Metatron

HTML Meta tag parser, with emphasis on OpenGraph/Twitter Cards, and complex (and custom) meta tag schemes. Supports Python 3.x and up. The Metatron object extends dict, and all the meta tag data is set within it.

Installation

You know the drill:

pip install metatron

Straight to an example

Simple meta tag collector

This example collects top level meta tags without a schema:

> mt = Metatron(url='https://www.bbc.co.uk')
> mt.traverse()

{
    'x-country': 'gb',
    'x-audience': 'Domestic',
    'CPS_AUDIENCE': 'Domestic',
    'CPS_CHANGEQUEUEID': '115204091',
    'theme-color': 'bb1919',
    'msapplication-TileColor': '#bb1919'
}

> mt['x-country']
> 'gb'

Collect structures OpenGraph meta tags

> mt = Metatron(url='https://www.bbc.co.uk', schemas=['og'])
> mt.traverse()

{
    'og': {
        'title': 'Home - BBC News',
        'description': 'Visit BBC News for up-to-the-minute news....',
        'site_name': 'BBC News',
        'locale': 'en_GB',
        'article': {
            'author': 'https://www.facebook.com/bbcnews',
            'section': 'Home'
        },
        'url': 'http://www.bbc.co.uk/news',
        'image': '//m.files.bbci.co.uk/modules/bbc-morph-news-waf-page-meta/2.1.0/bbc_news_logo.png'
    }
}

Supports opengraph arrays (and can receive content as input)

> content = """
    <meta property="og:title" content="First title tag" />
    <meta property="og:title" content="Second title tag" />
    <meta property="og:description" content="description tag" />
    <meta property="og:image" content="http://example.com/image.jpg" />
    <meta property="og:image:secure_url" content="https://secure.example.com/image.jpg" />
    <meta property="og:image:type" content="image/jpeg" />
    <meta property="og:image:width" content="400" />
    <meta property="og:image:height" content="300" />
    <meta property="og:image:alt" content="First image description" />
    <meta property="og:image" content="http://example.com/image2.jpg" />
    <meta property="og:image:secure_url" content="https://secure.example.com/image.jpg" />
    <meta property="og:image:type" content="image/jpeg" />
    <meta property="og:image:width" content="500" />
    <meta property="og:image:height" content="600" />
    <meta property="og:image:alt" content="Second image description" />
"""

> mt = Metatron(content=content, schemas=['og'])
> mt.traverse()

{
    'og': {
        'description': 'description tag',
        'image': [
            {
                'alt': 'First image descriptoin',
                'height': '300',
                'image': 'http://example.com/image.jpg',
                'secure_url': 'https://secure.example.com/image.jpg',
                'type': 'image/jpeg',
                'width': '400'
            },
            {
                'alt': 'A shiny green apple with a bite taken out',
                'height': '600',
                'image': 'http://example.com/image2.jpg',
                'secure_url': 'https://secure.example.com/ogp.jpg',
                'type': 'image/jpeg',
                'width': '500'
            }
        ],
        'title': [
            'First title tag',
            'Second title tag'
        ]
    }
}

Add your own meta tag schema

You can provide Metatron with your own meta tag schema spec

> from metatron import add_schema_spec
> my_spec = {
    'name': 'boom',
    'attribute': 'name',
    'value': 'value'
}
> add_schema_spec(my_spec)

You can then parse tags like:

<meta name="boom:title" value="Boom title" />
<meta name="boom:description" value="Boom description" />

Using

> mt = Metatron(url='http://example.com', schemas=['boom'])
> mt.traverse()
> {
    'boom': {
        'title': 'Boom title',
        'description': 'Boom description'
    }
}

Note that the add_schema_spec call allows for an additional argument to tag the schema as. This is useful if you have multiple schema definitions with the same name (or no name).

Custom schema - custom tag

It is possible to add a tag key to the custom schema spec, which can then lookup different tags other than meta. If not provided, meta is used by default.

Added in 0.4

> from metatron import add_schema_spec
> my_spec = {
    'tag': 'link',
    'name': '',
    'attribute': 'rel',
    'value': 'href'
}
> # Register the schema as 'link' to avoid duplicating existing no-name schema
> add_schema_spec(my_spec, 'link')
> content = """
    <meta property="og:title" content="First title tag" />
    <meta property="og::title" content="Second title tag" />
    <meta property="og:description" content="description tag" />
    <meta property="og:image" content="http://example.com/image.jpg" />

    <link rel="canonical" href="https://example.com/cannonicallink/" />
"""
> mt = Metatron(content=content, schemas=['link'])
> mt.traverse()

Can run from the command line

$ make run URL=http://bbc.co.uk/news SCHEMA=og

or

$ python -m metatron.metatron http://bbc.co.uk/news og

$ Getting: http://bbc.co.uk/news (schemas: og)
{'og': {'description': 'Visit BBC News for up-to-the-minute news, breaking '
                   'news, video, audio and feature stories. BBC News '
                   'provides trusted World and UK news as well as local '
                   'and regional perspectives. Also entertainment, '
                   'business, science, technology and health news.',
    'image': '//m.files.bbci.co.uk/modules/bbc-morph-news-waf-page-meta/2.2.1/bbc_news_logo.png',
    'locale': 'en_GB',
    'section': 'Home',
    'site_name': 'BBC News',
    'title': 'Home - BBC News',
    'type': 'website',
    'url': 'http://www.bbc.co.uk/news'}}

Dependencies

• requests
• beautifulsoup4