Added sentence support

Author: Thomas Hambach

CSharpDiff.Tests/DiffSentenceTests.cs

@@ -0,0 +1,27 @@
+using CSharpDiff.Convert;
+using CSharpDiff.Diff;
+
+namespace CSharpDiff.Tests
+{
+
+    public class DiffSentenceTests
+    {
+        [Fact]
+        public void ShouldDiffSentences()
+        {
+            var diff = new DiffSentence();
+            var diffResult = diff.diff("New Value.", "New ValueMoreData.");
+            var converted = DiffConvertXml.Convert(diffResult);
+            Assert.Equal("<del>New Value.</del><ins>New ValueMoreData.</ins>", converted);
+        }
+
+        [Fact]
+        public void ShouldDiffLastSentence()
+        {
+            var diff = new DiffSentence();
+            var diffResult = diff.diff("Here im. Rock you like old man.", "Here im. Rock you like hurricane.");
+            var converted = DiffConvertXml.Convert(diffResult);
+            Assert.Equal("Here im. <del>Rock you like old man.</del><ins>Rock you like hurricane.</ins>", converted);
+        }
+    }
+}

CSharpDiff/Diff/Diff.cs

@@ -3,7 +3,7 @@
 
 namespace CSharpDiff.Diff
 {
-    public class Diff
+    public class Diff : IDiff
     {
         public bool UseLongestToken { get; set; } = true;
         public bool IgnoreWhiteSpace { get; set; } = false;
@@ -12,10 +12,10 @@ public IList<DiffResult> diff(string oldString, string newString)
         {
             var cleanOldString = removeEmpty(tokenize(oldString));
             var cleanNewString = removeEmpty(tokenize(newString));
-            return determineDiff(oldString, newString, cleanOldString, cleanNewString);
+            return determineDiff(cleanOldString, cleanNewString);
         }
 
-        public IList<DiffResult> determineDiff(string oldString, string newString, string[] cleanOldString, string[] cleanNewString)
+        public IList<DiffResult> determineDiff(string[] cleanOldString, string[] cleanNewString)
         {
 
             var diffs = new List<DiffResult>();

CSharpDiff/Diff/DiffArray.cs

@@ -0,0 +1,19 @@
+using CSharpDiff.Diff.Models;
+
+namespace CSharpDiff.Diff
+{
+    public class DiffArray : Diff
+    {
+        public IList<DiffResult> diff(string[] oldArray, string[] newArray)
+        {
+            var cleanOldString = removeEmpty(oldArray);
+            var cleanNewString = removeEmpty(newArray);
+            return determineDiff(cleanOldString, cleanNewString);
+        }
+
+        public new string join(string[] strings)
+        {
+            return "";
+        }
+    }
+}

CSharpDiff/Diff/DiffLines.cs

@@ -8,7 +8,7 @@ public class DiffLines : Diff
         {
             var cleanOldString = removeEmpty(tokenize(oldString));
             var cleanNewString = removeEmpty(tokenize(newString));
-            return determineDiff(oldString, newString, cleanOldString, cleanNewString);
+            return determineDiff(cleanOldString, cleanNewString);
         }
 
         public new string[] tokenize(string value)

CSharpDiff/Diff/DiffSentence.cs

@@ -0,0 +1,21 @@
+using System.Text.RegularExpressions;
+using CSharpDiff.Diff.Models;
+
+namespace CSharpDiff.Diff
+{
+    public class DiffSentence : Diff
+    {
+        public new IList<DiffResult> diff(string oldString, string newString)
+        {
+            var cleanOldString = removeEmpty(tokenize(oldString));
+            var cleanNewString = removeEmpty(tokenize(newString));
+            return determineDiff(cleanOldString, cleanNewString);
+        }
+
+        public new string[] tokenize(string value)
+        {
+            var regex = new Regex(@"(\S.+?[.!?])(?=\s+|$)");
+            return regex.Split(value);
+        }
+    }
+}

CSharpDiff/Diff/IDiff.cs

@@ -0,0 +1,55 @@
+using CSharpDiff.Diff.Models;
+
+namespace CSharpDiff.Diff
+{
+    public interface IDiff
+    {
+        bool UseLongestToken { get; set; }
+
+        /// <summary>
+        /// Defines whether we ignore white space.
+        /// </summary>
+        bool IgnoreWhiteSpace { get; set; }
+
+        List<DiffResult> buildValues(List<DiffResult> components, string[] newString, string[] oldString, bool useLongestToken);
+
+        BestPath clonePath(BestPath path);
+
+        IList<DiffResult> determineDiff(string[] cleanOldString, string[] cleanNewString);
+
+        IList<DiffResult> diff(string oldString, string newString);
+
+        /// <summary>
+        /// Check whether left matches right.
+        /// </summary>
+        /// <param name="left"></param>
+        /// <param name="right"></param>
+        /// <returns></returns>
+        bool equals(char left, char right);
+
+        /// <summary>
+        /// Check whether left matches right.
+        /// </summary>
+        /// <param name="left"></param>
+        /// <param name="right"></param>
+        /// <returns></returns>
+        bool equals(string left, string right);
+        int extractCommon(BestPath basePath, string[] newString, string[] oldString, int diagonalPath);
+        string join(string[] strings);
+        List<DiffResult> pushComponent(List<DiffResult> components, bool? added, bool? removed);
+
+        /// <summary>
+        /// Remove empty values from the array. Differs per diff type.
+        /// </summary>
+        /// <param name="array"></param>
+        /// <returns></returns>
+        string[] removeEmpty(string[] array);
+
+        /// <summary>
+        /// Tokenize string, differs per diff type.
+        /// </summary>
+        /// <param name="value"></param>
+        /// <returns></returns>
+        string[] tokenize(string value);
+    }
+}

README.md

@@ -10,11 +10,29 @@ C# Diff with Unified Diff Support, this codebase is a port from the popular JS l
 
 ## Usage Example
 
+### Diff
+
+#### Sentence
+
+```c#
+using CSharpDiff.Diff;
+
+var text1 = "Here im. Rock you like old man.";
+var text2 = "Here im. Rock you like hurricane.";
+
+var diff = new DiffSentence();
+string patch = diff.diff(text1, text2);
+```
+
+### Patch
+
+#### Create
+
 ```c#
-using CSharpDiff;
+using CSharpDiff.Patch;
 
-var text1 = "";
-var text2 = "";
+var text1 = "...";
+var text2 = "...";
 
 var ps = new Patch();
 string patch = ps.create("filename1", "filename2", text1, text2, "header1", "header2", new PatchServiceOptions());
@@ -33,13 +51,13 @@ string patch = ps.create("filename1", "filename2", text1, text2, "header1", "hea
   - [ ] Merge
   - [ ] Parse
 - [ ] Diff
-  - [ ] Array
+  - [ ] Array (difficult, JS allows type mixing)
   - [x] Base
   - [ ] Character
   - [ ] CSS
   - [ ] JSON
   - [x] Line
-  - [ ] Sentence
+  - [x] Sentence
   - [ ] Word
 - [ ] Convert
   - [x] XML