ReadItLater v3 (Several new features!)

[tBunnyMan]

I added a bunch of functionality to the baked in RIL Plugin! Version three since it seems I'm the third person to work on it.

Hilights

• Displays articles from oldest to newest
• Uses max_articles_per_feed to determine how many articles to download per eBook (Default 50)
• Has a minimum_articles setting at the top. Plugin fails to run if you don't have enough articles (Default 10, set to 0 to disable)
• Marks downloaded articles read

For the LATEST version please go to https://github.com/tbunnyman/ReadItLater-Calibre-Plugin

If anyone can think of other cool features I will consider them.

(I was considering redoing the way it formats the book but wanted to release as is for now.)

[kajko]

Registered just to say thanks! New to Calibre and the recipe functionality is awesome. Anyway, I'm a big user of Read It Later and I like your new update since I don't want to mark the articles as read automatically, but..

When I try to load the custom recipe into Calibre from file, I receive an error message
"calibre, version 0.8.44
ERROR: Invalid input: Could not create recipe. Error:invalid syntax (<string>, line 4)"

Any tips? (I know next to nothing about coding)

[dsip]

This is really awesome. It has increased my productivity immensely, I'm using both my phone and my Kindle for more useful stuff now. I actually get to read all the interesting stuff that I find on the web now, instead of just quick news. Thanks a lot!

One feature request, that should be quite easy to implement..: Since I sometimes create more than one RIL "issue" per day, would it be possible, to have a creation time displayed somewhere? Right now on the cover displays the date, but no time. Would there be an easy solution for that?

[tBunnyMan]

Originally Posted by kajko

When I try to load the custom recipe into Calibre from file, I receive an error message
"calibre, version 0.8.44
ERROR: Invalid input: Could not create recipe. Error:invalid syntax (<string>, line 4)"

Any tips? (I know next to nothing about coding)

I can't seem to replicate the problem. Have you tried updating to 0.8.45? Also make sure you are loading the _v3 version. Otherwise you may need to contact calibre support

[tBunnyMan]

Originally Posted by dsip

would it be possible, to have a creation time displayed somewhere? Right now on the cover displays the date, but no time. Would there be an easy solution for that?

Thanks for the kind words!!!

It is a little bit of a hack but it works great per my testing. Download the latest version from github for the time under the date on the cover!

[tBunnyMan]

Read It Later has been rebranded to Pocket!

I am currently rebranding and auditing the plugin to make sure I can future proof this and maybe add features from their current site.

While I will try and remember to post updates make sure to go to the github repository for the latest version.

[arraintxo]

Hi,

I've just been working on this last week and using the new REST API to fetch the article list, so I have this almos working.

I took the base script from Keith M. Callenberg's blog on:
http://mem.bio.pitt.edu/calibre-readitlater
and improved it to get all the functionallity we allready got on the previous calibre readitlater recipe.

I also made the script get the data from the new "Article View" on getpocket, so it can download article's images and so...

You can get the recipe from:
http://bazaar.launchpad.net/~arraint...itlater.recipe

I hope it helps...

[tBunnyMan]

Originally Posted by arraintxo

Hi,

I've just been working on this last week and using the new REST API to fetch the article list, so I have this almos working.

I hope it helps...

Thanks! I'm still using the old method of scraping the Pocket minimal interface because I find it produces more constant results. I'm currently trying to get access to the v3 article rendering method mentioned in the API documents since it is what Pocket uses now. They are holding most of it off until everything settles.
I've written something up similar but am waiting until I can perfect it to commit it to my github, especially if they are just about to change their API.


I was looking at your code and I was wondering, Line 66-68

        index = self.API_INDEX + 'v2/get?' + self.get_auth_params()
        index += '&state=unread'
        index += '&count=' + str(self.max_articles_per_feed)

Won't this pull the NEWEST max_articles_per_feed? The API doesn't have an option to present in reverse so you need to take the list and reverse it yourself before lopping off the feed count.

The biggest reason for modding the RIL plugin in the first place was to add the functionality of presenting oldest to newest, there is a few feature regressions in your version.

[tBunnyMan]

Pocket v1.0 is now up on the github posted above.
MOST of it's changes are cosmetic with the rebranding but I did update the URL pulled from to the new pocket.

As mentioned above I'm holding off on the REST API until I hear back from Pocket's api development team on the new rendering engine and a few other questions.

[arraintxo]

Originally Posted by tBunnyMan

I was looking at your code and I was wondering, Line 66-68

        index = self.API_INDEX + 'v2/get?' + self.get_auth_params()
        index += '&state=unread'
        index += '&count=' + str(self.max_articles_per_feed)

Won't this pull the NEWEST max_articles_per_feed? The API doesn't have an option to present in reverse so you need to take the list and reverse it yourself before lopping off the feed count.

The biggest reason for modding the RIL plugin in the first place was to add the functionality of presenting oldest to newest, there is a few feature regressions in your version.

You are right on this one. I think I will be able to get around this by asking about stats before downloading the article list.

Thank you.

[Spiny]

Firstly, thank you for your efforts in updating this plugin tBunnyMan.

However, I'm having problems getting it to run:

Here's my error log:


calibre, version 0.8.50
ERROR: Conversion Error: <b>Failed</b>: Fetch news from Pocket

Fetch news from Pocket
Resolved conversion options
calibre version: 0.8.50
{<snip>}
InputFormatPlugin: Recipe Input running
Python function terminated unexpectedly
'NoneType' object has no attribute 'findAll' (Error Code: 1)
Traceback (most recent call last):
File "site.py", line 132, in main
File "site.py", line 109, in run_entry_point
File "site-packages\calibre\utils\ipc\worker.py", line 191, in main
File "site-packages\calibre\gui2\convert\gui_conversion.py", line 25, in gui_convert
File "site-packages\calibre\ebooks\conversion\plumber.py", line 972, in run
File "site-packages\calibre\customize\conversion.py", line 208, in __call__
File "site-packages\calibre\ebooks\conversion\plugins\recipe_ input.py", line 105, in convert
File "site-packages\calibre\web\feeds\news.py", line 861, in download
File "site-packages\calibre\web\feeds\news.py", line 1005, in build_index
File "<string>", line 76, in parse_index
AttributeError: 'NoneType' object has no attribute 'findAll'

[tBunnyMan]

Originally Posted by Spiny

Firstly, thank you for your efforts in updating this plugin tBunnyMan.

However, I'm having problems getting it to run:

You are welcome!

I'm having problems replicating this error. I had someone else report an error like this before and the problem was with a specific story in Pocket, likely a video saved in. While I try to generically patch around in can you run your Fetch and watch the status? It should list each article as it is fetching it. If I can get an example of what article is causing the crash it would be of great help!

[tBunnyMan]

I believe I have fixed this issue. There is a few things that can cause this but I have a work around for all of them according to my testing.

Download the latest v1.1 version to get the fixes!
http://github.com/tbunnyman/ReadItLater-Calibre-Plugin

[Spiny]

Didn't work I'm afraid, see attached error log. I also tried removing all videos from my pocket. It seems to be marking unread articles as read successfully though.

[tBunnyMan]

Originally Posted by Spiny

Didn't work I'm afraid, see attached error log. I also tried removing all videos from my pocket. It seems to be marking unread articles as read successfully though.

This is REALLY WEIRD but I'll get to the bottom of it. The function to mark articles as read isn't called until cleanup, IE after the articles have been passed back to calibre's renderer for making the epub file. AFAIK cleanup should never get called since the book was never generated. If this isn't the case then I should be able to put in a work around to make sure articles aren't marked if the book isn't successful.

The actual exception you are getting thrown is the "Not enough articles". This means that the number of articles that it is able to pull is less than the amount you have set in for minimum_articles (it defaults to ten).


Can you set minimum_articles to 1 and then find out what these two articles are? They are the two articles marked as read, or at least it claims it marked as read.
https://getpocket.com/textview/12378810
https://getpocket.com/textview/163547104