import os import json import tqdm import re from clean import clean authors = { 'Alice': 'Alice Margatroid', 'Hourai': 'Hourai', 'Shanghai': 'Shanghai', 'Marisa': 'Marisa Kirisame', 'Patchouli': 'Patchouli Knowledge', 'Yuuka': 'Yuuka Kazami', 'Cirno': 'Cirno', 'Remilia': 'Remilia Scarlet', 'Koakuma': 'Koakuma', 'Medicine': 'Medicine Melancholy', 'Shinki': 'Shinki', '???': '???', 'Flandre': 'Flandre Scarlet', 'Rumia': 'Rumia', 'Meiling': 'Hong Meiling', 'Sakuya': 'Sakuya Izayoi', 'Reimu': 'Reimu Hakurei', 'Yukari': 'Yukari Yakumo', 'Chen': 'Chen', 'Sanae': 'Sanae Kochiya', 'Ran': 'Ran Yakumo', 'Letty': 'Letty Whiterock', 'Lyrica': 'Lyrica Prismriver', 'Lunasa': 'Lunasa Prismriver', 'Merlin': 'Merlin Prismriver', 'Sisters': 'Lyrica, Lunasa, and Merlin', 'Youmu': 'Youmu Konpaku', 'Yuyuko': 'Yuyuko Saigyouji', 'Marisa\nYuyuko': 'Marisa and Yuyuko', 'Reimu\nYuyuko': 'Reimu and Yuyuko', 'Sakuya\nYuyuko': 'Sakuya and Yuyuko', 'Keine': 'Keine Kamishirasawa', 'Kaguya': 'Kaguya Houraisan', 'Mokou': 'Fujiwara no Mokou', 'Wriggle': 'Wriggle Nightbug', 'Mystia': 'Mystia Lorelei', 'Reisen': 'Reisen Udongein Inaba', 'Eirin': 'Eirin Yagokoro', '\xa0': '???', 'Aya': 'Aya Shameimaru', 'Tewi': 'Tewi Inaba', 'Komachi': 'Komachi Onozuka', 'Eiki': 'Eiki Shiki', 'Kanako': 'Kanako Yasaka', 'Suwako': 'Suwako Moriya', 'Minoriko': 'Minoriko Aki', 'Hina': 'Hina Kagiyama', 'Nitori': 'Nitori Kawashiro', 'Iku': 'Iku Nagae', 'Tenshi': 'Tenshi Hinanawi', 'Suika': 'Suika Ibuki', 'Koishi': 'Koishi Komeiji', 'Satori': 'Satori Komeiji', 'Yamame': 'Yamame Kurodani', 'Parsee': 'Parsee Mizuhashi', 'Yuugi': 'Yuugi Hoshiguma', 'Rin': 'Rin Kaenbyou', 'Utsuho': 'Utsuho Reiuji', 'Kogasa': 'Kogasa Tatara', 'Nue': 'Nue Houjuu', 'Nazrin': 'Nazrin', 'Ichirin': 'Ichirin Kumoi', 'Minamitsu': 'Minamitsu Murasa', 'Shou': 'Shou Toramaru', 'Byakuren': 'Byakuren Hijiri', 'Master Big Catfish': 'Master Big Catfish', 'Hatate': 'Hatate Himekaidou', 'Luna': 'Luna Child', 'Star': 'Star Sapphire', 'Sunny': 'Sunny Milk', 'Mamizou': 'Mamizou Futatsuiwa', 'Kyouko': 'Kyouko Kasodani', 'Yoshika': 'Yoshika Miyako', 'Seiga': 'Seiga Kaku', 'Futo': 'Mononobe no Futo', 'Miko': 'Toyosatomimi no Miko', 'Hijiri': 'Byakuren Hijiri', 'Kokoro': 'Hata no Kokoro', '<miko2>': 'Toyosatomimi no Miko\'s Clone', 'Benben': 'Benben Tsukumo', 'Yatsuhashi': 'Yatsuhashi Tsukumo', 'Raiko': 'Raiko Horikawa', 'Wakasagihime': 'Wakasagihime', 'Sekibanki': 'Sekibanki', 'Kagerou': 'Kagerou Imaizumi', 'Seija': 'Seija Kijin', 'Shinmyoumaru': 'Shinmyoumaru', 'Kasen': 'Kasen Ibaraki', 'Usami': 'Sumireko Usami', 'Doremy': 'Doremy Sweet', 'Junko': 'Junko', 'Hecatia': 'Hecatia Lapislazuli', 'Seiran': 'Seiran', 'Ringo': 'Ringo', 'Sagume': 'Sagume Kishin', 'Clownpiece': 'Clownpiece', 'Udonge': 'Reisen Udongein Inaba', 'Joon': 'Joon Yorigami', 'Shion': 'Shion Yorigami', 'Joon\nShion': 'Joon and Shion', 'Usami2': 'Sumireko Usami\'s Clone', 'Usami3': 'Sumireko Usami\'s Other Clone', 'Shinmyoumaru2': 'Shinmyoumaru\'s Clone', 'Mai': 'Mai', 'Satono': 'Satono Nishida', 'Okina': 'Okina Matara', 'Eternity': 'Eternity Larva', 'Nemuno': 'Nemuno Sakata', 'Aunn': 'Aunn Komano', 'Narumi': 'Narumi Yatadera', 'Sumireko': 'Sumireko Usami', 'Kutaka': 'Kutaka Niwatari', 'Marisa(Wolf)': 'Marisa Kirisame', 'Saki': 'Saki Kurokuma', 'Eika': 'Eika Ebisu', 'Urumi': 'Urumi Ushizaki', 'Yachie': 'Yachie Kicchou', 'Mayumi': 'Mayumi Joutouguu', 'Keiki': 'Keiki Haniyasushin', 'Marisa(Otter)': 'Marisa Kirisame', 'Marisa(Eagle)': 'Marisa Kirisame', 'Reimu(Wolf)': 'Reimu Hakurei', 'Reimu(Otter)': 'Reimu Hakurei', 'Reimu(Eagle)': 'Reimu Hakurei', 'Youmu(Wolf)': 'Youmu Konpaku', 'Youmu(Otter)': 'Youmu Konpaku', 'Youmu(Eagle)': 'Youmu Konpaku', 'Tsukasa': 'Tsukasa Kudamaki', 'Momoyo': 'Momoyo Himemushi', 'Misumaru': 'Misumaru Tamatsukuri', 'Mike': 'Mike Goutokuji', 'Takane': 'Takane Yamashiro', 'Sannyo': 'Sannyo Komakusa', 'Megumu': 'Megumu Iizunamaru', 'Chimata': 'Chimata Tenkyuu', } # regex to remove strings like <balloon$a11x3> exp = re.compile(r'<.*?>') def worker_parse_hourai(filename, out_filename=None): # parse csv file messages = [] with open(filename, 'r', encoding='utf-8') as f: data = f.read() lines = data.split('\n') for line in lines: if line.startswith('name,line'): continue line = line.split(',') if len(line) < 2: continue author = line[0] text = clean(line[1].replace('\"', '')) messages.append(f'{authors[author]}: {text}') # write to file with open(out_filename, 'w', encoding='utf-8') as f: txt = '\n'.join(messages) f.write('⁂\n[Title: Hourai; Genre: Touhou Project]\n⁂\n') f.write(txt) return len(messages) def worker_parse_dialogue(filename, out_filename=None): # open json messages = [] with open(filename, 'r', encoding='utf-8') as f: data = json.load(f) for message in data: # first, check if author is in dict values author = message['Name'] # iterate over values for key, value in authors.items(): if value == author: author = key break author = authors[author] text = clean(message['Text']).replace(' ?"@', '').replace('&', '').replace(' ', '').replace('\ ', '').replace('\. ', '\n') for key, value in authors.items(): text = text.replace(key, value) text = exp.sub('', text) messages.append(f'{author}: {text}') # write to file with open(out_filename, 'w', encoding='utf-8') as f: txt = '\n'.join(messages) f.write('⁂\n[Title: The Touhou Project]\n⁂\n') f.write(txt.replace('\n\n', '\n')) return len(messages) def dump_stats(length, name): stats = {'misc': {}} if os.path.exists('stats.json'): stats = json.load(open('stats.json', 'r', encoding='utf-8')) if 'misc' not in stats: stats['misc'] = {} stats['misc'][name] = length with open('stats.json', 'w', encoding='utf-8') as f: json.dump(stats, f) def parse(): if not os.path.exists('./data/misc'): os.mkdir('./data/misc') dump_stats(worker_parse_hourai('./raw/misc/lines.csv', './data/misc/lines.txt'), 'hourai') dump_stats(worker_parse_dialogue('./raw/misc/dialogue.json', './data/misc/dialogue.txt'), 'dialogue') if __name__ == '__main__': parse()