On Fri, 2010-06-04 at 14:54 +0300, Tanel Tammik wrote: > "Ashley Sheridan" <ash@xxxxxxxxxxxxxxxxxxxx> wrote in message > news:1275652342.2217.51.camel@xxxxxxxxxxxx > > On Fri, 2010-06-04 at 14:44 +0300, Tanel Tammik wrote: > > > >> "Ashley Sheridan" <ash@xxxxxxxxxxxxxxxxxxxx> wrote in message > >> news:1275651371.2217.46.camel@xxxxxxxxxxxx > >> > On Fri, 2010-06-04 at 14:12 +0300, Tanel Tammik wrote: > >> > > >> >> Hello, > >> >> > >> >> if there is some webpage content with html tags in database is it > >> >> possible > >> >> to search it without tags? > >> >> > >> >> data : '<div style="">you need some styling!</div>' > >> >> > >> >> when i now search for 'you style' i don't want to get any rows! is it > >> >> possible? > >> >> when i search 'you styling' i get the row! > >> >> > >> >> Br > >> >> Tanel > >> >> > >> >> > >> >> > >> > > >> > > >> > Use a second field in the DB that stores the content without any HTML > >> > tags. That way, you can search and not worry about tags and attribute > >> > values getting in the way. > >> > > >> > Thanks, > >> > Ash > >> > http://www.ashleysheridan.co.uk > >> > > >> > > >> > > >> > >> Is this the only way? Couldn't i do it in mysql query? Seems much > >> cleaner... > >> > >> Br, > >> Tanel > >> > >> > >> > > > > > > You could try and do it in MySQL with a regex to filter out the HTML > > tags. The regex would be real complex though, and prone to failure if > > the HTML wasn't perfectly formed. And it would be a *lot* slower than > > searching a plain text field. I think it's far cleaner to use a second > > field like that. > > > > Thanks, > > Ash > > http://www.ashleysheridan.co.uk > > > > > > > OK! then i should use preg_replace before making the serch entry for DB > storage? What would be the regular expression for that? Basically i need to > get rid everything between the html tags with tags included? > > Br > Tanel > > > No, you'd have to use a regex within MySQL, not PHP. Like I said, it would be very complex, and I wouldn't know where to begin writing a query that would search for specific strings and ignore any content within the < & > without writing sub-queries. Also, you did see that I said it would be a lot slower didn't you? Imagine at the moment a query is taking a second to complete. With this sort of complex regex it could take maybe 5 seconds. That's 5 seconds per person searching. Are you not able to make a second field in the DB? Thanks, Ash http://www.ashleysheridan.co.uk
