pdf: move rendering image and resolution to the pdf renderer only.

pdf: tests add lib leptonica dependency in the make toolchain
phymbert · Apr 20, 2024 · 09b6875 · 09b6875
1 parent 94b95b1
commit 09b6875
Show file tree

Hide file tree

Showing 4 changed files with 92 additions and 98 deletions.
diff --git a/Makefile.am b/Makefile.am
@@ -1472,7 +1472,7 @@ progress_test_LDADD = $(GTEST_LIBS) $(GMOCK_LIBS) $(TESS_LIBS) $(LEPTONICA_LIBS)
 
 pdfrenderer_test_SOURCES = unittest/pdfrenderer_test.cc
 pdfrenderer_test_CPPFLAGS = $(unittest_CPPFLAGS)
-pdfrenderer_test_LDADD = $(TESS_LIBS) $(TRAINING_LIBS)
+pdfrenderer_test_LDADD = $(TESS_LIBS) $(TRAINING_LIBS) $(LEPTONICA_LIBS)
 
 qrsequence_test_SOURCES = unittest/qrsequence_test.cc
 qrsequence_test_CPPFLAGS = $(unittest_CPPFLAGS)

diff --git a/include/tesseract/renderer.h b/include/tesseract/renderer.h
@@ -106,23 +106,6 @@ class TESS_API TessResultRenderer {
     return imagenum_;
   }
 
-  /**
-   * Specifies an alternate image to render with the extracted text.
-   * It must be called after BeginDocument and before AddImage.
-   */
-  void SetRenderingImage(Pix *rendering_image) {
-    rendering_image_ = rendering_image;
-  }
-
-  /**
-   * Specifies the expected rendering resolution.
-   * If not set, rendering_dpi api params will be used, else the source image
-   * resolution.
-   */
-  void SetRenderingResolution(int rendering_dpi) {
-    rendering_dpi_ = rendering_dpi;
-  }
-
 protected:
   /**
    * Called by concrete classes.
@@ -156,21 +139,6 @@ class TESS_API TessResultRenderer {
   // This method will grow the output buffer if needed.
   void AppendData(const char *s, int len);
 
-  // Renderers can call this to get the actual image to render with extracted
-  // text. This method returns:
-  //  - the rendering image set by the caller or
-  //  - the input image scaled to the rendering_dpi field if defined or
-  //  - the input image from the api otherwise
-  Pix *GetRenderingImage(TessBaseAPI *api);
-
-  // Resolution of the rendering image either set manually by the caller or with
-  // the rendering_dpi api parameter.
-  int GetRenderingResolution(TessBaseAPI *api);
-
-  // Reset rendering image and dpi to previous state. Destroy scaled rendered
-  // image if exists.
-  void ResetRenderingState(Pix *rendering_image_prev, int rendering_dpi_prev);
-
   template <typename T>
   auto AppendData(T &&d) {
     AppendData(d.data(), d.size());
@@ -183,8 +151,6 @@ class TESS_API TessResultRenderer {
   const char *file_extension_; // standard extension for generated output
   std::string title_;          // title of document being rendered
   int imagenum_;               // index of last image added
-  Pix *rendering_image_;       // Image to render with the extracted text
-  int rendering_dpi_;          // Resolution of the rendering_image
   bool happy_;                 // I get grumpy when the disk fills up, etc.
 };
 
@@ -275,6 +241,26 @@ class TESS_API TessPDFRenderer : public TessResultRenderer {
   // we load a custom PDF font from this location.
   TessPDFRenderer(const char *outputbase, const char *datadir,
                   bool textonly = false);
+    // Reset rendering image and dpi to previous state. Destroy scaled rendered
+    // image if exists.
+    void ResetRenderingState(Pix *rendering_image_prev, int rendering_dpi_prev);
+
+    /**
+     * Specifies an alternate image to render with the extracted text.
+     * It must be called after BeginDocument and before AddImage.
+     */
+    void SetRenderingImage(Pix *rendering_image) {
+        rendering_image_ = rendering_image;
+    }
+
+    /**
+     * Specifies the expected rendering resolution.
+     * If not set, rendering_dpi api params will be used, else the source image
+     * resolution.
+     */
+    void SetRenderingResolution(int rendering_dpi) {
+        rendering_dpi_ = rendering_dpi;
+    }
 
 protected:
   bool BeginDocumentHandler() override;
@@ -292,12 +278,24 @@ class TESS_API TessPDFRenderer : public TessResultRenderer {
   std::vector<long int> pages_;   // object number for every /Page object
   std::string datadir_;           // where to find the custom font
   bool textonly_;                 // skip images if set
+  Pix *rendering_image_;          // Image to render with the extracted text
+  int rendering_dpi_;             // Resolution of the rendering_image
   // Bookkeeping only. DIY = Do It Yourself.
   void AppendPDFObjectDIY(size_t objectsize);
   // Bookkeeping + emit data.
   void AppendPDFObject(const char *data);
   // Create the /Contents object for an entire page.
   char *GetPDFTextObjects(TessBaseAPI *api, double width, double height);
+  // Renderers can call this to get the actual image to render with extracted
+  // text. This method returns:
+  //  - the rendering image set by the caller or
+  //  - the input image scaled to the rendering_dpi field if defined or
+  //  - the input image from the api otherwise
+  Pix *GetRenderingImage(TessBaseAPI *api);
+  // Resolution of the rendering image either set manually by the caller or with
+  // the rendering_dpi api parameter.
+  int GetRenderingResolution(TessBaseAPI *api);
+
   // Turn an image into a PDF object. Only transcode if we have to.
   static bool imageToPDFObj(Pix *pix, const char *filename, long int objnum,
                             char **pdf_object, long int *pdf_object_size,

diff --git a/src/api/pdfrenderer.cpp b/src/api/pdfrenderer.cpp
@@ -192,7 +192,10 @@ static const int kMaxBytesPerCodepoint = 20;
  * PDF Renderer interface implementation
  **********************************************************************/
 TessPDFRenderer::TessPDFRenderer(const char *outputbase, const char *datadir, bool textonly)
-    : TessResultRenderer(outputbase, "pdf"), datadir_(datadir) {
+    : TessResultRenderer(outputbase, "pdf")
+    , datadir_(datadir)
+    , rendering_image_(nullptr)
+    , rendering_dpi_(0) {
   obj_ = 0;
   textonly_ = textonly;
   offsets_.push_back(0);
@@ -834,7 +837,60 @@ bool TessPDFRenderer::imageToPDFObj(Pix *pix, const char *filename, long int obj
   return true;
 }
 
+void TessPDFRenderer::ResetRenderingState(Pix *rendering_image_prev,
+                                             int rendering_dpi_prev) {
+  if (rendering_image_ != rendering_image_prev) {
+    pixDestroy(&rendering_image_);
+    rendering_image_ = rendering_image_prev;
+  }
+  rendering_dpi_ = rendering_dpi_prev;
+}
+
+Pix *TessPDFRenderer::GetRenderingImage(TessBaseAPI *api) {
+  if (!rendering_image_) {
+    Pix *source_image = api->GetInputImage();
+    int source_dpi = api->GetSourceYResolution();
+    if (!source_image || source_dpi <= 0) {
+      return nullptr;
+    }
+
+    int rendering_dpi = GetRenderingResolution(api);
+    if (rendering_dpi != source_dpi) {
+      float scale = (float)rendering_dpi / (float)source_dpi;
+      rendering_image_ = pixScale(source_image, scale, scale);
+    } else {
+      return source_image;
+    }
+  }
+  return rendering_image_;
+}
+
+int TessPDFRenderer::GetRenderingResolution(tesseract::TessBaseAPI *api) {
+  if (rendering_dpi_) {
+    return rendering_dpi_;
+  }
+  int source_dpi = api->GetSourceYResolution();
+  int rendering_dpi;
+  if (api->GetIntVariable("rendering_dpi", &rendering_dpi) &&
+      rendering_dpi > 0 && rendering_dpi != source_dpi) {
+    if (rendering_dpi < kMinCredibleResolution ||
+        rendering_dpi > kMaxCredibleResolution) {
+#if !defined(NDEBUG)
+      tprintf(
+          "Warning: User defined rendering dpi %d is outside of expected range "
+          "(%d - %d)!\n",
+          rendering_dpi, kMinCredibleResolution, kMaxCredibleResolution);
+#endif
+    }
+    rendering_dpi_ = rendering_dpi;
+    return rendering_dpi_;
+  }
+  return source_dpi;
+}
+
 bool TessPDFRenderer::AddImageHandler(TessBaseAPI *api) {
+  Pix *rendering_image_prev = rendering_image_;
+  int rendering_dpi_prev = rendering_dpi_;
   Pix *pix = GetRenderingImage(api);
   const char *filename = api->GetInputName();
   int ppi = GetRenderingResolution(api);
@@ -920,12 +976,14 @@ bool TessPDFRenderer::AddImageHandler(TessBaseAPI *api) {
     int jpg_quality;
     api->GetIntVariable("jpg_quality", &jpg_quality);
     if (!imageToPDFObj(pix, filename, obj_, &pdf_object, &objsize, jpg_quality)) {
+      ResetRenderingState(rendering_image_prev, rendering_dpi_prev);
       return false;
     }
     AppendData(pdf_object, objsize);
     AppendPDFObjectDIY(objsize);
     delete[] pdf_object;
   }
+  ResetRenderingState(rendering_image_prev, rendering_dpi_prev);
   return true;
 }
 

diff --git a/src/api/renderer.cpp b/src/api/renderer.cpp
@@ -18,14 +18,12 @@
 #ifdef HAVE_CONFIG_H
 #  include "config_auto.h"
 #endif
-#include <allheaders.h>
 #include <tesseract/baseapi.h>
 #include <tesseract/renderer.h>
 #include <cstring>
 #include <memory>     // std::unique_ptr
 #include <string>     // std::string
 #include "serialis.h" // Serialize
-#include "tprintf.h"
 
 namespace tesseract {
 
@@ -38,8 +36,6 @@ TessResultRenderer::TessResultRenderer(const char *outputbase, const char *exten
     , file_extension_(extension)
     , title_("")
     , imagenum_(-1)
-    , rendering_image_(nullptr)
-    , rendering_dpi_(0)
     , happy_(true) {
   if (strcmp(outputbase, "-") && strcmp(outputbase, "stdout")) {
     std::string outfile = std::string(outputbase) + "." + extension;
@@ -94,71 +90,13 @@ bool TessResultRenderer::AddImage(TessBaseAPI *api) {
     return false;
   }
   ++imagenum_;
-  Pix *rendering_image_prev = rendering_image_;
-  int rendering_dpi_prev = rendering_dpi_;
   bool ok = AddImageHandler(api);
-  ResetRenderingState(rendering_image_prev, rendering_dpi_prev);
   if (next_) {
     ok = next_->AddImage(api) && ok;
   }
   return ok;
 }
 
-void TessResultRenderer::ResetRenderingState(Pix *rendering_image_prev,
-                                             int rendering_dpi_prev) {
-  if (rendering_image_ != rendering_image_prev) {
-    pixDestroy(&rendering_image_);
-    rendering_image_ = rendering_image_prev;
-  }
-  if (rendering_dpi_ != rendering_dpi_prev) {
-    rendering_dpi_ = rendering_dpi_prev;
-  }
-}
-
-Pix *TessResultRenderer::GetRenderingImage(TessBaseAPI *api) {
-  if (!rendering_image_) {
-    Pix *source_image = api->GetInputImage();
-    int source_dpi = api->GetSourceYResolution();
-    if (!source_image || source_dpi <= 0) {
-      happy_ = false;
-      return nullptr;
-    }
-
-    int rendering_dpi = GetRenderingResolution(api);
-    if (rendering_dpi != source_dpi) {
-      float scale = (float)rendering_dpi / (float)source_dpi;
-
-      rendering_image_ = pixScale(source_image, scale, scale);
-    } else {
-      return source_image;
-    }
-  }
-  return rendering_image_;
-}
-
-int TessResultRenderer::GetRenderingResolution(tesseract::TessBaseAPI *api) {
-  if (rendering_dpi_) {
-    return rendering_dpi_;
-  }
-  int source_dpi = api->GetSourceYResolution();
-  int rendering_dpi;
-  if (api->GetIntVariable("rendering_dpi", &rendering_dpi) &&
-      rendering_dpi > 0 && rendering_dpi != source_dpi) {
-    if (rendering_dpi < kMinCredibleResolution ||
-        rendering_dpi > kMaxCredibleResolution) {
-#if !defined(NDEBUG)
-      tprintf(
-          "Warning: User defined rendering dpi %d is outside of expected range "
-          "(%d - %d)!\n",
-          rendering_dpi, kMinCredibleResolution, kMaxCredibleResolution);
-#endif
-    }
-    rendering_dpi_ = rendering_dpi;
-    return rendering_dpi_;
-  }
-  return source_dpi;
-}
-
 bool TessResultRenderer::EndDocument() {
   if (!happy_) {
     return false;