Accept static as a declaration modifier, not only a function modifier

[ghedwards]

Closes #63.

component { static myStatic = "v"; }

threw:

java.lang.NullPointerException: Cannot invoke "org.antlr.v4.runtime.tree.TerminalNode.getSymbol()"
because the return value of "cfml.CFSCRIPTParser$FunctionDeclarationContext.FUNCTION()" is null

It threw rather than reporting an error, so one such line took the whole file with it. Through CFLint that surfaces as PARSE_ERROR: Unable to parse and no lint results at all for the file.

Cause

STATIC was reachable only as a functionModifier. functionDeclaration starts functionModifier* accessType? functionModifier* typeSpec? FUNCTION identifier, so prediction committed on seeing static, no FUNCTION token arrived, and the visitor dereferenced ctx.FUNCTION() unconditionally.

localAssignmentExpression already carried FINAL for exactly this reason — which is why final x = 1 worked and static x = 1 did not, despite the two sitting side by side in both componentModifier and functionModifier. STATIC now sits beside it.

The visitor is taught the shape

Per CLAUDE.md, a green parse proves nothing here — this is the rule where visitChildren silently folds an unhandled shape into whatever is nearby. So CFVarDeclExpression gains staticDecl alongside finalDecl, and Decompile emits static rather than falling through to var .

Checked by reading the output rather than trusting the parse:

component { static myStatic = "v"; final myConst = "c"; }
  ->  static myStatic = 'v';
      final myConst = 'c';

Both markers survive the round trip.

Pre-existing faults left alone

Confirmed against master before this change, so none is a regression introduced here:

var final x = 1;   ->  { var final; x = 1; }
var static x = 1;  ->  { var static; x = 1; }
static var x = 1;  ->  { static public var function x() ;; 1; }

FINAL and STATIC are both valid identifiers since #46, so var final parses as VAR plus an identifier named final and splits into two statements. static var produces outright garbage. Separate fault in the same rule; the fixture deliberately stays off these so it tests #63 and nothing else. Worth its own issue — say the word and I'll file it.

Also unrelated and unchanged: any multi-element component decompiles with an extra { } around the body, including component { x = 1; y = 2; } with no static in sight.

Verification

• 323 tests, ./gradlew build
• Differential harness unchanged at 1, nothing newly broken case-by-case
• CFLint 675 against a clean build
• components/static_assignment_63.cfc fails with cfml.parsing/src/main stashed — the expected tree does not match

Not covered by this PR

#64, the access modifier inside a static block, is the other half of cfmleditor/CFLint#50 and is a separate fault in staticBlock. Coming next.

---

Generated by Claude Code