LibriVox Forum
https://forum.librivox.org/

LibriVox API Discussion Thread
https://forum.librivox.org/viewtopic.php?f=24&t=44129
Page 12 of 13

Author:  msfry [ January 13th, 2016, 2:17 pm ]
Post subject:  Re: LibriVox API Discussion Thread

IA's fund drive, as I recall, was BLATENT! A big BLACK banner across the top of the screen saying they needed $2 million to reach their goals and if everybody would stop and contribute $50 RIGHT NOW they could end this fund drive TODAY and get back to doing the coolest thing in the universe for the human race (or some such verbiage that meant the same thing). So I did what they told me to do. Wikipedia did the same basic thing, and so did I, and I notice their fund drive is also over. I'm just saying, there's a template if we ever need it for a successful fund drive. :P

Author:  TriciaG [ January 13th, 2016, 2:35 pm ]
Post subject:  Re: LibriVox API Discussion Thread

And I very much dislike those blatant, in-your-face fundraising campaigns. They may work, but they also create negative feelings.

And I don't think we have a flash banner option on our website to blare out a fundraising campaign like IA and Wikipedia have. We'd need development for that (if we decided to do it that way) - for which we'd need money - for which we'd need the banner, for which we'd need development money, for which we'd need a banner... :lol:

IMHO, of course. (insert angel smiley here)

Author:  PetrKryze [ March 2nd, 2016, 7:08 am ]
Post subject:  Re: LibriVox API Discussion Thread

ScottLawton wrote:
I don't know if there's a solution. LibriVox doesn't have a budget for updating the API so many of us just maintain our own version of the catalog and search that.


Thanks for the reply! Could you maybe give some tips on how to get the whole catalog so I can do the search myself as you suggested? I don't know if I should suck it out of the API somehow (cause I am slowly starting to hate the API), or if there is some better way. Thanks! :help:

Author:  msfry [ March 2nd, 2016, 8:31 am ]
Post subject:  Re: LibriVox API Discussion Thread

TriciaG wrote:
And I very much dislike those blatant, in-your-face fundraising campaigns. They may work, but they also create negative feelings.
Not necessarily if done well, Tricia. Many people set aside a fixed amount each year for charitable giving, then go about looking for the best places to put it. They want to feel their money is well placed to do the most good (without doing mountains of research), and appreciate a bold and compelling appeal for assistance. I want to know:

  • The type work you're doing
  • Who you are helping, and how (in as few words as possible).
  • GOALS: Ultimate, and already reached.
  • The goal you've set for this year (i.e., tangible, stepped, realizable goals).
  • How much you need to raise to make this happen.

No compelling appeal that makes me feel useful to your cause, and that you have a definite plan, most likely no assistance. The benefits of feeling useful outweighs the blatant ads, in my opinion, especially if the campaign doesn't last long.

The squeaky wheel does get the grease. There are many TED talks on the subject of philanthropy. Here's one I saw recently:
http://www.ted.com/talks/dan_pallotta_the_way_we_think_about_charity_is_dead_wrong

Author:  ScottLawton [ March 2nd, 2016, 10:58 am ]
Post subject:  Re: LibriVox API Discussion Thread

PetrKryze wrote:
Could you maybe give some tips on how to get the whole catalog so I can do the search myself as you suggested? I don't know if I should suck it out of the API somehow (cause I am slowly starting to hate the API), or if there is some better way. Thanks! :help:

First: get what you can from the API ... see the rest of this thread for details.

Then, if that doesn't have all the data you want, do one or more of:
- investigate the alternate catalogs (e.g. from ekzemplaro) to see what additional data they include
- get the json file of each book from archive.org; that's easy and 'code-friendly'
- get and parse the HTML of individual books, authors & readers
- get and parse the HTML of search results ... tedious, but the only way to get every piece of info

It's several weeks of work if you really want everything.

Author:  ekzemplaro [ March 3rd, 2016, 3:16 am ]
Post subject:  Re: LibriVox API Discussion Thread

DobrĂ½ den, Petr san

Welcome to LibriVox. I hope you enjoy it here.
PetrKryze wrote:
Could you maybe give some tips on how to get the whole catalog so I can do the search myself as you suggested?

Information for the alternative catalog is here.

For the webpage
https://github.com/ekzemplaro/librivox_catalog

How to gather data
https://github.com/ekzemplaro/librivox_database

I think the first step is to impliment the catalog on your machine.
If you have any questions, please feel free to ask.

Cheers,
Masa

Author:  PetrKryze [ March 3rd, 2016, 8:12 am ]
Post subject:  Re: LibriVox API Discussion Thread

ScottLawton wrote:
PetrKryze wrote:
...
- get the json file of each book from archive.org; that's easy and 'code-friendly'
...


Hi Scott and thanks for replying!
I just spent an hour reading through the whole thread and the other thread about Masa's alternate catalogue (which seems cool, but I have no idea how to use it for my purposes - I am really confused by github and I have never touched JSON in my life yet - but thanks for replying too, Masa! I really admire your dedication for work with Librivox). I got the idea about current API state, no funding :cry: etc. :)

I can get all the info I need from the Librivox API, and I have developed a workaround to get the sections download links from Archive.org, but the thing that bothers me most is (as I have posted here before) the broken search, i.e. how the API won't find Romeo and Juliet if I query "Juliet" etc.

I like the idea about using the Archive.org API for searches, I found it here: https://archive.org/advancedsearch.php#raw . Do you think it is possible to use it so it only return Librivox books? If so, could you (or anyone) give an specific example? I am confused by it. :)

Thanks!

Author:  msfry [ March 3rd, 2016, 9:50 am ]
Post subject:  Re: LibriVox API Discussion Thread

PetrKryze wrote:
how the API won't find Romeo and Juliet if I query "Juliet" etc.
One thing that I see as underutilized is the keywords submitted on projects. BC's should try and add as many as they can before the project is catalogued, maybe MC's could remind them. This list of keywords can generally get longer and better after the book has been fully read. One thing for sure, very few of us are going to create an "alternate catalog" for ourselves!

Author:  ScottLawton [ March 3rd, 2016, 9:52 am ]
Post subject:  Re: LibriVox API Discussion Thread

This seems to work:

title:(juliet) AND subject:(librivox)

https://archive.org/search.php?query=title%3A%28juliet%29%20AND%20subject%3A%28librivox%29

Author:  PetrKryze [ March 3rd, 2016, 11:01 am ]
Post subject:  Re: LibriVox API Discussion Thread

ScottLawton wrote:


Great, thanks!! :thumbs:

Author:  ekzemplaro [ March 4th, 2016, 5:27 am ]
Post subject:  Re: LibriVox API Discussion Thread

Hello,

msfry wrote:
One thing for sure, very few of us are going to create an "alternate catalog" for ourselves!

Everybody is welcome to use "alternate catalog". I'm also open to any requests. If it is easy, I'll implement it.

If you don't have fund, why don't you do it by yourself?

Cheers,
Masa

Author:  mkaflowski [ August 24th, 2016, 6:01 am ]
Post subject:  Re: LibriVox API Discussion Thread

How to get books in some language? I see that tbook achived that in his app.

Author:  csbubbles [ January 2nd, 2017, 9:14 pm ]
Post subject:  Re: LibriVox API Discussion Thread

Hey guys,

I am trying to approach LibriVox API, and I have found that the audio track URLs returned are not working. It looks like they point to a wrong domain.

For instance, this request https://librivox.org/api/feed/audiotracks/project_id/47?format=json returns the following URLs:
  • http://www.greenkri.com/librivox/montecristo/count_of_monte_cristo_001_dumas.mp3
  • http://www.greenkri.com/librivox/montecristo/count_of_monte_cristo_002_dumas.mp3
etc., each of which returns a 404 Not Found error.

While all the audio files seem to be actually located on a different domain (archive.org, not greenkri.com):
http://archive.org/download/count_monte_cristo_0711_librivox/

Could please someone take a look and tell whether it's a bug on LibriVox side and needs to get fixed there, or I am doing something wrong?

Thanks!

Author:  kayray [ January 2nd, 2017, 11:41 pm ]
Post subject:  Re: LibriVox API Discussion Thread

Sorry I don't know how to fix this, but as an old-timer I thought I'd just mention that Kri was a volunteer admin many years ago and greenkri.com was her personal domain. Perhaps we stored in-progress books on her server at some point (I don't remember doing this, but it's not impossible...)

Completed books have always been hosted at archive.org.

Author:  RuthieG [ January 3rd, 2017, 4:43 am ]
Post subject:  Re: LibriVox API Discussion Thread

I know absolutely zilch about APIs, but as far as I can see, "audiotracks" is not a supported field. (See https://librivox.org/api/info) Having a mooch around, it appears that all the audiotracks field will yield is where the tracks were originally uploaded. These locations are never permanent and are deleted as soon as the project is catalogued. The only files that should be accessed from outside LibriVox are the completed files from archive.org.

Ruth

Page 12 of 13 All times are UTC - 8 hours [ DST ]
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/