Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Nice Usage Example: What's that song?

438 views
Skip to first unread message

Robert Synnott

unread,
Mar 29, 2004, 1:38:33 PM3/29/04
to
Hi,
Here's an example of Google Web API in action; a page which, when
given some song lyrics, returns the name and artist of the song.
http://www.synnottsoftware.com/songsearch Try it out! Not foolproof
yet; please report any problems you have.
Thanks,
Rob.

Frederik De Bleser

unread,
Apr 9, 2004, 5:42:45 AM4/9/04
to
> given some song lyrics, returns the name and artist of the song.

Interesting example. Care to comment on the algorithm? What sites do
you search lyrics for?

Regards,

Frederik

Robert Synnott

unread,
Apr 13, 2004, 11:56:17 AM4/13/04
to
fred...@pandora.be (Frederik De Bleser) wrote in message news:<d5f54400.04040...@posting.google.com>...

Is very simple; it prepends "lyrics " to the search string supplied,
and knows how to parse the titles of about 20 big lyrics sites (they
tend to be in a format along the lines of "<lyricssite> - <title> -
<artist>" or somesuch. It only shows results from sites it knows
about, so it's a bit easier than wading through the results of the raw
google search. Was very quick to make.
Rob

xun

unread,
Apr 14, 2004, 5:02:42 PM4/14/04
to
what do you mean by "the titles of about 20 big lyrics sites"?

What constitutes title of a resultelement

By the way, why Google would turn out some pages that are not on web server at all?

thanks,

xun


synn...@tcd.ie (Robert Synnott) wrote in message news:<a0e22b2c.04041...@posting.google.com>...

Robert Synnott

unread,
Apr 15, 2004, 6:39:16 AM4/15/04
to
xd...@mcic.org (xun) wrote in message news:<f08e80b0.0404...@posting.google.com>...

> what do you mean by "the titles of about 20 big lyrics sites"?
>
> What constitutes title of a resultelement
>
> By the way, why Google would turn out some pages that are not on web server at all?
>
> thanks,
>
> xun
>
>
The API returns the page title, the url and a bit of content for each
result. These titles are generally in a standard format, so the
program looks at the URL, sees is it a site it knows; if so, it parses
the title, otherwise it ignores it.
Rob.
0 new messages