Merge pull request #51 from WolframResearch/feature/add-rewrites-for-named-character-names

Add rewrites for named character names (to named characters)

WolframLanguageForJupyter/Resources/CompletionUtilities.wl

@@ -0,0 +1,84 @@
+(************************************************
+				CompletionUtilities.wl
+*************************************************
+Description:
+	Utilities for the aiding in the
+		(auto-)completion of Wolfram Language
+		code
+Symbols defined:
+	rewriteNamedCharacters
+*************************************************)
+
+(************************************
+	Get[] guard
+*************************************)
+
+If[
+	!TrueQ[WolframLanguageForJupyter`Private`$GotCompletionUtilities],
+	
+	WolframLanguageForJupyter`Private`$GotCompletionUtilities = True;
+
+(************************************
+	load required
+		WolframLanguageForJupyter
+		files
+*************************************)
+
+	Get[FileNameJoin[{DirectoryName[$InputFileName], "Initialization.wl"}]]; (* unicodeNamedCharactersReplacements,
+																					verticalEllipsis *)
+
+(************************************
+	private symbols
+*************************************)
+
+	(* begin the private context for WolframLanguageForJupyter *)
+	Begin["`Private`"];
+
+(************************************
+	utilities for rewriting
+		Wolfram Language code
+*************************************)
+
+	(* rewrite names (in a code string) into named characters *)
+	rewriteNamedCharacters[codeToAnalyze_?StringQ] :=
+		Module[
+			{codeUsingFullReplacements},
+			codeUsingFullReplacements =
+				StringReplace[
+					codeToAnalyze,
+					Normal @ unicodeNamedCharactersReplacements
+				];
+			If[
+				StringCount[
+					codeUsingFullReplacements,
+					verticalEllipsis | "\\["
+				] != 1,
+				Return[{codeUsingFullReplacements}];
+			];
+			Return[
+				Flatten[
+					StringCases[
+						codeUsingFullReplacements,
+						before___ ~~ name : ((verticalEllipsis | "\\[") ~~ rest__ ~~ EndOfString) :>
+							(
+								(StringJoin[before, #1] &) /@
+									Values[
+										KeySelect[
+											unicodeNamedCharactersReplacements,
+											StringMatchQ[#1, name ~~ ___] &
+										]
+									]
+							)
+					]
+				]
+			];
+		];
+
+	(* end the private context for WolframLanguageForJupyter *)
+	End[]; (* `Private` *)
+
+(************************************
+	Get[] guard
+*************************************)
+
+] (* WolframLanguageForJupyter`Private`$GotCompletionUtilities *)

WolframLanguageForJupyter/Resources/Initialization.wl

@@ -16,6 +16,8 @@ Symbols defined:
 	keyString,
 	baseString,
 	heartbeatString,
+	verticalEllipsis,
+	unicodeNamedCharactersReplacements,
 	ioPubString,
 	controlString,
 	inputString,
@@ -61,6 +63,112 @@ If[
 	$Messages = {};
 	$Output = {}; *)
 
+(************************************
+	discover the named unicode
+		characters and their names
+*************************************)
+
+	(* the vertical ellipsis character *)
+	verticalEllipsis = FromCharacterCode[8942, "Unicode"];
+
+	(* pre-define the association of names and named characters as empty *)
+	unicodeNamedCharactersReplacements = Association[];
+
+	Block[
+		{
+			(* the absolute file name for "UnicodeCharacters.tr" *)
+			unicodeCharactersTRFileName,
+
+			(* raw data extracted from "UnicodeCharacters.tr" *)
+			charactersAndTheirNames
+		},
+
+		(* attempt to get the full location of "UnicodeCharacters.tr" *)
+		unicodeCharactersTRFileName = UsingFrontEnd[System`Dump`unicodeCharactersTR];
+
+		(* try again if using System`Dump`unicodeCharactersTR does not work *)
+		If[
+			!StringQ[unicodeCharactersTRFileName],
+			unicodeCharactersTRFileName =
+				UsingFrontEnd[
+					ToFileName[
+						FrontEnd`FileName[
+							{
+								$InstallationDirectory,
+								"SystemFiles",
+								"FrontEnd",
+								"TextResources"
+							},
+							"UnicodeCharacters.tr"
+						]
+					]
+				];
+		];
+
+		If[
+			StringQ[unicodeCharactersTRFileName],
+			charactersAndTheirNames =
+				(
+					(* parse the third item of a row (for a named character) into a list *)
+					ReplacePart[
+						#1,
+						3 ->
+							StringCases[
+								(* remove extraneous parentheses *)
+								StringTrim[
+									#1[[3]],
+									"(" | ")"
+								],
+								(* extract the escape sequences for the named character *)
+								Longest[escSeq : (Except[WhitespaceCharacter] ..)] :>
+									StringJoin[verticalEllipsis, StringTrim[escSeq, "$"], verticalEllipsis]
+							]
+					] &
+				) /@
+					(* only use lists with at least three items *)
+					Select[
+						(* parse the rows into their items (where each item is separated by two tabs) *)
+						(StringSplit[#1, "\t\t"] &) /@
+							(
+								(* split unicodeCharactersTRFileName into its lines *)
+								StringSplit[
+									Import[unicodeCharactersTRFileName, "String"],
+									"\n"
+								(* drop the first row *)
+								][[2 ;;]]
+							),
+						(Length[#1] >= 3) &
+					];
+
+			(* parse the data into an association of names and the named characters they correspond to *)
+			unicodeNamedCharactersReplacements =
+				(* sort the keys by string length *)
+				KeySort[
+					(* sort the keys in the default manner *)
+					KeySort[
+						(* drop "empty names" *)
+						KeyDrop[
+							(* make an association *)
+							Association[
+								(* create a list of rules of names and named characters *)
+								Thread[
+									Rule[
+										(Prepend[#1[[3]], #1[[2]]]),
+										FromCharacterCode[FromDigits[StringDrop[#1[[1]], 2], 16], "Unicode"]
+									]
+								] & /@ charactersAndTheirNames
+							],
+							{
+								StringJoin[Table[verticalEllipsis, {2}]],
+								"\\[]"
+							}
+						]
+					],
+					(StringLength[#1] < StringLength[#2]) &
+				];
+		];
+	];
+
 (************************************
 	various important symbols
 		for use by
@@ -101,13 +209,10 @@ If[
 	$outputSetToTraditionalForm := (Lookup[Options[$Output], FormatType] === TraditionalForm);
 	$outputSetToTeXForm := (Lookup[Options[$Output], FormatType] === TeXForm);
 	$trueFormatType :=
-		Which[
+		If[
 			$outputSetToTraditionalForm,
 			TraditionalForm,
-			$outputSetToTeXForm,
-			TeXForm,
-			True,
-			Identity
+			If[$outputSetToTeXForm, TeXForm, #&]
 		];
 
 	(* obtain details on how to connect to Jupyter, from Jupyter's invocation of "KernelForWolframLanguageForJupyter.wl" *)

WolframLanguageForJupyter/Resources/KernelForWolframLanguageForJupyter.wl

@@ -36,7 +36,7 @@ Get[FileNameJoin[{DirectoryName[$InputFileName], "Initialization.wl"}]]; (* init
 Get[FileNameJoin[{DirectoryName[$InputFileName], "SocketUtilities.wl"}]]; (* sendFrame *)
 Get[FileNameJoin[{DirectoryName[$InputFileName], "MessagingUtilities.wl"}]]; (* getFrameAssoc, createReplyFrame *)
 
-Get[FileNameJoin[{DirectoryName[$InputFileName], "RequestHandlers.wl"}]]; (* executeRequestHandler *)
+Get[FileNameJoin[{DirectoryName[$InputFileName], "RequestHandlers.wl"}]]; (* executeRequestHandler, completeRequestHandler *)
 
 (************************************
 	private symbols
@@ -96,6 +96,10 @@ loop[] :=
 					"execute_request",
 					(* executeRequestHandler will read and update loopState *)
 					executeRequestHandler[];,
+					(* use the tab-completion functionality to rewrite named character names *)
+					"complete_request",
+					(* completeRequestHandler will read and update loopState *)
+					completeRequestHandler[];,
 					(* if asking the kernel to shutdown, set doShutdown to True *)
 					"shutdown_request",
 					loopState["replyMsgType"] = "shutdown_reply";

WolframLanguageForJupyter/Resources/OutputHandlingUtilities.wl

@@ -209,10 +209,7 @@ If[
 						be identical to the string result of an InputForm-wrapped expression itself *)
 					({"&#", ToString[#1], ";"} & /@ 
 						ToCharacterCode[
-							(* toStringUsingOutput[result] *)
-							Quiet[
-								ToString[If[!isTeXWrapped, $trueFormatType[result], result]]
-							],
+							(* toStringUsingOutput[result] *) ToString[If[!isTeXWrapped, $trueFormatType[result], result]],
 							"Unicode"
 						]),
 

WolframLanguageForJupyter/Resources/RequestHandlers.wl

@@ -5,7 +5,8 @@ Description:
 	Handlers for message frames of type
 		"x_request" arriving from Jupyter
 Symbols defined:
-	executeRequestHandler
+	executeRequestHandler,
+	completeRequestHandler
 *************************************************)
 
 (************************************
@@ -30,7 +31,10 @@ If[
 
 	Get[FileNameJoin[{DirectoryName[$InputFileName], "EvaluationUtilities.wl"}]]; (* simulatedEvaluate *)
 
-	Get[FileNameJoin[{DirectoryName[$InputFileName], "OutputHandlingUtilities.wl"}]]; (* textQ, toOutText, toOutImage *)
+	Get[FileNameJoin[{DirectoryName[$InputFileName], "OutputHandlingUtilities.wl"}]]; (* textQ, toOutText, toOutImage,
+																							containsPUAQ *)
+
+	Get[FileNameJoin[{DirectoryName[$InputFileName], "CompletionUtilities.wl"}]]; (* rewriteNamedCharacters *)
 
 (************************************
 	private symbols
@@ -69,7 +73,6 @@ If[
 			(* set the content of the reply to information about WolframLanguageForJupyter's execution of the input *)
 			loopState["replyContent"] = 
 				ExportString[
-					(* kind of self-explanatory *)
 					Association[
 						"status" -> "ok",
 						"execution_count" -> loopState["executionCount"],
@@ -278,6 +281,54 @@ If[
 			loopState["executionCount"] += totalResult["ConsumedIndices"];
 		];
 
+(************************************
+	handler for complete_requests
+*************************************)
+
+	(* handle complete_request messages frames received on the shell socket *)
+	completeRequestHandler[] :=
+		Module[
+			{
+				(* for storing the code string to offer completion suggestions on *)
+				codeStr
+			},
+			(* get the code string to rewrite the named characters of, ending at the cursor *)
+			codeStr =
+				StringTake[
+					loopState["frameAssoc"]["content"]["code"],
+					{
+						1,
+						loopState["frameAssoc"]["content"]["cursor_pos"]
+					}
+				];
+			(* set the appropriate reply type *)
+			loopState["replyMsgType"] = "complete_reply";
+			(* set the content of the reply to a list of rewrites for any named characters in the code string *)
+			loopState["replyContent"] = 
+				ByteArrayToString[
+					ExportByteArray[
+						Association[
+							"matches" ->
+								DeleteDuplicates[
+									Prepend[
+										Select[
+											rewriteNamedCharacters[codeStr],
+											(!containsPUAQ[#1])&
+										],
+										codeStr
+									]
+								],
+							"cursor_start" -> 0,
+							"cursor_end" -> StringLength[codeStr],
+							"metadata" -> {},
+							"status" -> "ok"
+						], 
+						"JSON",
+						"Compact" -> True
+					]
+				];
+		];
+
 	(* end the private context for WolframLanguageForJupyter *)
 	End[]; (* `Private` *)
 

extras/custom.js

@@ -0,0 +1,50 @@
+/* (adapted) from https://stackoverflow.com/a/19961519 by Erik Aigner */
+HTMLTextAreaElement.prototype.insertAtCaret = function (text) {
+	text = text || '';
+	if(document.selection) {
+		// IE
+		this.focus();
+		var sel = document.selection.createRange();
+		sel.text = text;
+	}
+	else if(this.selectionStart || this.selectionStart === 0) {
+		// Others
+		var startPos = this.selectionStart;
+		var endPos = this.selectionEnd;
+		this.value = this.value.substring(0, startPos) + text + this.value.substring(endPos, this.value.length);
+		this.selectionStart = startPos + text.length;
+		this.selectionEnd = startPos + text.length;
+	}
+	else {
+		this.value += text;
+	}
+};
+
+/* (adapted) from https://stackoverflow.com/a/51114347 by bambam */
+function redirectEsc(event) {
+	if(event.which == 27)
+	{
+		event.target.insertAtCaret(
+				/* the vertical ellipsis character */
+				String.fromCharCode(8942)
+			);
+		event.stopImmediatePropagation();
+	}
+}
+
+/* (adapted) from https://stackoverflow.com/a/51114347 by bambam */
+var observer = new MutationObserver(function(mutations) {
+
+	Array.from(
+			document.querySelectorAll('.input_area')
+		).forEach(
+				textarea =>
+				{
+					textarea.removeEventListener('keydown', redirectEsc);
+					textarea.addEventListener('keydown', redirectEsc);
+				}
+		);
+
+});
+
+observer.observe(document, {childList:true, subtree:true});