ratebeer.py

What is this?

RateBeer is a database of user-created reviews about beers and breweries. However, their API has been down for some time, making it difficult to get that information programmatically. This simplifies that process, allowing you to access it in the most painless way possible. Data is returned to you in a friendly, Pythonic way:

>>> import ratebeer
>>> rb = ratebeer.RateBeer()
>>> rb.search('Summit')
{'beers': [<Beer('/beer/21st-amendment-summit-ipa/61118/')>,
           <Beer('/beer/3-daughters-summitennial-double-ipa/427367/')>,
           ...
           <Beer('/beer/elland-summit-for-the-weekend/197690/')>,
           <Beer('/beer/elland-summit-from-another-galaxy/197112/')>],
 'breweries': [<Brewery('/brewers/summit-cider/22197/')>,
               <Brewery('/brewers/summit-city-brewerks/18972/')>,
               ...
               <Brewery('/brewers/summit-hard-cider-and-perry/18260/')>,
               <Brewery('/brewers/summit-station-restaurant-brewery/346/')>]}

Why not BeerAdvocate.com?

Because they're evil, and they issue takedown notices left and right. We like RateBeer. Scratch that, we love RateBeer.

Requirements

Requires requests[security], beautifulsoup4, and lxml.

Installation

Use pip:

pip install ratebeer

Or clone the package:

git clone https://github.com/alilja/ratebeer.git

Usage

Because ratebeer.py does not use an API, since one is not provided, no key is required. Simply:

>>> import ratebeer
>>> rb = ratebeer.RateBeer()
>>> rb.search("summit extra pale ale")

RateBeer Class

Methods

• get_beer -- Pass in the URL for a beer page and this function will return a Beer object containing information about the beer. In addition the the URL, it accepts an optional fetch argument (default: False), which can be set to true to immediately download the object's attributes. See the Beer class below. You can replicate the RateBeer.beer(URL) functionality using RateBeer.get_beer(URL, True).__dict__.
• beer -- Returns a dictionary with information about that beer.

>>> rb.beer("/beer/new-belgium-tour-de-fall/279122/")
{'_has_fetched': True,
 'abv': 6.0,
 'brewed_at': None,
 'brewery': <Brewery('/brewers/new-belgium-brewing-company/77/')>,
 'calories': 180,
 'description': "New Belgium's love for beer, bikes and benefits is best "
                'described by being at Tour de Fat. Our love for Cascade and '
                'Amarillo hops is best tasted in our Tour de Fall Pale Ale. '
                "We're cruising both across the country during our favorite "
                'time of year. Hop on and find Tour de Fall Pale Ale in fall '
                '2014.',
 'ibu': 38,
 'img_url': 'https://res.cloudinary.com/ratebeer/image/upload/w_120,c_limit/beer_279122.jpg',
 'mean_rating': None,
 'name': 'New Belgium Tour de Fall',
 'num_ratings': 261,
 'overall_rating': 72,
 'retired': False,
 'seasonal': 'Autumn',
 'style': 'American Pale Ale',
 'style_rating': 70,
 'style_url': '/beerstyles/american-pale-ale/18/',
 'tags': ['cascade', 'amarillo'],
 'url': '/beer/new-belgium-tour-de-fall/279122/',
 'weighted_avg': 3.34}

• get_brewery -- Pass in the URL for a brewery page and this function will return a Brewery object containing information about that brewery. In addition the the URL, it accepts an optional fetch argument (default: False), which can be set to true to immediately download the object's attributes. See the Brewery class below. You can replicate the RateBeer.brewery(URL) functionality using RateBeer.get_brewery(URL, True).__dict__.
• brewery -- Returns a dictionary with information about the brewery. Includes a 'get_beers()' generator that provides information about the brewery's beers.

>>> rb.brewery("/brewers/deschutes-brewery/233/")
{'_has_fetched': True,
 'city': 'Bend',
 'country': 'USA',
 'name': 'Deschutes Brewery',
 'postal_code': '97702',
 'state': 'Oregon',
 'street': '901 SW Simpson Ave',
 'telephone': '(541) 385-8606',
 'type': 'Microbrewery',
 'url': '/brewers/deschutes-brewery/233/',
 'web': 'https://www.facebook.com/deschutes.brewery'}

• search -- A generic search. A dictionary with two keys: beers and breweries. Each of those contains a list of objects, beers and breweries, respectively.

>>> rb = RateBeer()
>>> results = rb.search("summit extra pale ale")
>>> results
{'beers': [<Beer('/beer/summit-extra-pale-ale/7344/')>,
           <Beer('/beer/summit-extra-pale-ale--rose-petals/317841/')>],
 'breweries': []}
>>> results['beers'][0].__dict__
{'_has_fetched': False,
 'name': 'Summit Extra Pale Ale',
 'num_ratings': 721,
 'overall_rating': 60,
 'url': '/beer/summit-extra-pale-ale/7344/'}

• beer_style_list -- Returns a dictionary containing the beer style name and the style id.

>>> rb.beer_style_list()
{'Abbey Dubbel': 71,
 'Abbey Tripel': 72,
 ...
 'Witbier': 48,
 'Zwickel/Keller/Landbier': 74}

• beer_style -- Returns a generator of Beer objects from the beer style page. Takes a ident for a beer style (see output of beer_style_list()) and optional sort_type ("score" (default), "count", or "abv") and sort_order ("ascending" (low-to-high) or "descending" (high-to-low, default)).

>>> [b for b in rb.beer_style(71)]
[<Beer('/beer/st-bernardus-prior-8/2531/')>,
 <Beer('/beer/westmalle-dubbel/2205/')>,
 ...
 <Beer('/beer/belgh-brasse-mons-abbey-dubbel/187593/')>,
 <Beer('/beer/new-glarus-thumbprint-series-dubbel/254781/')>]

Beer Class

Beer requires the url of the beer you're looking for, like RateBeer.beer and RateBeer.get_beer.

Attributes

• abv (float): percentage alcohol
• brewery (string): the name of the beer's brewery
• brewery (Brewery object): the beer's brewery
• brewed_at (Brewery object): actual brewery if contract brewed
• calories (float): estimated calories for the beer
• description (string): the beer's description
• img_url (string): a url to an image of the beer
• mean_rating (float): the mean rating for the beer (out of 5)
• name (string): the full name of the beer (may include the brewery name)
• num_ratings (int): the number of reviews
• overall_rating (int): the overall rating (out of 100)
• retired (boolean): True if the beer is retired, otherwise False
• seasonal (string): Summer, Winter, Autumn, Spring, Series, Special, None
• style (string): beer style
• style_url (string): beer style URL
• style_rating (int): rating of the beer within its style (out of 100)
• url (string): the url of the beer's ratebeer page
• tags (list of strings): tags given to the beer
• weighted_avg (float): the beer rating average, weighted using some unknown algorithm (out of 5)

Any attributes not available will be returned as None

Methods

• get_reviews -- Returns a generator of Review objects for all the reviews in the beer. Takes a review_order argument, which can be "most recent", "top raters", or "highest score".

Review Class

Review returns a datatype that contains information about a specific review. For efficiency reasons, it requires the soup of the individual review. Probably best to not try to make one yourself: use beer.get_reviews instead.

Attributes

• appearance (int): rating for appearance (out of 5)
• aroma (int): aroma rating (out of 10)
• date (datetime): review date
• overall (int): overall rating (out of 20, for some reason)
• palate (int): palate rating (out of 5)
• rating (float): another overall rating provided in the review. Not sure how this different from overall.
• taste (int): taste rating (out of 10)
• text (string): actual text of the review.
• user_location (string): writer's location
• user_name (string): writer's username

Brewery Class

Brewery requires the url of the brewery you want information on.

Attributes

• city (string): the brewery's city
• country (string): the brewery's country
• name (string): the brewery's name
• postal_code (string): the brewery's postal code
• state (string): the brewery's state/municipality/province
• street (string): the street address of the brewery
• telephone (string): the brewery's telephone number
• type (string): the type of brewery. Typically "microbrewery" or "macrobrewery"
• url (string): the url of the brewery's ratebeer page
• web (string): the url of the brewery's homepage

Methods

• get_beers -- Returns a generator of Beer objects for every beer produced by the brewery. Some brewery pages list beers that are produced by do not have any pages, ratings, or information besides a name. For now, these beers are omitted from the results.

Tests

ratebeer uses the standard Python unit testing library.

This can be run via python test.py.

Changes

Note that the nature of web scraping means this might break at any time.

v2.3.1

• Overhauled the Beer object so that it will be a little easier to fix with future changes. Beer object now also returns Brewery objects rather than strings for the brewery and brewed_at attributes. Also returns the url for the image of the beer and a list of user-assigned tags. The test.py file has been updated to be a bit clearer about where failures occur.

v2.3

• Fixes to work with the new RateBeer search page.

v2.2.1

• Beer and Brewery objects are now "lazy", meaning they will not fetch the RateBeer page unless the requested attributes are not available. This should help minimize unnecessary requests.
• RateBeer.search() now returns two lists of Beer and Brewery objects.
• RateBeer.beer_style_list() now returns Beer and Brewery objects.
• Beer and Brewery objects now allow custom attributes to be set.

v2.1

• Bugfixes and performance enhancements.
• Python 3 compatibility.

v2.0

Major changes.

• New Beer, Review, and Brewery classes.
• Substantial overhaul in ratebeer.py, addition of new files including separation of responsibilities
• New generator functions in new classes.

v1.4

• reviews is now a generator.

v1.3.5

• Several improvements to results, particularly for edge cases and situations where search results are not in the expected order.

v1.3.4

• Metadata for beers returns floats when appropriate.

v1.3.3

• Captures more meta data.
• Plays better with foreign beers.
• Now if information is missing from a beer entry, its key is not added to the beer output.

v1.3.2

• Captures aliases for beer names.

v1.3

• Added beer_style_list and beer_style.

v1.2

• Everything conforms to PEP8 now. Thanks to the fine folks here.
• Minor refactoring.

v1.1

• Added reviews.
• Better exceptions (no more LookupError for 404s)

v1.0

• Initial release.

License

Creator: Andrew Lilja

Contributors: * Vincent Castellano (Surye) - Python 2 and 3 compatability * Steven A. Cholewiak - General bug squishing * parryc - Scraping updates and general bug squishing

All code released under the Unlicense (a.k.a. Public Domain).