XML parser

smilesmita

hi there,
I have a very complex xml doc which if convert into array gives you a lot of nested array.
i wanted some good parser which would parse this complex xml data.
any idea about a good parser or code..which would work on php 4.4.4
I want to extract values from one offset tag and get those extracted values to the DB.
please help.

NogDog

Have you looked at the [man]dom_xml[/man] functions?

Bobulous

Hello, smilesmita.

If you just want to grab the content of one element, take a look at my page about that subject:

http://www.bobulous.org.uk/coding/php-xml-regex.html

and an example of extracting data from an RSS XML feed:

http://www.bobulous.org.uk/coding/php-xml-feeds.html

Make sure to read about the limitations of using these methods, though, as some XML structures don't make it possible to isolate what you want.

Weedpacket

Bobulous wrote:
some XML structures don't make it possible to isolate what you want.

Unless you use the tools developed for the purpose (Google for "Desperate Perl hacker").

Bobulous

When I said that some XML structures can't be isolated, I meant that the functions I offer aren't able to isolate elements within certain structures. For instance, nested elements that have the same name.

Customised regular expressions can very likely isolate any element regardless of the enclosing structure. But sometimes a simple, generic function will do the job just fine.

Weedpacket

Bobulous wrote:
Customised regular expressions can very likely isolate any element regardless of the enclosing structure.

But like I said, why not use the tools developed for the purpose?

Bobulous wrote:
But sometimes a simple, generic function will do the job just fine.

Exactly. So why not use one instead of mucking around with regular expressions?

--
Find all tables that contain cells with a colspan greater than one contained in rows of class "wide". Note that tables may be nested. And the entire matching table is wanted.

//table[tr[@class='wide']/td[@colspan>1]]

--
Oh, you just want an img element? Well, I don't know if the element appears more than once or not, and I don't know if you want all of them or just the one (or which). So I'll play it safe and and retrieve all of them.

//img

Ah, you just want the first.
code[1][/code]
...Sorry, the last.
code[last()][/code]
Er, the last to not have an alt attribute

(//img[not(@alt)])[last()]

Bobulous

What library are you demonstrating there, Weedpacket? (I did a search for Desperate Perl Hacker, but I didn't find an answer.)

Shrike

Bobulous wrote:
What library are you demonstrating there, Weedpacket? (I did a search for Desperate Perl Hacker, but I didn't find an answer.)

XSL maybe 😉

Bobulous

Right, but what built-in libraries permit the expression of XPATH in PHP 4?

I'm on a shared hosting package, so I can't rebuild PHP with any extra toys. If there had been a powerful, built-in option, I'd have used that.

NogDog

You might take a look at the XML_Serializer PEAR package. I've not used it, but it's supposed to be able to parse XML pretty much the same as the PHP5 SimpleXML class.

Bobulous

You need to be able to build packages on the command line for that, right?

NogDog

Bobulous wrote:
You need to be able to build packages on the command line for that, right?

Not necessarily. See the "Installing a local copy of PEAR through ftp/ftps/sftp" section on this page.

Bobulous

That looks like it requires PHP 5 and command line access, so I'm out of luck on both counts.