Rev 297 | Go to most recent revision | Details | Compare with Previous | Last modification | View Log | RSS feed
| Rev | Author | Line No. | Line |
|---|---|---|---|
| 293 | PointedEar | 1 | """ |
| 2 | Created on 2014-10-20 |
||
| 3 | |||
| 4 | @author: Thomas 'PointedEars' Lahn <mail@PointedEars.de> |
||
| 5 | |||
| 6 | """ |
||
| 7 | |||
| 8 | from os import chdir, stat |
||
| 9 | from sys import stderr |
||
| 10 | from os.path import dirname, realpath, basename |
||
| 11 | from pickle import dump, load |
||
| 12 | from re import match, DOTALL, search, sub, split, compile |
||
| 13 | |||
| 14 | debug_level = 2 |
||
| 15 | |||
| 16 | def dmsg(*args, **kwargs): |
||
| 17 | if not hasattr(kwargs, 'min_level') or kwargs['min_level'] is None: |
||
| 18 | kwargs['min_level'] = 1 |
||
| 19 | |||
| 20 | if not hasattr(kwargs, 'file'): |
||
| 21 | kwargs['file'] = stderr |
||
| 22 | |||
| 23 | if debug_level >= kwargs['min_level']: |
||
| 24 | del kwargs['min_level'] |
||
| 25 | print(*args, **kwargs) |
||
| 26 | |||
| 27 | def sort_dict_alnum_english_key(phrase): |
||
| 28 | return sub(r'\{(.+)\}', r'\1', phrase[0]).lower() |
||
| 29 | |||
| 30 | class Dictionary(dict): |
||
| 31 | """ |
||
| 297 | PointedEar | 32 | A Dictionary (not to be confused with its ancestor, dict) |
| 33 | represents a word dictionary stored in a file. |
||
| 293 | PointedEar | 34 | |
| 35 | """ |
||
| 296 | PointedEar | 36 | _language_key = 'en' |
| 293 | PointedEar | 37 | _keys = "ipa|en|lit|pos|com|tag|ex" |
| 38 | _expressions = {} |
||
| 39 | |||
| 294 | PointedEar | 40 | def load (self, dictionary_file, language_key='en'): |
| 297 | PointedEar | 41 | """ |
| 42 | Loads a word dictionary from a file. |
||
| 43 | :param dictionary_file: |
||
| 44 | :type dictionary_file: |
||
| 45 | :param language_key: |
||
| 46 | :type language_key: |
||
| 47 | """ |
||
| 296 | PointedEar | 48 | self._language_key = language_key |
| 49 | |||
| 293 | PointedEar | 50 | dmsg('Loading dictionary '.format(dictionary_file), end='', min_level=1) |
| 51 | |||
| 52 | chdir(dirname(realpath(__file__))) |
||
| 53 | |||
| 54 | pickle_file = basename(dictionary_file) + '.pickle' |
||
| 55 | |||
| 56 | try: |
||
| 57 | pickle_mtime = stat(pickle_file).st_mtime |
||
| 58 | except FileNotFoundError: |
||
| 59 | pickle_mtime = None |
||
| 60 | |||
| 61 | if pickle_mtime is None or stat(dictionary_file).st_mtime > pickle_mtime: |
||
| 62 | dmsg('from {0} ...'.format(dictionary_file), end='', min_level=1) |
||
| 63 | phrase = None |
||
| 64 | key = None |
||
| 65 | value = None |
||
| 66 | with open(dictionary_file) as f: |
||
| 67 | indent = None |
||
| 68 | |||
| 69 | for line in f: |
||
| 296 | PointedEar | 70 | m = match(r'^\s*{0}:\s*(?P<phrase>.+)'.format(self._language_key), line) |
| 293 | PointedEar | 71 | if m is not None: |
| 72 | phrase = m.group("phrase") |
||
| 295 | PointedEar | 73 | self[phrase] = {} |
| 293 | PointedEar | 74 | indent = None |
| 75 | else: |
||
| 76 | m = match(r'(?P<indent>\s*)(?P<key>{0}):\s*(?P<value>.+)'.format(self._keys), line) |
||
| 77 | if m is not None: |
||
| 78 | # join previous value if necessary |
||
| 79 | if type(value) == list: |
||
| 295 | PointedEar | 80 | self[phrase][key] = ' '.join(value) |
| 293 | PointedEar | 81 | |
| 82 | indent = m.group("indent") |
||
| 83 | key = m.group("key") |
||
| 84 | value = m.group("value") |
||
| 85 | # assign a string for memory efficiency |
||
| 295 | PointedEar | 86 | self[phrase][key] = value |
| 293 | PointedEar | 87 | elif indent is not None: |
| 88 | m = match(r'(?P<indent>\s+)(?P<continuation>\S.*)', line) |
||
| 89 | if m is not None: |
||
| 90 | if len(m.group("indent")) == len(indent) + 2: |
||
| 91 | continuation = m.group("continuation") |
||
| 92 | if type(value) == str: |
||
| 93 | # when a continuation is first found, convert to a list |
||
| 94 | # because there could be more continuations |
||
| 295 | PointedEar | 95 | value = self[phrase][key] = [value, continuation] |
| 293 | PointedEar | 96 | else: |
| 97 | value.append(continuation) |
||
| 98 | |||
| 99 | # join last value if necessary |
||
| 100 | if type(value) == list: |
||
| 295 | PointedEar | 101 | self[phrase][key] = ' '.join(value) |
| 293 | PointedEar | 102 | |
| 103 | dmsg('\nSaving pickle {0} ...'.format(pickle_file), end='', min_level=1) |
||
| 104 | # TODO: Pickle should only contain strings to be small |
||
| 295 | PointedEar | 105 | with open(pickle_file, mode='wb') as f: dump(self, f) |
| 293 | PointedEar | 106 | dmsg(' done.', min_level=1) |
| 107 | else: |
||
| 108 | dmsg('from {0} ...'.format(pickle_file), end='', min_level=1) |
||
| 109 | with open(pickle_file, mode='rb') as f: pickle = load(f) |
||
| 110 | for key, value in pickle.items(): |
||
| 295 | PointedEar | 111 | self[key] = value |
| 293 | PointedEar | 112 | |
| 295 | PointedEar | 113 | dmsg(' done ({0} entries).'.format(len(self)), min_level=1) |
| 293 | PointedEar | 114 | |
| 115 | def clean (self): |
||
| 297 | PointedEar | 116 | """ |
| 117 | Cleans dictionary entries |
||
| 118 | """ |
||
| 296 | PointedEar | 119 | re_parens = compile(r'\(.+\)', DOTALL) |
| 297 | PointedEar | 120 | re_parens_no_alt = compile(r'\(([^\|]+)\)', DOTALL) |
| 296 | PointedEar | 121 | re_braces = compile( |
| 295 | PointedEar | 122 | r'^\s*\{(?P<phrase>.+)\}(?:\s*\((?P<variant>.+?)\))?\s*$', |
| 123 | DOTALL) |
||
| 296 | PointedEar | 124 | re_semicolon = compile(r'\s*;\s*') |
| 293 | PointedEar | 125 | |
| 295 | PointedEar | 126 | for orig_phrase, data in list(self.items()): |
| 293 | PointedEar | 127 | # if there are optional or alternating parts |
| 296 | PointedEar | 128 | if search(re_parens, orig_phrase): |
| 293 | PointedEar | 129 | if orig_phrase.find('|') > -1: |
| 130 | # TODO alternation |
||
| 131 | pass |
||
| 132 | else: |
||
| 133 | # TODO optional parts |
||
| 134 | pass |
||
| 135 | |||
| 136 | if orig_phrase.find(';') > -1: |
||
| 137 | synonyms = map( |
||
| 296 | PointedEar | 138 | lambda x: sub(re_braces, r'\1', x), |
| 139 | split(re_semicolon, orig_phrase)) |
||
| 293 | PointedEar | 140 | |
| 141 | for synonym in synonyms: |
||
| 295 | PointedEar | 142 | self[synonym] = data |
| 293 | PointedEar | 143 | |
| 295 | PointedEar | 144 | del self[orig_phrase] |
| 293 | PointedEar | 145 | else: |
| 296 | PointedEar | 146 | m = match(re_braces, orig_phrase) |
| 293 | PointedEar | 147 | if m is not None: |
| 297 | PointedEar | 148 | phrase = m.group('phrase') |
| 149 | |||
| 150 | if callable(getattr(self, 'clean_entry', None)): |
||
| 151 | phrase = self.clean_entry(phrase) |
||
| 152 | |||
| 296 | PointedEar | 153 | m_parens = search(re_parens, phrase) |
| 154 | if m_parens is not None: |
||
| 155 | # alternation and optional parts |
||
| 156 | expr = sub(re_parens_no_alt, r'(?:\1)?', phrase) |
||
| 157 | expr = sub('~', '(?=.)', expr) |
||
| 158 | self._expressions[expr] = data |
||
| 159 | else: |
||
| 160 | # remove braces |
||
| 161 | self[phrase] = data |
||
| 295 | PointedEar | 162 | |
| 163 | del self[orig_phrase] |
||
| 296 | PointedEar | 164 | |
| 165 | def translate (self, phrase): |
||
| 297 | PointedEar | 166 | """ |
| 167 | Translate a phrase according to this dictionary. |
||
| 168 | For language-specific processing, this method should be |
||
| 169 | called/overridden by inheriting classes. |
||
| 170 | :param phrase: |
||
| 171 | :type phrase: str |
||
| 172 | """ |
||
| 296 | PointedEar | 173 | translation = self.get(phrase.lower(), None) |
| 174 | if translation is not None: |
||
| 175 | translation[self._language_key] = phrase |
||
| 176 | return translation |
||
| 177 | |||
| 178 | return None |
||
| 179 | |||
| 180 | def translate_expression (self, phrase): |
||
| 297 | PointedEar | 181 | """ |
| 182 | Translate a phrase according entries in this dictionary |
||
| 183 | based on regular expressions. |
||
| 184 | :param phrase: |
||
| 185 | :type phrase: |
||
| 186 | """ |
||
| 187 | for expression, data in sorted(self._expressions.items(), key=lambda item:-len(item[1])): |
||
| 298 | PointedEar | 188 | expression_match = match(r'{0}$'.format(expression), phrase) |
| 296 | PointedEar | 189 | if expression_match is not None: |
| 190 | data[self._language_key] = expression_match.group(0) |
||
| 191 | return data |
||
| 192 | |||
| 193 | return None |