Important fix to RTL languages saves last space on each line, which w…

…as previously lost
tesseract-ocr · Jul 20, 2017 · 4e8018d · 4e8018d
1 parent 3f77354
commit 4e8018d
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 2 deletions.
diff --git a/ccmain/resultiterator.cpp b/ccmain/resultiterator.cpp
@@ -549,6 +549,12 @@ bool ResultIterator::IsAtFinalElement(PageIteratorLevel level,
   return true;
 }
 
+// Returns the number of blanks before the current word.
+int ResultIterator::BlanksBeforeWord() const {
+  if (CurrentParagraphIsLtr()) return LTRResultIterator::BlanksBeforeWord();
+  return IsAtBeginningOf(RIL_TEXTLINE) ? 0 : 1;
+}
+
 /**
  * Returns the null terminated UTF-8 encoded text string for the current
  * object at the given level. Use delete [] to free after use.
@@ -585,7 +591,7 @@ char* ResultIterator::GetUTF8Text(PageIteratorLevel level) const {
         if (at_beginning_of_minor_run_) {
           text += reading_direction_is_ltr ? kLRM : kRLM;
         }
-        text = it_->word()->BestUTF8(blob_index_, !reading_direction_is_ltr);
+        text = it_->word()->BestUTF8(blob_index_, false);
         if (IsAtFinalSymbolOfWord()) AppendSuffixMarks(&text);
       }
       break;
@@ -608,7 +614,7 @@ void ResultIterator::AppendUTF8WordText(STRING *text) const {
   GenericVector<int> blob_order;
   CalculateBlobOrder(&blob_order);
   for (int i = 0; i < blob_order.size(); i++) {
-    *text += it_->word()->BestUTF8(blob_order[i], !reading_direction_is_ltr);
+    *text += it_->word()->BestUTF8(blob_order[i], false);
   }
   AppendSuffixMarks(text);
 }
@@ -643,6 +649,9 @@ void ResultIterator::IterateAndAppendUTF8TextlineText(STRING *text) {
     }
     AppendUTF8WordText(text);
     words_appended++;
+    if (BidiDebug(2)) {
+      tprintf("Num spaces=%d, text=%s\n", numSpaces, text->string());
+    }
   } while (Next(RIL_WORD) && !IsAtBeginningOf(RIL_TEXTLINE));
   if (BidiDebug(1)) {
     tprintf("%d words printed\n", words_appended);

diff --git a/ccmain/resultiterator.h b/ccmain/resultiterator.h
@@ -82,6 +82,10 @@ class TESS_API ResultIterator : public LTRResultIterator {
   virtual bool IsAtFinalElement(PageIteratorLevel level,
                                 PageIteratorLevel element) const;
 
+  // ============= Functions that refer to words only ============.
+  // Returns the number of blanks before the current word.
+  int BlanksBeforeWord() const;
+
   // ============= Accessing data ==============.
 
   /**