(https://github.com/ruby/reline/pull/649)
* Fix waiting_proc precedence
* Fix waiting_operator bugs
* Add waiting_proc and vi_waiting_operator test
* Fix vi waiting operator arg number
vi_arg and vi_waiting_operator_arg should be multiplied
* Implement `yy` copies whole line in vi_command mode
* Simplify incremental search cancel test
* Add complex vi test with waiting_proc and vi_waiting_operator, split test input
https://github.com/ruby/reline/commit/777dffae1c
(https://github.com/ruby/irb/pull/886)
This is a feature that has been requested for a long time. It is now
possible to define custom commands in IRB.
Example usage:
```ruby
require "irb/command"
class HelloCommand < IRB::Command::Base
description "Prints hello world"
category "My commands"
help_message "It doesn't do more than printing hello world."
def execute
puts "Hello world"
end
end
IRB::Command.register(:hello, HelloCommand)
```
https://github.com/ruby/irb/commit/888643467c
navigation
(https://github.com/ruby/reline/pull/677)
Fixes https://github.com/ruby/reline/pull/675
This commit extracts the upward navigation condition in `LineEditor#input_key` to a new private method, and adds a new alias. This change allows Reline to support upward navigation in when a user has configured `inputrc` to map Shift-Tab to `menu-complete-backward`, a common setting in Bash (>= 4.x).
Instead of special-casing upward navigation in `LineEditor#input_key`, we now allow it to be processed by the branch that calls `process_key`. The extracted method no longer includes the editing mode check since this check is already made by `#wrap_method_call` by the time `#completion_journey_up` (or `#menu_complete_backward`) is called. Since upward navigation is happening in a method other than `#input_key` now, the `completion_occurs` variable that used to be local to `#input_key` is changed to an instance variable so that the new method can change its value. (I see many examples of mutating such instance variables in `LineEditor`, so I assumed this would be an uncontroversial change consistent with the coding practices already in place.)
Test coverage of this change has been added to the emacs and vi `KeyActor` tests.
Many thanks to @ima1zumi for their very helpful comments on #675 which encouraged me to contribute this work!
https://github.com/ruby/reline/commit/2ccdb374a4
In cases where `rb_ary_sort_bang` is called with a block and
tmp is an embedded array, we need to account for the block
potentially impacting the capacity of ary.
ex:
```
var_0 = (1..70).to_a
var_0.sort! do |var_0_block_129, var_1_block_129|
var_0.pop
var_1_block_129 <=> var_0_block_129
end.shift(3)
```
The above example can put the array into a corrupted state
resulting in a heap buffer overflow and possible segfault:
```
ERROR: AddressSanitizer: heap-buffer-overflow on address [...]
WRITE of size 560 at 0x60b0000034f0 thread T0 [...]
```
This commit adds a conditional to determine when the capacity
of ary has been modified by the provided block. If this is
the case, ensure that the capacity of ary is adjusted to
handle at minimum the len of tmp.
Given this example:
```rb
<<~HEREDOC
#{x}
HEREDOC
```
Both the parser gem and Prism's translation layer would generate the following AST:
```
s(:dstr,
s(:begin,
s(:int, 1)),
s(:str, " a\n"))
```
However, the parser gem inserts a empty string node into this node's location, like:
```
<Parser::Source::Map::Heredoc:0x0000000104ce73b8
@expression=#<Parser::Source::Range (string) 0...10>,
@heredoc_body=#<Parser::Source::Range (string) 11...20>,
@heredoc_end=#<Parser::Source::Range (string) 20...27>,
@node=s(:dstr,
s(:str, ""),
s(:begin,
s(:int, 1)),
s(:str, " a\n"))>
```
This is required to calculate the correct whitespace for the heredoc body.
We need to adjust the translation layer to account for this.
With this fix, we also won't need to ignore the tilde heredoc fixture anymore.
https://github.com/ruby/prism/commit/e7372e3ba5
(https://github.com/ruby/irb/pull/920)
This has a few benefits:
- We can keep hiding the evaluation logic inside the Context level, which
has always been the convention until #824 was merged recently.
- Although not an official API, gems like `debug` and `mission_control-jobs`
patch `Context#evaluate` to wrap their own logic around it. This implicit
contract was broken after #824, and this change restores it.
In addition to the refactor, I also converted some context-level evaluation
tests into integration tests, which are more robust and easier to maintain.
https://github.com/ruby/irb/commit/b32aee4068
../prism_compile.c: In function ‘pm_compile_node’:
../compile.c:583:24: warning: ‘retry_end_l’ may be used uninitialized in this function [-Wmaybe-uninitialized]
583 | anchor->last->next = elem;
| ~~~~~~~~~~~~~~~~~~~^~~~~~
In file included from ../compile.c:14256:
../prism_compile.c:5796:16: note: ‘retry_end_l’ was declared here
5796 | LABEL *retry_end_l;
| ^~~~~~~~~~~
../compile.c:255:42: warning: ‘retry_label’ may be used uninitialized in this function [-Wmaybe-uninitialized]
255 | #define LABEL_REF(label) ((label)->refcnt++)
| ^~
In file included from ../compile.c:14256:
../prism_compile.c:5795:16: note: ‘retry_label’ was declared here
5795 | LABEL *retry_label;
| ^~~~~~~~~~~
../prism_compile.c:5919:52: warning: ‘previous_block’ may be used uninitialized in this function [-Wmaybe-uninitialized]
5919 | ISEQ_COMPILE_DATA(iseq)->current_block = previous_block;
| ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~
This is best understood by looking at the change to the output:
```diff
# Insn: 0002 opt_and (stack_size: 2)
- mov rax, rsi
- and rax, rdi
- mov rsi, rax
+ and rsi, rdi
```
It's a bit awkward to match against due to how stack operands are
lowered, but hey, it's nice to save the 2 unnecessary MOVs.
The `rb_fstring(rb_enc_str_new())` pattern is inefficient because:
- It passes a mutable string to `rb_fstring` so if it has to be interned it will first be duped.
- It an equivalent interned string already exists, we allocated the string for nothing.
With `rb_enc_interned_str` we either directly get the pre-existing string with 0 allocations,
or efficiently directly intern the one we create without first duping it.
The backtick method recieves a frozen string unless it is interpolated.
Otherwise the string held in the ISeq could be mutated by a custom
backtick method.
yui-knk
Mari Imaizumi
tomoya ishida
Nobuyoshi Nakada
Stan Lo
Michael J. Giarlo
Zack Deveau
Kevin Newton
Peter Zhu
Stan Lo
David Marshall
Satoshi Tagomori
Takashi Kokubun
Cody Cutrer
Alan Wu
Jean Boussier
Naoto Ono