Add script to regenerate side identifiers

This commit is contained in:
Yohann D'ANELLO 2020-09-28 10:51:21 +02:00
parent be76bf4857
commit 9ecd876923
3 changed files with 121 additions and 70 deletions

View File

@ -1,5 +1,5 @@
# -*- mode: python; coding: utf-8 -*-
# Copyright (C) 2017-2019 by BDE ENS Paris-Saclay
# Copyright (C) 2017-2020 by BDE ENS Paris-Saclay
# SPDX-License-Identifier: GPL-3.0-or-later
import json
@ -7,6 +7,7 @@ import re
import unicodedata
import urllib.request
from django.db.models import QuerySet
from django.forms import ModelForm
from django.utils.translation import gettext_lazy as _
@ -14,6 +15,60 @@ from .models import Auteur, BD
from .scraper import BedetequeScraper
def generate_side_identifier(title, authors, subtitle=None):
if isinstance(authors, QuerySet):
authors = list(authors)
title_normalized = title.upper()
title_normalized = title_normalized.replace('', '\'')
title_normalized = ''.join(
char
for char in unicodedata.normalize(
'NFKD', title_normalized.casefold())
if all(not unicodedata.category(char).startswith(cat)
for cat in {'M', 'P', 'Z', 'C'}) or char == ' '
).casefold().upper()
title_normalized = re.sub(r'^DE ', '', title_normalized)
title_normalized = re.sub(r'^LE ', '', title_normalized)
title_normalized = re.sub(r'^LA ', '', title_normalized)
title_normalized = re.sub(r'^LES ', '', title_normalized)
title_normalized = re.sub(r'^L\'', '', title_normalized)
title_normalized = re.sub(r'^THE ', '', title_normalized)
title_normalized = re.sub(r'Œ', 'OE', title_normalized)
title_normalized = title_normalized.replace(' ', '')
title_normalized = re.sub("[^A-Z0-9$]", "", title_normalized)
authors = authors.copy()
def sort(author):
return str(-author.note) + author.name.split(" ")[-1] + ".{:05d}".format(author.pk)
authors.sort(key=sort)
primary_author = authors[0]
author_name = primary_author.name.upper()
if ',' not in author_name and ' ' in author_name:
author_name = author_name.split(' ')[-1]
author_name = ''.join(
char for char in unicodedata.normalize('NFKD', author_name.casefold())
if all(not unicodedata.category(char).startswith(cat) for cat in {'M', 'P', 'Z', 'C'}) or char == ' '
).casefold().upper()
author_name = re.sub("[^A-Z]", "", author_name)
side_identifier = "{:.3} {:.3}".format(author_name, title_normalized, )
if subtitle:
subtitle = re.sub(r'</span>', '', subtitle)
subtitle = re.sub(r'<span.*>', '', subtitle)
start = subtitle.split(' ')[0].replace('.', '')
if start.isnumeric():
side_identifier += " {:0>2}".format(start, )
# Normalize side identifier, in order to remove accents
side_identifier = ''.join(char for char in unicodedata.normalize('NFKD', side_identifier.casefold())
if all(not unicodedata.category(char).startswith(cat) for cat in {'M', 'P', 'Z', 'C'})
or char == ' ').casefold().upper()
return side_identifier
class MediaAdminForm(ModelForm):
def __init__(self, *args, **kwargs):
super().__init__(*args, **kwargs)
@ -188,54 +243,12 @@ class MediaAdminForm(ModelForm):
)
if self.cleaned_data['authors']:
authors = self.cleaned_data['authors']
old_authors = authors.copy()
def sort(author):
return str(-author.note) + "." \
+ str(old_authors.index(author)) \
+ "." + author.name
authors.sort(key=sort)
author_name = self.cleaned_data['authors'][0].name
if ',' not in author_name and ' ' in author_name:
author_name = author_name.split(' ')[-1]
title_normalized = self.cleaned_data['title'].upper()
title_normalized = re.sub(r'^LE ', '', title_normalized)
title_normalized = re.sub(r'^LA ', '', title_normalized)
title_normalized = re.sub(r'^LES ', '', title_normalized)
title_normalized = re.sub(r'^L\'', '', title_normalized)
title_normalized = re.sub(r'^THE ', '', title_normalized)
title_normalized = re.sub(r'Œ', 'OE', title_normalized)
side_identifier = "{:.3} {:.3}".format(
author_name,
title_normalized.replace(' ', ''), )
if self.cleaned_data['subtitle']:
self.cleaned_data['subtitle'] = re.sub(
r'</span>',
'',
self.cleaned_data['subtitle']
side_identifier = generate_side_identifier(
self.cleaned_data["title"],
self.cleaned_data["authors"],
self.cleaned_data["subtitle"],
)
self.cleaned_data['subtitle'] = re.sub(
r'<span.*>',
'',
self.cleaned_data['subtitle']
)
start = self.cleaned_data['subtitle'].split(' ')[0] \
.replace('.', '')
if start.isnumeric():
side_identifier += " {:0>2}".format(start, )
# Normalize side identifier, in order to remove accents
side_identifier = ''.join(
char
for char in unicodedata.normalize(
'NFKD', side_identifier.casefold())
if all(not unicodedata.category(char).startswith(cat)
for cat in {'M', 'P', 'Z', 'C'}) or char == ' '
).casefold().upper()
self.cleaned_data['side_identifier'] = side_identifier
return self.cleaned_data

View File

@ -1,10 +1,10 @@
import re
import unicodedata
from argparse import FileType
from sys import stdin
from django.core.management import BaseCommand
from media.models import Auteur, Roman
from media.forms import generate_side_identifier
from media.models import Roman, Auteur
class Command(BaseCommand):
@ -29,27 +29,9 @@ class Command(BaseCommand):
continue
title = book[1]
title_normalized = title.upper()
title_normalized = title_normalized.replace('', '\'')
title_normalized = ''.join(
char
for char in unicodedata.normalize(
'NFKD', title_normalized.casefold())
if all(not unicodedata.category(char).startswith(cat)
for cat in {'M', 'P', 'Z', 'C'}) or char == ' '
).casefold().upper()
title_normalized = re.sub(r'^DE ', '', title_normalized)
title_normalized = re.sub(r'^LE ', '', title_normalized)
title_normalized = re.sub(r'^LA ', '', title_normalized)
title_normalized = re.sub(r'^LES ', '', title_normalized)
title_normalized = re.sub(r'^L\'', '', title_normalized)
title_normalized = re.sub(r'^THE ', '', title_normalized)
title_normalized = re.sub(r'Œ', 'OE', title_normalized)
title_normalized = title_normalized.replace(' ', '')
authors = [Auteur.objects.get_or_create(name=n)[0]
for n in book[0].split(';')]
side_identifier = "{:.3} {:.3}" \
.format(authors[0].name.upper(), title_normalized, )
side_identifier = generate_side_identifier(title, authors)
roman = Roman.objects.create(
title=title,
side_identifier=side_identifier,

View File

@ -0,0 +1,56 @@
from django.core.management import BaseCommand
from django.db import transaction
from media.forms import generate_side_identifier
from media.models import BD, Manga, Roman
class Command(BaseCommand):
def add_arguments(self, parser):
parser.add_argument('--type', '-t',
type=str,
default='bd',
choices=['bd', 'manga', 'roman'],
help="Type of medium where the sides need to be regenerated.")
parser.add_argument('--noninteractivemode', '-n', action="store_true",
help="Disable the interaction mode and replace existing side identifiers.")
@transaction.atomic
def handle(self, *args, **options):
t = options["type"]
medium_class = None
if t == "bd":
medium_class = BD
elif t == "manga":
medium_class = Manga
elif t == "roman":
medium_class = Roman
interactive_mode = not options["noninteractivemode"]
replaced = 0
for obj in medium_class.objects.all():
current_side_identifier = obj.side_identifier
if not obj.authors.all():
self.stdout.write(str(obj))
subtitle = obj.subtitle if hasattr(obj, "subtitle") else None
generated_side_identifier = generate_side_identifier(obj.title, obj.authors.all(), subtitle)
if current_side_identifier != generated_side_identifier:
answer = 'y'
if interactive_mode:
answer = ''
while answer != 'y' and answer != 'n':
answer = input(f"For medium {obj}, current side: {current_side_identifier}, generated side: "
f"{generated_side_identifier}, would you like to replace ? [y/n]").lower()[0]
if answer == 'y':
self.stdout.write(self.style.WARNING(f"Replace side of {obj} from {current_side_identifier} "
f"to {generated_side_identifier}..."))
obj.side_identifier = generated_side_identifier
obj.save()
replaced += 1
if replaced:
self.stdout.write(self.style.SUCCESS(f"{replaced} side identifiers were replaced."))
else:
self.stdout.write(self.style.WARNING("Nothing changed."))