIS filter.
Fixed bug #723: Enable EXSLT for alvis. We use whatever YAZ finds out for Unix confiugre (YAZ_HAVE_EXSLT & YAZ_HAVE_XSLT).
Common stream reader interface for record filters (struct ZebraRecStream).