Hacker News Reader

Comment by flobosg | Hacker News Reader

To name a few:

* A script that emails me every day a list of new articles published by a selection of academic journals. Using an RSS reader for this turned out to be too messy.

* A script that generates Spotify playlists based on my liked songs, optionally according to some criteria. The playlists can be sorted according to specific song features (higher to lower energy, for example).

* A zettelkasten engine, including search, rendering, and backup of entries.

* A system to generate and upload backups of my photos from my cameras’ SD cards.

View on HN

I'd be curious to see the first script, if it's shareable?

Sure! It is very rudimentary and some values are still hardcoded, but I could do bit of clean-up and then upload it.

EDIT: Here it is – https://github.com/flobosg/journalert

Thanks! On the fence about whether to use as an excuse to try something small in python for the second time in my life... or just steal the idea and pretty much copy it line by line into something my non-programmer mind will find easier (php). Though maybe it'll be so simple to run and not need future tweaking that it won't even force me into learning by keeping it in python.

Though also, I was hoping for the interesting sources too :P

The sources are not that special, but I will post them here anyway. They lean towards biochemistry, bioinformatics and structural biology:

  feeds = ("http://feeds.nature.com/nbt/rss/current",
           "http://feeds.nature.com/nchembio/rss/current",
           "http://feeds.nature.com/nsmb/rss/current",
           "http://feeds2.feedburner.com/plosbiology/NewArticles",
           "http://feeds.plos.org/ploscompbiol/NewArticles",
           "http://feeds.plos.org/plosgenetics/NewArticles",
           "https://www.frontiersin.org/journals/bioinformatics/rss",
           "https://www.frontiersin.org/journals/molecular-biosciences/rss",
           "https://www.nature.com/subjects/computational-biology-and-bioinformatics.rss",
           "http://www.nature.com/subjects/protein-design.rss",
           "http://www.nature.com/subjects/protein-folding.rss",
           "https://www.nature.com/subjects/structural-biology.rss",
           "http://online.liebertpub.com/action/showFeed?mi=cjwv&ai=su&jc=CMB&type=etoc&feed=rss",
           "https://onlinelibrary.wiley.com/action/showFeed?jc=1469896x&type=etoc&feed=rss",
           "https://onlinelibrary.wiley.com/action/showFeed?jc=10970134&type=etoc&feed=rss",
           "https://onlinelibrary.wiley.com/action/showFeed?jc=15585646&type=etoc&feed=rss",
           "https://science.sciencemag.org/rss/current.xml")
  
  no_pubdate = ("http://rss.sciencedirect.com/publication/science/09581669",
                "http://rss.sciencedirect.com/publication/science/0959440X",
                "http://rss.sciencedirect.com/publication/science/13675931")

Most academic journals have RSS feeds, so it should be easier to find ones that fit your interests.

Did you share 2?

No, but I could share it as well. That one is a mess right now, though. It needs a lot of refactoring to make it readable. Let me see what I can do.

It uses spotipy (https://github.com/plamere/spotipy) to fetch my liked songs and generate the recommendations.

What criterias do you extract from your liked songs to pick new songs?

I pick five tracks at random as a seed to generate the recommendations (https://spotipy.readthedocs.io/en/2.19.0/#spotipy.client.Spo...). If I want I can also filter recommendations according to certain filters (audio features such as danceability, track duration, etc.).

Reply on news.ycombinator.com