factor of 3 speedup for the translation without slowing down the grammar loading. The parser still seems to be 8 times slower than before I introduced the BIND. At least the Android UI now feels a lot more comfortable

src/runtime/c/pgf/parser.c

@@ -556,6 +556,27 @@ cmp_item_prob(GuOrder* self, const void* a, const void* b)
 static GuOrder
 pgf_item_prob_order[1] = { { cmp_item_prob } };
 
+static int
+cmp_item_production_idx_entry(GuOrder* self, const void* a, const void* b)
+{
+	PgfProductionIdxEntry *entry1 = (PgfProductionIdxEntry *) a;
+	PgfProductionIdxEntry *entry2 = (PgfProductionIdxEntry *) b;
+
+	if (entry1->ccat->fid < entry2->ccat->fid)
+		return -1;
+	else if (entry1->ccat->fid > entry2->ccat->fid)
+		return 1;
+	else if (entry1->lin_idx < entry2->lin_idx)
+		return -1;
+	else if (entry1->lin_idx > entry2->lin_idx)
+		return 1;
+	else
+		return 0;
+}
+
+static GuOrder
+pgf_production_idx_entry_order[1] = { { cmp_item_production_idx_entry } };
+
 static PgfItemContss*
 pgf_parsing_get_contss(PgfParseState* state, PgfCCat* cat, GuPool *pool)
 {
@@ -1321,18 +1342,21 @@ pgf_parsing_td_predict(PgfParsing* ps,
 				pgf_new_parse_state(ps, lentry->offset, lentry->bind_type);
 
 			if (state != NULL) {
-				size_t n_entries = gu_buf_length(lentry->idx);
+				PgfProductionIdxEntry key;
-				for (size_t j = 0; j < n_entries; j++) {
+				key.ccat    = ccat;
-					PgfProductionIdxEntry* pentry =
+				key.lin_idx = lin_idx;
-						gu_buf_index(lentry->idx, PgfProductionIdxEntry, j);
+				key.papp    = NULL;
+				PgfProductionIdxEntry* value =
+					gu_seq_binsearch(gu_buf_data_seq(lentry->idx),
+									 pgf_production_idx_entry_order,
+									 PgfProductionIdxEntry, &key);
 
-					if (pentry->ccat == ccat && pentry->lin_idx == lin_idx) {
+				if (value != NULL) {
-						GuVariantInfo i = { PGF_PRODUCTION_APPLY, pentry->papp };
+					GuVariantInfo i = { PGF_PRODUCTION_APPLY, value->papp };
-						PgfProduction prod = gu_variant_close(i);
+					PgfProduction prod = gu_variant_close(i);
-						pgf_parsing_add_production(ps, state, state->next, 
+					pgf_parsing_add_production(ps, state, state->next, 
-												   conts, prod, 
+											   conts, prod, 
-												   pentry->papp->fun->absfun->ep.prob);
+											   value->papp->fun->absfun->ep.prob);
-					}
 				}
 			}
 		}
@@ -2611,18 +2635,18 @@ pgf_parser_index(PgfConcr* concr,
                  PgfCCat* ccat, PgfProduction prod,
                  GuPool *pool)
 {
-	GuPool* tmp_pool = gu_local_pool();
+	GuVariantInfo i = gu_variant_open(prod);
-	GuChoice* choice = gu_new_choice(tmp_pool); // we need this for the pres
+	switch (i.tag) {
+	case PGF_PRODUCTION_APPLY: {
+		PgfProductionApply* papp = i.data;
 
-	for (size_t lin_idx = 0; lin_idx < ccat->cnccat->n_lins; lin_idx++) {
+		if (gu_seq_length(papp->args) > 0)
-		GuVariantInfo i = gu_variant_open(prod);
+			break;
-		switch (i.tag) {
-		case PGF_PRODUCTION_APPLY: {
-			PgfProductionApply* papp = i.data;
 
-			if (gu_seq_length(papp->args) > 0)
+		GuPool* tmp_pool = gu_local_pool();
-				break;
+		GuChoice* choice = gu_new_choice(tmp_pool); // we need this for the pres
 
+		for (size_t lin_idx = 0; lin_idx < papp->fun->n_lins; lin_idx++) {
 			PgfSequence* seq = papp->fun->lins[lin_idx];
 			if (seq->idx == NULL) {
 				seq->idx = gu_new_buf(PgfProductionIdxEntry, pool);
@@ -2634,16 +2658,16 @@ pgf_parser_index(PgfConcr* concr,
 			entry->lin_idx = lin_idx;
 			entry->papp    = papp;
 		}
+		
+		gu_pool_free(tmp_pool);
 		break;
-		case PGF_PRODUCTION_COERCE:
-			// Nothing to be done here
-			break;
-		default:
-			gu_impossible();
-		}
 	}
-	
+	case PGF_PRODUCTION_COERCE:
-	gu_pool_free(tmp_pool);
+		// Nothing to be done here
+		break;
+	default:
+		gu_impossible();
+	}
 }
 
 prob_t

src/runtime/c/pgf/reader.c

@@ -929,6 +929,16 @@ pgf_read_pargs(PgfReader* rdr, PgfConcr* concr)
 	return pargs;
 }
 
+extern void
+pgf_parser_index(PgfConcr* concr, 
+                 PgfCCat* ccat, PgfProduction prod,
+                 GuPool *pool);
+
+extern void
+pgf_lzr_index(PgfConcr* concr, 
+              PgfCCat* ccat, PgfProduction prod,
+              GuPool *pool);
+
 static void
 pgf_read_production(PgfReader* rdr, PgfConcr* concr, 
                     PgfCCat* ccat, size_t* top, size_t* bot)
@@ -970,6 +980,9 @@ pgf_read_production(PgfReader* rdr, PgfConcr* concr,
 	default:
 		pgf_read_tag_error(rdr);
 	}
+
+	pgf_parser_index(concr, ccat, prod, rdr->opool);
+	pgf_lzr_index(concr, ccat, prod, rdr->opool);
 }
 
 static void
@@ -1105,25 +1118,11 @@ typedef struct {
 	GuPool* pool;
 } PgfIndexFn;
 
-extern void
-pgf_parser_index(PgfConcr* concr, 
-                 PgfCCat* ccat, PgfProduction prod,
-                 GuPool *pool);
-
-void
-pgf_lzr_index(PgfConcr* concr, 
-              PgfCCat* ccat, PgfProduction prod,
-              GuPool *pool);
-
 static void
 pgf_read_ccat_cb(GuMapItor* fn, const void* key, void* value, GuExn* err)
 {
-	(void) (key && err);
+	(void) (fn && key && err);
-	
-	PgfIndexFn* clo = (PgfIndexFn*) fn;
     PgfCCat* ccat = *((PgfCCat**) value);
-    PgfConcr *concr = clo->concr;
-    GuPool *pool = clo->pool;
 
 	if (ccat->prods == NULL)
 		return;
@@ -1136,9 +1135,6 @@ pgf_read_ccat_cb(GuMapItor* fn, const void* key, void* value, GuExn* err)
 		if (!ccat->cnccat) {
 			pgf_ccat_set_cnccat(ccat, prod);
 		}
-
-		pgf_parser_index(concr, ccat, prod, pool);
-		pgf_lzr_index(concr, ccat, prod, pool);
 	}
 
 //	pgf_ccat_set_viterbi_prob(ccat);
@@ -1184,8 +1180,8 @@ pgf_read_concrete(PgfReader* rdr, PgfAbstr* abstr, PgfAbsFun* abs_lin_fun)
 	concr->callbacks = pgf_new_callbacks_map(concr, rdr->opool); 
 	concr->total_cats = pgf_read_int(rdr);
 
-	PgfIndexFn clo1 = { { pgf_read_ccat_cb }, concr, rdr->opool };
+	GuMapItor clo1 = { pgf_read_ccat_cb };
-	gu_map_iter(concr->ccats, &clo1.fn, NULL);
+	gu_map_iter(concr->ccats, &clo1, NULL);
 
 	return concr;
 }