worker.py

import urllib
import urlparse
import urllib2
import json

API_KEY = "0chdxJar9tkwkFEAfoyf"
SM_KEY = "2B269D5CE2"
BASE_URL = "http://thesaurus.altervista.org/"
PATH = "thesaurus/v1"
KEY_WORD = "word"
KEY_LANGUAGE = "language"
KEY = 'key'
KEY_OUTPUT = 'output'
OUTPUT_TYPE = 'json'

WIKI_BASE_URL = "https://en.wikipedia.org/"
WIKI_PATH = "w/api.php"
wiki_args_dict = {"action": "opensearch", "limit": 1, "format": "json"}
SM_BASE_URL = "http://api.smmry.com/"

sm_args_dict = {"SM_API_KEY": SM_KEY}

def encode_obj(in_obj):
    '''
    return the unicode of a string
    '''
    def encode_list(in_list):
        out_list = []
        for el in in_list:
            out_list.append(encode_obj(el))
        return out_list

    def encode_dict(in_dict):
        out_dict = {}
        for k, v in in_dict.iteritems():
            out_dict[k] = encode_obj(v)
        return out_dict

    if isinstance(in_obj, unicode):
        return in_obj.encode('utf-8')
    elif isinstance(in_obj, list):
        return encode_list(in_obj)
    elif isinstance(in_obj, tuple):
        return tuple(encode_list(in_obj))
    elif isinstance(in_obj, dict):
        return encode_dict(in_obj)

    return in_obj

def _build_url(baseurl, path, args_dict):
    '''return a list in the structure of urlparse.ParseResult

    args:
        baseurl -- the base url to use
        path -- the path off of the base url
        args_dict -- the key, value pairs of the keywords and parameters supplied
    '''
    url_parts = list(urlparse.urlparse(baseurl))
    url_parts[2] = path
    url_parts[4] = urllib.urlencode(encode_obj(args_dict))
    return urlparse.urlunparse(url_parts)

def request(word, key=API_KEY, output_type = 'json', language = 'en_US'):
    '''send a request to the thesaurus API

    args:
        word -- the word to lookup in the thesaurus
        key -- the API key
        output_type -- the format of the response data
        language -- the language used
    '''
    args_dict = {}
    args_dict[KEY_WORD] = word
    args_dict[KEY] = key
    args_dict[KEY_OUTPUT] = output_type
    args_dict[KEY_LANGUAGE] = language
    url = _build_url(BASE_URL, PATH, args_dict)

    response = urllib2.urlopen(url).next()
    return response

def wiki_request(word):
    '''
    Send GET request to wiki API which auto complete a search term (e.g. Alzheimer's to Alzheimer's disease)

    args:
        word -- the word (can be incomplete) that is searched
    '''
    word = word.replace(' ', '_')
    wiki_args_dict["search"] = word
    url = _build_url(WIKI_BASE_URL,WIKI_PATH, wiki_args_dict)
    try:
        f = urllib2.urlopen(url)
        j = json.loads(f.read())
        link_url = j[-1][0].encode('utf-8')
        return link_url
    except urllib2.HTTPError, e:
        print(e.code)
    except urllib2.URLError, e:
        print(e.args)
    except:
        return None

def smmry_request(paragraph):
    '''
    Send request to smmry api

    args:
        paragraph -- the paragraph that needs to be summarized
    '''
    url = _build_url(SM_BASE_URL,"", sm_args_dict)
    data = {'sm_api_input': paragraph}
    post = urllib.urlencode(encode_obj(data))
    resp = urllib.urlopen(url, data=post)
    return resp