Plugin: Read XMPP PubSub node items as Atom Syndication feed

[sjehuda]

Greetings!

I have made a proof of concept of an HTTP software that conveys XMPP PubSub nodes and node items as Atom Syndication feed.

The software is written in Python and utilizes slixmpp to communicate with the XMPP protocol.

Pros

• View PubSub content from desktop using Liferea instead of web browser interfaces (i.e. Jappix, Movim, Libervia.

Cons

• Jabber ID account is required.

Pros (Jabber ID)

• Anonymous Jabber ID (fixed "hostname" part and random "local" part) is compatible.
• There are over ten thousand XMPP server, so you can open as many accounts as you need anywhere you want and also set a certain account for a specific purpose.
• Because Jabber ID account is used, the other side will only see the Jabber ID, which means that you IP is always concealed.

Here is the code:

import datetime
import feedgenerator
from fastapi import FastAPI, Request, Response
from slixmpp import ClientXMPP

app = FastAPI()

class XmppInstance(ClientXMPP):
    def __init__(self, jid, password):
        super().__init__(jid, password)
        self.register_plugin('xep_0060')
        self.connect()
        # self.process(forever=False)

xmpp = None

@app.get('/atom')
async def view_pubsub(request: Request):
    global xmpp
    if not xmpp:
        xmpp = XmppInstance('JABBER_ID', 'PASSWORD')
        # xmpp.connect()

    pubsub = request.query_params.get('pubsub', '')
    node = request.query_params.get('node', '')
    item_id = request.query_params.get('item', '')
    if pubsub and node and item_id:
        iq = await xmpp.plugin['xep_0060'].get_item(pubsub, node, item_id)
        link = 'xmpp:{pubsub}?;node={node};item={item}'.format(
            pubsub=pubsub, node=node, item=item_id)
        print(link)
        result = pubsub_to_atom(iq, link)
    elif pubsub and node:
        iq = await xmpp.plugin['xep_0060'].get_items(pubsub, node)
        link = 'xmpp:{pubsub}?;node={node}'.format(pubsub=pubsub, node=node)
        result = pubsub_to_atom(iq, link)
    elif pubsub:
        iq = await xmpp.plugin['xep_0060'].get_nodes(pubsub)
        link = 'xmpp:{pubsub}'.format(pubsub=pubsub)
        result = pubsub_to_opml(iq)
    elif node:
        result = 'PubSub parameter is missing.'
    else:
        result = ('Mandatory parameter PubSub and '
                  'optional parameter Node are missing.')
    return Response(content=result, media_type="application/xml")

def pubsub_to_atom(iq,link):
    """Convert XEP-0060: Publish-Subscribe to RFC 4287: The Atom Syndication Format."""
    feed = feedgenerator.Atom1Feed(
        description = ('This is a syndication feed generated with PubSub to '
                       'Atom, which conveys XEP-0060: Publish-Subscribe nodes '
                       'to standard RFC 4287: The Atom Syndication Format.'),
        language = iq['pubsub']['items']['lang'],
        link = link,
        subtitle = 'Syndication Feed generated with PubSub to Atom',
        title = iq['pubsub']['items']['node'])
    # See also iq['pubsub']['items']['substanzas']
    entries = iq['pubsub']['items']
    for entry in entries:
        item = entry['payload']
        namespace = '{http://www.w3.org/2005/Atom}'
        title = item.find(namespace + 'title')
        title = None if title == None else title.text
        updated = item.find(namespace + 'updated')
        updated = None if updated == None else updated.text
        # if updated: updated = datetime.datetime.fromisoformat(updated)
        content = item.find(namespace + 'content')
        content = 'No content' if content == None else content.text
        link = item.find(namespace + 'link')
        link = '' if link == None else link.attrib['href']
        author = item.find(namespace + 'author')
        try:
            print(author.attrib)
        except:
            pass
        author = 'None' if author == None else author.text
        # create entry
        feed.add_item(
            description = content,
            # enclosure = feedgenerator.Enclosure(enclosure, enclosure_size, enclosure_type) if args.entry_enclosure else None,
            link = link,
            # pubdate = updated,
            title = title,
            unique_id = link)
    return feed.writeString('utf-8')

• Save the code into a file named main.py.
• Execute it with $ uvicorn main:app --reload
• Try it:

To view node items

PubSub: xmpp:pubsub.movim.eu
node: berlin-xmpp-meetup
http://127.0.0.1:8000/atom?pubsub=pubsub.movim.eu&node=berlin-xmpp-meetup

To view a node item

PubSub: xmpp:pubsub.movim.eu
node: berlin-xmpp-meetup
item: let-s-talk-about-the-xsf-and-possibly-other-things-6A8eV4
http://127.0.0.1:8000/atom?pubsub=pubsub.movim.eu&node=berlin-xmpp-meetup&item=let-s-talk-about-the-xsf-and-possibly-other-things-6A8eV4

[sjehuda]

The important parts here are:

• Class XmppInstance
• Function pubsub_to_atom

I need to incorporate them into a plugin to read URIs like this xmpp:pubsub.movim.eu?;node=berlin-xmpp-meetup.

I guess a userscript (i.e. filter) would be even easier to write, and it would also be possible to use with other feed readers such as snownews, yet if you think a plugin is feasible, I will gladly make one.

A userscript is sensible but lesser good than a plugin because the userscript (i.e. filter) is turned off and on or being called multiple times which means that upon scanning 20 feeds, there would be 20 instances of an XMPP account on an XMPP server.

[sjehuda]

The proof of concept script has been developed into an HTTP gateway.
I encourage you to view the video demonstration (length 7:44).

https://video.xmpp-it.net/w/vNqcMooy3pqRAZ8Yb8grr1
https://git.xmpp-it.net/sch/Rivista

[sjehuda]

Mr. Martin Rotter, the developer of RSS Guard, has added a plugin to handle XMPP.

https://github.com/martinrotter/rssguard/releases/tag/devbuild5

martinrotter/rssguard#1548

He has also extended it to handle messages, which potentially qualifies RSS Guard as an answering machine.