-
Notifications
You must be signed in to change notification settings - Fork 13
Description
npm run lint fails - I'm not convinced it has passed any time recently.
Please can we run it in CI and either fix/ignore the errors or adjust the rules to suit the team?
If you're happy with the existing rules I can take a first pass through fixing and flag controversial bits here. It seems likely to mostly be uncontroversial fixes except for a few places where some Blockly naming conventions appear to differ from the rules. I think it would make sense to ignore those rules on a file or line basis in a first pass.
I tend to have eslint diagnostics enabled in my editor and find all the lint errors distract me from the new errors I'm introducing 😄
ASTNode createXXX return type
There's an annoyingly prevalent pattern of a non-null assertion (not allowed by the rules) in code like ASTNode.createBlockNode(newBlock)! that could do with a separate fix. Why can it (and similar, but mainly this one in practice) return null? The implementation seems to be undecided about its types:
static createBlockNode(block: Block): ASTNode | null {
if (!block) {
return null;
}
return new ASTNode(ASTNode.types.BLOCK, block);
}