Feature domain tags

[alvaroreig]

Hi there @mmagnus,

I've recently changed my workflow and I don't use tags anymore. Therefore, all the articles appear under "untagged", which I didn't like.

Therefore, I've developed a new feature that organizes the articles under sections based on the first level domain of the URL of each article. Check this screenshot:

However, there are two drawbacks:

First of all, this feature is incompatible with the pocket tags normal functionality (TAGS, TAGS_EXCEPTIONS, INCLUDE_UNTAGGED, etc.). This is because the recipe ignores duplicates, and therefore an article can't belong in a section inherited from a pocket tag and also belong in a section created from its domain. Therefore, if SECTIONS_BY_DOMAIN=True pocket tags will be ignored. Because of that, I have left it disabled by default, so it won't break current use cases.

And second, because I use the tld module for extracting the domain, now the recipe has a new dependency (pip install tld). Therefore, even if someone is not going to use it python would check the import line and raise an exception. To avoid that, I have commented the related import. This can be resolved in a way better elegant way (lazy load, use a regular expressión, etc.) but I don't have the time right now.

Therefore, if someone wants to use this functionality you have to:

• uncomment "from tld import get_fld" in line 64
• install the tld module (pip install tld or whatever)
• set the SECTIONS_BY_DOMAIN to True in line 46

https://github.com/mmagnus/Pocket-Plus-Calibre-Plugin/tree/feature_domain_tags

[mmagnus]

Hi @alvaroreig,

Yeah, this is pretty cool. I had a similar problem to this, I was super lazy on tagging, so in PocketX (some more experimental version of the plugin) I coded a way to automatically Tag given urls ;-)

URLS_TO_TAGS = {'investing':
                ['fool.com',
                 'finance',
                 'marketwatch.com',
                 'benzinga'],
                 'rna':
                ['rnajournal']

so tag RNA will be added to URLs with "rnajournal" in it.

I like your approach. I think this should be an option.

Worked for me as well*

• I have one problem that Calibre's python does not see my libs, and I have to always help it to find it.

if SECTIONS_BY_DOMAIN:
    # maybe this is only me? @mmagnus but I need explicit append like this
    # sys.path.append("/Users/magnus/miniconda2/envs/py37/lib/python3.7/site-packages") 
    from tld import get_fld