README.md
In rmflight/staticReader: Managing a static RSS reader

staticReader

StaticReader is an R package that provides a static RSS reader using the python RSS reader rawdog, some saving functionality using javascript and github gists, and other github manipulation using the github R package. This makes it it possible to easily save rss item links somewhere else when hosting the feed on static hosting such as github. The templates, combined with some rawdog plugins, github hosting and gists and javascript cross-site requests let us do some interesting things.

You will need to download and install rawdog, and its requirement feedparser. The following steps assume that this has been done.

Clone this repo to a directory where you will keep everything.

git clone https://github.com/rmflight/staticReader.git

Assuming you want to save URL's from the rawdog output somewhere, you will also need a personal github oauth token and a gist to save them to. Oauth tokens can be generated from the github account settings, under Applications -> Personal Access Tokens -> Create new token. Copy this string and put it in a text file, this will be used to connect the web-page with your gist.

Create a gist with at least one file (savedURL.md is the assumed file name by staticReader) with a line of text, and a return. This is the file where URL's will be saved. You will need the gist id, this is the last part of the numeric part of the gist url, i.e. gist.github.com/username/gistid, gistid is that part you want. And the name of the file in the gist you want to store the URL's in. You can use savedURL.md to incur the least changes to the javascript code.

With this information in hand, you need to modify the saveDataGist.js file. Specifically, you need to modify the variable gistID on line 70, and if you are using a different file name in the gist, then also modify lines 144 and 149.

You also need to define the file where the rawdog output should go, line 151 in .rawdog/config.

If you were previously using Google Reader, and have your feeds stored in an OPML file, you can also use the opml_to_text.R function to extract your feed URLs into a format suitable for rawdog.

To verify that everything is working properly, you should now run rawdog once. The included config file has the reddit rss feed for new included to allow for testing.

rawdog -d pathtodir/.rawdog -uw

Now start a simple web server and point it to your staticReader directory.

python -m SimpleHTTPServer

library(servr)
httd(dir=".", port=8000)

And open localhost:8000/index.html in your browser. You should see your rawdog output, and there should be two buttons, Submit and Fetch Gist Data. Submit should be red. Paste in your token you generated and saved earlier, and hit Submit, it should turn green. Your token is now saved as a cookie. Next, hit Fetch Gist Data, it should also turn green. If it doesn't, open up a console and look at error messages. Finally, hit save on one of the items in the feed. The save button should turn red while the request is processed, then green when it gets the confirmation back. Finally, open the gist where you were keeping the data, and you should see a new link. Congratulations, it works!.

I am not a regular RSS reader, I like to read things in bulk, when I get a chance. Therefore, the rawdog configuration and plugins are what are useful to me. My use case is to have continuous updates of the rawdog state using cron and rawdog -d staticReaderDir/.rawdog -u so that I'm not overloading RSS feed servers, and then when I am ready to read, do rawdog -d staticReaderDir/.rawdog -w to write the output, and then push the output to static hosting in GitHub. Therefore, I have the since-last plugin installed to just write the entries since the last write. So I assume that I won't write again until I'm done reading a given set of entries. If this is not what you want, then you may want to delve deeper into the rawdog config and plugins and change things up for yourself.

Currently only tested on Chrome, have noticed one bug in Firefox that has not been investigated.

If you want to host this on GitHub, then you will need to create a gh-pages branch of your repo, and make sure to include the html, js, and css files on that branch.