Merge pull request #147 from constructive-io/devin/1775609815-fix-scan-json-escaping

fix: escape control characters in build_scan_json() for multi-line tokens

Author: pyramation

full/src/wasm_wrapper.c

@@ -379,8 +379,19 @@ static char* build_scan_json(PgQuery__ScanResult *scan_result, const char* origi
             char c = token_text[j];
             if (c == '"' || c == '\\') {
                 escaped_text[escaped_pos++] = '\\';
+                escaped_text[escaped_pos++] = c;
+            } else if (c == '\n') {
+                escaped_text[escaped_pos++] = '\\';
+                escaped_text[escaped_pos++] = 'n';
+            } else if (c == '\r') {
+                escaped_text[escaped_pos++] = '\\';
+                escaped_text[escaped_pos++] = 'r';
+            } else if (c == '\t') {
+                escaped_text[escaped_pos++] = '\\';
+                escaped_text[escaped_pos++] = 't';
+            } else {
+                escaped_text[escaped_pos++] = c;
             }
-            escaped_text[escaped_pos++] = c;
         }
         escaped_text[escaped_pos] = '\0';
 

full/test/scan.test.js

@@ -225,5 +225,55 @@ describe("Query Scanning", () => {
       assert.equal(typeof result1.version, "number");
       assert.ok(result1.version > 0);
     });
+
+    it("should handle multi-line dollar-quoted strings without JSON errors", () => {
+      // Without the fix, scanSync throws:
+      //   "Bad control character in string literal"
+      // because build_scan_json() doesn't escape \n in the token text.
+      const sql = `CREATE FUNCTION test() RETURNS void AS $$
+BEGIN
+  RAISE NOTICE 'hello';
+END;
+$$ LANGUAGE plpgsql`;
+
+      const result = query.scanSync(sql);
+      assert.equal(typeof result, "object");
+      assert.ok(Array.isArray(result.tokens));
+      assert.ok(result.tokens.length > 0);
+
+      // The dollar-quoted body spans multiple lines
+      const dollarToken = result.tokens.find(t => t.text.includes('BEGIN'));
+      assert.ok(dollarToken, "should have a token containing the function body");
+      assert.ok(dollarToken.text.includes('\n'), "token text should preserve newlines");
+    });
+
+    it("should handle dollar-quoted tokens with tabs", () => {
+      // Tab characters also break JSON.parse when unescaped.
+      const sql = `SELECT $$line1
+	indented
+line3$$`;
+
+      const result = query.scanSync(sql);
+      assert.equal(typeof result, "object");
+      assert.ok(Array.isArray(result.tokens));
+
+      const dollarToken = result.tokens.find(t => t.text.includes('indented'));
+      assert.ok(dollarToken, "should have a token containing the tabbed content");
+    });
+
+    it("should handle multi-line block comments", () => {
+      // C-style block comments spanning multiple lines hit the same bug.
+      const sql = `SELECT 1; /* multi
+line
+comment */ SELECT 2`;
+
+      const result = query.scanSync(sql);
+      assert.equal(typeof result, "object");
+      assert.ok(Array.isArray(result.tokens));
+
+      const commentToken = result.tokens.find(t => t.tokenName === "C_COMMENT");
+      assert.ok(commentToken, "should have a C_COMMENT token");
+      assert.ok(commentToken.text.includes('\n'), "comment text should preserve newlines");
+    });
   });
-});
+});