update Tessract.h & Tessract.mm, adding clear method to avoid memory lea...

External/Tesseract-Wrapper/Tesseract.h

@@ -9,19 +9,65 @@
 
 #import <Foundation/Foundation.h>
 
-@interface Tesseract : NSObject {    
+extern NSString * const OcrEngineModeTesseractOnly;
+extern NSString * const OcrEngineModeCubeOnly;
+extern NSString * const OcrEngineModeTesseractCubeCombined;
+extern NSString * const OcrEngineModeDefault;
+
+@interface Tesseract : NSObject {
     NSString* _dataPath;
     NSString* _language;
     NSMutableDictionary* _variables;
 }
 
 + (NSString *)version;
 
+/**
+ * Returns <code>NO</code> or <code>nil</code> on failure. On success, Returns
+ * an initialized <code>Tesseract</code> object set up to use the given tessdata
+ * path, language, and OCR engine mode, configured with the given variables and
+ * configuration filenames.
+ *
+ * @param dataPath
+ *  The name of the parent directory of tessdata. Must end in / . Any name after
+ *  the last / will be stripped.
+ * @param language
+ *  An ISO 639-3 string or nil, which will default to eng.
+ *  The language may be a string of the form [~]<lang>[+[~]<lang>]* indicating
+ *  that multiple languages are to be loaded. Eg hin+eng will load Hindi and
+ *  English. Languages may specify internally that they want to be loaded with
+ *  one or more other languages, so the ~ sign is available to override that.
+ *  Eg if hin were set to load eng by default, then hin+~eng would force loading
+ *  only hin. The number of loaded languages is limited only by memory, with the
+ *  caveat that loading additional languages will impact both speed and
+ *  accuracy, as there is more work to do to decide on the applicable language,
+ *  and there is more chance of hallucinating incorrect words.
+ * @param mode
+ *  The OcrEngineMode to use or nil, which is equivalent to OcrEngineModeDefault
+ * @param configFilenames
+ *  The names (including paths) of config files to use. Example config files can
+ *  be found at
+ *  http://code.google.com/p/tesseract-ocr/source/browse/#svn/trunk/tessdata/configs
+ * @param variables
+ *  A dictionary of tesseract variables to set. Each key must be an NSString
+ *  instance corresponding to a tesseract variable, and each value must be an
+ *  NSString instance containing a valid value for that variable.
+ *  Eg @{@"tessedit_char_whitelist" : @"0123456789.-"}
+ * @param setOnlyNonDebugParams
+ *  If true, only params that do not contain "debug" in the name will be set.
+ */
+- (id)initWithDataPath:(NSString *)dataPath
+              language:(NSString *)language
+         ocrEngineMode:(NSString *)mode
+       configFilenames:(NSArray*)configFilenames
+             variables:(NSDictionary*)variables
+ setOnlyNonDebugParams:(BOOL)setOnlyNonDebugParams;
 - (id)initWithDataPath:(NSString *)dataPath language:(NSString *)language;
 - (void)setVariableValue:(NSString *)value forKey:(NSString *)key;
 - (void)setImage:(UIImage *)image;
 - (BOOL)setLanguage:(NSString *)language;
 - (BOOL)recognize;
 - (NSString *)recognizedText;
+- (void)clear;
 
 @end

External/Tesseract-Wrapper/Tesseract.mm

@@ -13,13 +13,22 @@
 #import "environ.h"
 #import "pix.h"
 
+NSString * const OcrEngineModeTesseractOnly = @"OcrEngineModeTesseractOnly";
+NSString * const OcrEngineModeCubeOnly = @"OcrEngineModeCubeOnly";
+NSString * const OcrEngineModeTesseractCubeCombined = @"OcrEngineModeTesseractCubeCombined";
+NSString * const OcrEngineModeDefault = @"OcrEngineModeDefault";
+
 namespace tesseract {
     class TessBaseAPI;
 };
 
 @interface Tesseract () {
     tesseract::TessBaseAPI* _tesseract;
     uint32_t* _pixels;
+    bool _setOnlyNonDebugParams;
+    tesseract::OcrEngineMode _ocrEngineMode;
+
+    NSArray* _configFilenames;
 }
 
 @end
@@ -30,26 +39,99 @@ + (NSString *)version {
     return [NSString stringWithFormat:@"%s", tesseract::TessBaseAPI::Version()];
 }
 
-- (id)initWithDataPath:(NSString *)dataPath language:(NSString *)language {
+- (id)initWithDataPath:(NSString *)dataPath
+              language:(NSString *)language
+         ocrEngineMode:(NSString *)mode
+       configFilenames:(NSArray*)configFilenames
+             variables:(NSDictionary*)variables
+ setOnlyNonDebugParams:(BOOL)setOnlyNonDebugParams {
     self = [super init];
     if (self) {
         _dataPath = dataPath;
         _language = language;
-        _variables = [[NSMutableDictionary alloc] init];
+
+        if (mode) {
+            if ([mode isEqualToString:OcrEngineModeTesseractOnly]) {
+                _ocrEngineMode = tesseract::OEM_TESSERACT_ONLY;
+            } else if ([mode isEqualToString:OcrEngineModeCubeOnly]) {
+                _ocrEngineMode = tesseract::OEM_CUBE_ONLY;
+            } else if ([mode isEqualToString:OcrEngineModeTesseractCubeCombined]) {
+                _ocrEngineMode = tesseract::OEM_TESSERACT_CUBE_COMBINED;
+            } else {
+                _ocrEngineMode = tesseract::OEM_DEFAULT;
+            }
+        } else {
+            _ocrEngineMode = tesseract::OEM_DEFAULT;
+        }
+
+        _configFilenames = configFilenames;
+
+        if (variables) {
+            _variables = [variables mutableCopyWithZone:NULL];
+        } else {
+            _variables = [[NSMutableDictionary alloc] init];
+        }
+
+        if (setOnlyNonDebugParams) {
+            _setOnlyNonDebugParams = true;
+        } else {
+            _setOnlyNonDebugParams = false;
+        }
 
         [self copyDataToDocumentsDirectory];
         _tesseract = new tesseract::TessBaseAPI();
 
         BOOL success = [self initEngine];
         if (!success) {
-            return NO;
+            return nil;
         }
     }
     return self;
 }
 
+- (id)initWithDataPath:(NSString *)dataPath language:(NSString *)language {
+    return [self initWithDataPath:dataPath
+                         language:language
+                    ocrEngineMode:OcrEngineModeDefault
+                  configFilenames:nil
+                        variables:nil
+            setOnlyNonDebugParams:NO];
+}
+
 - (BOOL)initEngine {
-    int returnCode = _tesseract->Init([_dataPath UTF8String], [_language UTF8String]);
+    char **configs = NULL;
+    int configs_size = 0;
+    GenericVector<STRING> keys;
+    GenericVector<STRING> values;
+
+    if (_configFilenames && _configFilenames.count > 0) {
+        configs = (char**) malloc(_configFilenames.count * sizeof(char*));
+        for (; configs_size < _configFilenames.count; configs_size++) {
+            configs[configs_size] = (char*) malloc((((NSString*)(_configFilenames[configs_size])).length + 1) * sizeof(char));
+            strncpy(configs[configs_size], [_configFilenames[configs_size] UTF8String], ((NSString*)(_configFilenames[configs_size])).length + 1);
+        }
+    }
+
+    for (NSString* key in _variables) {
+        keys.push_back([key UTF8String]);
+        values.push_back([_variables[key] UTF8String]);
+    }
+
+    int returnCode = _tesseract->Init(_dataPath ? [_dataPath UTF8String] : NULL,
+                                      _language ? [_language UTF8String] : NULL,
+                                      _ocrEngineMode,
+                                      configs,
+                                      configs_size,
+                                      &keys,
+                                      &values,
+                                      _setOnlyNonDebugParams);
+    if (configs) {
+        for (int filenameIndex = 0; filenameIndex < configs_size; filenameIndex++) {
+            free(configs[filenameIndex]);
+        }
+        free(configs);
+    }
+
     return (returnCode == 0) ? YES : NO;
 }
 
@@ -113,7 +195,17 @@ - (BOOL)recognize {
 
 - (NSString *)recognizedText {
     char* utf8Text = _tesseract->GetUTF8Text();
-    return [NSString stringWithUTF8String:utf8Text];
+    NSString *text = utf8Text ? [NSString stringWithUTF8String: utf8Text] : @"";
+    delete[] utf8Text;
+    return text;
+}
+
+- (void)clear
+{
+    free(_pixels);
+    _tesseract->Clear();
+    _tesseract->End();
+    delete _tesseract;
 }
 
 - (void)setImage:(UIImage *)image

OCR-Example.xcodeproj/project.pbxproj

@@ -7,6 +7,7 @@
 	objects = {
 
 /* Begin PBXBuildFile section */
+		48E5A2F419AE0ECC005BBE59 /* Images.xcassets in Resources */ = {isa = PBXBuildFile; fileRef = 48E5A2F319AE0ECC005BBE59 /* Images.xcassets */; };
 		7462256B173DCADA003EE853 /* UIKit.framework in Frameworks */ = {isa = PBXBuildFile; fileRef = 7462256A173DCADA003EE853 /* UIKit.framework */; };
 		7462256D173DCADA003EE853 /* Foundation.framework in Frameworks */ = {isa = PBXBuildFile; fileRef = 7462256C173DCADA003EE853 /* Foundation.framework */; };
 		7462256F173DCADA003EE853 /* CoreGraphics.framework in Frameworks */ = {isa = PBXBuildFile; fileRef = 7462256E173DCADA003EE853 /* CoreGraphics.framework */; };
@@ -28,6 +29,7 @@
 /* End PBXBuildFile section */
 
 /* Begin PBXFileReference section */
+		48E5A2F319AE0ECC005BBE59 /* Images.xcassets */ = {isa = PBXFileReference; lastKnownFileType = folder.assetcatalog; path = Images.xcassets; sourceTree = "<group>"; };
 		74622567173DCADA003EE853 /* OCR-Example.app */ = {isa = PBXFileReference; explicitFileType = wrapper.application; includeInIndex = 0; path = "OCR-Example.app"; sourceTree = BUILT_PRODUCTS_DIR; };
 		7462256A173DCADA003EE853 /* UIKit.framework */ = {isa = PBXFileReference; lastKnownFileType = wrapper.framework; name = UIKit.framework; path = System/Library/Frameworks/UIKit.framework; sourceTree = SDKROOT; };
 		7462256C173DCADA003EE853 /* Foundation.framework */ = {isa = PBXFileReference; lastKnownFileType = wrapper.framework; name = Foundation.framework; path = System/Library/Frameworks/Foundation.framework; sourceTree = SDKROOT; };
@@ -255,6 +257,7 @@
 				7462258E173DCD74003EE853 /* ResultsViewController.mm */,
 				746226E4173DDDC9003EE853 /* ImageProcessing.h */,
 				746226E5173DDDC9003EE853 /* ImageProcessing.mm */,
+				48E5A2F319AE0ECC005BBE59 /* Images.xcassets */,
 				74622571173DCADA003EE853 /* Supporting Files */,
 			);
 			path = "OCR-Example";
@@ -530,6 +533,7 @@
 			buildActionMask = 2147483647;
 			files = (
 				74622575173DCADA003EE853 /* InfoPlist.strings in Resources */,
+				48E5A2F419AE0ECC005BBE59 /* Images.xcassets in Resources */,
 				7462257D173DCADA003EE853 /* Default.png in Resources */,
 				7462257F173DCADA003EE853 /* Default@2x.png in Resources */,
 				74622581173DCADA003EE853 /* Default-568h@2x.png in Resources */,
@@ -636,8 +640,11 @@
 		7462258B173DCADA003EE853 /* Debug */ = {
 			isa = XCBuildConfiguration;
 			buildSettings = {
+				ASSETCATALOG_COMPILER_APPICON_NAME = AppIcon;
+				ASSETCATALOG_COMPILER_LAUNCHIMAGE_NAME = LaunchImage;
 				CLANG_CXX_LANGUAGE_STANDARD = "compiler-default";
 				CLANG_CXX_LIBRARY = "compiler-default";
+				CODE_SIGN_IDENTITY = "iPhone Developer";
 				GCC_C_LANGUAGE_STANDARD = "compiler-default";
 				GCC_PRECOMPILE_PREFIX_HEADER = YES;
 				GCC_PREFIX_HEADER = "OCR-Example/OCR-Example-Prefix.pch";
@@ -647,16 +654,21 @@
 					"\"$(SRCROOT)/OCR-Example/tesseract-ios-lib/lib\"",
 					"\"$(SRCROOT)/External/tesseract-ios-lib/lib\"",
 				);
+				ONLY_ACTIVE_ARCH = NO;
 				PRODUCT_NAME = "$(TARGET_NAME)";
+				VALID_ARCHS = "armv7 armv7s";
 				WRAPPER_EXTENSION = app;
 			};
 			name = Debug;
 		};
 		7462258C173DCADA003EE853 /* Release */ = {
 			isa = XCBuildConfiguration;
 			buildSettings = {
+				ASSETCATALOG_COMPILER_APPICON_NAME = AppIcon;
+				ASSETCATALOG_COMPILER_LAUNCHIMAGE_NAME = LaunchImage;
 				CLANG_CXX_LANGUAGE_STANDARD = "compiler-default";
 				CLANG_CXX_LIBRARY = "compiler-default";
+				CODE_SIGN_IDENTITY = "iPhone Developer";
 				GCC_C_LANGUAGE_STANDARD = "compiler-default";
 				GCC_PRECOMPILE_PREFIX_HEADER = YES;
 				GCC_PREFIX_HEADER = "OCR-Example/OCR-Example-Prefix.pch";
@@ -667,6 +679,7 @@
 					"\"$(SRCROOT)/External/tesseract-ios-lib/lib\"",
 				);
 				PRODUCT_NAME = "$(TARGET_NAME)";
+				VALID_ARCHS = "armv7 armv7s";
 				WRAPPER_EXTENSION = app;
 			};
 			name = Release;

OCR-Example/Images.xcassets/AppIcon.appiconset/Contents.json

@@ -0,0 +1,38 @@
+{
+  "images" : [
+    {
+      "idiom" : "iphone",
+      "scale" : "1x",
+      "size" : "57x57"
+    },
+    {
+      "idiom" : "iphone",
+      "scale" : "2x",
+      "size" : "57x57"
+    },
+    {
+      "idiom" : "iphone",
+      "scale" : "2x",
+      "size" : "60x60"
+    },
+    {
+      "idiom" : "iphone",
+      "scale" : "1x",
+      "size" : "29x29"
+    },
+    {
+      "idiom" : "iphone",
+      "scale" : "2x",
+      "size" : "29x29"
+    },
+    {
+      "idiom" : "iphone",
+      "scale" : "2x",
+      "size" : "40x40"
+    }
+  ],
+  "info" : {
+    "version" : 1,
+    "author" : "xcode"
+  }
+}

OCR-Example/Images.xcassets/LaunchImage.launchimage/Contents.json

@@ -0,0 +1,42 @@
+{
+  "images" : [
+    {
+      "orientation" : "portrait",
+      "idiom" : "iphone",
+      "filename" : "Default@2x.png",
+      "minimum-system-version" : "7.0",
+      "scale" : "2x"
+    },
+    {
+      "orientation" : "portrait",
+      "idiom" : "iphone",
+      "filename" : "Default-568h@2x.png",
+      "minimum-system-version" : "7.0",
+      "subtype" : "retina4",
+      "scale" : "2x"
+    },
+    {
+      "orientation" : "portrait",
+      "idiom" : "iphone",
+      "filename" : "Default.png",
+      "scale" : "1x"
+    },
+    {
+      "orientation" : "portrait",
+      "idiom" : "iphone",
+      "filename" : "Default@2x.png",
+      "scale" : "2x"
+    },
+    {
+      "orientation" : "portrait",
+      "idiom" : "iphone",
+      "filename" : "Default-568h@2x.png",
+      "subtype" : "retina4",
+      "scale" : "2x"
+    }
+  ],
+  "info" : {
+    "version" : 1,
+    "author" : "xcode"
+  }
+}

OCR-Example/OCR-Example-Info.plist

@@ -8,8 +8,12 @@
 	<string>${PRODUCT_NAME}</string>
 	<key>CFBundleExecutable</key>
 	<string>${EXECUTABLE_NAME}</string>
+	<key>CFBundleIcons</key>
+	<dict/>
+	<key>CFBundleIcons~ipad</key>
+	<dict/>
 	<key>CFBundleIdentifier</key>
-	<string>Christopher-Constable.${PRODUCT_NAME:rfc1034identifier}</string>
+	<string>com.glory.${PRODUCT_NAME:rfc1034identifier}</string>
 	<key>CFBundleInfoDictionaryVersion</key>
 	<string>6.0</string>
 	<key>CFBundleName</key>

OCR-Example/ResultsViewController.mm

@@ -64,6 +64,7 @@ - (void)viewDidAppear:(BOOL)animated
                              completion:^(BOOL finished) {
                                  self.loadingView.hidden = YES;
                                  [self.resultsTextView setText:[tesseract recognizedText]];
+                                 [tesseract clear];
                              }];
         });
     });