Would it be 'easy' to turn youtube-dl (written in python) into a dillo 'plugin'?? Bopb -- robert w hall
robert w hall wrote:
Would it be 'easy' to turn youtube-dl (written in python) into a dillo 'plugin'??
That's a fun idea! I'm a big fan of youtube-dl. But I'm not sure what you're suggesting. My dream would be to click on an embedded youtube video and have the download dpi (or something else like it) pull down the video. The problem is that, last time I looked at any rate, dpis are invoked by specific URL schemes. What I'm thinking of would require dillo to invoke dpis when it sees particular bits of HTML. That might be non-trivial to implement. As a side question, how much work would it be to get dillo to render <embed> and <object> elements in a useful way? Now there's a thought! How about if dillo rendered them as clickable boxes linked to some URL like embed://?html=<escaped_html>? Then they could be handled by a dpi that might eg. extract Youtube video URLs and run youtube-dl! Does this sound feasible? Regards, Jeremy Henty
Jeremy wrote:
The problem is that, last time I looked at any rate, dpis are invoked by specific URL schemes. What I'm thinking of would require dillo to invoke dpis when it sees particular bits of HTML. That might be non-trivial to implement.
Yeah, everything's awfully hardcoded in capi.c. I imagine it would require a fair amount of untangling and thinking. Maybe it would be nice to have the dpis able to register interest in resources meeting certain criteria or something. Or maybe the trick is to simply :) implement the NPAPI. BTW, for <object>s, I remember from the html4 spec that one is _supposed_ to use the data attribute, but nearly everyone ignores that and just makes up a... <param movie=[url]> tag instead, IIRC.
corvid wrote:
Maybe it would be nice to have the dpis able to register interest in resources meeting certain criteria or something. Or maybe the trick is to simply :) implement the NPAPI.
??? What is this "NPAPI" of which you speak, Earth being? :-) I guess I've not been keeping up. Jeremy Henty
Jeremy wrote:
corvid wrote:
Maybe it would be nice to have the dpis able to register interest in resources meeting certain criteria or something. Or maybe the trick is to simply :) implement the NPAPI.
??? What is this "NPAPI" of which you speak, Earth being? :-) I guess I've not been keeping up.
Sorry, I wrote more and then I guess I trimmed it down too much. Netscape Plugin API The mighty wikipedia claims: The following web browsers support NPAPI plugins: o Epiphany o Google Chrome o Konqueror o Mozilla project applications, including o Camino o Firefox o Mozilla Application Suite o Seamonkey o Netscape Navigator and Communicator o Opera
In article <20100704010103.GB2019@local.gobigwest.com>, corvid <corvid@lavabit.com> writes
Jeremy wrote:
corvid wrote:
Maybe it would be nice to have the dpis able to register interest in resources meeting certain criteria or something. Or maybe the trick is to simply :) implement the NPAPI.
??? What is this "NPAPI" of which you speak, Earth being? :-) I guess I've not been keeping up.
Sorry, I wrote more and then I guess I trimmed it down too much. Netscape Plugin API
Sounds interesting - but prone to be a long labour in the night... All I really want is a crisp way to get the alphanumeric file- descriptor safely away from the youtube window and into a youtube-dl call - cut and paste would be a significant improvement on copying it, usually incorrectly, character by character Bob -- robert w hall
robert wrote:
All I really want is a crisp way to get the alphanumeric file- descriptor safely away from the youtube window and into a youtube-dl call - cut and paste would be a significant improvement on copying it, usually incorrectly, character by character
I just tried it, copying and pasting the url from some youtube page, and youtube-dl was able to get the file from it with no problem. Why do you have to enter them character by character?
In article <20100706164411.GB1894@local.gobigwest.com>, corvid <corvid@lavabit.com> writes
robert wrote: I just tried it, copying and pasting the url from some youtube page, and youtube-dl was able to get the file from it with no problem.
Ah yes! - my bad (I can never usually get the paste keys to work with my 2-button mouse) Bob -- robert w hall
participants (3)
-
bobh＠n-cantrell.demon.co.uk -
corvid＠lavabit.com -
onepoint＠starurchin.org