Handle odd ligatures names and fix #945

Author: BobLd

Diff for: src/UglyToad.PdfPig.Fonts/GlyphList.cs

@@ -38,6 +38,16 @@ public class GlyphList
 
         private static readonly Lazy<GlyphList> LazyZapfDingbatsGlyphList = new Lazy<GlyphList>(() => GlyphListFactory.Get("zapfdingbats"));
 
+        private static readonly HashSet<string> OddLigaturesNames =
+        [
+            // See https://en.wikipedia.org/wiki/Ligature_(writing)
+            "f_f", "f_i", "f_j", "f_l", "f_a", "f_e", "f_o", "f_r", "f_s", "f_t", "f_b", "f_h",
+            "f_u", "f_y", "f_.", "f_,", "f_-",
+            "f_f_i", "f_f_l",
+            // Sometimes, ligatures for ⟨st⟩ (st), ⟨ſt⟩ (ſt), ⟨ch⟩, ⟨ck⟩, ⟨ct⟩, ⟨Qu⟩ and ⟨Th⟩ are used (e.g. in the typeface Linux Libertine).
+            "s_t", "ſ_t", "c_h", "c_k", "c_t", "Q_u", "T_h"
+        ]; // TODO - Go use for FrozenSet
+
         /// <summary>
         /// Zapf Dingbats.
         /// </summary>
@@ -159,6 +169,15 @@ public string NameToUnicode(string name)
                 System.Diagnostics.Debug.Assert(codePoint > 0);
                 unicode = char.ConvertFromUtf32(codePoint);
             }
+            else if (name.IndexOf('_') > 0 && OddLigaturesNames.Contains(name))
+            {
+                /*
+                 * MOZILLA-3136-0.pdf
+                 * 68-1990-01_A.pdf
+                 * TIKA-2054-0.pdf
+                 */
+                unicode = name.Replace("_", "");
+            }
             else
             {
                 return null;

Diff for: src/UglyToad.PdfPig.Tests/Integration/Documents/TIKA-2054-0.pdf

@@ -6,6 +6,41 @@
 
     public class GithubIssuesTests
     {
+        [Fact]
+        public void Issue945()
+        {
+            // Odd ligatures names
+            var path = IntegrationHelpers.GetDocumentPath("MOZILLA-3136-0.pdf");
+            using (var document = PdfDocument.Open(path))
+            {
+                var page = document.GetPage(2);
+                Assert.Contains("ff", page.Letters.Select(l => l.Value));
+            }
+
+            path = IntegrationHelpers.GetDocumentPath("68-1990-01_A.pdf");
+            using (var document = PdfDocument.Open(path))
+            {
+                var page = document.GetPage(7);
+                Assert.Contains("fi", page.Letters.Select(l => l.Value));
+            }
+
+            path = IntegrationHelpers.GetDocumentPath("TIKA-2054-0.pdf");
+            using (var document = PdfDocument.Open(path))
+            {
+                var page = document.GetPage(3);
+                Assert.Contains("fi", page.Letters.Select(l => l.Value));
+
+                page = document.GetPage(4);
+                Assert.Contains("ff", page.Letters.Select(l => l.Value));
+
+                page = document.GetPage(6);
+                Assert.Contains("fl", page.Letters.Select(l => l.Value));
+
+                page = document.GetPage(16);
+                Assert.Contains("ffi", page.Letters.Select(l => l.Value));
+            }
+        }
+
         [Fact]
         public void Issue943()
         {