Highlight DATA statement contents better (#128)

Fixes #43
mattgodbolt · Jan 27, 2025 · 085dda4 · 085dda4
1 parent 68fc71a
commit 085dda4
Show file tree

Hide file tree

Showing 2 changed files with 66 additions and 0 deletions.
diff --git a/src/bbcbasic.js b/src/bbcbasic.js
@@ -118,6 +118,7 @@ export function registerBbcBasicLanguage() {
                 [":", "symbol", "@pop"],
                 [/(\bREM|\xf4)/, {token: "keyword", next: "@remStatement"}], // A REM consumes to EOL
                 [/(FN|PROC|PRO\.|\xa4|\xf2)/, {token: "keyword", next: "@fnProcName"}],
+                [/(\bDATA|\bDAT\.|\bDA\.|\bD\.|\xdc)/, {token: "keyword", next: "@dataStatement"}], // DATA consumes to EOL
                 [
                     /THEN|THE\.|TH\.|ELSE|ELS\.|EL\.|ERROR|ERRO\.|ERR\.|\u018c|\u018b|\u0185/,
                     "keyword",
@@ -177,6 +178,13 @@ export function registerBbcBasicLanguage() {
                 [/["\u201c\u201d]/, {token: "string.quote", next: "@pop"}],
             ],
             remStatement: [[/.*/, "comment", "@pop"]],
+            dataStatement: [
+                [/(?=\n)/, "", "@pop"],
+                [",", "symbol"],
+                {include: "@whitespace"},
+                [/["\u201c\u201d]/, {token: "string.quote", next: "@string"}],
+                [/[^,\n]*[^,\n ]/, "string.unquoted"],
+            ],
             asm: [
                 [
                     /ADC|AND|ASL|B(CC|CS|EQ|MI|NE|PL|VC|VS)|BIT|BRK|CL[CDIV]|CMP|CP[XY]|DE[CXY]|EOR|IN[CXY]|JMP|JSR|LD[AXY]|LSR|NOP|ORA|PH[AP]|PL[AP]|RO[LR]|RTI|RTS|SBC|SE[CDI]|ST[AXY]|TA[XY]|TSX|TX[AS]|TYA/,

diff --git a/test/bbcbasic_test.js b/test/bbcbasic_test.js
@@ -276,6 +276,64 @@ describe("Tokenisation", () => {
             [{offset: 0, type: "keyword"}],
         );
     });
+    it("should suitably highlight DATA statements", () => {
+        checkTokens(
+            [
+                "DATA",
+                "D.a b,c",
+                "DA.,,,7,",
+                "DAT.PI,COS0,SQRLN@%",
+                "DATA3 ,\"comma,quoted\", \"esc\"\"ape\", qu\"te,q\", 3x",
+            ],
+            [
+                {offset: 0, type: "keyword"},
+            ],
+            [
+                {offset: 0, type: "keyword"},
+                {offset: 2, type: "string.unquoted"},
+                {offset: 5, type: "symbol"},
+                {offset: 6, type: "string.unquoted"},
+            ],
+            [
+                {offset: 0, type: "keyword"},
+                {offset: 3, type: "symbol"},
+                {offset: 6, type: "string.unquoted"},
+                {offset: 7, type: "symbol"},
+            ],
+            [
+                {offset: 0, type: "keyword"},
+                {offset: 4, type: "string.unquoted"},
+                {offset: 6, type: "symbol"},
+                {offset: 7, type: "string.unquoted"},
+                {offset: 11, type: "symbol"},
+                {offset: 12, type: "string.unquoted"},
+            ],
+            [
+                {offset: 0, type: "keyword"},
+                {offset: 4, type: "string.unquoted"},
+                {offset: 5, type: "white"},
+                {offset: 6, type: "symbol"},
+                {offset: 7, type: "string.quote"},
+                {offset: 8, type: "string"},
+                {offset: 20, type: "string.quote"},
+                {offset: 21, type: "symbol"},
+                {offset: 22, type: "white"},
+                {offset: 23, type: "string.quote"},
+                {offset: 24, type: "string"},
+                {offset: 27, type: "string.quote"},
+                {offset: 29, type: "string"},
+                {offset: 32, type: "string.quote"},
+                {offset: 33, type: "symbol"},
+                {offset: 34, type: "white"},
+                {offset: 35, type: "string.unquoted"},
+                {offset: 40, type: "symbol"},
+                {offset: 41, type: "string.unquoted"},
+                {offset: 43, type: "symbol"},
+                {offset: 44, type: "white"},
+                {offset: 45, type: "string.unquoted"},
+            ],
+        );
+    });
     it("should handle symbols and operators", () => {
         checkTokens(["~"], [{offset: 0, type: "operator"}]);
         checkTokens(["#"], [{offset: 0, type: "symbol"}]);