Merge branch 'master' of https://github.com/Sicos1977/TesseractOCR

Sicos1977 · May 4, 2022 · cfa1abc · cfa1abc
2 parents be6e86d + 7dd6c14
commit cfa1abc
Show file tree

Hide file tree

Showing 14 changed files with 375 additions and 158 deletions.
diff --git a/README.md b/README.md
@@ -14,7 +14,7 @@ You can get them at https://github.com/tesseract-ocr/tessdata or https://github.
 
 ## Microsoft Visual C++ runtimes
 
-The DLL's Tesseract51.dll (and exe) and leptonica-1.82.0.dll are compiled with Visual Studio 2022 you need these C++ runtimes for it on your server
+The DLL's Tesseract51.dll (and exe) and leptonica-1.82.0.dll are compiled with Visual Studio 2022 you need these C++ runtimes for it on your computer
 
 - X86: https://aka.ms/vs/17/release/vc_redist.x86.exe
 - X64: https://aka.ms/vs/17/release/vc_redist.x64.exe
@@ -32,7 +32,7 @@ Console.WriteLine("Text: \r\n{0}", page.Text);
 ## Iterate through the layout of a page
 
 ```c#
-using var engine = CreateEngine();
+using var engine = new Engine(@"./tessdata", Language.English, EngineMode.Default);
 using var img = Pix.Image.LoadFromFile(testImagePath);
 using var page = engine.Process(img);
 
@@ -126,6 +126,21 @@ Tesseract uses the Leptonica library to read images with one of these formats:
 
 **I have dropped support for the Windows.Drawing.Image namespace since this only works good on Windows and not on other systems. You should be fine with Leptonica**
 
+Logging
+=======
+
+TesseractOCR uses the Microsoft ILogger interface (https://docs.microsoft.com/en-us/dotnet/api/microsoft.extensions.logging.ilogger?view=dotnet-plat-ext-5.0). You can use any logging library that uses this interface.
+
+TesseractOCR has some build in loggers that can be found in the ```TesseractOCR.Logger``` namespace. 
+
+For example
+
+```csharp
+var logger = !string.IsNullOrWhiteSpace(<some logfile>)
+                ? new TesseractOCR.Loggers.Stream(File.OpenWrite(<some logfile>))
+                : new TesseractOCR.Loggers.Console();
+```
+
 Installing via NuGet
 ====================
 

diff --git a/Tesseract.Drawing/Tesseract.Drawing.csproj b/Tesseract.Drawing/Tesseract.Drawing.csproj
diff --git a/Tesseract.sln.DotSettings b/Tesseract.sln.DotSettings
@@ -104,6 +104,8 @@
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=pageseg/@EntryIndexedValue">True</s:Boolean>
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=pangle/@EntryIndexedValue">True</s:Boolean>
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=Panjabi/@EntryIndexedValue">True</s:Boolean>
+	<s:Boolean x:Key="/Default/UserDictionary/Words/=paraids/@EntryIndexedValue">True</s:Boolean>
+
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=Pashto/@EntryIndexedValue">True</s:Boolean>
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=pconf/@EntryIndexedValue">True</s:Boolean>
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=phototest/@EntryIndexedValue">True</s:Boolean>
@@ -148,6 +150,7 @@
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=Tesseract_0027s/@EntryIndexedValue">True</s:Boolean>
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=testregion/@EntryIndexedValue">True</s:Boolean>
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=TEXTLINE/@EntryIndexedValue">True</s:Boolean>
+	<s:Boolean x:Key="/Default/UserDictionary/Words/=Textlines/@EntryIndexedValue">True</s:Boolean>
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=textonly/@EntryIndexedValue">True</s:Boolean>
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=textord/@EntryIndexedValue">True</s:Boolean>
 	<s:Boolean x:Key="/Default/UserDictionary/Words/=Thaana/@EntryIndexedValue">True</s:Boolean>

diff --git a/TesseractOCR.Net45Tests/TesseractOCR.Net48Tests.csproj b/TesseractOCR.Net45Tests/TesseractOCR.Net48Tests.csproj
@@ -35,8 +35,8 @@
   <ItemGroup>
     <PackageReference Include="Microsoft.Extensions.Logging.Abstractions" Version="6.0.1" />
     <PackageReference Include="Microsoft.NET.Test.Sdk" Version="17.1.0" />
-    <PackageReference Include="MSTest.TestAdapter" Version="2.2.8" />
-    <PackageReference Include="MSTest.TestFramework" Version="2.2.8" />
+    <PackageReference Include="MSTest.TestAdapter" Version="2.2.10" />
+    <PackageReference Include="MSTest.TestFramework" Version="2.2.10" />
   </ItemGroup>
 
   <ItemGroup>

diff --git a/TesseractOCR.NetCore31Tests/TesseractOCR.NetCore31Tests.csproj b/TesseractOCR.NetCore31Tests/TesseractOCR.NetCore31Tests.csproj
@@ -34,8 +34,8 @@
   <ItemGroup>
     <PackageReference Include="Microsoft.Extensions.Logging.Abstractions" Version="6.0.1" />
     <PackageReference Include="Microsoft.NET.Test.Sdk" Version="17.1.0" />
-    <PackageReference Include="MSTest.TestAdapter" Version="2.2.8" />
-    <PackageReference Include="MSTest.TestFramework" Version="2.2.8" />
+    <PackageReference Include="MSTest.TestAdapter" Version="2.2.10" />
+    <PackageReference Include="MSTest.TestFramework" Version="2.2.10" />
   </ItemGroup>
 
   <ItemGroup>

diff --git a/TesseractOCR.Tests/BaseApiTests.cs b/TesseractOCR.Tests/BaseApiTests.cs
@@ -6,11 +6,23 @@ namespace Tesseract.Tests
     public class BaseApiTests : TesseractTestBase
     {
         [TestMethod]
-        public void GetVersion_Is500()
+        public void GetVersion_Is510()
         {
             using var engine = CreateEngine();
             var version = engine.Version;
-            Assert.IsTrue(version.StartsWith("5.0.0"));
+            Assert.IsTrue(version.StartsWith("5.1.0"));
         }
+
+        [TestMethod]
+        public void LoadedLanguages()
+        {
+            using var engine = CreateEngine();
+            var dp = engine.DataPath;
+            engine.ClearAdaptiveClassifier();
+            engine.ClearPersistentCache();
+            var languages = engine.AvailableLanguages;
+            //Assert.IsTrue(version.StartsWith("5.0.0"));
+        }
+
     }
 }
diff --git a/TesseractOCR.Tests/EngineTests.cs b/TesseractOCR.Tests/EngineTests.cs
@@ -141,27 +141,6 @@ public void CanProcessDifferentRegionsInSameImage()
             Assert.AreEqual(region2Text, expectedTextRegion2);
         }
 
-        [TestMethod]
-        public void CanGetSegmentedRegions()
-        {
-            const int expectedCount = 8; // number of text lines in test image
-
-            using var engine = CreateEngine();
-            var imgPath = TestFilePath(TestImagePath);
-            using var img = TesseractOCR.Pix.Image.LoadFromFile(imgPath);
-            using var page = engine.Process(img);
-            var boxes = page.GetSegmentedRegions(PageIteratorLevel.TextLine);
-
-            for (var i = 0; i < boxes.Count; i++)
-            {
-                var box = boxes[i];
-                Console.WriteLine("Box[{0}]: x={1}, y={2}, w={3}, h={4}", i, box.X, box.Y, box.Width,
-                    box.Height);
-            }
-
-            Assert.AreEqual(boxes.Count, expectedCount);
-        }
-
         [TestMethod]
         public void CanProcessEmptyPixUsingResultIterator()
         {
@@ -265,6 +244,7 @@ public void CanProcessPixUsingResultIterator()
 
                 foreach (var paragraph in block.Paragraphs)
                 {
+                    var regions = block.SegmentedRegions;
                     result.AppendLine($"Paragraph confidence: {paragraph.Confidence}");
                     if (paragraph.BoundingBox != null)
                     {

diff --git a/TesseractOCR/Engine.cs b/TesseractOCR/Engine.cs
@@ -57,6 +57,31 @@ public class Engine : DisposableBase
         ///     Gets or sets default <see cref="PageSegMode" /> mode used by one of the Process methods
         /// </summary>
         public PageSegMode DefaultPageSegMode { get; set; }
+
+        /// <summary>
+        ///     Returns the current engine mode
+        /// </summary>
+        public EngineMode CurrentEngineMode => TessApi.Native.BaseAPIOem(_handle);
+
+        /// <summary>
+        ///     Returns the <see cref="Language"/> used in the last valid initialization
+        /// </summary>
+        public Language InitLanguage => LanguageHelper.StringToEnum(MarshalHelper.PtrToString(TessApi.Native.BaseApiGetDatapath(_handle)));
+
+        /// <summary>
+        ///     Returns the data path
+        /// </summary>
+        public string DataPath => MarshalHelper.PtrToString(TessApi.Native.BaseApiGetDatapath(_handle)).TrimEnd('/');
+
+        /// <summary>
+        ///     Returns a list of loaded <see cref="Language"/>'s
+        /// </summary>
+        public List<Language> LoadedLanguages => TessApi.BaseApiLoadedLanguages(_handle);
+
+        /// <summary>
+        ///     Returns a list of available <see cref="Language"/>'s
+        /// </summary>
+        public List<Language> AvailableLanguages => TessApi.BaseAPIGetAvailableLanguagesAsVector(_handle);
         #endregion
 
         #region Constructors
@@ -630,7 +655,35 @@ private void Initialize(string dataPath, Language language, EngineMode engineMod
         }
         #endregion
 
-        #region BaseApiSetDebugVariable
+        #region ClearAdaptiveClassifier
+        /// <summary>
+        ///     Call between pages or documents etc to free up memory and forget adaptive data
+        /// </summary>
+        public void ClearAdaptiveClassifier()
+        {
+            TessApi.Native.BaseAPIClearAdaptiveClassifier(_handle);
+        }
+        #endregion
+
+        #region ClearPersistentCache
+        /// <summary>
+        ///     Clear any library-level memory caches. There are a variety of expensive-to-load constant data structures
+        ///     (mostly language dictionaries) that are cached globally -- surviving the Init() and End() of individual TessBaseAPI's.
+        ///     This function allows the clearing of these caches
+        /// </summary>
+        public void ClearPersistentCache()
+        {
+            TessApi.Native.BaseAPIClearPersistentCache(_handle);
+        }
+        #endregion
+
+        #region SetDebugVariable
+        /// <summary>
+        ///     Sets a debug variable.
+        /// </summary>
+        /// <param name="name"></param>
+        /// <param name="value"></param>
+        /// <returns></returns>
         public bool SetDebugVariable(string name, string value)
         {
             return TessApi.BaseApiSetDebugVariable(_handle, name, value) != 0;

diff --git a/TesseractOCR/Enums/Language.cs b/TesseractOCR/Enums/Language.cs
@@ -383,6 +383,12 @@ public enum Language
         /// </summary>
         [String("jpn")] 
         Japanese,
+
+        /// <summary>
+        ///     Japanese (vertical)
+        /// </summary>
+        [String("jpn_vert")] 
+        JapaneseVertical,
 
         /// <summary>
         ///     Kannada