enforce v5

Author: pachadotdev

.github/workflows/clang19.yaml

@@ -4,7 +4,7 @@ on:
   pull_request:
     branches: [main, master]
 
-name: "clang 19"
+name: "clang19-rhub"
 
 jobs:
   check:

DESCRIPTION

@@ -25,7 +25,7 @@ License: Apache License (>= 2)
 URL: https://pacha.dev/cpp11tesseract/
 BugReports: https://github.com/pachadotdev/cpp11tesseract/issues
 SystemRequirements:
-    Tesseract OCR (
+    Tesseract OCR (>= 5.0.0) (
         deb: libtesseract-dev libleptonica-dev tesseract-ocr-eng,
         rpm: tesseract-devel leptonica-devel tesseract-langpack-eng,
         brew: tesseract leptonica

configure

@@ -1,3 +1,4 @@
+# filepath: /home/pacha/github/cpp11tesseract/configure
 #!/bin/sh
 
 # Anticonf (tm) script by Pacha (2024)
@@ -8,8 +9,8 @@
 
 # Library settings
 PKG_CONFIG_NAME="tesseract"
-PKG_DEB_NAME="libtesseract-dev libleptonica-dev"
-PKG_RPM_NAME="tesseract-devel leptonica-devel"
+PKG_DEB_NAME="libtesseract-dev libleptonica-dev tesseract-ocr-eng"
+PKG_RPM_NAME="tesseract-devel leptonica-devel tesseract-langpack-eng"
 PKG_BREW_NAME="tesseract leptonica"
 PKG_CFLAGS="-I/usr/include/leptonica"
 PKG_LIBS="-ltesseract"
@@ -50,12 +51,16 @@ echo "PKG_LIBS: $PKG_LIBS"
 
 # Check if the compiler is clang
 if [ "$CC" = "clang" ] || [ "$CXX" = "clang++" ]; then
-  CXX=clang++ -stdlib=libstdc++
-  CXXFLAGS += -stdlib=libstdc++
-  LDFLAGS += -stdlib=libstdc++
+  PKG_LIBS="$PKG_LIBS -stdlib=libc++"
 fi
 
-# Tesseract 4 enforces C++11 in the headers but R does not have CXX11CPP
+# Stop if Tesseract < 5.0.0
+if ! tesseract --version | grep -q "tesseract 5"; then
+  echo "Tesseract >= 5.0.0 is required"
+  exit 1
+fi
+
+# Tesseract 5 enforces C++11 in the headers but R does not have CXX11CPP
 CPPFLAGS=`${R_HOME}/bin/R CMD config CPPFLAGS`
 CXX11STD=`${R_HOME}/bin/R CMD config CXX11STD`
 if [ $? -eq 0 ]; then

src/cpp11tesseract.cpp

@@ -1,38 +1,17 @@
 #include "cpp11tesseract_types.h"
 
-#if TESSERACT_MAJOR_VERSION < 5
-#include <tesseract/genericvector.h>
-#else
 #define STRING std::string
 #define GenericVector std::vector
-#endif
 
 [[cpp11::register]] int tesseract_major_version() {
   return TESSERACT_MAJOR_VERSION;
 }
 
 using namespace cpp11;
 
-/* libtesseract 4.0 insisted that the engine is initiated in 'C' locale.
- * We do this as exemplified in the example code in the libc manual:
- * https://www.gnu.org/software/libc/manual/html_node/Setting-the-Locale.html
- * Full discussion: https://github.com/tesseract-ocr/tesseract/issues/1670
- */
-#if TESSERACT_MAJOR_VERSION == 4 && TESSERACT_MINOR_VERSION == 0
-#define TESSERACT40
-#endif
-
 static tesseract::TessBaseAPI *make_analyze_api() {
-#ifdef TESSERACT40
-  char *old_ctype = strdup(setlocale(LC_ALL, NULL));
-  setlocale(LC_ALL, "C");
-#endif
   tesseract::TessBaseAPI *api = new tesseract::TessBaseAPI();
   api->InitForAnalysePage();
-#ifdef TESSERACT40
-  setlocale(LC_ALL, old_ctype);
-  free(old_ctype);
-#endif
   return api;
 }
 
@@ -70,11 +49,6 @@ static tesseract::TessBaseAPI *make_analyze_api() {
     configs.push_back(&config_strings.back()[0]);
   }
 
-#ifdef TESSERACT40
-  char *old_ctype = strdup(setlocale(LC_ALL, NULL));
-  setlocale(LC_ALL, "C");
-#endif
-
   tesseract::TessBaseAPI *api = new tesseract::TessBaseAPI();
 
   GenericVector<STRING> params, values;
@@ -85,11 +59,6 @@ static tesseract::TessBaseAPI *make_analyze_api() {
   int err = api->Init(path, lang, tesseract::OEM_DEFAULT, configs.data(),
                       configs.size(), &params, &values, false);
 
-#ifdef TESSERACT40
-  setlocale(LC_ALL, old_ctype);
-  free(old_ctype);
-#endif
-
   if (err) {
     delete api;
     throw std::runtime_error(
@@ -124,13 +93,7 @@ tesseract::TessBaseAPI *get_engine(TessPtr engine) {
   writable::logicals out(params.size());
   STRING str;
   for (int i = 0; i < params.size(); i++) {
-#if TESSERACT_MAJOR_VERSION >= 5
     out[i] = api->GetVariableAsString(std::string(params.at(i)).c_str(), &str);
-#else
-    const char *value =
-        api->GetStringVariable(std::string(params.at(i)).c_str());
-    out[i] = (value != nullptr);
-#endif
   }
   api->End();
   delete api;
@@ -142,15 +105,9 @@ tesseract::TessBaseAPI *get_engine(TessPtr engine) {
   GenericVector<STRING> langs;
   api->GetAvailableLanguagesAsVector(&langs);
   writable::strings available;
-#if TESSERACT_MAJOR_VERSION >= 5
   for (const auto &lang : langs) {
     available.push_back(lang);
   }
-#else
-  for (int i = 0; i < langs.size(); i++) {
-    available.push_back(langs.get(i).c_str());
-  }
-#endif
   langs.clear();
   api->GetLoadedLanguagesAsVector(&langs);
   writable::strings loaded;
@@ -175,15 +132,8 @@ tesseract::TessBaseAPI *get_engine(TessPtr engine) {
   std::vector<std::string> values;
   for (int i = 0; i < params.size(); ++i) {
     STRING str;
-#if TESSERACT_MAJOR_VERSION >= 5
     if (api->GetVariableAsString(std::string(params.at(i)).c_str(), &str)) {
       values.push_back(str);
-#else
-    const char *value =
-        api->GetStringVariable(std::string(params.at(i)).c_str());
-    if (value) {
-      values.push_back(value);
-#endif
     } else {
       values.push_back("");
     }

src/cpp11tesseract_types.h

@@ -1,13 +1,12 @@
-#include <tesseract/baseapi.h> // tesseract
-#include <allheaders.h> // leptonica
+#include <allheaders.h>         // leptonica
+#include <tesseract/baseapi.h>  // tesseract
 
-#include <memory>
+#include <cpp11.hpp>
 #include <list>
+#include <memory>
 #include <string>
 #include <vector>
 
-#include <cpp11.hpp>
-
 inline void tess_finalizer(tesseract::TessBaseAPI* engine) {
   engine->End();
   delete engine;