RAII: Result of ResultIterator::GetUTF8Text() was leaked inside TessB…

…aseAPI::GetUTF8Text()
tesseract-ocr · May 9, 2017 · ec165b3 · ec165b3
1 parent b86b4fa
commit ec165b3
Show file tree

Hide file tree

Showing 6 changed files with 17 additions and 21 deletions.
diff --git a/api/baseapi.cpp b/api/baseapi.cpp
@@ -1267,9 +1267,7 @@ char* TessBaseAPI::GetUTF8Text() {
   ResultIterator *it = GetIterator();
   do {
     if (it->Empty(RIL_PARA)) continue;
-    char *para_text = it->GetUTF8Text(RIL_PARA);
-    text += para_text;
-    delete []para_text;
+    text += it->GetUTF8Text(RIL_PARA).get();
   } while (it->Next(RIL_PARA));
   char* result = new char[text.length() + 1];
   strncpy(result, text.string(), text.length() + 1);
@@ -1539,11 +1537,10 @@ char* TessBaseAPI::GetHOCRText(ETEXT_DESC* monitor, int page_number) {
     if (bold) hocr_str += "<strong>";
     if (italic) hocr_str += "<em>";
     do {
-      const char *grapheme = res_it->GetUTF8Text(RIL_SYMBOL);
+      const std::unique_ptr<const char[]> grapheme = res_it->GetUTF8Text(RIL_SYMBOL);
       if (grapheme && grapheme[0] != 0) {
-        hocr_str += HOcrEscape(grapheme);
+        hocr_str += HOcrEscape(grapheme.get());
       }
-      delete []grapheme;
       res_it->Next(RIL_SYMBOL);
     } while (!res_it->Empty(RIL_BLOCK) && !res_it->IsAtBeginningOf(RIL_WORD));
     if (italic) hocr_str += "</em>";
@@ -1661,7 +1658,7 @@ char* TessBaseAPI::GetTSVText(int page_number) {
     if (res_it->IsAtFinalElement(RIL_BLOCK, RIL_WORD)) bcnt++;
 
     do {
-      tsv_str += res_it->GetUTF8Text(RIL_SYMBOL);
+      tsv_str += res_it->GetUTF8Text(RIL_SYMBOL).get();
       res_it->Next(RIL_SYMBOL);
     } while (!res_it->Empty(RIL_BLOCK) && !res_it->IsAtBeginningOf(RIL_WORD));
     tsv_str += "\n";  // end of row

diff --git a/api/capi.cpp b/api/capi.cpp
@@ -733,7 +733,7 @@ TESS_API BOOL  TESS_CALL TessResultIteratorNext(TessResultIterator* handle, Tess
 
 TESS_API char* TESS_CALL TessResultIteratorGetUTF8Text(const TessResultIterator* handle, TessPageIteratorLevel level)
 {
-    return handle->GetUTF8Text(level);
+    return handle->GetUTF8Text(level).release();
 }
 
 TESS_API float TESS_CALL TessResultIteratorConfidence(const TessResultIterator* handle, TessPageIteratorLevel level)

diff --git a/api/pdfrenderer.cpp b/api/pdfrenderer.cpp
@@ -460,10 +460,10 @@ char* TessPDFRenderer::GetPDFTextObjects(TessBaseAPI* api,
     STRING pdf_word("");
     int pdf_word_len = 0;
     do {
-      const char *grapheme = res_it->GetUTF8Text(RIL_SYMBOL);
+      const std::unique_ptr<const char[]> grapheme = res_it->GetUTF8Text(RIL_SYMBOL);
       if (grapheme && grapheme[0] != '\0') {
         GenericVector<int> unicodes;
-        UNICHAR::UTF8ToUnicode(grapheme, &unicodes);
+        UNICHAR::UTF8ToUnicode(grapheme.get(), &unicodes);
         char utf16[kMaxBytesPerCodepoint];
         for (int i = 0; i < unicodes.length(); i++) {
           int code = unicodes[i];
@@ -473,7 +473,6 @@ char* TessPDFRenderer::GetPDFTextObjects(TessBaseAPI* api,
           }
         }
       }
-      delete []grapheme;
       res_it->Next(RIL_SYMBOL);
     } while (!res_it->Empty(RIL_BLOCK) && !res_it->IsAtBeginningOf(RIL_WORD));
     if (word_length > 0 && pdf_word_len > 0 && fontsize > 0) {

diff --git a/ccmain/paragraphs.cpp b/ccmain/paragraphs.cpp
@@ -2446,8 +2446,8 @@ void InitializeRowInfo(bool after_recognition,
     return;
   }
   info->text = "";
-  char *text = it.GetUTF8Text(RIL_TEXTLINE);
-  int trailing_ws_idx = strlen(text);  // strip trailing space
+  const std::unique_ptr<const char[]> text = it.GetUTF8Text(RIL_TEXTLINE);
+  int trailing_ws_idx = strlen(text.get());  // strip trailing space
   while (trailing_ws_idx > 0 &&
          // isspace() only takes ASCII
          ((text[trailing_ws_idx - 1] & 0x80) == 0) &&
@@ -2460,7 +2460,6 @@ void InitializeRowInfo(bool after_recognition,
     for (int i = 0; i < trailing_ws_idx; i++)
       info->text += text[i];
   }
-  delete []text;
 
   if (info->text.size() == 0) {
     return;

diff --git a/ccmain/resultiterator.cpp b/ccmain/resultiterator.cpp
@@ -551,10 +551,10 @@ bool ResultIterator::IsAtFinalElement(PageIteratorLevel level,
 
 /**
  * Returns the null terminated UTF-8 encoded text string for the current
- * object at the given level. Use delete [] to free after use.
+ * object at the given level.
  */
-char* ResultIterator::GetUTF8Text(PageIteratorLevel level) const {
-  if (it_->word() == NULL) return NULL;  // Already at the end!
+std::unique_ptr<char[]> ResultIterator::GetUTF8Text(PageIteratorLevel level) const {
+  if (it_->word() == NULL) return nullptr;  // Already at the end!
   STRING text;
   switch (level) {
     case RIL_BLOCK:
@@ -591,8 +591,8 @@ char* ResultIterator::GetUTF8Text(PageIteratorLevel level) const {
       break;
   }
   int length = text.length() + 1;
-  char* result = new char[length];
-  strncpy(result, text.string(), length);
+  std::unique_ptr<char[]> result(new char[length]);
+  strncpy(result.get(), text.string(), length);
   return result;
 }
 

diff --git a/ccmain/resultiterator.h b/ccmain/resultiterator.h
@@ -22,6 +22,7 @@
 #ifndef TESSERACT_CCMAIN_RESULT_ITERATOR_H_
 #define TESSERACT_CCMAIN_RESULT_ITERATOR_H_
 
+#include <memory> // std::unique_ptr
 #include "platform.h"
 #include "ltrresultiterator.h"
 
@@ -86,9 +87,9 @@ class TESS_API ResultIterator : public LTRResultIterator {
 
   /**
    * Returns the null terminated UTF-8 encoded text string for the current
-   * object at the given level. Use delete [] to free after use.
+   * object at the given level.
   */
-  virtual char* GetUTF8Text(PageIteratorLevel level) const;
+  virtual std::unique_ptr<char[]> GetUTF8Text(PageIteratorLevel level) const;
 
   /**
    * Return whether the current paragraph's dominant reading direction