shuffling some stuff around, added a very in-depth doc on xterm parsing

README.md

@@ -2,6 +2,7 @@
 
 ## Dev notes
 - VT codes doc: https://docs.microsoft.com/en-us/windows/console/console-virtual-terminal-sequences
+- xterm: http://invisible-island.net/xterm/ctlseqs/ctlseqs.html
 - Sample for ConPTY: https://github.com/Microsoft/console/tree/master/samples/ConPTY/EchoCon
 - Win32 APIs: WinBase.h, ProcessEnv.h, consoleapi.h
 - Calculating size of a textbox: https://stackoverflow.com/questions/9264398/how-to-calculate-wpf-textblock-width-for-its-known-font-size-and-characters

src/ConsoleBuffer/Buffer.cs

@@ -8,7 +8,7 @@
 
     public sealed class Buffer : INotifyPropertyChanged
     {
-        private readonly AnsiParser parser = new AnsiParser();
+        private readonly SequenceParser parser = new SequenceParser();
         private readonly List<Line> lines = new List<Line>();
         private readonly object renderLock = new object();
 

src/ConsoleBuffer/AnsiParser.cs → src/ConsoleBuffer/SequenceParser.cs

@@ -36,7 +36,15 @@ enum ParserCommand
         CR,
     }
 
-    sealed class AnsiParser
+    // Notes on the parser:
+    // - I elected to hand-roll this instead of generating a parser. This is primarily for performance purposes.
+    // - The names I've chosen to given to various "areas" of parsing are made up and not based on good research on my
+    //   part.  The areas themselves may be silly/erroneous. Please call me out on my bullshit as desired.
+    // - The parser is not greedy. I have seen parsers which work to varying levels of greed, we will stop on the first
+    //   invalid character for whatever sequence we are in (and not emit that character, nor any preceding). So for
+    //   example the sequence '\e[32[33m hello' will emit an unmodified '33m hello' string as we gave up at the invalid
+    //   '[' character.
+    sealed class SequenceParser
     {
         enum SequenceType
         {
@@ -53,12 +61,13 @@ enum SequenceType
             /// </summary>
             None,
         }
+
         private SequenceType sequenceType = SequenceType.None;
         private bool inSequence = false;
 
         public ParserCommand CurrentCommand { get; private set; }
 
-        public AnsiParser()
+        public SequenceParser()
         {
         }