add option to specify trie number in marisa-trie

Author: nokuno

README.md

@@ -49,6 +49,7 @@ You can provide queries through standard input line by line.
     options:
     -d directory: specify data directory or prefix (default: ./)
     -m model: specify model file name (default: ngram)
+    -t trie_num: specify trie number in marisa-trie (default: 3)
     -r: build reverse model
 
 ### jsc-server

src/builder.cpp

@@ -2,7 +2,7 @@
 
 namespace jsc {
 
-bool Builder::Build(const char *filename, const char *prefix, bool reverse) {
+bool Builder::Build(const char *filename, const char *prefix, bool reverse, int trie_num) {
 
   // open text file
   fstream fs(filename);
@@ -96,7 +96,7 @@ bool Builder::Build(const char *filename, const char *prefix, bool reverse) {
   }
   // build and save source trie
   marisa::Trie source_trie;
-  source_trie.build(source_keyset);
+  source_trie.build(source_keyset, trie_num);
   {
     string file(prefix);
     file.append(SOURCE_TRIE);
@@ -106,7 +106,7 @@ bool Builder::Build(const char *filename, const char *prefix, bool reverse) {
   // build and save ngram trie
   // Notice: ngram trie is used later for building entry array.
   marisa::Trie ngram_trie;
-  ngram_trie.build(ngram_keyset);
+  ngram_trie.build(ngram_keyset, trie_num);
   {
     string file(prefix);
     file.append(NGRAM_TRIE);

src/builder.h

@@ -18,7 +18,7 @@ class Builder {
   //   5) Sort source array by source_id
   //   6) Extract offset of source_id
   //   7) Save offset and entry array as binary
-  bool Build(const char *filename, const char *prefix, bool reverse=false);
+  bool Build(const char *filename, const char *prefix, bool reverse=false, int trie_num=3);
 };
 
 }

tools/jsc-build.cpp

@@ -6,8 +6,9 @@ int main(int argc, char **argv) {
   string model = "ngram";
 
   bool reverse = false;
+  int trie_num = 3;
   int c;
-  while ((c = getopt (argc, argv, "d:m:r")) != -1) {
+  while ((c = getopt (argc, argv, "d:m:t:r")) != -1) {
     switch (c) {
       case 'd':
         prefix = optarg;
@@ -18,6 +19,9 @@ int main(int argc, char **argv) {
       case 'r':
         reverse = true;
         break;
+      case 't':
+        trie_num = atoi(optarg);
+        break;
       case '?':
         cerr << "Unknown option -" << optopt << endl;
         return 1;
@@ -28,7 +32,7 @@ int main(int argc, char **argv) {
   {
     Builder builder;
     cerr << "Building model..." << endl;
-    bool result = builder.Build(model.c_str(), prefix.c_str(), reverse);
+    bool result = builder.Build(model.c_str(), prefix.c_str(), reverse, trie_num);
     if (result) {
       cerr << "Succeeded to build model." << endl;
     } else {