to it
(https://github.com/ruby/irb/pull/856)
* Test IRB's behaviour with empty input
* Handle empty input and pass it to debugger
Since `rdbg` accepts empty input to repeat the previous command, IRB
should take empty input in `irb:rdbg` sessions and pass them to the
debugger.
Currently, IRB simply ignores empty input and does nothing. This commit
creates `EmptyInput` to represent empty input so it can fit into the
current IRB's input processing flow in `Irb#eval_input`.
0e9db419be
When creating the Ruby AST, we were previously allocating Location
objects for every node and every inner location. Instead, this
commit changes it to pack both the start offset and length into a
single u64 and pass that into the nodes. Then, when the locations
are requested via a reader method, we lazily allocate the Location
objects.
de203dca83
Co-Authored-By: Aaron Patterson <tenderlove@ruby-lang.org>
```
test.rb:1:in 'Object#toplevel_meth': unhandled exception
from test.rb:4:in 'Foo.class_meth'
from test.rb:6:in 'Foo#instance_meth'
from test.rb:11:in 'singleton_meth'
from test.rb:13:in '<main>'
```
[Feature #19117]
Previously, our compile time check rejected dynamic symbols (e.g. what
String#to_sym could return) even though we could handle them just fine.
The runtime guards for the type of method name was also overly
restrictive and didn't accept dynamic symbols.
Fold the type check into the rb_get_symbol_id() and take advantage of
the guard already checking for 0. This also avoids generating the same
call twice in case the same method name is presented as different
types.
LSPs need this because the protocol dictates that you return code
units for offsets. None of our existing APIs provided that
information, and since we hid the source it's not nearly as useful
for them. Now they can pass an encoding directly to:
* Location#start_code_units_offset
* Location#end_code_units_offset
* Location#start_code_units_column
* Location#end_code_units_column
4757a2cc06
Co-Authored-By: Vinicius Stock <vinicius.stock@shopify.com>
This PR fixes following error when using `version: latest` argument.
```console
$ ruby -rprism -e "p Prism.parse('-> { it }', version: 'latest')"
-e:1:in `parse': invalid version: latest (ArgumentError)
p Prism.parse('-> { it }', version: 'latest')
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
from -e:1:in `<main>'
```
The argument `version: latest` in the added test is commented as potentially being
better replaced with `version: 3.4.0` in the future.
27b5c933cb
(https://github.com/ruby/irb/pull/869)
Currently, if the signature's constant part is not defined, a NameError
would be raised.
```
irb(main):001> show_source Foo
(eval):1:in `<top (required)>': uninitialized constant Foo (NameError)
Foo
^^^
from (irb):1:in `<main>'
```
This commit fixes the issue and simplifies the `edit` command's implementation.
8c16e029d1
(https://github.com/ruby/irb/pull/870)
* Assign `@eval_path` through `irb_path=` method
This simplifies the original caching logic for the `eval_path` method
and makes it easier to understand.
* Refactor SourceFinder::Source
c63e4c4035
(https://github.com/ruby/irb/pull/862)
* Powerup show_source by enabling RubyVM.keep_script_lines
* Add file_content field to avoid reading file twice while show_source
* Change path passed to eval, don't change irb_path.
* Encapsulate source coloring logic and binary file check insode class Source
* Add edit command testcase when irb_path does not exist
* Memoize irb_path existence to reduce file existence check calculating eval_path
239683a937
Previously, `**nil` by itself worked, but if you add a block argument,
it raised a conversion error. The presence of the block argument
shouldn't change how keyword splat works.
See: <https://bugs.ruby-lang.org/issues/20064>
The following code previously caused a crash:
```ruby
h = {}
1000000.times{|i| h[i.to_s.to_sym] = i}
def f(kw: 1, **kws) end
f(**h)
```
Inside a thread or fiber, the size of the keyword splat could be much smaller
and still cause a crash.
I found this issue while optimizing method calling by reducing implicit
allocations. Given the following code:
```ruby
def f(kw: , **kws) end
kw = {kw: 1}
f(**kw)
```
The `f(**kw)` call previously allocated two hashes callee side instead of a
single hash. This is because `setup_parameters_complex` would extract the
keywords from the keyword splat hash to the C stack, to attempt to mirror
the case when literal keywords are passed without a keyword splat. Then,
`make_rest_kw_hash` would build a new hash based on the extracted keywords
that weren't used for literal keywords.
Switch the implementation so that if a keyword splat is passed, literal keywords
are deleted from the keyword splat hash (or a copy of the hash if the hash is
not mutable).
In addition to avoiding the crash, this new approach is much more
efficient in all cases. With the included benchmark:
```
1
miniruby: 5247879.9 i/s
miniruby-before: 2474050.2 i/s - 2.12x slower
1_mutable
miniruby: 1797036.5 i/s
miniruby-before: 1239543.3 i/s - 1.45x slower
10
miniruby: 1094750.1 i/s
miniruby-before: 365529.6 i/s - 2.99x slower
10_mutable
miniruby: 407781.7 i/s
miniruby-before: 225364.0 i/s - 1.81x slower
100
miniruby: 100992.3 i/s
miniruby-before: 32703.6 i/s - 3.09x slower
100_mutable
miniruby: 40092.3 i/s
miniruby-before: 21266.9 i/s - 1.89x slower
1000
miniruby: 21694.2 i/s
miniruby-before: 4949.8 i/s - 4.38x slower
1000_mutable
miniruby: 5819.5 i/s
miniruby-before: 2995.0 i/s - 1.94x slower
```
(https://github.com/ruby/irb/pull/867)
* Remove IRB.irb_exit! method
It's not necessary to introduce a new method just for the exit! command
at this moment.
* Rename ExitForcedAction to ForceExit
* Move force exit tests to a dedicated file
* Fix nested history saving with exit! command
Because we switched to use `Kernel#exit` instead of `exit!`, the outer
session's ensure block in `Irb#run` will be run, which will save the
history. This means the separate check to save history when force exiting
is no longer necessary.
* execute_lines helper should also capture IRB setup's output
This prevents setup warnings from being printed to test output
while allowing those output to be tested.
* Update readme
899d10ade1
(https://github.com/ruby/irb/pull/851)
* Added failing test for when writing history on exit
* Save history on exit
* Exit early when calling Kernel.exit
* use status 0 for kernel.exit
* Added test for nested sessions
* Update lib/irb.rb
---------
c0a5f31679
Co-authored-by: Stan Lo <stan001212@gmail.com>
`dest` of `const_decl_path` is `NODE_COLON2` or `NODE_COLON3` in some cases.
For example, `B::C ||= [“Not ” + “shareable”]` passes `NODE_COLON2`
and `::C ||= [“Not ” + “shareable”]` passes `NODE_COLON3`.
This commit fixes `Ractor::IsolationError` message for such case.
```
# shareable_constant_value: literal
::C ||= ["Not " + "shareable"]
# Before
# => cannot assign unshareable object to C (Ractor::IsolationError)
# After
# => cannot assign unshareable object to ::C (Ractor::IsolationError)
```
If lhs of assignment is top-level constant reference, the first
constant name is omitted from error message.
This commit fixes it.
```
# shareable_constant_value: literal
::C = ["Not " + "shareable"]
# Before
# => cannot assign unshareable object to (Ractor::IsolationError)
# After
# => cannot assign unshareable object to ::C (Ractor::IsolationError)
```
