mathblogging-base

A new base for mathblogging.org

How it works

This code base is a very simple "feed planet". It takes a bunch of feeds (organized by categories), fetches them, creates a (Jekyll) page and feed merging the feeds of each category. It also grabs some Twitter data to pull in "picks" from editors. Finally, there are some automation tools for running a cron job to build the site continuously.

The various parts

data

• data/feeds.json: a json file consisting of feed URLs and category names for each URL

the main parts

• lib/feedFetchTrimmer.js: fetches feeds, filters recent posts and saves them in feed/
• lib/feedmerger.js: merges a bunch of feeds into a new feed
• lib/pagewriter.js: takes feed object and writes out a jekyll page
• lib/editors.js: fetches twitter data and generates the frontpage
• lib/specialPages.js: generate special jekyll content
• lib/feedJsonHelpers.js: helpers for wrangling data/feeds.json
• lib/helpers: a few simple helper functions
• lib/app.js: puts the pieces together

tools

• tools/csv_to_json.js: fetches csv data from Google Sheets and generates data/feeds.json from it
• tools/cse.js: logs out the homepage URLs for all feeds to feed into a Google Custom Search Engine.

automation

• bin/run_feedFetchTrimmer.js: a wrapper for queuing feed aggregation
• bin/run.sh: simple script (assuming a bunch of things) for a cron job which runs all the things and pushes updates out to https://github.com/mathblogging/mathblogging.org.

How to re-use

First, you need a simple spreadsheet with entries like

feedURL, category1, category2, category3

Then you'll need to remove all mathblogging specific data.

For example,

• lib/editors.js: change the array of Twitter handles to match your group of editors and set up your Twitter app secrets.
• Replace all references to ./mathblogging.org to match the name of your Jekyll site (cloned into a subfolder).
• Replace all other mentions of mathblogging.org

Not perfect, we know. If you need help, file an issue.