Merge pull request #1479 from stweil/version

training: Add initial support for --version argument and check library version

api/capi.h

@@ -18,6 +18,10 @@
 #ifndef API_CAPI_H_
 #define API_CAPI_H_
 
+#if defined(TESSERACT_API_BASEAPI_H_) && !defined(TESS_CAPI_INCLUDE_BASEAPI)
+# define TESS_CAPI_INCLUDE_BASEAPI
+#endif
+
 #ifdef TESS_CAPI_INCLUDE_BASEAPI
 #   include "baseapi.h"
 #   include "pageiterator.h"

training/ambiguous_words.cpp

@@ -24,12 +24,14 @@
 #include <stdio.h>
 
 #include "baseapi.h"
+#include "commontraining.h"     // CheckSharedLibraryVersion
 #include "helpers.h"
 #include "strngs.h"
 #include "dict.h"
 #include "tesseractclass.h"
 
 int main(int argc, char** argv) {
+  tesseract::CheckSharedLibraryVersion();
 
   // Parse input arguments.
   if (argc != 4 && (argc != 6 || strcmp(argv[1], "-l") != 0)) {

training/classifier_tester.cpp

@@ -111,6 +111,7 @@ static tesseract::ShapeClassifier* InitializeClassifier(
 // full     : Tesseract full classifier.
 //            with an input trainer.)
 int main(int argc, char **argv) {
+  tesseract::CheckSharedLibraryVersion();
   ParseArguments(&argc, &argv);
   STRING file_prefix;
   tesseract::MasterTrainer* trainer =

training/cntraining.cpp

@@ -41,13 +41,6 @@
 
 DECLARE_STRING_PARAM_FLAG(D);
 
-/*----------------------------------------------------------------------------
-          Public Function Prototypes
-----------------------------------------------------------------------------*/
-int main (
-     int  argc,
-     char  **argv);
-
 /*----------------------------------------------------------------------------
           Private Function Prototypes
 ----------------------------------------------------------------------------*/
@@ -131,6 +124,8 @@ CLUSTERCONFIG  CNConfig =
 * @note History: Fri Aug 18 08:56:17 1989, DSJ, Created.
 */
 int main(int argc, char *argv[]) {
+  tesseract::CheckSharedLibraryVersion();
+
   // Set the global Config parameters before parsing the command line.
   Config = CNConfig;
 

training/combine_lang_model.cpp

@@ -15,7 +15,9 @@
 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 // See the License for the specific language governing permissions and
 // limitations under the License.
+
 #include "commandlineflags.h"
+#include "commontraining.h"     // CheckSharedLibraryVersion
 #include "lang_model_helpers.h"
 #include "tprintf.h"
 #include "unicharset_training_utils.h"
@@ -38,6 +40,7 @@ BOOL_PARAM_FLAG(pass_through_recoder, false,
                 " unicharset. Otherwise, potentially a compression of it");
 
 int main(int argc, char** argv) {
+  tesseract::CheckSharedLibraryVersion();
   tesseract::ParseCommandLineFlags(argv[0], &argc, &argv, true);
 
   // Check validity of input flags.

training/combine_tessdata.cpp

@@ -18,6 +18,7 @@
 //
 ///////////////////////////////////////////////////////////////////////
 
+#include "commontraining.h"     // CheckSharedLibraryVersion
 #include "lstmrecognizer.h"
 #include "tessdatamanager.h"
 
@@ -65,6 +66,8 @@
 // components from tessdata/eng.traineddata.
 //
 int main(int argc, char **argv) {
+  tesseract::CheckSharedLibraryVersion();
+
   int i;
   tesseract::TessdataManager tm;
   if (argc == 2) {

training/commandlineflags.cpp

@@ -7,6 +7,8 @@
 // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 // See the License for the specific language governing permissions and
 // limitations under the License.
+
+#include "baseapi.h"            // TessBaseAPI::Version
 #include "commandlineflags.h"
 
 #ifndef GOOGLE_TESSERACT
@@ -55,7 +57,6 @@ bool StringFlagExists(const char* flag_name, const char** value) {
   return p != nullptr;
 }
 
-
 void SetIntFlagValue(const char* flag_name, const int32_t new_val) {
   STRING full_flag_name("FLAGS_");
   full_flag_name += flag_name;
@@ -149,7 +150,6 @@ void PrintCommandLineFlags() {
   }
 }
 
-
 void ParseCommandLineFlags(const char* usage,
                            int* argc, char*** argv,
                            const bool remove_flags) {
@@ -157,6 +157,9 @@ void ParseCommandLineFlags(const char* usage,
     printf("USAGE: %s\n", usage);
     PrintCommandLineFlags();
     exit(0);
+  } else if (*argc > 1 && !strcmp((*argv)[1], "--version")) {
+    printf("%s\n", TessBaseAPI::Version());
+    exit(0);
   }
 
   unsigned int i = 1;
@@ -174,7 +177,7 @@ void ParseCommandLineFlags(const char* usage,
     // If this is asking for usage, print the help message and abort.
     if (!strcmp(current_arg, "help") ||
         !strcmp(current_arg, "helpshort")) {
-      tprintf("USAGE: %s\n", usage);
+      printf("USAGE: %s\n", usage);
       PrintCommandLineFlags();
       exit(0);
     }

training/commontraining.cpp

@@ -87,6 +87,8 @@ void ParseArguments(int* argc, char ***argv) {
   STRING usage;
   if (*argc) {
     usage += (*argv)[0];
+    usage += " --version | ";
+    usage += (*argv)[0];
   }
   usage += " [.tr files ...]";
   tesseract::ParseCommandLineFlags(usage.c_str(), argc, argv, true);

training/commontraining.h

@@ -14,6 +14,10 @@
 #ifndef TESSERACT_TRAINING_COMMONTRAINING_H_
 #define TESSERACT_TRAINING_COMMONTRAINING_H_
 
+#ifdef HAVE_CONFIG_H
+#include "config_auto.h"
+#include "baseapi.h"
+#endif
 #include "cluster.h"
 #include "commandlineflags.h"
 #include "featdefs.h"
@@ -63,6 +67,21 @@ void ParseArguments(int* argc, char*** argv);
 
 namespace tesseract {
 
+// Check whether the shared tesseract library is the right one.
+// This function must be inline because otherwise it would be part of
+// the shared library, so it could not compare the versions.
+static inline void CheckSharedLibraryVersion()
+{
+#ifdef HAVE_CONFIG_H
+  if (!!strcmp(TESSERACT_VERSION_STR, TessBaseAPI::Version())) {
+    tprintf("ERROR: shared library version mismatch (was %s, expected %s\n"
+            "Did you use a wrong shared tesseract library?\n",
+            TessBaseAPI::Version(), TESSERACT_VERSION_STR);
+    exit(1);
+  }
+#endif
+}
+
 // Helper loads shape table from the given file.
 ShapeTable* LoadShapeTable(const STRING& file_prefix);
 // Helper to write the shape_table.

training/dawg2wordlist.cpp

@@ -17,6 +17,7 @@
 //
 ///////////////////////////////////////////////////////////////////////
 
+#include "commontraining.h"     // CheckSharedLibraryVersion
 #include "dawg.h"
 #include "host.h"
 #include "serialis.h"
@@ -72,6 +73,8 @@ int WriteDawgAsWordlist(const UNICHARSET &unicharset,
 }
 
 int main(int argc, char *argv[]) {
+  tesseract::CheckSharedLibraryVersion();
+
   if (argc != 4) {
     tprintf("Print all the words in a given dawg.\n");
     tprintf("Usage: %s <unicharset> <dawgfile> <wordlistfile>\n",

training/lstmeval.cpp

@@ -36,6 +36,7 @@ INT_PARAM_FLAG(verbosity, 1,
                "Amount of diagnosting information to output (0-2).");
 
 int main(int argc, char **argv) {
+  tesseract::CheckSharedLibraryVersion();
   ParseArguments(&argc, &argv);
   if (FLAGS_model.empty()) {
     tprintf("Must provide a --model!\n");

training/lstmtraining.cpp

@@ -71,6 +71,7 @@ const int kNumPagesPerBatch = 100;
 // The program iterates over the inputs, feeding the data to the network,
 // until the error rate reaches a specified target or max_iterations is reached.
 int main(int argc, char **argv) {
+  tesseract::CheckSharedLibraryVersion();
   ParseArguments(&argc, &argv);
   // Purify the model name in case it is based on the network string.
   if (FLAGS_model_output.empty()) {

training/merge_unicharsets.cpp

@@ -17,10 +17,12 @@
 //
 ///////////////////////////////////////////////////////////////////////
 
-#include <stdio.h>
+#include "commontraining.h"     // CheckSharedLibraryVersion
 #include "unicharset.h"
 
 int main(int argc, char** argv) {
+  tesseract::CheckSharedLibraryVersion();
+
   // Print usage
   if (argc < 4) {
     printf("Usage: %s unicharset-in-1 ... unicharset-in-n unicharset-out\n",

training/mftraining.cpp

@@ -76,14 +76,6 @@ const int kMaxShapeLabelLength = 10;
 
 DECLARE_STRING_PARAM_FLAG(test_ch);
 
-/*----------------------------------------------------------------------------
-          Public Function Prototypes
-----------------------------------------------------------------------------*/
-int main (
-     int  argc,
-     char  **argv);
-
-
 /*----------------------------------------------------------------------------
             Public Code
 -----------------------------------------------------------------------------*/
@@ -236,6 +228,8 @@ static void SetupConfigMap(ShapeTable* shape_table, IndexMapBiDi* config_map) {
  * @note History: Mon May 18 1998, Christy Russson, Revistion started.
  */
 int main (int argc, char **argv) {
+  tesseract::CheckSharedLibraryVersion();
+
   ParseArguments(&argc, &argv);
 
   ShapeTable* shape_table = nullptr;

training/set_unicharset_properties.cpp

@@ -12,11 +12,8 @@
 // object, fills it with properties about the unichars it contains and writes
 // the result back to a file.
 
-#include <stdlib.h>
-#include <string.h>
-#include <string>
-
 #include "commandlineflags.h"
+#include "commontraining.h"     // CheckSharedLibraryVersion
 #include "tprintf.h"
 #include "unicharset_training_utils.h"
 
@@ -30,6 +27,7 @@ DECLARE_STRING_PARAM_FLAG(O);
 DECLARE_STRING_PARAM_FLAG(X);
 
 int main(int argc, char** argv) {
+  tesseract::CheckSharedLibraryVersion();
   tesseract::ParseCommandLineFlags(argv[0], &argc, &argv, true);
 
   // Check validity of input flags.

training/shapeclustering.cpp

@@ -45,6 +45,8 @@ STRING_PARAM_FLAG(canonical_class2, "", "Class to show ambigs for");
 // Otherwise, if FLAGS_canonical_class1 is set, prints a table of font-wise
 // cluster distances between FLAGS_canonical_class1 and FLAGS_canonical_class2.
 int main(int argc, char **argv) {
+  tesseract::CheckSharedLibraryVersion();
+
   ParseArguments(&argc, &argv);
 
   STRING file_prefix;

training/text2image.cpp

@@ -40,6 +40,7 @@
 #include "allheaders.h"  // from leptonica
 #include "boxchar.h"
 #include "commandlineflags.h"
+#include "commontraining.h"     // CheckSharedLibraryVersion
 #include "degradeimage.h"
 #include "errcode.h"
 #include "fileio.h"
@@ -412,7 +413,7 @@ using tesseract::SpanUTF8NotWhitespace;
 using tesseract::SpanUTF8Whitespace;
 using tesseract::StringRenderer;
 
-int Main() {
+static int Main() {
   if (FLAGS_list_available_fonts) {
     const std::vector<std::string>& all_fonts = FontUtils::ListAvailableFonts();
     for (unsigned int i = 0; i < all_fonts.size(); ++i) {
@@ -672,6 +673,7 @@ int Main() {
 }
 
 int main(int argc, char** argv) {
+  tesseract::CheckSharedLibraryVersion();
   tesseract::ParseCommandLineFlags(argv[0], &argc, &argv, true);
-  Main();
+  return Main();
 }

training/unicharset_extractor.cpp

@@ -24,11 +24,11 @@
 #include <cstdlib>
 #include "boxread.h"
 #include "commandlineflags.h"
+#include "commontraining.h"     // CheckSharedLibraryVersion
 #include "genericvector.h"
 #include "lang_model_helpers.h"
 #include "normstrngs.h"
 #include "strngs.h"
-#include "tprintf.h"
 #include "unicharset.h"
 #include "unicharset_training_utils.h"
 
@@ -61,7 +61,7 @@ static void AddStringsToUnicharset(const GenericVector<STRING>& strings,
   }
 }
 
-int Main(int argc, char** argv) {
+static int Main(int argc, char** argv) {
   UNICHARSET unicharset;
   // Load input files
   for (int arg = 1; arg < argc; ++arg) {
@@ -95,6 +95,7 @@ int Main(int argc, char** argv) {
 }  // namespace tesseract
 
 int main(int argc, char** argv) {
+  tesseract::CheckSharedLibraryVersion();
   if (argc > 1) {
     tesseract::ParseCommandLineFlags(argv[0], &argc, &argv, true);
   }

training/wordlist2dawg.cpp

@@ -20,9 +20,8 @@
 // Given a file that contains a list of words (one word per line) this program
 // generates the corresponding squished DAWG file.
 
-#include <stdio.h>
-
 #include "classify.h"
+#include "commontraining.h"     // CheckSharedLibraryVersion
 #include "dawg.h"
 #include "dict.h"
 #include "emalloc.h"
@@ -32,6 +31,8 @@
 #include "unicharset.h"
 
 int main(int argc, char** argv) {
+  tesseract::CheckSharedLibraryVersion();
+
   if (!(argc == 4 || (argc == 5 && strcmp(argv[1], "-t") == 0) ||
       (argc == 6 && strcmp(argv[1], "-r") == 0))) {
     printf("Usage: %s [-t | -r [reverse policy] ] word_list_file"