Getting html content of google search

[dedndave]

 :redface:

JZ and JE are the same opcode

[Siekmanski]

Oooh yeahhh   ::)

I really need a break it's 5 AM and i need to go to bed 

[dedndave]

me too - see you tomorrow, Marinus   :t

[Siekmanski]

Thanks guys for testing and helping out.  :t
I'm done now with the search routines.
Now i can use it for searching album covers in my program.

It's also nice to search for fixed sizes, in this case images 512 by 512 pixels.

final version,

[Magnum]

Excellent job.  :t

What kind of download time is average for that file size ?

Andy

[Siekmanski]

Thanks 

It depends on the speed of your Internet connection i think.
But you can search for different image sizes.

[Magnum]

Does the code search for rammstein.jpg or just download that file if it finds it ?

Andy

[Siekmanski]

if you want to search for album cover art,

artist = Rammstein
album = Sehnsucht

the search frase is then this one, it searches for album art images with exact sizes (512 by 512)

/search?q=rammstein sehnsucht album cover&tbm=isch&tbs=isz:ex,iszw:512,iszh:512

But you can search for any image you want,

search for panda?

/search?q=panda&tbm=isch

in the source code the save name  "rammstein.jpg" is fixed and saves the first found image ( for testing only )

[Antariy]

Hi Marinus :t You get the large HTTP answer from google when you used the UserAgent string that is used currently?

[Siekmanski]

Yes.

But nothing mentioned by microsoft or MSDN. I was playing with it because ""Mozilla/5.0" returned 70 Kb instead of 36 Kb.
So i searched the net for useragent examples as i used it before in the winsock example.

[traphunter]

maybe your own real user-agent works: http://www.viewmyuseragent.com/

[Antariy]

Yes.

But nothing mentioned by microsoft or MSDN. I was playing with it because ""Mozilla/5.0" returned 70 Kb instead of 36 Kb.
So i searched the net for useragent examples as i used it before in the winsock example.

I think Google servers try to filter automated requests by checking UserAgent, and if it looks like not very similar to the real browser's string, it returns not full answer. Also I noticed that if the're is too many / too frequent requests from one IP, then google blocks the request and provides a captcha to verify that the request was done by people, so the program should not ask for searches too frequently - that is not looks like the people do the search. But you may use the proxies as well - specify one external proxy in there:

invoke InternetOpen,CTXT("ASM example"),INTERNET_OPEN_TYPE_PROXY,CTXT("proxyaddress:proxyport"),CTXT("<local>"),0

so your request will be routed through external proxy with its address "seeing" to google :t If one IP address gets blocked after frequent searches, you may change the proxy and continue seaches

[Siekmanski]

Thanks Antariy, i'll keep this "proxy address" trick in mind.  :t

[Evan_]

Weird stuff. Make a fake browser or something. Fork another; idk.
Run it so you don't even have to look at the web pages anymore.

Or just script it with your creepy requests.

[dedndave]

yes.....
Marinus is very "creepy" - lol
(those creepy Nederlanders)
that's just how he rolls   8)

we've all learned something from Marinus, though   :t