add -H option to lt-comp so it doesn't eat ε (#92)

apertium · Nov 14, 2021 · 1ca0c28 · 1ca0c28
1 parent 98273a6
commit 1ca0c28
Show file tree

Hide file tree

Showing 3 changed files with 20 additions and 3 deletions.
diff --git a/lttoolbox/att_compiler.cc b/lttoolbox/att_compiler.cc
@@ -59,7 +59,7 @@ AttCompiler::convert_hfst(UString& symbol)
 {
   if (symbol == Transducer::HFST_EPSILON_SYMBOL_SHORT ||
       symbol == Transducer::HFST_EPSILON_SYMBOL_LONG ||
-      symbol == Transducer::LTTB_EPSILON_SYMBOL) {
+      (!hfstSymbols && symbol == Transducer::LTTB_EPSILON_SYMBOL)) {
     symbol.clear();
   } else if (symbol == Transducer::HFST_SPACE_SYMBOL) {
     symbol = " "_u;
@@ -469,3 +469,9 @@ AttCompiler::write(FILE *output)
   }
 //  fclose(output);
 }
+
+void
+AttCompiler::setHfstSymbols(bool b)
+{
+  hfstSymbols = b;
+}
diff --git a/lttoolbox/att_compiler.h b/lttoolbox/att_compiler.h
@@ -84,8 +84,12 @@ class AttCompiler
 
   void write(FILE *fd) ;
 
+  void setHfstSymbols(bool b);
+
 private:
 
+  bool hfstSymbols = false;
+
   /** The final state(s). */
   map<int, double> finals;
   /**

diff --git a/lttoolbox/lt_comp.cc b/lttoolbox/lt_comp.cc
@@ -48,12 +48,14 @@ void endProgram(char *name)
     cout << "  -a, --alt:                 set alternative (monodix)" << endl;
     cout << "  -l, --var-left:            set left language variant (bidix)" << endl;
     cout << "  -r, --var-right:           set right language variant (bidix)" << endl;
+    cout << "  -H, --hfst:                expect HFST symbols" << endl;
 #else
     cout << "  -m:     keep morpheme boundaries" << endl;
     cout << "  -v:     set language variant" << endl;
     cout << "  -a:     set alternative (monodix)" << endl;
     cout << "  -l:     set left language variant (bidix)" << endl;
     cout << "  -r:     set right language variant (bidix)" << endl;
+    cout << "  -H:     expect HFST symbols" << endl;
 #endif
     cout << "Modes:" << endl;
     cout << "  lr:     left-to-right compilation" << endl;
@@ -89,14 +91,15 @@ int main(int argc, char *argv[])
       {"var-left",  required_argument, 0, 'l'},
       {"var-right", required_argument, 0, 'r'},
       {"keep-boundaries",      no_argument,       0, 'm'},
+      {"hfst",      no_argument,       0, 'H'},
       {"help",      no_argument,       0, 'h'},
       {"verbose",   no_argument,       0, 'V'},
       {0, 0, 0, 0}
     };
 
-    int cnt=getopt_long(argc, argv, "a:v:l:r:mhV", long_options, &option_index);
+    int cnt=getopt_long(argc, argv, "a:v:l:r:mHhV", long_options, &option_index);
 #else
-    int cnt=getopt(argc, argv, "a:v:l:r:mhV");
+    int cnt=getopt(argc, argv, "a:v:l:r:mHhV");
 #endif
     if (cnt==-1)
       break;
@@ -125,6 +128,10 @@ int main(int argc, char *argv[])
         c.setKeepBoundaries(true);
         break;
 
+      case 'H':
+        a.setHfstSymbols(true);
+        break;
+
       case 'V':
         c.setVerbose(true);
         break;