set unlv_tilde_crunching to false; fixes #1449 #948

tesseract-ocr · Oct 23, 2018 · 3d508a6 · stweil · Oct 23, 2018 · amitdo
1 parent 5a68b7f
commit 3d508a6
Show file tree

Hide file tree

Showing 4 changed files with 4 additions and 2 deletions.
diff --git a/src/api/tesseractmain.cpp b/src/api/tesseractmain.cpp
@@ -456,6 +456,7 @@ static void PreloadRenderers(
 
     api->GetBoolVariable("tessedit_write_unlv", &b);
     if (b) {
+      api->SetVariable("unlv_tilde_crunching", "true");
       tesseract::TessUnlvRenderer* renderer =
         new tesseract::TessUnlvRenderer(outputbase);
       if (renderer->happy()) {

diff --git a/src/ccmain/tesseractclass.cpp b/src/ccmain/tesseractclass.cpp
@@ -271,7 +271,7 @@ Tesseract::Tesseract()
                   this->params()),
       double_MEMBER(quality_rowrej_pc, 1.1,
                     "good_quality_doc gte good char limit", this->params()),
-      BOOL_MEMBER(unlv_tilde_crunching, true,
+      BOOL_MEMBER(unlv_tilde_crunching, false,
                   "Mark v.bad words for tilde crunch", this->params()),
       BOOL_MEMBER(hocr_font_info, false, "Add font info to hocr output",
                   this->params()),

diff --git a/src/ccmain/tesseractclass.h b/src/ccmain/tesseractclass.h
@@ -963,7 +963,7 @@ class Tesseract : public Wordrec {
   BOOL_VAR_H(bland_unrej, false, "unrej potential with no checks");
   double_VAR_H(quality_rowrej_pc, 1.1,
                "good_quality_doc gte good char limit");
-  BOOL_VAR_H(unlv_tilde_crunching, true,
+  BOOL_VAR_H(unlv_tilde_crunching, false,
              "Mark v.bad words for tilde crunch");
   BOOL_VAR_H(hocr_font_info, false,
              "Add font info to hocr output");

diff --git a/tessdata/configs/unlv b/tessdata/configs/unlv
@@ -1 +1,2 @@
 tessedit_write_unlv 1
+unlv_tilde_crunching T