dawg2wordlist.cpp File Reference

#include "commontraining.h"
#include "dawg.h"
#include <tesseract/serialis.h>
#include "trie.h"
#include "unicharset.h"

Go to the source code of this file.

Classes
class	WordOutputter

Functions
int	main (int argc, char *argv[])

Function Documentation

main()

int main	(	int	argc,
		char *	argv[]
	)

Definition at line 69 of file dawg2wordlist.cpp.

                                 {
  tesseract::CheckSharedLibraryVersion();
 
  if (argc > 1 && (!strcmp(argv[1], "-v") || !strcmp(argv[1], "--version"))) {
    printf("%s\n", tesseract::TessBaseAPI::Version());
    return 0;
  } else if (argc != 4) {
    tprintf("Print all the words in a given dawg.\n");
    tprintf("Usage: %s -v | --version | %s <unicharset> <dawgfile> <wordlistfile>\n",
            argv[0], argv[0]);
    return 1;
  }
  const char *unicharset_file = argv[1];
  const char *dawg_file = argv[2];
  const char *wordlist_file = argv[3];
  UNICHARSET unicharset;
  if (!unicharset.load_from_file(unicharset_file)) {
    tprintf("Error loading unicharset from %s.\n", unicharset_file);
    return 1;
  }
  tesseract::Dawg *dict = LoadSquishedDawg(unicharset, dawg_file);
  if (dict == nullptr) {
    tprintf("Error loading dictionary from %s.\n", dawg_file);
    return 1;
  }
  int retval = WriteDawgAsWordlist(unicharset, dict, wordlist_file);
  delete dict;
  return retval;
}