INDEX

2024-01-21 17:56 - RSS Repeaters

I want to make a tool to take a folder and host an rss feed of it
Since I already have that in bash, it should have a config to allow multiple feeds
Was doing it 1 port per feed but can have one big python website for many at once
FastAPI or flask? Probably Flask since I know fastapi at this point
	# Simple compose script with python and volume mount
	docker compose run python bash

Firstly need to setup environment - can install pip globally in docker
But also why not put in a venv
	pip install -r requirements.txt # Install from list of python requirements
	pip freeze > requirements.txt # Export requirements

	flask --app main run -p 1234 # Weirdly the order of this matters

	network_mode: "host" # Set this in docker-compose to use host network
	# Weirdly sets hostname to same as my host too (might be due to image choice)

Anyway exposed 1234 and website seems to run fine. Just need arbitrary endpoints
	# From the documentation
	from markupsafe import escape
	@app.route('/user/<username>')
	def show_user_profile(username):
	   # show the user profile for that user
	   return f'User {escape(username)}'

Seems to work. Now just need to make a list of folders to copy
Why not write in yaml - using pyyaml - can do basic parsing just reads as dict/json
	for key, value in d.items(): # Always useful


2024-01-22 15:35

So before I wrote a basic website to scan yaml and read it - gonna clean that up
Plan is have /serve and /track - serve is local and track is remote
It's not going to be an rss tracker - more a generator for awkward sites that don't use rss
Can generate rss feeds for youtube playlists and just link back to original videos
Essentially replacing the "using a website" bit of everything - just link to the source

Trying to figure out how to structure this yaml - don't necessarily need an ID for each item
Just do "- ITEM" to add a list item - but you need a list label
	# Using template from redhat site
	- name: Check URL
	  hosts: all
	  gather_facts: True
	  tasks:
	  - uri:
	      url: "{{ urlpath }}"
	      method: GET
	    register: result
	    until: result.status == 200
	    retries: 30
	    delay: 20
	    delegate_to: localhost
	  - debug:
	      msg: "{{ result.status }} - {{ result.msg }}"
	# This way the ID is not "uri" - you just have 2 tasks indexed by 0 and 1

Need to figure out how to set up expected yaml structures in python
Might have to use typing for this - strictyaml or schema which is best?
Features I want to use in strictyaml are "in alpha" so looking at schema now
https://www.andrewvillazon.com/validate-yaml-python-schema/
https://github.com/keleshev/schema

Can do things like ["directory": os.path.exists] to see if a file given exists
	# Current version (untested)
	schema = Schema({
	   "local": {
	       "defaults": [{
	           Optional("category", default=""): str,
	           Optional("enabled", default=False): bool,
	           Optional("itemtype", default="mp3"): Or("mp3","mp4")
	       }],
	       "feeds": [{
	           "directory": os.path.exists,
	           "title": str,
	           Optional("category", default=""): str,
	           Optional("enabled", default=False): bool,
	           Optional("itemtype", default="mp3"): Or("mp3","mp4")
	   }]}
	   "track": {
	       "defaults": [{
	           Optional("category", default=""): str,
	           Optional("quality", default="mid"): Or("low","mid","high"),
	           Optional("selfgenerate", default=False): bool
	       }],
	       "feeds": [{
	           "source": str,
	           "title": str,
	           "id": str,
	           Optional("category", default=""): str,
	           Optional("quality", default="mid"): Or("low","mid","high"),
	           Optional("selfgenerate", default=False): bool
	}]}}

Gets ugly when you try to figure out how to host a file server inside this
I'll figure it out a bit later. For now, got proper yaml reading and a plan for structure

For now it's just going to be an rss relay and generator - no actual hosting

Okay so in the time I've been working I've added a few things
Made it scan a folder for relevant files and save all this as an rss feed
Add a basic http server to run script (running in background) so feed link as valid
Added formatting checking to config file - with defaults applied