/** * @file mkcpair.c * * * @brief 文法からカテゴリ対制約を抽出する * * DFA文法から,認識処理の第1パスで用いるカテゴリ間の接続制約を抽出します. * 抽出したカテゴリ対情報は文法データ内に格納されます. * * ショートポーズ単語が文法上で指定されている場合,接続制約はこの * ショートポーズ単語のスキップを考慮して抽出されます. * * なお,ここでは処理上の制限から,このスキップ可能なショートポーズ単語が * 文頭およぶ文末に出現することは許されていません. * 入力開始・終端の無音部分にあたる単語は,silB, silE など * このスキップ可能なショートポーズ以外の単語を割り当てるようにしてください * * * * @brief Extract category-pair constraint from DFA grammar * * These functions extract whether the each grammar category can be connected * or not in the given DFA grammar, and store the extracted data to the DFA * grammar information. This category-pair constraint will be used at the * first pass of recognition as a degenerated linguistic constraint. * * If a short pause word is defined in the grammar, the connection constraint * will be extracted considering the skipping of this pause model, since the * pause word may not appear on the specified location in the actual utterance. * * Please note that a grammar rule that allows such skippable short-pause word * to appear at the beginning and end of sentence is prohibited. Instead, * you should use another (non-skippable) silence word like "sil" as the * beginning and ending of sentence to match the head and tail silence. * * * * @author Akinobu LEE * @date Tue Feb 15 14:35:33 2005 * * $Revision: 1.6 $ * */ /* * Copyright (c) 1991-2012 Kawahara Lab., Kyoto University * Copyright (c) 2000-2005 Shikano Lab., Nara Institute of Science and Technology * Copyright (c) 2005-2012 Julius project team, Nagoya Institute of Technology * All rights reserved */ #include #include /** * Extract category-pair constraint from DFA grammar and newly set the category * pair matrix of the give DFA. * * @param dinfo [i/o] DFA grammar, in which the category-pair matrix will be created. */ boolean extract_cpair(DFA_INFO *dinfo) { int i; DFA_ARC *arc_l, *arc_r, *arc_r2; int left, right; int size; /* initialize */ /* initial size = average fun-out num per state */ size = dinfo->arc_num / dinfo->state_num; if (size < DFA_CP_MINSTEP) size = DFA_CP_MINSTEP; malloc_dfa_cp(dinfo, dinfo->term_num, size); /* extract cpair info */ for (i=0;istate_num;i++) { if ((dinfo->st[i].status & INITIAL_S) != 0) { /* arc from initial state */ for (arc_r = dinfo->st[i].arc; arc_r; arc_r = arc_r->next) { if (dinfo->is_sp[arc_r->label]) { jlog("Error: mkcpair: skippable sp should not appear at end of sentence\n"); return FALSE; } set_dfa_cp_end(dinfo, arc_r->label, TRUE); } } for(arc_l = dinfo->st[i].arc; arc_l; arc_l = arc_l->next) { left = arc_l->label; if ((dinfo->st[arc_l->to_state].status & ACCEPT_S) != 0) {/* arc to accept state */ if (dinfo->is_sp[left]) { jlog("Error: mkcpair: skippable sp should not appear at beginning of sentence\n"); return FALSE; } set_dfa_cp_begin(dinfo, left, TRUE); } /* others */ for (arc_r = dinfo->st[arc_l->to_state].arc; arc_r; arc_r = arc_r->next) { right = arc_r->label; set_dfa_cp(dinfo, right, left, TRUE); if (dinfo->is_sp[right]) { for (arc_r2 = dinfo->st[arc_r->to_state].arc; arc_r2; arc_r2 = arc_r2->next) { if (dinfo->is_sp[arc_r2->label]) { /* sp model continues twice */ jlog("Error: mkcpair: skippable sp should not repeat\n"); return FALSE; } set_dfa_cp(dinfo, arc_r2->label, left, TRUE); } } } } } return TRUE; } /** * Append the category pair matrix at the last. * * @param dst [i/o] DFA grammar * @param src [in] DFA grammar to be appended to @a dst * @param coffset [in] category id in @a dst where the new data should be stored */ boolean cpair_append(DFA_INFO *dst, DFA_INFO *src, int coffset) { /* dst info must be already appended */ /* [coffset..dst->term_num-1] is the new categories */ if (dst->term_num - coffset != src->term_num) { jlog("Error: mkcpair: append term num not match!: %d, %d, %d\n", dst->term_num, src->term_num, coffset); return FALSE; } /* allocate appended area */ if (dfa_cp_append(dst, src, coffset) == FALSE) return FALSE; return TRUE; }