Search results

This turns parseVariableReference( final BasicScope scope, final Variable var ) into parseVariableReference( final BasicScope scope, final VariableReference var ). This is because VariableReferences will hold their current location, while Variables only have the location of their initial...

Small patch which makes it so that foreach in x generates Key variable name expected rather than Reserved word 'in' cannot be a key variable name as you clearly had just forgotten to put a variable name, rather than trying to name it "in". The impact of this patch is currently small, but will...

That's all of those submitted so far. You're up to date.

Currently, if you submit both a File and an InputStream to Parser( File, InputStream, Map ) (not that it's currently done), the InputStream will be discarded in favour of the File (in part because it is assumed that both are never submitted together). With the language server, this assumption...

"_prToday" shouldn't be updated if the response text of trying to visit a PR zone is "That isn't anywhere."

The methods parseRecord, parseFunction, parseAggregateLiteral, parseBasicScript, parseNewRecord and parseParameters all do something like this: while ( !this.currentToken().equals( ")" ) ) { [...] } this.readToken(); // read ) which assumes that the loop will always be exited while in...

Patch which delegates the discovery/reading of the initial token of parseNewRecord and parseTypedConstant to these very methods, rather than having to do it *before* going in, conforming to the rest of the method calls in parseValue.

Patch which streamlines parseStatic. This is possible because of how parseCommandOrDeclaration actually simply returns the same scope it was supplied as his first parameter, making Scope scope = [...]; Type type = [...]; return this.parseCommandOrDeclaration( scope, type ); and Scope scope =...

Patch which streamlines parseCatchValue.

Patch which streamlines parseSingleCommandScope, removing the need for Scope( final ParseTreeNode command, final BasicScope parentScope ).

Patch which removes the second parameter of parseType( final BasicScope scope, final boolean aggregates, final boolean records ), i.e. "aggregates". This parameter was used to determine whether or not declaring aggregates was allowed, but ever since the parsePluralConstant change, cases where...

Patch which adds documentation to parseAggregateLiteral, and which moves the reading of the opening bracket ("{") inside of it (it was previously necessary to read it *before* entering the method). This is done because, later on, every value/symbol/...thing will be assigned a "location", and...

@heeheehee The coverage... it's over NINE T(y p)E(r)CEEEEEEEEEEN(t) (94,2% :) )

oh, ok... Anyway, here's some more tests/coverage

Small patch which removes a redundant condition in parseInvoke if ( !this.parseIdentifier( current.content ) ) { throw this.parseException( "Variable reference expected for function name" ); } name = this.parseVariableReference(...

@heeheehee clover doesn't like multiline strings :( { "Basic function with forward reference", // vvvv line 478 vvvv """ void foo(); void bar() { foo(); } void foo() { return; }...

Here's the 2nd one, which is way shorter. One thing you'll probably notice is something in parseAggregateType: else if ( this.parseIdentifier( this.currentToken().content ) ) { indexType = scope.findType( this.currentToken().content ); if ( indexType !=...

Sure, here... It doesn't conflict with the error_conditionals_1 It's a simple addition of a null check. (The other conditions *still check for null*. That's intentional, because, again, the code will end up moving forward instead of throwing an error.)

Yeah, sorry about that. I put the null check earlier in the method, but didn't add it because I wanted this patch to not have (too many) extraneous stuff, since it's pretty big. It'll be in the next one, which will be way shorter.

Well, that's... kind of mean... (._. )