Add additional tests

Author: Enkidu93

src/SIL.Machine/Corpora/UsfmVersificationMismatchDetector.cs

@@ -208,7 +208,7 @@ public override void Chapter(
             string pubNumber
         )
         {
-            if (_currentChapter != 0)
+            if (_currentBook > 0 && Canon.IsCanonical(_currentBook) && _currentChapter > 0)
             {
                 var versificationMismatch = new UsfmVersificationMismatch(
                     _currentBook,
@@ -234,16 +234,19 @@ string pubNumber
         )
         {
             _currentVerse = state.VerseRef;
-            var versificationMismatch = new UsfmVersificationMismatch(
-                _currentBook,
-                _currentChapter,
-                _currentVerse.AllVerses().Last().VerseNum,
-                _currentChapter,
-                _currentVerse.AllVerses().Last().VerseNum,
-                _currentVerse
-            );
-            if (versificationMismatch.CheckMismatch())
-                _errors.Add(versificationMismatch);
+            if (_currentBook > 0 && Canon.IsCanonical(_currentBook) && _currentChapter > 0)
+            {
+                var versificationMismatch = new UsfmVersificationMismatch(
+                    _currentBook,
+                    _currentChapter,
+                    _currentVerse.AllVerses().Last().VerseNum,
+                    _currentChapter,
+                    _currentVerse.AllVerses().Last().VerseNum,
+                    _currentVerse
+                );
+                if (versificationMismatch.CheckMismatch())
+                    _errors.Add(versificationMismatch);
+            }
         }
     }
 }

tests/SIL.Machine.Tests/Corpora/ParatextProjectVersificationMismatchDetector.cs

@@ -1,5 +1,7 @@
+using System.Text;
 using System.Text.Json;
 using NUnit.Framework;
+using SIL.Scripture;
 
 namespace SIL.Machine.Corpora;
 
@@ -196,6 +198,99 @@ public void GetUsfmVersificationMismatches_ExtraVerseSegment()
         Assert.That(mismatches[0].Type, Is.EqualTo(UsfmVersificationMismatchType.ExtraVerseSegment));
     }
 
+    [Test]
+    public void GetUsfmVersificationMismatches_MissingVerseSegment()
+    {
+        var env = new TestEnvironment(
+            settings: new MemoryParatextProjectFileHandler.DefaultParatextProjectSettings(
+                versification: GetCustomVersification(@"*3JN 1:13,a,b")
+            ),
+            files: new Dictionary<string, string>()
+            {
+                {
+                    "653JNTest.SFM",
+                    @"\id 3JN
+        \c 1
+        \v 1
+        \v 2
+        \v 3
+        \v 4
+        \v 5
+        \v 6
+        \v 7
+        \v 8
+        \v 9
+        \v 10
+        \v 11
+        \v 12
+        \v 13
+        \v 14
+        \v 15
+        "
+                }
+            }
+        );
+        IReadOnlyList<UsfmVersificationMismatch> mismatches = env.GetUsfmVersificationMismatches();
+        Assert.That(mismatches, Has.Count.EqualTo(1), JsonSerializer.Serialize(mismatches));
+        Assert.That(mismatches[0].Type, Is.EqualTo(UsfmVersificationMismatchType.MissingVerseSegment));
+    }
+
+    [Test]
+    public void GetUsfmVersificationMismatches_IgnoreNonCanonicals()
+    {
+        var env = new TestEnvironment(
+            files: new Dictionary<string, string>()
+            {
+                {
+                    "98XXETest.SFM",
+                    @"\id XXE
+        \c 1
+        \v 3-2
+        "
+                }
+            }
+        );
+        IReadOnlyList<UsfmVersificationMismatch> mismatches = env.GetUsfmVersificationMismatches();
+        Assert.That(mismatches, Has.Count.EqualTo(0), JsonSerializer.Serialize(mismatches));
+    }
+
+    [Test]
+    public void GetUsfmVersificationMismatches_ExtraVerse_ExcludedInCustomVrs()
+    {
+        var env = new TestEnvironment(
+            settings: new MemoryParatextProjectFileHandler.DefaultParatextProjectSettings(
+                versification: GetCustomVersification(@"-3JN 1:13")
+            ),
+            files: new Dictionary<string, string>()
+            {
+                {
+                    "653JNTest.SFM",
+                    @"\id 3JN
+        \c 1
+        \v 1
+        \v 2
+        \v 3
+        \v 4
+        \v 5
+        \v 6
+        \v 7
+        \v 8
+        \v 9
+        \v 10
+        \v 11
+        \v 12
+        \v 13
+        \v 14
+        \v 15
+        "
+                }
+            }
+        );
+        IReadOnlyList<UsfmVersificationMismatch> mismatches = env.GetUsfmVersificationMismatches();
+        Assert.That(mismatches, Has.Count.EqualTo(1), JsonSerializer.Serialize(mismatches));
+        Assert.That(mismatches[0].Type, Is.EqualTo(UsfmVersificationMismatchType.ExtraVerse));
+    }
+
     private class TestEnvironment(ParatextProjectSettings? settings = null, Dictionary<string, string>? files = null)
     {
         public ParatextProjectVersificationMismatchDetector Detector { get; } =
@@ -206,4 +301,21 @@ public IReadOnlyList<UsfmVersificationMismatch> GetUsfmVersificationMismatches()
             return Detector.GetUsfmVersificationMismatches();
         }
     }
+
+    private static ScrVers GetCustomVersification(string customVrsContents, ScrVers? baseVersification = null)
+    {
+        baseVersification ??= ScrVers.English;
+        ScrVers customVersification = baseVersification;
+        using (var reader = new StreamReader(new MemoryStream(Encoding.UTF8.GetBytes(customVrsContents))))
+        {
+            customVersification = Versification.Table.Implementation.Load(
+                reader,
+                "custom.vrs",
+                baseVersification,
+                baseVersification.ToString() + "-" + customVrsContents.GetHashCode()
+            );
+        }
+        Versification.Table.Implementation.RemoveAllUnknownVersifications();
+        return customVersification;
+    }
 }