Tags: mvidner/sorbet
Tags
Feature/1001 avoid segfault on empty literal mergestring (sorbet#1012) * test(ast/desugar): Add EmptyTree Desuger Test * test: Add empty_string_concatenation test * fix: check if expression is EmptyTree or not on mergeStrings * fix(ast/desuger/Desuger): Use ""sv * fix(ast/desuger/test): Remove redundant test
Changes for sorbet#979 - Document Prereq's (sorbet#1007) * Changes for sorbet#979 Changes for sorbet#979 * Move to FAQ and make it look better
Fix Float#truncate signature. (sorbet#988) It can get an (optional) Integer as an argument and return Integer or Float depending on the arg value.
Enable more [[nodiscard]] in stdlib (sorbet#999) Found in https://libcxx.llvm.org/docs/UsingLibcxx.html#extended-applications-of-nodiscard
Add comment on concurrent queue choice. (sorbet#1008)
Desugar multiple assignment such that it returns rhs (sorbet#984) * Desugar multiple assignment such that it returns rhs We were not being faithful to what Ruby does with our multiple assignment desugar. We were returning the result of a call to our magic expand splat method which is almost always not what actually happens at runtime. Here is the result of the attached test case before this commit: ```ruby extend T::Sig sig {returns(T.nilable(T::Array[Integer]))} def foo nil end T.reveal_type((a, b = foo)) # Revealed type: T.any(T::Array[Integer], [NilClass, NilClass]) T.reveal_type((c, *, d = 100)) # Revealed type: [Integer(100), NilClass] (2-tuple) ``` * Insert into the front instead of reserving space and poking * Add example from sorbet#981 * Add ast.exp file * Add helpful comment Co-Authored-By: Jake Zimmerman <zimmerman.jake@gmail.com>
Docs: Clarify what should be comitted to source control (sorbet#943) * Docs: Clarify what should be comitted to source control * Remove incorrect mention of Makefile * Add cross reference to reasoning
Docs: Fix description of `typed: strong` (sorbet#942) * Docs: Fix description of `typed: strong` * Explain that typed: strong is rarely used * Fix prettier
Make LSP Hover non-experimental (removes `--enable-experimental-lsp-h… …over`) (sorbet#990) Note: This is going to break those who passed --enable-experimental-lsp-hover before. Hover works well enough to be enabled for everyone. If there is a bug, it is easy to add a new test case to Sorbet and iterate on a solution. In real-world uses, hover does not appear to cause any issues with typechecking speed. In particular, VS Code cancels hover if you move your mouse away, so mousing over many things during a slow path won't interrupt edit batching.
PreviousNext