How can I get Google Page Rank with Php?

[tunayilmaz]

I am using Php, I want to show some webpages' PR to my clients.
Can I do this?

[Brett_Tabke]

Can't be done. Closest would be to cherry pick pages from the google odp directory.

[andreasfriedrich]

Configure your browser to use some kind of proxy for http. When using squid you need to change strip_query_terms to off in your squid.conf. This causes the query string to be logged.

Request the page you want to check once with your browser and the google toolbar enabled.

In your proxy log file look for a URL that looks like http://www.google.com/search?client=navclient-auto&ch=51313900179&q=info:http%3A%2F%2F [google.com]

Request that URL using php and parse the returned xml document looking for the element gsp/res/r/rk. Its contents is the page rank as displayed by the google toolbar.

Beware that you cannot change the q parameter in the query string to just any URL as the ch parameter seems to be connected to it. I havenīt tried to figure out what it is exactly.

Hope this helps.

[edited by: andreasfriedrich at 1:13 pm (utc) on July 22, 2002]

[edited by: jatar_k at 5:46 pm (utc) on Sep. 21, 2002]
[edit reason] fixed side scroll [/edit]

[Brett_Tabke]

Well, it _was_ a secret anyway ;)

[EliteWeb]

Packet Sniffing and Logs ;) Ah what fun!

[andreasfriedrich]

The above was my first post and I didnīt know the [ url ] code, so the above URL got messed up with the point at the end.

http://www.google.com/search?client=navclient-auto&ch=51313900179&q=info:http%3A%2F%2F [google.com]

[edited by: jatar_k at 5:44 pm (utc) on Sep. 21, 2002]
[edit reason] fixed side scroll [/edit]

[martin]

Wow, great info. Thanks very much for that.

Very useful really as I only use Linux.

Where were you when I posted several requests for PageRank check through the Google API?

[martin]

I was in a hurry to test that. Got a 403.

Is the ch site specific? Anybody with IE on Windows?

[andreasfriedrich]

In my ealier post I wrote

Beware that you cannot change the q parameter in the query string to just any URL as the ch parameter seems to be connected to it. I havenīt tried to figure out what it is exactly.

So short of having hacked that algo you need to request the url once from a windows box with the toolbar enabled. I believe the ch not only to be site specific but specific to each URL, some sort of hash value.

Andreas

[martin]

I hope it is site specific, I'll have some friend check that.

Thanks again for the valuable info.

[martin]

What about this one?

[216.239.33.102...]

Results in something like:

Rank_1:1:4

Btw does ch change between updates.

(I think that Google can update the toolbar algo without requesting user confirmation.)

/edit
Removed non-necessary parts of the URI.

[martin]

What user-agent, HTTP headers, etc. should I send to Google?

Should I act like it's the toolbar or try to identify myself (according to Google's robots.txt it is disallowed).

Whatever of these two I decide to do I think it will be illegal :-(

[martin]

I found that all pages with a query string generate the same request for PageRank from the toolbar, ie.

www.example.com/?key=value1
www.example.com/?key=value2

have the same PageRank returned by the toolbar. It just requests the PageRank for www.example.com/?