Harmless improvements from 3.00 going in to 2.04

git-svn-id: https://tesseract-ocr.googlecode.com/svn/trunk@209 d0cd1f9f-072b-0410-8dd7-cf729c803f20
lzpfmh · Dec 30, 2008 · 93c8e5d · 93c8e5d
1 parent de46248
commit 93c8e5d
Show file tree

Hide file tree

Showing 12 changed files with 139 additions and 103 deletions.
diff --git a/ccutil/boxread.cpp b/ccutil/boxread.cpp
@@ -1,8 +1,8 @@
 /**********************************************************************
  * File:        boxread.cpp
  * Description: Read data from a box file.
- * Author:		Ray Smith
- * Created:		Fri Aug 24 17:47:23 PDT 2007
+ * Author:      Ray Smith
+ * Created:     Fri Aug 24 17:47:23 PDT 2007
  *
  * (C) Copyright 2007, Google Inc.
  ** Licensed under the Apache License, Version 2.0 (the "License");
@@ -23,12 +23,25 @@
 #include "unichar.h"
 #include "tprintf.h"
 
+// Box files are used ONLY DURING TRAINING, but by both processes of
+// creating tr files with tesseract, and unicharset_extractor.
+// read_next_box factors out the code to interpret a line of a box
+// file so that applybox and unicharset_extractor interpret the same way.
+// This function returns the next valid box file utf8 string and coords
+// and returns true, or false on eof (and closes the file).
+// It ignores the uft8 file signature, checks for valid utf-8 and allows
+// space or tab between fields.
+// utf8_str must be at least kBoxReadBufSize in length.
+// If there are page numbers in the file, it reads them all.
 bool read_next_box(FILE* box_file, char* utf8_str,
                    int* x_min, int* y_min, int* x_max, int* y_max) {
   return read_next_box(-1, box_file, utf8_str,
                        x_min, y_min, x_max, y_max);
 }
 
+// As read_next_box above, but get a specific page number. (0-based)
+// Use -1 to read any page number. Files without page number all
+// read as if they are page 0.
 bool read_next_box(int target_page, FILE* box_file, char* utf8_str,
                    int* x_min, int* y_min, int* x_max, int* y_max) {
   static int line = 0;

diff --git a/ccutil/boxread.h b/ccutil/boxread.h
@@ -17,8 +17,8 @@
  *
  **********************************************************************/
 
-#ifndef THIRD_PARTY_TESSERACT_CCUTIL_BOXREAD_H__
-#define THIRD_PARTY_TESSERACT_CCUTIL_BOXREAD_H__
+#ifndef TESSERACT_CCUTIL_BOXREAD_H__
+#define TESSERACT_CCUTIL_BOXREAD_H__
 
 #include <stdio.h>
 
@@ -41,5 +41,4 @@ bool read_next_box(FILE* box_file, char* utf8_str,
 bool read_next_box(int page, FILE* box_file, char* utf8_str,
                    int* x_min, int* y_min, int* x_max, int* y_max);
 
-#endif  // THIRD_PARTY_TESSERACT_CCUTIL_BOXREAD_H__
-
+#endif  // TESSERACT_CCUTIL_BOXREAD_H__
diff --git a/ccutil/errcode.cpp b/ccutil/errcode.cpp
@@ -31,7 +31,6 @@
 
 const ERRCODE BADERRACTION = "Illegal error action";
 #define MAX_MSG       1024
-extern inT16 global_abort_code;
 
 /**********************************************************************
  * error
@@ -76,27 +75,29 @@ const char *format, ...          //special message
                                  //no specific
     msgptr += sprintf (msgptr, "\n");
 
-  tprintf(msg);
-  if ((strstr (message, "File") != NULL) ||
+  fprintf(stderr, msg);
+  /*if ((strstr (message, "File") != NULL) ||
     (strstr (message, "file") != NULL))
-    global_abort_code = FILE_ABORT;
   else if ((strstr (message, "List") != NULL) ||
     (strstr (message, "list") != NULL))
-    global_abort_code = LIST_ABORT;
   else if ((strstr (message, "Memory") != NULL) ||
     (strstr (message, "memory") != NULL))
     global_abort_code = MEMORY_ABORT;
   else
     global_abort_code = NO_ABORT_CODE;
+    */
 
+  int* p = NULL;
   switch (action) {
     case DBG:
     case TESSLOG:
       return;                    //report only
     case EXIT:
-      err_exit();
+      //err_exit();
     case ABORT:
-      abort();
+      // Create a deliberate segv as the stack trace is more useful that way.
+      if (!*p)
+        abort();
     default:
       BADERRACTION.error ("error", ABORT, NULL);
   }

diff --git a/ccutil/globaloc.cpp b/ccutil/globaloc.cpp
@@ -22,18 +22,18 @@
 #include          "errcode.h"
 #include          "tprintf.h"
 
-inT16 global_loc_code = LOC_INIT;//location code
+/*inT16 global_loc_code = LOC_INIT;//location code
 inT16 global_subloc_code = SUBLOC_NORM;
                                  //pass2 subloc code
 inT16 global_subsubloc_code = SUBSUBLOC_OTHER;
                                  //location code
 inT16 global_abort_code = NO_ABORT_CODE;
                                  //Prog abort code
-
+*/
 void signal_exit(                 //
                  int signal_code  //Signal which
                 ) {
-  int exit_status;
+  /*int exit_status;
 
   if ((global_loc_code == LOC_PASS2) || (global_loc_code == LOC_FUZZY_SPACE))
     global_loc_code += global_subloc_code + global_subsubloc_code;
@@ -49,7 +49,8 @@ void signal_exit(                 //
       exit_status, global_loc_code, signal_code);
   }
 
-  exit(exit_status);
+  exit(exit_status);*/
+  exit(signal_code);
 }
 
 
@@ -95,19 +96,20 @@ void signal_termination_handler(  //The real signal
 
 //};                                                                                                            //end extern "C"
 
+
 void set_global_loc_code(int loc_code) {
-  global_loc_code = loc_code;
+  // global_loc_code = loc_code;
 
 }
 
 
 void set_global_subloc_code(int loc_code) {
-  global_subloc_code = loc_code;
+  // global_subloc_code = loc_code;
 
 }
 
 
 void set_global_subsubloc_code(int loc_code) {
-  global_subsubloc_code = loc_code;
+  // global_subsubloc_code = loc_code;
 
 }
diff --git a/ccutil/host.h b/ccutil/host.h
@@ -171,4 +171,10 @@ typedef unsigned char BOOL8;
 #define NULL            0L
 #endif
 
+// Return true if x is within tolerance of y
+template<class T> bool NearlyEqual(T x, T y, T tolerance) {
+  T diff = x - y;
+  return diff <= tolerance && -diff <= tolerance;
+}
+
 #endif
diff --git a/ccutil/mainblk.cpp b/ccutil/mainblk.cpp
@@ -40,7 +40,10 @@ EXTERN DLLSYM STRING imagebasename;
 EXTERN BOOL_VAR (m_print_variables, FALSE,
 "Print initial values of all variables");
 EXTERN STRING_VAR (m_data_sub_dir, "tessdata/", "Directory for data files");
-EXTERN INT_VAR (memgrab_size, 0, "Preallocation size for batch use");
+/*
+EXTERN INT_VAR (memgrab_size, 0, "Preallocation size for batch use");*/
+
+
 const ERRCODE NO_PATH =
 "Warning:explicit path for executable will not be used for configs";
 static const ERRCODE USAGE = "Usage";
@@ -51,7 +54,7 @@ static const ERRCODE USAGE = "Usage";
  * Main for mithras demo program. Read the arguments and set up globals.
  **********************************************************************/
 
-void main_setup(                         /*main demo program */
+void main_setup(                 /*main demo program */
                 const char *argv0,       //program name
                 const char *basename,    //name of image
                 int argc,                /*argument count */
@@ -65,21 +68,29 @@ void main_setup(                         /*main demo program */
 
   imagebasename = basename;      /*name of image */
 
-  if(!getenv("TESSDATA_PREFIX")) {
-  #ifdef TESSDATA_PREFIX
-  #define _STR(a) #a
-  #define _XSTR(a) _STR(a)
-  datadir = _XSTR(TESSDATA_PREFIX);
-  #undef _XSTR
-  #undef _STR
-  #else
-  if (getpath (argv0, datadir) < 0)
-  #ifdef __UNIX__
-    CANTOPENFILE.error ("main", ABORT, "%s to get path", argv[0]);
-  #else
-  NO_PATH.error ("main", DBG, NULL);
-  #endif
-  #endif
+  // TESSDATA_PREFIX Environment variable overrules everything.
+  // Compiled in -DTESSDATA_PREFIX is next.
+  // NULL goes to current directory.
+  // An actual value of argv0 is used if getpath is successful.
+  if (!getenv("TESSDATA_PREFIX")) {
+#ifdef TESSDATA_PREFIX
+#define _STR(a) #a
+#define _XSTR(a) _STR(a)
+    datadir = _XSTR(TESSDATA_PREFIX);
+#undef _XSTR
+#undef _STR
+#else
+    if (argv0 != NULL) {
+      if (getpath(argv0, datadir) < 0)
+#ifdef __UNIX__
+        CANTOPENFILE.error("main", ABORT, "%s to get path", argv0);
+#else
+        NO_PATH.error("main", DBG, NULL);
+#endif
+    } else {
+      datadir = "./";
+    }
+#endif
   } else {
     datadir = getenv("TESSDATA_PREFIX");
   }
@@ -114,18 +125,4 @@ void main_setup(                         /*main demo program */
 
 
   datadir += m_data_sub_dir;     /*data directory */
-
-  #ifdef __UNIX__
-  if (memgrab_size > 0) {
-    void *membuf;                //test virtual mem
-                                 //test memory
-    membuf = malloc (memgrab_size);
-    if (membuf == NULL) {
-      raise(SIGTTOU);  //hangup for jobber
-      sleep (10);
-    }
-    else
-      free(membuf);
-  }
-  #endif
 }
diff --git a/ccutil/memry.cpp b/ccutil/memry.cpp
@@ -114,7 +114,8 @@ DLLSYM char *alloc_string(             //allocate string
   }
   return &string[1];             //string for user
 #else
-  return static_cast<char*>(malloc(count));
+  // Round up the amount allocated to a multiple of 4
+  return static_cast<char*>(malloc((count + 3) & ~3));
 #endif
 }
 

diff --git a/ccutil/ndminx.h b/ccutil/ndminx.h
@@ -21,11 +21,11 @@
 #define           NDMINX_H
 
 #ifndef MAX
-  #define MAX(a,b)        ( (a>b) ? a : b )
+#define MAX(x,y)    (((x) >= (y))?(x):(y))
 #endif
 
 #ifndef MIN
-  #define MIN(a,b)        ( (a<b) ? a : b )
+#define MIN(x,y)    (((x) <= (y))?(x):(y))
 #endif
 
 #endif
diff --git a/ccutil/unichar.h b/ccutil/unichar.h
@@ -17,8 +17,8 @@
 //
 ///////////////////////////////////////////////////////////////////////
 
-#ifndef THIRD_PARTY_TESSERACT_CCUTIL_UNICHAR_H__
-#define THIRD_PARTY_TESSERACT_CCUTIL_UNICHAR_H__
+#ifndef TESSERACT_CCUTIL_UNICHAR_H__
+#define TESSERACT_CCUTIL_UNICHAR_H__
 
 #include <memory.h>
 
@@ -29,6 +29,11 @@
 // A UNICHAR_ID is the unique id of a unichar.
 typedef int UNICHAR_ID;
 
+// A variable to indicate an invalid or uninitialized unichar id.
+static const int INVALID_UNICHAR_ID = -1;
+// A special unichar that corresponds to INVALID_UNICHAR_ID.
+static const char INVALID_UNICHAR[] = "__INVALID_UNICHAR__";
+
 // The UNICHAR class holds a single classification result. This may be
 // a single Unicode character (stored as between 1 and 4 utf8 bytes) or
 // multple Unicode characters representing the NFKC expansion of a ligature
@@ -76,4 +81,4 @@ class UNICHAR {
   char chars[UNICHAR_LEN];
 };
 
-#endif  // THIRD_PARTY_TESSERACT_CCUTIL_UNICHAR_H__
+#endif  // TESSERACT_CCUTIL_UNICHAR_H__
diff --git a/ccutil/unicharmap.h b/ccutil/unicharmap.h
@@ -17,8 +17,8 @@
 //
 ///////////////////////////////////////////////////////////////////////
 
-#ifndef THIRD_PARTY_TESSERACT_CCUTIL_UNICHARMAP_H__
-#define THIRD_PARTY_TESSERACT_CCUTIL_UNICHARMAP_H__
+#ifndef TESSERACT_CCUTIL_UNICHARMAP_H__
+#define TESSERACT_CCUTIL_UNICHARMAP_H__
 
 #include "unichar.h"
 
@@ -79,4 +79,4 @@ class UNICHARMAP {
   UNICHARMAP_NODE* nodes;
 };
 
-#endif  // THIRD_PARTY_TESSERACT_CCUTIL_UNICHARMAP_H__
+#endif  // TESSERACT_CCUTIL_UNICHARMAP_H__
diff --git a/ccutil/varable.cpp b/ccutil/varable.cpp
@@ -584,7 +584,7 @@ DLLSYM BOOL8 read_variables_file(const char *file  // name to read
 
       if (!foundit) {
         anyerr = TRUE;         // had an error
-        tprintf("read_variables_file:variable not found: %s",
+        tprintf("read_variables_file:variable not found: %s\n",
           line);
       }
     }