[Java.Interop.Tools.JavaSource] Handle common parsing exceptions (#1060)

Fixes the 91 instances of exceptions being thrown when generating API
docs for the API-33 version of Mono.Android.  The two most common
instances were:

	System.IndexOutOfRangeException: Index was outside the bounds of the array.
	  at Java.Interop.Tools.JavaSource.UnknownHtmlElementStartTerminal.TryMatch (Irony.Parsing.ParsingContext context, Irony.Parsing.ISourceStream source) [0x0003e] in /Users/builder/azdo/_work/4/s/external/Java.Interop/src/Java.Interop.Tools.JavaSource/Java.Interop.Tools.JavaSource/SourceJavadocToXmldocGrammar.HtmlBnfTerms.cs:390

	System.ArgumentOutOfRangeException: Index was out of range. Must be non-negative and less than the size of the collection.
	Parameter name: index
	  at System.Collections.Generic.List`1[T].get_Item (System.Int32 index) [0x00009] in <08f46039e5064c628bf7795f9b970b7b>:0
	  at Java.Interop.Tools.JavaSource.SourceJavadocToXmldocGrammar+BlockTagsBnfTerms+<>c__DisplayClass1_0.<CreateRules>b__8 (Irony.Ast.AstContext context, Irony.Parsing.ParseTreeNode parseNode) [0x00025] in /Users/builder/azdo/_work/4/s/external/Java.Interop/src/Java.Interop.Tools.JavaSource/Java.Interop.Tools.JavaSource/SourceJavadocToXmldocGrammar.BlockTagsBnfTerms.cs:151

35 of the 118 href element parsing errors have also been fixed.

Author: pjcollins

src/Java.Interop.Tools.JavaSource/Java.Interop.Tools.JavaSource/SourceJavadocToXmldocGrammar.BlockTagsBnfTerms.cs

@@ -145,8 +145,12 @@ internal void CreateRules (SourceJavadocToXmldocGrammar grammar)
 					if (!grammar.ShouldImport (ImportJavadoc.ReturnTag)) {
 						return;
 					}
-					// When encountering multiple @return keys in a line, append subsequent @return key content to the original <returns> element.
 					var jdi = FinishParse (context, parseNode);
+					// If we have no return value, continue
+					if (parseNode.ChildNodes.Count < 2) {
+						return;
+					}
+					// When encountering multiple @return keys in a line, append subsequent @return key content to the original <returns> element.
 					if (jdi.Returns.Count == 0) {
 						var r = new XElement ("returns",
 							AstNodeToXmlContent (parseNode.ChildNodes [1]));

src/Java.Interop.Tools.JavaSource/Java.Interop.Tools.JavaSource/SourceJavadocToXmldocGrammar.HtmlBnfTerms.cs

@@ -2,6 +2,8 @@
 using System.Collections.Generic;
 using System.IO;
 using System.Linq;
+using System.Net;
+using System.Xml;
 using System.Xml.Linq;
 
 using Irony.Ast;
@@ -111,20 +113,17 @@ internal void CreateRules (SourceJavadocToXmldocGrammar grammar)
 						}
 					}
 
-					XNode astNodeElement = new XText (unparsedAElementValue);
-					try {
-						var seeElement = XElement.Parse ($"<see href={unparsedAElementValue}</see>");
-						var hrefValue = seeElement.Attribute ("href")?.Value ?? string.Empty;
-						if (!string.IsNullOrEmpty (hrefValue) &&
+					var seeElement = TryParseHRef (unparsedAElementValue);
+					if (seeElement == null)
+						seeElement = TryParseHRef (WebUtility.HtmlDecode (unparsedAElementValue), logError: true);
+
+					var hrefValue = seeElement?.Attribute ("href")?.Value ?? string.Empty;
+					if (!string.IsNullOrEmpty (hrefValue) &&
 							(hrefValue.StartsWith ("http", StringComparison.OrdinalIgnoreCase) || hrefValue.StartsWith ("www", StringComparison.OrdinalIgnoreCase))) {
-							parseNode.AstNode = seeElement;
-						} else {
-							// TODO: Need to convert relative paths or code references to appropriate CREF value.
-							parseNode.AstNode = astNodeElement;
-						}
-					} catch (Exception) {
-						Console.Error.WriteLine ($"# Unable to parse HTML element: <see href={unparsedAElementValue}</see>");
-						parseNode.AstNode = astNodeElement;
+						parseNode.AstNode = seeElement;
+					} else {
+						// TODO: Need to convert relative paths or code references to appropriate CREF value.
+						parseNode.AstNode = new XText (unparsedAElementValue);
 					}
 				};
 
@@ -185,6 +184,17 @@ static IEnumerable<XElement> GetParagraphs (ParseTreeNodeList children)
 				}
 			}
 
+			static XElement? TryParseHRef (string unparsedAElementValue, bool logError = false)
+			{
+				try {
+					return XElement.Parse ($"<see href={unparsedAElementValue}</see>");
+				} catch (Exception x) {
+					if (logError)
+						Console.Error.WriteLine ($"## Unable to parse HTML element: <see href={unparsedAElementValue}</see>\n{x.GetType ()}: {x.Message}");
+					return null;
+				}
+			}
+
 			public  readonly    NonTerminal AllHtmlTerms               = new NonTerminal (nameof (AllHtmlTerms), ConcatChildNodes);
 
 			public  readonly    NonTerminal TopLevelInlineDeclaration  = new NonTerminal (nameof (TopLevelInlineDeclaration), ConcatChildNodes);
@@ -387,7 +397,7 @@ public override void Init (GrammarData grammarData)
 			source.PreviewPosition += 1;
 			int start = source.Location.Position;
 			int stop  = start;
-			while (source.Text [stop] != '>' && stop < source.Text.Length)
+			while (stop < source.Text.Length && source.Text [stop] != '>' )
 				stop++;
 			if (addingRemarks) {
 				Console.Error.WriteLine ($"# Unsupported HTML element: {source.Text.Substring (start, stop - start)}");

tests/Java.Interop.Tools.JavaSource-Tests/SourceJavadocToXmldocGrammar.HtmlBnfTermsTests.cs

@@ -62,6 +62,11 @@ public void HyperLinkDeclaration ()
 			Assert.AreEqual ("<see href=\"https://developer.android.com/guide/topics/manifest/application-element.html\">application</see>",
 					r.Root.AstNode.ToString ());
 
+			r = p.Parse ("<a href=\"http://www.ietf.org/rfc/rfc2396.txt\">RFC&nbsp;2396: Uniform Resource Identifiers (URI): Generic Syntax</a>");
+			Assert.IsFalse (r.HasErrors (), DumpMessages (r, p));
+			Assert.AreEqual ("<see href=\"http://www.ietf.org/rfc/rfc2396.txt\">RFC 2396: Uniform Resource Identifiers (URI): Generic Syntax</see>",
+					r.Root.AstNode.ToString ());
+
 			r = p.Parse ("<a href=\"AutofillService.html#FieldClassification\">field classification</a>");
 			Assert.IsFalse (r.HasErrors (), DumpMessages (r, p));
 			Assert.AreEqual ("\"AutofillService.html#FieldClassification\"&gt;field classification",