Microservice.

Provides news from various sources - blogs, portals, facebook pages and exposes via API.

Created for an anarchist portal, with aim to propagate the freedom and grass-root social movements, to automate things for building a better world. Viva la revolución.

• https://wolnosciowiec.net/#/flash
• http://iwa-ait.org
• http://zsp.net.pl
• http://federacja-anarchistyczna.pl/

Clone the repository, set up a MySQL database, execute make deploy

Application is fully manageable by the HTTP API, so you can run eg. a postman to add news boards and attach feed sources. There is an example postman collection exported to file postman.json

A news board is a board where your articles will appear, it's like a time line on facebook or twitter.

{{DOMAIN}}/api/news_board/create

{
    "id": "federacja",
    "name": "Federacja Anarchistyczna",
    "description": "Działająca na terenie Polski federacja grup anarchistycznych. Istniejąca od 1988 roku (początkowo jako Międzymiastówka Anarchistyczna). Dąży do „stworzenia samorządnego społeczeństwa, tworzonego na zasadzie dobrowolności”"
} 

{{DOMAIN}}/api/feed_source/newsBoard-federacja/create

{
    "id": "fa-wroclaw",
    "title": "FA Wrocław",
    "collectorName": "rss",
    "icon": "http://wolnywroclaw.pl/wp-content/uploads/2015/05/icon-555f99cdv1_site_icon-256x256.png",
    "sourceData": {
        "url": "http://wolnywroclaw.pl/feed/"
    },
    "defaultLanguage": "pl",
    "description": "Federacja Anarchistyczna - sekcja Wrocław",
    "enabled": true,
    "scrapingSpecification": {
        "removePaths": [
            "//*[@id='main-content']/article/div[1]/div"
        ],
        
        "contentPath": "//*[@id='main-content']/article/div[1]"
    }
}

Explanation:

• collectorName: Must point to a valid name of supported collector, actually only "rss" is bundled by default by the plans are to add also a "facebook_page"

• id: You will use it to filter by a source for example

• defaultLanguage: It's a fallback for articles that do not expose a language via eg. RSS

• sourceData: Describes parameters that are passed to the collector, in this example to "rss" collector which requires only URL

• scrapingSpecification: Scrapper is a tool that is able to attempt to extract full article content from linked article page, and remove ads everything is operated by xPaths (OPTIONAL)

You may want to create an option that will allow to select from which sources to show the content. There is an endpoint that shows this list and allows filtering it.

{{DOMAIN}}/api/feed_source/newsBoard-federacja/search

{
    "title": "FA"
}

{{DOMAIN}}/public/api/feed/federacja/browse/1/5

{
    "lanaguage": ["pl"],
    "exceptFeedSource": ["fa-krakow"]
}

Explanation:

• /browse/1/5 - the first number is a page number, the second is a count of elements to return per page
• language: Allows to filter by content language (OPTIONAL)
• exceptFeedSource: Excludes feed sources by id (OPTIONAL)
• See {{DOMAIN}}/public/api/feed/search/help endpoint for more options

• [ ] Complete test coverage
• [ ] Implement OAuth2 tokens as currently the service is for internal usage only
• [ ] Implement Facebook pages posts crawler
• [ ] Implement a distributed mode (a crawler that collects data from other instance)

Crawled content often contains attached images, and the article itself very often has a main image. To avoid hotlinking entries there is a way to fetch all of those images to a own server and replace all usages.

File Repository is a Wolnościowiec's microservice that exposes an API interface for file storage, the service itself handles de-duplication, caching and of course storing the files in right places.

To enable the integration simply only all environment variables needs to be filled up properly, the rest is handled by the ImageRepositoryBundle.

When links to the articles does not have SSL it is possible to use a Wolnościowiec Web-Proxy to generate urls secured with SSL. The web-proxy service is available here: https://github.com/Wolnosciowiec/web-proxy

Example case:

1. We have a source url to the article: http://wolnywroclaw.pl/relacje/protest-lokatorow-z-ul-slicznej-i-zaulka-rogozinskiego-wideo/
2. But... our site has SSL, to render the article in iframe we call the web-proxy
3. The webproxy returns: https://some-proxy.services.someservice.org/?__wp_one_time_token=YT323DEe21FAFAFAZSP1312161F234ewfADSy65Efre4RE23EDW312gtrhtr6
4. Embedding the URL from the web-proxy allows to render correctly the page through our web-proxy server

Enabling:

# web-proxy
NFP_WEB_PROXY_URL="https://some-proxy.services.someservice.org"
NFP_WEB_PROXY_PASSPHRASE="my-super-secret-passphrase-from-webproxy-service"
NFP_PROCESS_INTERNAL_LINKS=1 # process HTML and CSS content, so all static content will be rendered by the proxy
NFP_EXPIRATION_TIME_MINUTES=1
NFP_ENABLED_SSL=true

There is an endpoint that shows the application health, it's placed under {{DOMAIN}}/{{NFP_MONITORING_KEY}}/monitor/health/run NFP_MONITORING_KEY is a environment variable you can pass to nginx/apache/docker

There exists an image wolnosciowiec/news-feed-provider, it uses environment variables to configure the application. You may want to take a look at the list of variables with example values in the Dockerfile.

sudo docker run -e NFP_MAILER_HOST=somehost -e NFP_MAILER_USER=someuser -p 80:80 wolnosciowiec/news-feed-provider

There are a lot of configuration variables so you might consider using a docker-compose for simple usage, or docker-swarm or Kubernetes if scaling in a cluser.