#!/usr/bin/python # -*- coding: utf-8 -*- # --------------------------------------------------------------------------- # # Isida-NG Jabber Bot (fork of iSida) # # Original iSida Copyright (C) 2009-2015 diSabler # # Isida-NG Copyright (C) 2026 Luciferus # # Project home: https://git.ubunix.pro/Luciferus/isida-ng # # --------------------------------------------------------------------------- # import re import os turn_base = [] # --- ЗАГРУЗКА СЛОВАРЯ ИСКЛЮЧЕНИЙ --- IGNORE_FILE = data_folder % 'ignore_turn.txt' _IGNORE_WORDS = set() # --- ЗАГРУЗКА РУССКИХ БИГРАММ/ТРИГРАММ (КАК PYTHON-КОД) --- RUGRAMM_FILE = data_folder % 'rugramm.txt' _RUGRAMM = set() # --- ПРЕДЛОГИ, АРТИКЛИ, МЕЖДОМЕТИЯ --- STOP_WORDS = { 'a', 'an', 'the', 'and', 'or', 'but', 'for', 'nor', 'on', 'at', 'to', 'by', 'in', 'of', 'off', 'over', 'under', 'up', 'down', 'into', 'through', 'across', 'between', 'among', 'without', 'within', 'upon', 'toward', 'until', 'since', 'after', 'before', 'about', 'against', 'along', 'around', 'behind', 'below', 'beneath', 'beside', 'beyond', 'during', 'except', 'from', 'like', 'near', 'onto', 'opposite', 'outside', 'per', 'plus', 'round', 'save', 'since', 'than', 'till', 'unto', 'via', 'with', 'so', 'as', 'than', 'that', 'these', 'those', 'some', 'any', 'no', 'all', 'each', 'every', 'both', 'neither', 'either', 'then', 'now', 'just', 'only', 'very', 'too', 'also', 'well', 'oh', 'ah', 'ok', 'okay', 'yes', 'yeah', 'no', 'nope', 'maybe', 'perhaps', 'probably', 'definitely', 'certainly', 'rain', '*' } # --- АНТИ-БИГРАММЫ (жёсткие) --- ANTI_BIGRAMS = { 'yfg', 'hbvth', 'pf,sd', 'ftim', 'crbq', 'zpsr', 'кфт', 'вщь', 'цщ', 'квы', 'gbitim', 'gbit', 'yfghb', 'djn', 'nfr', 'lfkmit', 'ghjcnj', 'sdftim', 'gthtrk', 'xbnmcz', 'heccrbq', 'zpsr', 'yfghbvth', 'ghb', 'gtht', 'gtrk', 'xbn', 'cz', 'hecc', 'rbq', 'zps', 'yfg', 'hbv', 'th', 'zpsr', 'pf,sd', 'zxc', 'qwe', 'asd', 'rty', 'fgh', 'vbn', 'uio', 'jkl', 'xcv', 'bnm', 'егы', 'шщз', 'щзх', 'зхъ', 'фыв', 'джэ', 'ячс', 'бюё', 'nmcz', 'k.xb', 'gthtr', 'кфт', 'вщь', 'цщ', 'квы', 'pfrjy', 'xbkcz', 'kbvbn', ',scnh', 'jltq', 'cndbt', 'pf,', 'sdft', 'im?', 'cvtyb', 'nm&', 'heccr', 'ccrbq', 'vexf', 'xf.cm', '.cm', 'cj cv', 'tifyy', 'yysv', 'utyf', 'kexituj', ',tp', 'tot', 'to`', ',kz', 'егктб', 'Utyf-', ',elt', 'j,y', 'fhe;', 'e;bk', 'ckexf', 'xfqyj?', 'elfkz', 'kznm?' } # --- ВСТРОЕННЫЕ СЛОВАРИ (НЕ ЗАВИСЯТ ОТ ВНЕШНИХ ФАЙЛОВ) --- # Частотные английские слова: если слово есть здесь — это настоящее # английское слово, и его НЕ надо перекладывать в русскую раскладку. ENGLISH_WORDS = frozenset(""" the be to of and a in that have i it for not on with he as you do at this but his by from they we say her she or an will my one all would there their what so up out if about who get which go me when make can like time no just him know take people into year your good some could them see other than then now look only come its over think also back after use two how our work first well way even new want because any these give day most us was are had has were been being is am hello world morning evening night today tomorrow yesterday thank thanks please help sorry answer question test message maybe perhaps probably definitely certainly hats hat cat dog house home car computer phone internet chat chatroom link email mail code program programming language python jabber bot user name nick username password pass login signup sign login account create delete add remove edit change settings config configuration command commands botowner admin owner moder moderator help me you him her them us we they you're she's he's it's that's let's don't can't won't isn't aren't wasn't weren't didn't doesn't could'll should would may might must need shall ought used ought get go came going goes gone takes took made make let put keep hold find found look looking see seen show shown tell told say said speak spoke talk asks asked answer answer ask want wanted like liked love hate play played playing game keyboard layout letter word words sentence character number numbers figure figures correct wrong write wrote writing read reading know known knew remember forget think thinking thought feel feeling think line work works working job money time hour minute second clock day week month year season weather rain snow heat cold warm cool sun moon star sky cloud wind road street city town village country world map street way path route man woman child children friend friends family mother father sister brother house home room kitchen bed table chair door window wall floor ground thing things stuff place places problem problems question questions whole every almost between near behind beside beyond under above across through without within about against along around except from into off on to whether whereas while wherever whichever whoever who whom whose why where when how then such each own same different other others another whole too two both few more most many much several enough plenty one two three four five six seven eight nine ten hundred thousand million billion apple banana orange lemon berry cherry bread milk water coffee tea juice sugar salt butter cheese meat fish egg food drink eat drank drunk eats drinking fast slow quick early late soon far near short long big small little huge tiny strong weak hard soft easy difficult simple complex important necessary useful bring bought brought carry cat dog horse bird fish mouse rabbit bear lion wolf change chance choice choose chose chosen lose lost lose win won beat defeat victory please plant play point pull push put reach read ready real reason record red rich ride road roll room round run safe said same save say school score sea search seat second see seem seen sell send sense serve set seven shall shape share she shop short should show side sign simple sing sir sister sit six size skill skin sky sleep slip slow small smile smoke so soft some son song soon sorry sort sound source south space speak special speed spell spend sport spring square staff stage stand star start state station stay steel step stick still stone stop store story straight strange street strong student study subject such sudden summer sun support sure surface system table take talk tape teach team tell ten tend term test than thank that their them then there these they thin thing think third this those though thought thousand three through throw thus time tip together told tomorrow tonight too took tool top total touch toward town track trade train travel tree trip trouble true trust truth try turn twelve twenty twice two type under unit until unto up upon use usual value various very view voice walk wall want war warm was wash watch water wave way we wear week well went were west wet what wheel when where which while white who whole why wide wife wild will win wind window wine wing winter wire wish with within without woman wonder wood word work world worry would write wrong wrote yard year yellow yes yet you young your book read story news article user file folder directory server database table row field record index tree node graph list store view update insert select delete drop create alter grant revoke transaction commit rollback cursor execute query result """.split()) # Русские буквосочетания для проверки русской осмысленности # (используются, если файл rugramm.txt недоступен или не загрузился) RU_BIGRAMS_EMBED = frozenset(""" ст но ен то на по ро ов ни ра ко ос ет ер ал ан ом од ел ог лол ат ит ор пр во за бо ди го до ег ие ия ка ла ль ме мо не ри са се та те ту ус че ши ща ью ья жи ши ча ща чу щу цы дл со об от зе же жд зн наи пре при суть сть тель чек чик ка бе ве не пер рес соб ор аб ап ар ас та те сит """.split()) RU_VOWELS = frozenset('аеёиоуыэюяАЕЁИОУЫЭЮЯ') EN_VOWELS = frozenset('aeiouyAEIOUY') # --- ЗАГРУЗКА СЛОВАРЯ ИГНОРА --- def _load_ignore_words(): global _IGNORE_WORDS ignore_set = set() if os.path.exists(IGNORE_FILE): try: with open(IGNORE_FILE, 'r', encoding='utf-8') as f: for line in f: word = line.strip().lower() if word and not word.startswith('#'): ignore_set.add(word) except Exception as e: pprint('*** ERROR loading ignore_turn.txt: %s' % str(e), 'red') _IGNORE_WORDS = ignore_set return ignore_set _load_ignore_words() def _reload_ignore(): return _load_ignore_words() def _save_ignore_word(word, action='add'): if not word: return False lines = [] if os.path.exists(IGNORE_FILE): try: with open(IGNORE_FILE, 'r', encoding='utf-8') as f: lines = f.readlines() except: pass word_lower = word.lower() if action == 'add': for line in lines: if line.strip().lower() == word_lower: return False lines.append(word_lower + '\n') else: lines = [line for line in lines if line.strip().lower() != word_lower] try: with open(IGNORE_FILE, 'w', encoding='utf-8') as f: f.writelines(lines) _reload_ignore() return True except: return False # --- ЗАГРУЗКА РУССКИХ БИГРАММ (КАК PYTHON-КОД) --- def _load_rugramm(): global _RUGRAMM rugramm_set = set() if os.path.exists(RUGRAMM_FILE): try: with open(RUGRAMM_FILE, 'r', encoding='utf-8') as f: content = f.read() # Выполняем содержимое файла как Python-код # Ожидаем, что это словарь или множество data = eval(content) if isinstance(data, dict): # Если это словарь, берём ключи rugramm_set = set(data.keys()) elif isinstance(data, (set, list, tuple)): rugramm_set = set(data) else: pprint('*** ERROR: rugramm.txt must contain a dict or set', 'red') rugramm_set = set() except Exception as e: pprint('*** ERROR loading rugramm.txt: %s' % str(e), 'red') pprint('*** Check the syntax in data/rugramm.txt', 'red') rugramm_set = set() _RUGRAMM = rugramm_set return rugramm_set _load_rugramm() def _reload_rugramm(): return _load_rugramm() # --------------------------------------------------------------------------- # # ТАБЛИЦЫ РАСКЛАДОК # # --------------------------------------------------------------------------- # EN_TO_RU = { 'q': 'й', 'w': 'ц', 'e': 'у', 'r': 'к', 't': 'е', 'y': 'н', 'u': 'г', 'i': 'ш', 'o': 'щ', 'p': 'з', 'a': 'ф', 's': 'ы', 'd': 'в', 'f': 'а', 'g': 'п', 'h': 'р', 'j': 'о', 'k': 'л', 'l': 'д', 'z': 'я', 'x': 'ч', 'c': 'с', 'v': 'м', 'b': 'и', 'n': 'т', 'm': 'ь', '[': 'х', ']': 'ъ', ';': 'ж', "'": 'э', ',': 'б', '.': 'ю', '/': '.', '\\': '/', '~': 'Ё', '{': 'Х', '+': 'Ё', ':': 'Ж', '"': 'Э', '<': 'Б', '>': 'Ю', '?': ',', '&': '?', '@': '"', '#': '№', '$': ';', '^': ':', '`': 'ё' } RU_TO_EN = {v: k for k, v in EN_TO_RU.items()} # --------------------------------------------------------------------------- # # ПРОВЕРКА НА ОСМЫСЛЕННОСТЬ (БИГРАММЫ + АНТИ) # # --------------------------------------------------------------------------- # # Английские биграммы (постоянные) ENGLISH_COMMON = { 'th', 'he', 'in', 'en', 'an', 're', 'er', 'on', 'at', 'nd', 'st', 'or', 'nt', 'ea', 'ti', 'is', 'ou', 'ar', 'es', 'te', 'of', 'it', 'al', 'ri', 'ha', 've', 'ed', 'se', 'me', 'de', 'le', 'co', 'ma', 'li', 'ca', 'be', 'ne', 'ra', 'ro', 've', 'el', 'me', 'ta', 'te', 'ic', 'al', 'om', 'ut', 'ur', 'et', 'em', 'ol', 'tion', 'ing', 'ent', 'ion', 'tio', 'and', 'her', 'for', 'you', 'tha', 'ter', 'con', 'com', 'dis', 'ex', 'im', 'in', 'non', 'pre', 'pro', 're', 'un', 'able', 'ible', 'tion', 'sion', 'ment', 'ness', 'ful', 'ous', 'ive', 'al', 'ic', 'cal', 'ary', 'ory', 'ery', 'ism', 'ist', 'ize', 'ise', 'ing', 'ed', 'er', 'or', 'ar', 'ly', 'ty', 'y', 'es', 'ies', 'ied', 'ing', 'ly', 'ful' } def _is_meaningful_word(word): if len(word) < 3: return True word_lower = word.lower() # Проверяем анти-биграммы (жёсткие) for anti in ANTI_BIGRAMS: if anti in word_lower: return False # Проверяем русские биграммы (из файла) for gram in _RUGRAMM: if gram in word_lower: return True # Проверяем английские биграммы bigrams = [word_lower[i:i+2] for i in range(len(word_lower)-1)] for bg in bigrams: if bg in ENGLISH_COMMON: return True return False # --------------------------------------------------------------------------- # # ВСПОМОГАТЕЛЬНЫЕ ФУНКЦИИ # # --------------------------------------------------------------------------- # def _is_command(word): if word.startswith(('.', '/', '#', '@')): return True if re.match(r'(svn|http[s]?|ftp)://|magnet:\?', word, re.S|re.I|re.U): return True return False def _is_abbreviation(word): if len(word) >= 2 and word.isupper() and word.isalpha(): return True return False def _should_skip(word): if len(word) < 2: return True if _is_command(word) or _is_abbreviation(word): return True if word.lower() in _IGNORE_WORDS: return True if word.lower() in STOP_WORDS: return True return False def fix_layout(text, direction='auto'): if not text: return text if direction == 'auto': en_count = sum(1 for c in text if c.lower() in EN_TO_RU) ru_count = sum(1 for c in text if c in RU_TO_EN) direction = 'en2ru' if en_count >= ru_count else 'ru2en' result = [] for char in text: if direction == 'en2ru': if char.lower() in EN_TO_RU: replacement = EN_TO_RU[char.lower()] if char.isupper(): replacement = replacement.upper() result.append(replacement) else: result.append(char) else: if char in RU_TO_EN: replacement = RU_TO_EN[char] if char.isupper(): replacement = replacement.upper() result.append(replacement) else: result.append(char) return ''.join(result) # --- ВЫДЕЛЕНИЕ РАСКЛАДКИ: ПОСИМВОЛЬНОЕ/ПОСЛОВНОЕ РЕШЕНИЕ (ДЛЯ АВТОТУРНА) --- def _en_readable(word): """Настоящее английское слово: v словаре, либо с гласными и типичными английскими биграммами.""" w = word.lower() if len(w) < 2 or any('\u0400' <= c <= '\u04ff' for c in w): return False if w in ENGLISH_WORDS: return True vowels = sum(1 for c in w if c in 'aeiouy') if vowels < 1: return False bigrams = [w[i:i+2] for i in range(len(w) - 1)] if not bigrams: return False hits = sum(1 for b in bigrams if b in ENGLISH_COMMON) return float(hits) / len(bigrams) >= 0.5 def _ru_readable(word): """Похоже на настоящее русское слово: есть гласные и привычные русские буквосочетания (анти-биграммы запрещены).""" w = word.lower() if len(w) < 2 or any('a' <= c <= 'z' for c in w): return False for anti in ANTI_BIGRAMS: if anti in w: return False vowels = sum(1 for c in w if c in 'аеёиоуыэюя') if vowels < 1: return False for gram in _RUGRAMM: if gram in w: return True for gram in RU_BIGRAMS_EMBED: if gram in w: return True return len(w) <= 3 def _convert_letters(word, direction): """Меняет раскладку ТОЛЬКО для букв, все остальные символы (пунктуацию, цифры) оставляет на своих местах.""" result = [] for char in word: if direction == 'en2ru': if 'a' <= char.lower() <= 'z': replacement = EN_TO_RU.get(char.lower(), char) if char.isupper(): replacement = replacement.upper() result.append(replacement) else: result.append(char) else: if '\u0400' <= char <= '\u04ff': replacement = RU_TO_EN.get(char, char) if char.isupper(): replacement = replacement.upper() result.append(replacement) else: result.append(char) return ''.join(result) def _convert_with_layout_punct(word, direction): """Перекладывает слово в заданную раскладку. Буквы перекладываются как обычно; ВНУТРЕННИЕ (между буквами) '.', ',' и '\'' — это тоже клавиши русской раскладки (ю, б, э), поэтому они тоже перекладываются. Ведущая/замыкающая пунктуация (',', '.', '!', '?' и т.п.) сохраняется.""" if direction == 'en2ru': # '.' и ',' между буквами — это 'ю' и 'б'; '\'' между буквами — 'э' word = re.sub(r'(?<=[A-Za-z])[.,\']+(?=[A-Za-z])', lambda m: ''.join(EN_TO_RU.get(c, c) for c in m.group(0)), word) else: word = re.sub(r'(?<=[\u0400-\u04ff])[.,\']+(?=[\u0400-\u04ff])', lambda m: ''.join(RU_TO_EN.get(c, c) for c in m.group(0)), word) return _convert_letters(word, direction) def _fix_layout_word(word): """Возвращает исправленное слово (с сохранением пунктуации) или None, если слово менять не нужно.""" if not word or not word.strip(): return None # Пропускаем URL и команды целиком if _is_command(word) or re.match(r'^(?:https?|ftp|svn)://|magnet:\?|www\.', word, re.I): return None # Пропускаем домены и слова с точкой (jabber.ru, xmpp.ru, 1.2.3, file.txt, ...) if '.' in word: return None # Извлекаем ТОЛЬКО буквы и проверяем раскладку по ним letters = re.findall(r'[A-Za-z\u0400-\u04ff]', word) if len(letters) < 2: return None core = ''.join(letters) if core.isupper(): return None # аббревиатура (LOL, USA, ...) low_core = core.lower() if low_core in _IGNORE_WORDS or low_core in STOP_WORDS: return None has_latin = any('a' <= c.lower() <= 'z' for c in core) has_cyr = any('\u0400' <= c <= '\u04ff' for c in core) if has_latin and has_cyr: return None # смешанное слово — оставляем как есть if has_latin: # латиница: могла быть напечатана на русской раскладке ru_core = _convert_letters(core, 'en2ru') if _ru_readable(ru_core) and not _en_readable(core): return _convert_with_layout_punct(word, 'en2ru') elif has_cyr: # кириллица: могла быть напечатана на английской раскладке en_core = _convert_letters(core, 'ru2en') if _en_readable(en_core) and not _ru_readable(core): return _convert_with_layout_punct(word, 'ru2en') return None def _turn_layout_words(text): """Поместный разбор: перекладывает только слова, явно набранные не в той раскладке. Возвращает None, если менять нечего.""" parts = [] changed = False for piece in re.split(r'(\s+)', text): if not piece.strip(): parts.append(piece) continue fixed = _fix_layout_word(piece) if fixed is not None: parts.append(fixed) changed = True else: parts.append(piece) return ''.join(parts) if changed else None # --------------------------------------------------------------------------- # # ОСНОВНАЯ ЛОГИКА ПЕРЕКЛЮЧЕНИЯ # # --------------------------------------------------------------------------- # def turner_raw(text, jid, nick): global turn_base to_turn = text if not text: for tmp in turn_base: if tmp[0] == jid and tmp[1] == nick: turn_base.remove(tmp) to_turn = tmp[2] break if not to_turn: return None msg = '' if to_turn[:3] == '/me': msg = '*%s' % nick to_turn = to_turn[3:] elif ': ' in to_turn and to_turn.split(': ', 1)[0] not in ['http', 'https', 'ftp']: parts = to_turn.split(': ', 1) msg = '%s:' % parts[0] to_turn = parts[1] words = re.findall(r'\S+', to_turn) result_parts = [] for word in words: if _should_skip(word): result_parts.append(word) continue ru_letters = all(c in RU_TO_EN for c in word) en_letters = all(c.lower() in EN_TO_RU for c in word) if not ru_letters and not en_letters: result_parts.append(word) continue if ru_letters: switched = fix_layout(word, 'ru2en') else: switched = fix_layout(word, 'en2ru') if _is_meaningful_word(switched): result_parts.append(switched) else: result_parts.append(word) result = ' '.join(result_parts) if msg: result = '%s %s' % (msg, result) result = result.strip() if get_config(getRoom(jid), 'censor'): result = to_censore(result, jid) return result # --------------------------------------------------------------------------- # # КОМАНДЫ НАСТРОЙКИ # # --------------------------------------------------------------------------- # def turn_reload(type, jid, nick, text): if get_level(jid, nick)[0] < 9: send_msg(type, jid, nick, L('You need bot owner access level.', '%s/%s' % (jid, nick))) return ignore_count = len(_reload_ignore()) rugramm_count = len(_reload_rugramm()) msg = L('Ignore list reloaded (%s words). Russian grams reloaded (%s items).', '%s/%s' % (jid, nick)) % (ignore_count, rugramm_count) send_msg(type, jid, nick, msg) def turner(type, jid, nick, text): if not text and type != 'groupchat': send_msg(type, jid, nick, L('Not allowed in private!', '%s/%s' % (jid, nick))) return to_turn = turner_raw(text, jid, nick) if to_turn: send_msg(type, jid, ['', nick][type == 'groupchat'], to_turn) else: send_msg(type, jid, nick, L('What?', '%s/%s' % (jid, nick))) def turn_ignore(type, jid, nick, text): if get_level(jid, nick)[0] < 9: send_msg(type, jid, nick, L('You need bot owner access level.', '%s/%s' % (jid, nick))) return parts = text.strip().split(' ', 1) if len(parts) != 2: send_msg(type, jid, nick, L('Usage: turn_ignore add|del ', '%s/%s' % (jid, nick))) return action, word = parts[0].lower(), parts[1].strip() if not word: send_msg(type, jid, nick, L('Word cannot be empty.', '%s/%s' % (jid, nick))) return if action not in ['add', 'del']: send_msg(type, jid, nick, L('Usage: turn_ignore add|del ', '%s/%s' % (jid, nick))) return if _save_ignore_word(word, action): if action == 'add': msg = L('Word "%s" added to ignore list.', '%s/%s' % (jid, nick)) % word else: msg = L('Word "%s" removed from ignore list.', '%s/%s' % (jid, nick)) % word else: if action == 'add': msg = L('Word "%s" already in ignore list.', '%s/%s' % (jid, nick)) % word else: msg = L('Word "%s" not found in ignore list.', '%s/%s' % (jid, nick)) % word send_msg(type, jid, nick, msg) def append_to_turner(room, jid, nick, type, text): global turn_base for tmp in turn_base: if tmp[0] == room and tmp[1] == nick: try: turn_base.remove(tmp) except: pass break turn_base.append((room, nick, text)) def remove_from_turner(room, jid, nick, type, text): global turn_base if type == 'unavailable': for tmp in turn_base: if tmp[0] == room and tmp[1] == nick: try: turn_base.remove(tmp) except: pass break def autoturn(room, jid, nick, type, text): if not (get_config(room, 'autoturn') and type == 'groupchat'): return if cur_execute_fetchone('select * from commonoff where room=%s and cmd=%s', (room, 'turn')): return nowname = get_xnick(room) if nick == nowname: return text = (text or '').strip() if not text: return # 1. Отбрасываем обращения к боту: "Bot, " / "Bot: " text = re.sub('^%s[,:]\s' % re.escape(nowname), '', text) text = text.strip() if not text: return # 2. Полностью пропускаем /me if text.lower().startswith('/me'): return # 3. Полностью пропускаем команды (.turn, !ping, #topic, ...) first = re.split(r'\s+', text, 1)[0] if first.startswith(('.', '/', '!', '#', '@')): return # 4. Полностью пропускаем ссылки if re.match(r'^(?:https?|ftp|svn)://|magnet:\?|www\.', text, re.I): return # 5. Пошаговый разбор (переводим только слова, очевидно набранные не в той раскладке): converted = _turn_layout_words(text) if converted and converted != text: pprint('Autoturn text: %s/%s [%s] %s > %s' % (room, nick, jid, text, converted), 'dark_gray') send_msg(type, room, '', converted) return True # --------------------------------------------------------------------------- # # РЕГИСТРАЦИЯ # # --------------------------------------------------------------------------- # global execute, message_control, presence_control, message_act_control message_control = [append_to_turner] presence_control = [remove_from_turner] message_act_control = [autoturn] execute = [ (3, 'turn', turner, 2, 'Turn text from one layout to another.\nturn - switch layout for text\nturn - switch layout for last message'), (9, 'turn_reload', turn_reload, 2, 'Reload ignore list from data/ignore_turn.txt and rugramm.txt'), (9, 'turn_ignore', turn_ignore, 2, 'Add or remove word from ignore list.\nturn_ignore add|del '), ]