pvdb/video_metadata/views/tmdb.py

import cgi
from contextlib import closing
from datetime import date
from enum import Enum
import json
from pprint import pprint
import re
import ssl
from urllib.error import URLError, HTTPError
import urllib.parse
import urllib.request

from flask import Blueprint, current_app, flash, request, render_template

from .. import util


####################
# constants

bp = Blueprint('tmdb', __name__, url_prefix='/tmdb')
TMDB_URL = 'https://api.themoviedb.org'
TMDB_SEARCH_MOVIE = '/3/search/movie'
TMDB_SEARCH_TV = '/3/search/tv'


####################
# data structures

class query_type(Enum):
    UNKNOWN = 0
    MOVIE   = 1
    TV_SHOW = 2

class query_data(object):
    def __init__(self):
        self.errors = []
        self.full_url = ''
        self.movie_year = 0
        self.parsed_query_str = ''
        self.query_params = {}
        self.query_type = query_type.UNKNOWN
        self.raw_query_str = ''
        self.response_body = ''
        self.response_headers = []
        self.response_object = {}
        self.response_status = 0
        self.tmdb_api_key = ''


####################
# flask routes

@bp.route('/query_movie')
def query_movie():
    view_data = query_data()
    view_data.tmdb_api_key = current_app.config['TMDB_API_KEY']

    # NOTE: query string can be submitted from a link or form input
    if 'query' in request.form:
        view_data.raw_query_str = request.form['query']
    elif 'query' in request.args:
        view_data.raw_query_str = request.args['query']

    if parseQueryStr(view_data) and makeQuery(view_data):
        try:
            p = view_data.response_object = json.loads(view_data.response_body)
            pprint(p)
        except json.JSONDecodeError as e:
            view_data.errors.append(e)
    else:
        # display errors
        pass

    if view_data.errors:
        flash(view_data.errors)

    return render_template(
            'tmdb/query.html',
            view_data=view_data,
            debug_str=util.dumpObjAttr(view_data)
        )


###################
# helpers

# TODO: need a way to correctly parse titles like: "2001: A Space Odyssey (1968)"
#   with legitimate dates in the title
def parseQueryStr(query_obj):
    if not query_obj.raw_query_str:
        query_obj.errors.append('%s: raw query string is empty' % util.getFunctionName())
        return False

    s = query_obj.raw_query_str.replace('.', ' ').strip()
    year = None
    current_year = date.today().year

    # NOTE: find all 4 digit numbers to match the year of queries with 4 digit numbers in the title
    #   ex) "2036.Origin.Unknown.2018"
    regex = re.compile('[0-9]{4}')

    for m in regex.finditer(s):
        i = int(m[0])
        if i > 1878 and i <= current_year:
            query_obj.parsed_query_str = s[0:m.start()].strip() # NOTE: slice query string to year match
            query_obj.movie_year = i
            query_obj.query_type = query_type.MOVIE
            break

    # TODO: parse tv show titles
    if query_obj.query_type == query_type.MOVIE:
        return True
    else:
        query_obj.errors.append('%s: unable to determine query type' % util.getFunctionName())
        return False

def makeQuery(query_obj):
    if query_obj.query_type == query_type.MOVIE:
        search_part = TMDB_SEARCH_MOVIE
    elif query_obj.query_type == query_type.TV_SHOW:
        search_part = TMDB_SEARCH_TV
    else:
        query_obj.errors.append('%s(), query type is not set' % util.getFunctionName())
        return False

    request_headers = {"Accept": "text/html,application/xhtml+xml,application/xml"}
    query_obj.query_params = urllib.parse.urlencode(
            {
                'api_key': query_obj.tmdb_api_key,
                'query': query_obj.parsed_query_str,
                'year': query_obj.movie_year,
            }
        )

    query_obj.full_url = TMDB_URL + search_part + '?' + query_obj.query_params
    context = ssl.create_default_context()
    req = urllib.request.Request(query_obj.full_url, None, request_headers)
    ct_header = ''

    try:
        response = urllib.request.urlopen(req, context=context)
    except HTTPError as e:
        query_obj.errors.append(e)
        return False
    except URLError as e:
        query_obj.errors.append(e)
        return False

    for k, v in response.info().items():
        query_obj.response_headers.append((k, v))
        if k == 'Content-Type':
            ct_header = str(k + ':' + v)

    query_obj.response_status = response.status

    if 'charset' in ct_header:
        charset = cgi.parse_header(ct_header)[1]['charset']
        query_obj.response_body = response.read().decode(charset)
    else:
        query_obj.errors.append('%s(), no charset key in ct_header' % util.getFunctionName())
        query_obj.response_body = response.read()
        return False

    return True