mirror of
https://github.com/Xevion/the-office.git
synced 2025-12-06 15:16:45 -06:00
add preprocess option to build final|algolia commands, make get_characters no file return None instead of empty list, README.md credit link text fix
This commit is contained in:
@@ -23,7 +23,7 @@ A Vue.js and Flask Web Application designed to provide a quick way to search for
|
|||||||
|
|
||||||
### Credit
|
### Credit
|
||||||
|
|
||||||
Credit to [officequotes.net/](https://www.officequotes.net/) for providing all quote data.
|
Credit to [officequotes.net](https://www.officequotes.net/) for providing all quote data.
|
||||||
|
|
||||||
Credit to [imdb.com](https://www.imdb.com/title/tt0386676/) for episode descriptions.
|
Credit to [imdb.com](https://www.imdb.com/title/tt0386676/) for episode descriptions.
|
||||||
|
|
||||||
|
|||||||
@@ -8,7 +8,7 @@ import os
|
|||||||
import re
|
import re
|
||||||
import sys
|
import sys
|
||||||
import time
|
import time
|
||||||
from typing import List, Tuple, Union
|
from typing import List, Optional, Tuple, Union
|
||||||
|
|
||||||
import click
|
import click
|
||||||
import enlighten
|
import enlighten
|
||||||
@@ -177,15 +177,15 @@ def preprocess(season: int, episode: int, all: bool, overwrite: bool, silent_ski
|
|||||||
@click.option('-s', '--season', type=int,
|
@click.option('-s', '--season', type=int,
|
||||||
help='Season to be fetched. Without --episode, will download all episodes in a season.')
|
help='Season to be fetched. Without --episode, will download all episodes in a season.')
|
||||||
@click.option('-e', '--episode', type=int, help='Specific episode to be fetched. Requires --season to be specified.')
|
@click.option('-e', '--episode', type=int, help='Specific episode to be fetched. Requires --season to be specified.')
|
||||||
@click.option('--all', is_flag=True, help='Fetch all episodes, regardless of previous specifications.')
|
@click.option('--all', 'all_', is_flag=True, help='Fetch all episodes, regardless of previous specifications.')
|
||||||
@click.option('-r', '--report', is_flag=True, help='Report quote statistics once processing completed')
|
@click.option('-r', '--report', is_flag=True, help='Report quote statistics once processing completed.')
|
||||||
def process(season: int, episode: int, all: bool, report: bool):
|
def process(season: Optional[int], episode: Optional[int], all_: bool, report: bool):
|
||||||
"""
|
"""
|
||||||
Processes manually processed raw quote data into JSON.
|
Processes manually processed raw quote data into JSON.
|
||||||
"""
|
"""
|
||||||
episodes: List[Tuple[int, int]]
|
episodes: List[Tuple[int, int]]
|
||||||
|
|
||||||
if all:
|
if all_:
|
||||||
episodes = list(get_episodes())
|
episodes = list(get_episodes())
|
||||||
elif season:
|
elif season:
|
||||||
if episode:
|
if episode:
|
||||||
@@ -231,8 +231,10 @@ def process(season: int, episode: int, all: bool, report: bool):
|
|||||||
sections.append(section)
|
sections.append(section)
|
||||||
except FileNotFoundError:
|
except FileNotFoundError:
|
||||||
logger.info(f'Skipped Season {_season}, Episode {_episode}, no file found.')
|
logger.info(f'Skipped Season {_season}, Episode {_episode}, no file found.')
|
||||||
|
continue
|
||||||
except:
|
except:
|
||||||
logger.exception(f'Skipped Season {_season}, Episode {_episode}: Malformed data.')
|
logger.exception(f'Skipped Season {_season}, Episode {_episode}: Malformed data.')
|
||||||
|
if quote:
|
||||||
logger.info(
|
logger.info(
|
||||||
f'Last quote seen "{quote if type(quote) is str else "|".join(quote)}" in section {section_num}')
|
f'Last quote seen "{quote if type(quote) is str else "|".join(quote)}" in section {section_num}')
|
||||||
else:
|
else:
|
||||||
@@ -268,11 +270,18 @@ def build():
|
|||||||
|
|
||||||
@build.command('algolia')
|
@build.command('algolia')
|
||||||
@click.option('-ss', '--silent-skip', is_flag=True, help='Skip existing files silently')
|
@click.option('-ss', '--silent-skip', is_flag=True, help='Skip existing files silently')
|
||||||
@click.option('--process', is_flag=True, help='Run processing before building final data.')
|
@click.option('--process', 'process_', is_flag=True, help='Run processing before building final data.')
|
||||||
def algolia(silent_skip: bool):
|
def algolia(silent_skip: bool, process_: bool):
|
||||||
"""
|
"""
|
||||||
Generates algolia.json, a all encompassing file for Algolia's search index.
|
Generates algolia.json, a all encompassing file for Algolia's search index.
|
||||||
"""
|
"""
|
||||||
|
if process_:
|
||||||
|
logger.info('Processing before building algolia.json')
|
||||||
|
try:
|
||||||
|
process(["--all", '--silent'])
|
||||||
|
except:
|
||||||
|
pass
|
||||||
|
|
||||||
data = []
|
data = []
|
||||||
episode_num_abs, section_num_abs, quote_num_abs = 0, 0, 0
|
episode_num_abs, section_num_abs, quote_num_abs = 0, 0, 0
|
||||||
for season, episode in get_episodes():
|
for season, episode in get_episodes():
|
||||||
@@ -310,11 +319,19 @@ def algolia(silent_skip: bool):
|
|||||||
|
|
||||||
@build.command('final')
|
@build.command('final')
|
||||||
@click.option('-ss', '--silent-skip', is_flag=True, help='Skip existing files silently')
|
@click.option('-ss', '--silent-skip', is_flag=True, help='Skip existing files silently')
|
||||||
@click.option('--process', is_flag=True, help='Run processing before building final data.')
|
@click.option('--process', 'process_', is_flag=True, help='Run processing before building final data.')
|
||||||
def final(silent_skip: bool):
|
def final(silent_skip: bool, process_: bool):
|
||||||
"""Generates the latest application static data.json file, used by the backend API."""
|
"""Generates the latest application static data.json file, used by the backend API."""
|
||||||
descriptions = load_file(os.path.join(DATA_DIR, 'descriptions.json'), True)
|
descriptions = load_file(os.path.join(DATA_DIR, 'descriptions.json'), True)
|
||||||
seasons = [{'season_id': season, 'episodes': []} for season in range(1, 10)]
|
seasons = [{'season_id': season, 'episodes': []} for season in range(1, 10)]
|
||||||
|
|
||||||
|
if process_:
|
||||||
|
logger.info('Processing before building final.json')
|
||||||
|
try:
|
||||||
|
process(["--all"])
|
||||||
|
except:
|
||||||
|
pass
|
||||||
|
|
||||||
for season_id, episode_id in get_episodes():
|
for season_id, episode_id in get_episodes():
|
||||||
# Load data file
|
# Load data file
|
||||||
try:
|
try:
|
||||||
|
|||||||
File diff suppressed because it is too large
Load Diff
@@ -9,7 +9,7 @@ import os
|
|||||||
import time
|
import time
|
||||||
from collections import defaultdict
|
from collections import defaultdict
|
||||||
from math import ceil
|
from math import ceil
|
||||||
from typing import Dict, Iterable, List, Tuple, Union
|
from typing import Dict, Iterable, List, Optional, Tuple, Union
|
||||||
|
|
||||||
import enlighten
|
import enlighten
|
||||||
import requests
|
import requests
|
||||||
@@ -103,14 +103,14 @@ def sleep_from(wait_time: float, moment: float, manager: enlighten.Manager = Non
|
|||||||
return 0
|
return 0
|
||||||
|
|
||||||
|
|
||||||
def get_characters(season, episode) -> List[Dict[str, Union[int, str]]]:
|
def get_characters(season, episode) -> Optional[List[Dict[str, Union[int, str]]]]:
|
||||||
"""
|
"""
|
||||||
Extracts all characters and their number of appearances from a specific episode.
|
Extracts all characters and their number of appearances from a specific episode.
|
||||||
Prepared in a list of dictionary, preferable storage/for loop method.
|
Prepared in a list of dictionary, preferable storage/for loop method.
|
||||||
"""
|
"""
|
||||||
filepath = get_filepath(season, episode, 'processed')
|
filepath = get_filepath(season, episode, 'processed')
|
||||||
if not os.path.exists(filepath):
|
if not os.path.exists(filepath):
|
||||||
return []
|
return
|
||||||
scenes = load_file(filepath, True)
|
scenes = load_file(filepath, True)
|
||||||
|
|
||||||
characters = defaultdict(int)
|
characters = defaultdict(int)
|
||||||
|
|||||||
Reference in New Issue
Block a user