You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Plex-Meta-Manager/modules/imdb.py

130 lines
5.8 KiB

import logging, math, re, time
4 years ago
from modules import util
from modules.util import Failed
from urllib.parse import urlparse, parse_qs
4 years ago
logger = logging.getLogger("Plex Meta Manager")
builders = ["imdb_list", "imdb_id"]
base_url = "https://www.imdb.com"
urls = {
"list": f"{base_url}/list/ls",
"search": f"{base_url}/search/title/",
"keyword": f"{base_url}/search/keyword/"
}
xpath = {
"imdb_id": "//div[contains(@class, 'lister-item-image')]//a/img//@data-tconst",
"list": "//div[@class='desc lister-total-num-results']/text()",
"search": "//div[@class='desc']/span/text()",
"keyword": "//div[@class='desc']/text()"
}
item_counts = {"list": 100, "search": 250, "keyword": 50}
3 years ago
class IMDb:
4 years ago
def __init__(self, config):
self.config = config
4 years ago
3 years ago
def validate_imdb_lists(self, imdb_lists, language):
valid_lists = []
for imdb_dict in util.get_list(imdb_lists, split=False):
if not isinstance(imdb_dict, dict):
imdb_dict = {"url": imdb_dict}
dict_methods = {dm.lower(): dm for dm in imdb_dict}
imdb_url = util.parse("url", imdb_dict, methods=dict_methods, parent="imdb_list").strip()
if not imdb_url.startswith((urls["list"], urls["search"], urls["keyword"])):
raise Failed(f"IMDb Error: {imdb_url} must begin with either:\n{urls['list']} (For Lists)\n{urls['search']} (For Searches)\n{urls['keyword']} (For Keyword Searches)")
self._total(imdb_url, language)
list_count = util.parse("limit", imdb_dict, datatype="int", methods=dict_methods, default=0, parent="imdb_list", minimum=0) if "limit" in dict_methods else 0
3 years ago
valid_lists.append({"url": imdb_url, "limit": list_count})
return valid_lists
4 years ago
def _total(self, imdb_url, language):
headers = util.header(language)
if imdb_url.startswith(urls["keyword"]):
page_type = "keyword"
elif imdb_url.startswith(urls["list"]):
page_type = "list"
else:
page_type = "search"
results = self.config.get_html(imdb_url, headers=headers).xpath(xpath[page_type])
total = 0
for result in results:
if "title" in result:
try:
total = int(re.findall("(\\d+) title", result.replace(",", ""))[0])
break
except IndexError:
pass
if total > 0:
return total, item_counts[page_type]
raise ValueError(f"IMDb Error: Failed to parse URL: {imdb_url}")
4 years ago
4 years ago
def _ids_from_url(self, imdb_url, language, limit):
total, item_count = self._total(imdb_url, language)
headers = util.header(language)
4 years ago
imdb_ids = []
parsed_url = urlparse(imdb_url)
params = parse_qs(parsed_url.query)
imdb_base = parsed_url._replace(query=None).geturl()
params.pop("start", None)
params.pop("count", None)
params.pop("page", None)
if limit < 1 or total < limit:
limit = total
remainder = limit % item_count
if remainder == 0:
remainder = item_count
num_of_pages = math.ceil(int(limit) / item_count)
4 years ago
for i in range(1, num_of_pages + 1):
start_num = (i - 1) * item_count + 1
util.print_return(f"Parsing Page {i}/{num_of_pages} {start_num}-{limit if i == num_of_pages else i * item_count}")
if imdb_base.startswith((urls["list"], urls["keyword"])):
params["page"] = i
else:
params["count"] = remainder if i == num_of_pages else item_count
params["start"] = start_num
ids_found = self.config.get_html(imdb_base, headers=headers, params=params).xpath(xpath["imdb_id"])
if imdb_base.startswith((urls["list"], urls["keyword"])) and i == num_of_pages:
ids_found = ids_found[:remainder]
imdb_ids.extend(ids_found)
time.sleep(2)
util.print_end()
if len(imdb_ids) > 0:
logger.debug(f"{len(imdb_ids)} IMDb IDs Found: {imdb_ids}")
return imdb_ids
raise ValueError(f"IMDb Error: No IMDb IDs Found at {imdb_url}")
4 years ago
def get_items(self, method, data, language, is_movie):
4 years ago
show_ids = []
movie_ids = []
3 years ago
fail_ids = []
def run_convert(imdb_id):
tvdb_id = self.config.Convert.imdb_to_tvdb(imdb_id) if not is_movie else None
tmdb_id = self.config.Convert.imdb_to_tmdb(imdb_id) if tvdb_id is None else None
4 years ago
if tmdb_id: movie_ids.append(tmdb_id)
3 years ago
elif tvdb_id: show_ids.append(tvdb_id)
else:
logger.error(f"Convert Error: No {'' if is_movie else 'TVDb ID or '}TMDb ID found for IMDb: {imdb_id}")
fail_ids.append(imdb_id)
if method == "imdb_id":
3 years ago
logger.info(f"Processing IMDb ID: {data}")
run_convert(data)
4 years ago
elif method == "imdb_list":
status = f"{data['limit']} Items at " if data['limit'] > 0 else ''
3 years ago
logger.info(f"Processing IMDb List: {status}{data['url']}")
4 years ago
imdb_ids = self._ids_from_url(data["url"], language, data["limit"])
4 years ago
total_ids = len(imdb_ids)
for i, imdb in enumerate(imdb_ids, 1):
util.print_return(f"Converting IMDb ID {i}/{total_ids}")
run_convert(imdb)
logger.info(util.adjust_space(f"Processed {total_ids} IMDb IDs"))
4 years ago
else:
raise Failed(f"IMDb Error: Method {method} not supported")
logger.debug("")
3 years ago
logger.debug(f"{len(fail_ids)} IMDb IDs Failed to Convert: {fail_ids}")
logger.debug(f"{len(movie_ids)} TMDb IDs Found: {movie_ids}")
logger.debug(f"{len(show_ids)} TVDb IDs Found: {show_ids}")
4 years ago
return movie_ids, show_ids