Re: [PEPr] +1 for XML::XML_Indexing
| From: | Olivier Guilyardi | Date: | Mon, 18 Oct 2004 12:33:46 +0000 |
| Subject: | Re: [PEPr] +1 for XML::XML_Indexing | ||
| References: | 1 | Groups: | php.pear.dev |
| Request: | Send a blank email to pear-dev+get-33927@lists.php.net to get a copy of this message | ||
Hi Christian,
Christian Stocker wrote:
Looks good to me. More XPath support would be nice, but hey, I can't have everything ;)Thanks for your vote, some more XPath will come :-) Actually, the last thing I've been thinking of, is full-text search, that is : indexing PCDATA. This is related to my ideas about XML as a database storage format, and emulating such things as "WHERE foo LIKE 'bar%'". I thought about MySQL like full-text indexing, where you specify a certain length in caracters for your index as in "INDEX(foo(16)). In this case, it will be used for patterns like 'bar%' but not for '%bar%'. I have another idea though : word-based indexing, very much like what you find at the end of books. For each word that was found, that would associate a set of nodes that contain it. This way you would not be limited to prefix-based search as in 'bar%'. -- og