Bob Nystrom
e20da2f070
Replace all occurrences of the old `fn` function literal syntax with the new `new Fn {|params| body}` block syntax across benchmark, test, and source files. Update the Wren compiler to handle the pipe operator as a block argument delimiter by removing `TOKEN_PIPE` from the newline-swallowing token list and adding explicit newline matching in `infixOp`. Adjust test expectations in `conditional/precedence.wren` to remove `fn`-based expressions and update `test/assignment/local.wren` to use the new syntax.
The benchmarks in here attempt to faithfully implement the exact same algorithm in a few different languages. We're using Lua, Python, and Ruby for comparison here because those are all in Wren's ballpark: dynamically-typed, object-oriented, bytecode-compiled.
A bit about each benchmark:
binary_trees
This benchmark stresses object creation and garbage collection. It builds a few big, deeply nested binaries and then traverses them.
fib
This is just a simple naïve Fibonacci number calculator. It was the first benchmark I wrote when Wren supported little more than function calls and arithmetic. It isn't particularly representative of real-world code, but it does stress function call and arithmetic.
for
This microbenchmark just tests the performance of for loops. Not too useful, but i used it when implementing for in Wren to make sure it wasn't too far off the mark.
method_call
This is the most useful benchmark: it tests dynamic dispatch and polymorphism. You'll note that the main iteration loop is unrolled in all of the implementations. This is to ensure that the loop overhead itself doesn't dwarf the method call time.