Add ISO 639-2 data

README.md

@@ -20,5 +20,6 @@ Packages published on [NuGet](https://www.nuget.org/packages/InsaneGenius.Utilit
 
 ## External Data
 
-- ISO 639-3 language data is sourced from [ISO 639-3 Registration Authority](https://iso639-3.sil.org/sites/iso639-3/files/downloads/iso-639-3.tab).
-- RFC 5646 / BCP 47 language data is sourced from [IANA](https://www.iana.org/assignments/language-subtag-registry/language-subtag-registry).
+- ISO 639-2 language data is sourced from the [ISO 639-2 Registration Authority](https://www.loc.gov/standards/iso639-2/langhome.html) [download](https://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt)
+- ISO 639-3 language data is sourced from the [ISO 639-3 Registration Authority](https://iso639-3.sil.org/) [download](https://iso639-3.sil.org/sites/iso639-3/files/downloads/iso-639-3.tab).
+- RFC 5646 / BCP 47 language data is sourced from the [IANA Tags for Identifying Languages RFC 5646](https://www.rfc-editor.org/rfc/rfc5646.html) [download](https://www.iana.org/assignments/language-subtag-registry/language-subtag-registry).

Utilities.sln

@@ -23,6 +23,7 @@ Project("{9A19103F-16F7-4668-BE54-9A1E7A4F7556}") = "Sandbox", "Sandbox\Sandbox.
 EndProject
 Project("{2150E333-8FDC-42A3-9474-1A3956D46DE8}") = "Data", "Data", "{7E189F23-48EE-44B2-A102-0FDF27BE9C88}"
 	ProjectSection(SolutionItems) = preProject
+		Data\ISO-639-2_utf-8.txt = Data\ISO-639-2_utf-8.txt
 		Data\iso-639-3.tab = Data\iso-639-3.tab
 		Data\language-subtag-registry = Data\language-subtag-registry
 	EndProjectSection

Utilities/Iso6392.cs

@@ -0,0 +1,126 @@
+using System;
+using System.Collections.Generic;
+using System.Diagnostics;
+using System.IO;
+using System.Linq;
+using System.Reflection;
+
+namespace InsaneGenius.Utilities;
+
+// ISO-639-2 reference
+// https://www.loc.gov/standards/iso639-2/langhome.html
+// https://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
+
+// T4 template
+// https://docs.microsoft.com/en-us/visualstudio/modeling/code-generation-and-t4-text-templates
+// https://github.com/mono/t4
+// wget -O ./Data/ISO-639-2_utf-8.txt https://www.loc.gov/standards/iso639-2/ISO-639-2_utf-8.txt
+// dotnet tool install -g dotnet-t4
+// dotnet publish ./Utilities/Utilities.csproj --self-contained=false --output=./bin/T4
+// t4 -P="./bin/T4" --out=./Utilities/Iso6392Gen.cs ./Utilities/Iso6392Gen.tt
+
+// ISO 639-2 class
+public partial class Iso6392
+{
+    public class Record
+    {
+        // Default to use 2B
+        public string Id { get { return Part2B; } }
+        // 639-2 Bibliographic
+        public string Part2B { get; set; } = "";
+        // 639-2 Terminology
+        public string Part2T { get; set; } = "";
+        // 639-1
+        public string Part1 { get; set; } = "";
+        // English name
+        public string RefName { get; set; } = "";
+    }
+
+    public List<Record> RecordList { get; private set; } = new();
+
+    // Create() method is generated from Iso6392Gen.tt
+    // public bool Create() { return true; }
+
+    public bool Load(string fileName)
+    {
+        try
+        {
+            // Open the file as a stream
+            using StreamReader lineReader = new(File.OpenRead(fileName));
+
+            // Init
+            RecordList.Clear();
+
+            // Read header
+            var line = lineReader.ReadLine();
+            Debug.Assert(!string.IsNullOrEmpty(line));
+
+            // Read line by line
+            while ((line = lineReader.ReadLine()) is not null)
+            {
+                // Parse using pipe character
+                var records = line.Split('|');
+                Debug.Assert(records.Length == 5);
+
+                // Populate record                
+                var record = new Record
+                {
+                    Part2B = records[0].Trim(),
+                    Part2T = records[1].Trim(),
+                    Part1 = records[2].Trim(),
+                    RefName = records[3].Trim(),
+                    // Ignore the French name, German name is not in file
+                };
+                RecordList.Add(record);
+            }
+        }
+        catch (Exception e) when (LogOptions.Logger.LogAndHandle(e, MethodBase.GetCurrentMethod()?.Name))
+        {
+            return false;
+        }
+        return true;
+    }
+
+    public Record Find(string languageTag, bool includeDescription)
+    {
+        ArgumentException.ThrowIfNullOrEmpty(nameof(languageTag));
+
+        // Find the matching language entry
+        Record record = null;
+
+        // 693 3 letter form
+        if (languageTag.Length == 3)
+        {
+            // Try the 639-2/B
+            record = RecordList.FirstOrDefault(item => item.Part2B.Equals(languageTag, StringComparison.OrdinalIgnoreCase));
+            if (record != null)
+                return record;
+
+            // Try the 639-2/T
+            record = RecordList.FirstOrDefault(item => item.Part2T.Equals(languageTag, StringComparison.OrdinalIgnoreCase));
+            if (record != null)
+                return record;
+        }
+
+        // 693 2 letter form
+        if (languageTag.Length == 2)
+        {
+            // Try 639-1
+            record = RecordList.FirstOrDefault(item => item.Part1.Equals(languageTag, StringComparison.OrdinalIgnoreCase));
+            if (record != null)
+                return record;
+        }
+
+        // Long form
+        if (includeDescription)
+        {
+            // Try long form
+            record = RecordList.FirstOrDefault(item => item.RefName.Equals(languageTag, StringComparison.OrdinalIgnoreCase));
+            if (record != null)
+                return record;
+        }
+
+        // Not found
+        return null;
+    }
+}