Lines Matching full:will
78 // When the last char 'o' is added, the parser will find "dao" is a valid
93 // nodes at 'x' only have Shengmu ids, the parser will not try "x ian" (and it
95 // always be "xian"; but if the parser uses continue in the loop, "xi an" will
109 // in these fixed lemmas, all fixed lemmas will be merged together into a unit
111 // phrase will be the first lemma in the sentence. Because it contains some
128 // will be limited to a reasonable number.
131 // If it is true, the engine will prefer long history based prediction,
136 // If it is true, prediction will only be based on user dictionary. this flag
166 // first of which is a FULL id too. So even it is true, "da" will never be
223 // If dmi_c_phrase_ is true, the decoder will try to match the
224 // composing phrase (And definitely it will match successfully). If it
225 // is false, the decoder will try to match lemmas items in dictionaries.
245 // not valid before calling it. This function only will be called in the
257 // input Pinyin is "an", reset_search(1) will reset the search space to the
262 // The DMI nodes will be kept.
274 // This function will update the lemma and spelling segmentation information.
279 // Get spelling start posistions and ids. The result will be stored in
281 // fixed_hzs_ will be also assigned.
312 // If dmi == NULL, this function will extend the root node of DictTrie
314 // This function will not change dmi_nd_pool_used_. Please change it after
412 // fixed lemmas or composing string, this function will do nothing and just
425 // Number of spellings will be returned. The number of valid elements in
431 // Get one candiate string. If full sentence candidate is available, it will
436 // retstr_len is not NULL, it will be used to return the string length.
437 // If only_unfixed is true, only unfixed part will be fetched.
441 // Choose a candidate. The decoder will do a search after the fixed position.
