ruby/ast.c

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

1019 строки
38 KiB
C
Исходник Обычный вид История

/* indent-tabs-mode: nil */
#include "internal.h"
#include "internal/ruby_parser.h"
#include "internal/symbol.h"
#include "internal/warnings.h"
#include "iseq.h"
#include "node.h"
#include "ruby.h"
#include "ruby/encoding.h"
#include "ruby/util.h"
#include "vm_core.h"
#include "builtin.h"
static VALUE rb_mAST;
static VALUE rb_cNode;
static VALUE rb_cLocation;
struct ASTNodeData {
VALUE ast_value;
2020-07-08 12:07:30 +03:00
const NODE *node;
};
static void
node_gc_mark(void *ptr)
{
struct ASTNodeData *data = (struct ASTNodeData *)ptr;
rb_gc_mark(data->ast_value);
}
2019-07-23 10:18:40 +03:00
static size_t
node_memsize(const void *ptr)
{
struct ASTNodeData *data = (struct ASTNodeData *)ptr;
rb_ast_t *ast = rb_ruby_ast_data_get(data->ast_value);
[Universal parser] Decouple IMEMO from rb_ast_t This patch removes the `VALUE flags` member from the `rb_ast_t` structure making `rb_ast_t` no longer an IMEMO object. ## Background We are trying to make the Ruby parser generated from parse.y a universal parser that can be used by other implementations such as mruby. To achieve this, it is necessary to exclude VALUE and IMEMO from parse.y, AST, and NODE. ## Summary (file by file) - `rubyparser.h` - Remove the `VALUE flags` member from `rb_ast_t` - `ruby_parser.c` and `internal/ruby_parser.h` - Use TypedData_Make_Struct VALUE which wraps `rb_ast_t` `in ast_alloc()` so that GC can manage it - You can retrieve `rb_ast_t` from the VALUE by `rb_ruby_ast_data_get()` - Change the return type of `rb_parser_compile_XXXX()` functions from `rb_ast_t *` to `VALUE` - rb_ruby_ast_new() which internally `calls ast_alloc()` is to create VALUE vast outside ruby_parser.c - `iseq.c` and `vm_core.h` - Amend the first parameter of `rb_iseq_new_XXXX()` functions from `rb_ast_body_t *` to `VALUE` - This keeps the VALUE of AST on the machine stack to prevent being removed by GC - `ast.c` - Almost all change is replacement `rb_ast_t *ast` with `VALUE vast` (sorry for the big diff) - Fix `node_memsize()` - Now it includes `rb_ast_local_table_link`, `tokens` and script_lines - `compile.c`, `load.c`, `node.c`, `parse.y`, `proc.c`, `ruby.c`, `template/prelude.c.tmpl`, `vm.c` and `vm_eval.c` - Follow-up due to the above changes - `imemo.{c|h}` - If an object with `imemo_ast` appears, considers it a bug Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
2024-04-16 12:42:42 +03:00
return sizeof(struct ASTNodeData) + rb_ast_memsize(ast);
2019-07-23 10:18:40 +03:00
}
static const rb_data_type_t rb_node_type = {
"AST/node",
2019-07-23 10:18:40 +03:00
{node_gc_mark, RUBY_TYPED_DEFAULT_FREE, node_memsize,},
0, 0,
RUBY_TYPED_FREE_IMMEDIATELY,
};
struct ASTLocationData {
int first_lineno;
int first_column;
int last_lineno;
int last_column;
};
static void
location_gc_mark(void *ptr)
{
}
static size_t
location_memsize(const void *ptr)
{
return sizeof(struct ASTLocationData);
}
static const rb_data_type_t rb_location_type = {
"AST/location",
{location_gc_mark, RUBY_TYPED_DEFAULT_FREE, location_memsize,},
0, 0,
RUBY_TYPED_FREE_IMMEDIATELY,
};
static VALUE rb_ast_node_alloc(VALUE klass);
static void
setup_node(VALUE obj, VALUE ast_value, const NODE *node)
{
struct ASTNodeData *data;
TypedData_Get_Struct(obj, struct ASTNodeData, &rb_node_type, data);
data->ast_value = ast_value;
data->node = node;
}
static VALUE
ast_new_internal(VALUE ast_value, const NODE *node)
{
VALUE obj;
obj = rb_ast_node_alloc(rb_cNode);
setup_node(obj, ast_value, node);
return obj;
}
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
static VALUE rb_ast_parse_str(VALUE str, VALUE keep_script_lines, VALUE error_tolerant, VALUE keep_tokens);
static VALUE rb_ast_parse_file(VALUE path, VALUE keep_script_lines, VALUE error_tolerant, VALUE keep_tokens);
static VALUE
ast_parse_new(void)
{
return rb_parser_set_context(rb_parser_new(), NULL, 0);
}
static VALUE
ast_parse_done(VALUE ast_value)
{
rb_ast_t *ast = rb_ruby_ast_data_get(ast_value);
[Universal parser] Decouple IMEMO from rb_ast_t This patch removes the `VALUE flags` member from the `rb_ast_t` structure making `rb_ast_t` no longer an IMEMO object. ## Background We are trying to make the Ruby parser generated from parse.y a universal parser that can be used by other implementations such as mruby. To achieve this, it is necessary to exclude VALUE and IMEMO from parse.y, AST, and NODE. ## Summary (file by file) - `rubyparser.h` - Remove the `VALUE flags` member from `rb_ast_t` - `ruby_parser.c` and `internal/ruby_parser.h` - Use TypedData_Make_Struct VALUE which wraps `rb_ast_t` `in ast_alloc()` so that GC can manage it - You can retrieve `rb_ast_t` from the VALUE by `rb_ruby_ast_data_get()` - Change the return type of `rb_parser_compile_XXXX()` functions from `rb_ast_t *` to `VALUE` - rb_ruby_ast_new() which internally `calls ast_alloc()` is to create VALUE vast outside ruby_parser.c - `iseq.c` and `vm_core.h` - Amend the first parameter of `rb_iseq_new_XXXX()` functions from `rb_ast_body_t *` to `VALUE` - This keeps the VALUE of AST on the machine stack to prevent being removed by GC - `ast.c` - Almost all change is replacement `rb_ast_t *ast` with `VALUE vast` (sorry for the big diff) - Fix `node_memsize()` - Now it includes `rb_ast_local_table_link`, `tokens` and script_lines - `compile.c`, `load.c`, `node.c`, `parse.y`, `proc.c`, `ruby.c`, `template/prelude.c.tmpl`, `vm.c` and `vm_eval.c` - Follow-up due to the above changes - `imemo.{c|h}` - If an object with `imemo_ast` appears, considers it a bug Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
2024-04-16 12:42:42 +03:00
if (!ast->body.root) {
rb_ast_dispose(ast);
rb_exc_raise(GET_EC()->errinfo);
}
return ast_new_internal(ast_value, (NODE *)ast->body.root);
}
static VALUE
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
ast_s_parse(rb_execution_context_t *ec, VALUE module, VALUE str, VALUE keep_script_lines, VALUE error_tolerant, VALUE keep_tokens)
{
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
return rb_ast_parse_str(str, keep_script_lines, error_tolerant, keep_tokens);
}
static VALUE
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
rb_ast_parse_str(VALUE str, VALUE keep_script_lines, VALUE error_tolerant, VALUE keep_tokens)
{
VALUE ast_value;
StringValue(str);
VALUE vparser = ast_parse_new();
[Universal parser] DeVALUE of p->debug_lines and ast->body.script_lines This patch is part of universal parser work. ## Summary - Decouple VALUE from members below: - `(struct parser_params *)->debug_lines` - `(rb_ast_t *)->body.script_lines` - Instead, they are now `rb_parser_ary_t *` - They can also be a `(VALUE)FIXNUM` as before to hold line count - `ISEQ_BODY(iseq)->variable.script_lines` remains VALUE - In order to do this, - Add `VALUE script_lines` param to `rb_iseq_new_with_opt()` - Introduce `rb_parser_build_script_lines_from()` to convert `rb_parser_ary_t *` into `VALUE` ## Other details - Extend `rb_parser_ary_t *`. It previously could only store `rb_parser_ast_token *`, now can store script_lines, too - Change tactics of building the top-level `SCRIPT_LINES__` in `yycompile0()` - Before: While parsing, each line of the script is added to `SCRIPT_LINES__[path]` - After: After `yyparse(p)`, `SCRIPT_LINES__[path]` will be built from `p->debug_lines` - Remove the second parameter of `rb_parser_set_script_lines()` to make it simple - Introduce `script_lines_free()` to be called from `rb_ast_free()` because the GC no longer takes care of the script_lines - Introduce `rb_parser_string_deep_copy()` in parse.y to maintain script_lines when `rb_ruby_parser_free()` called - With regard to this, please see *Future tasks* below ## Future tasks - Decouple IMEMO from `rb_ast_t *` - This lifts the five-members-restriction of Ruby object, - So we will be able to move the ownership of the `lex.string_buffer` from parser to AST - Then we remove `rb_parser_string_deep_copy()` to make the whole thing simple
2024-03-28 04:26:42 +03:00
if (RTEST(keep_script_lines)) rb_parser_set_script_lines(vparser);
if (RTEST(error_tolerant)) rb_parser_error_tolerant(vparser);
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
if (RTEST(keep_tokens)) rb_parser_keep_tokens(vparser);
ast_value = rb_parser_compile_string_path(vparser, Qnil, str, 1);
return ast_parse_done(ast_value);
}
static VALUE
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
ast_s_parse_file(rb_execution_context_t *ec, VALUE module, VALUE path, VALUE keep_script_lines, VALUE error_tolerant, VALUE keep_tokens)
{
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
return rb_ast_parse_file(path, keep_script_lines, error_tolerant, keep_tokens);
}
static VALUE
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
rb_ast_parse_file(VALUE path, VALUE keep_script_lines, VALUE error_tolerant, VALUE keep_tokens)
{
VALUE f;
VALUE ast_value = Qnil;
rb_encoding *enc = rb_utf8_encoding();
f = rb_file_open_str(path, "r");
rb_funcall(f, rb_intern("set_encoding"), 2, rb_enc_from_encoding(enc), rb_str_new_cstr("-"));
VALUE vparser = ast_parse_new();
[Universal parser] DeVALUE of p->debug_lines and ast->body.script_lines This patch is part of universal parser work. ## Summary - Decouple VALUE from members below: - `(struct parser_params *)->debug_lines` - `(rb_ast_t *)->body.script_lines` - Instead, they are now `rb_parser_ary_t *` - They can also be a `(VALUE)FIXNUM` as before to hold line count - `ISEQ_BODY(iseq)->variable.script_lines` remains VALUE - In order to do this, - Add `VALUE script_lines` param to `rb_iseq_new_with_opt()` - Introduce `rb_parser_build_script_lines_from()` to convert `rb_parser_ary_t *` into `VALUE` ## Other details - Extend `rb_parser_ary_t *`. It previously could only store `rb_parser_ast_token *`, now can store script_lines, too - Change tactics of building the top-level `SCRIPT_LINES__` in `yycompile0()` - Before: While parsing, each line of the script is added to `SCRIPT_LINES__[path]` - After: After `yyparse(p)`, `SCRIPT_LINES__[path]` will be built from `p->debug_lines` - Remove the second parameter of `rb_parser_set_script_lines()` to make it simple - Introduce `script_lines_free()` to be called from `rb_ast_free()` because the GC no longer takes care of the script_lines - Introduce `rb_parser_string_deep_copy()` in parse.y to maintain script_lines when `rb_ruby_parser_free()` called - With regard to this, please see *Future tasks* below ## Future tasks - Decouple IMEMO from `rb_ast_t *` - This lifts the five-members-restriction of Ruby object, - So we will be able to move the ownership of the `lex.string_buffer` from parser to AST - Then we remove `rb_parser_string_deep_copy()` to make the whole thing simple
2024-03-28 04:26:42 +03:00
if (RTEST(keep_script_lines)) rb_parser_set_script_lines(vparser);
if (RTEST(error_tolerant)) rb_parser_error_tolerant(vparser);
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
if (RTEST(keep_tokens)) rb_parser_keep_tokens(vparser);
ast_value = rb_parser_compile_file_path(vparser, Qnil, f, 1);
rb_io_close(f);
return ast_parse_done(ast_value);
}
static VALUE
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
rb_ast_parse_array(VALUE array, VALUE keep_script_lines, VALUE error_tolerant, VALUE keep_tokens)
{
VALUE ast_value = Qnil;
array = rb_check_array_type(array);
VALUE vparser = ast_parse_new();
[Universal parser] DeVALUE of p->debug_lines and ast->body.script_lines This patch is part of universal parser work. ## Summary - Decouple VALUE from members below: - `(struct parser_params *)->debug_lines` - `(rb_ast_t *)->body.script_lines` - Instead, they are now `rb_parser_ary_t *` - They can also be a `(VALUE)FIXNUM` as before to hold line count - `ISEQ_BODY(iseq)->variable.script_lines` remains VALUE - In order to do this, - Add `VALUE script_lines` param to `rb_iseq_new_with_opt()` - Introduce `rb_parser_build_script_lines_from()` to convert `rb_parser_ary_t *` into `VALUE` ## Other details - Extend `rb_parser_ary_t *`. It previously could only store `rb_parser_ast_token *`, now can store script_lines, too - Change tactics of building the top-level `SCRIPT_LINES__` in `yycompile0()` - Before: While parsing, each line of the script is added to `SCRIPT_LINES__[path]` - After: After `yyparse(p)`, `SCRIPT_LINES__[path]` will be built from `p->debug_lines` - Remove the second parameter of `rb_parser_set_script_lines()` to make it simple - Introduce `script_lines_free()` to be called from `rb_ast_free()` because the GC no longer takes care of the script_lines - Introduce `rb_parser_string_deep_copy()` in parse.y to maintain script_lines when `rb_ruby_parser_free()` called - With regard to this, please see *Future tasks* below ## Future tasks - Decouple IMEMO from `rb_ast_t *` - This lifts the five-members-restriction of Ruby object, - So we will be able to move the ownership of the `lex.string_buffer` from parser to AST - Then we remove `rb_parser_string_deep_copy()` to make the whole thing simple
2024-03-28 04:26:42 +03:00
if (RTEST(keep_script_lines)) rb_parser_set_script_lines(vparser);
if (RTEST(error_tolerant)) rb_parser_error_tolerant(vparser);
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
if (RTEST(keep_tokens)) rb_parser_keep_tokens(vparser);
ast_value = rb_parser_compile_array(vparser, Qnil, array, 1);
return ast_parse_done(ast_value);
}
[Universal parser] Decouple IMEMO from rb_ast_t This patch removes the `VALUE flags` member from the `rb_ast_t` structure making `rb_ast_t` no longer an IMEMO object. ## Background We are trying to make the Ruby parser generated from parse.y a universal parser that can be used by other implementations such as mruby. To achieve this, it is necessary to exclude VALUE and IMEMO from parse.y, AST, and NODE. ## Summary (file by file) - `rubyparser.h` - Remove the `VALUE flags` member from `rb_ast_t` - `ruby_parser.c` and `internal/ruby_parser.h` - Use TypedData_Make_Struct VALUE which wraps `rb_ast_t` `in ast_alloc()` so that GC can manage it - You can retrieve `rb_ast_t` from the VALUE by `rb_ruby_ast_data_get()` - Change the return type of `rb_parser_compile_XXXX()` functions from `rb_ast_t *` to `VALUE` - rb_ruby_ast_new() which internally `calls ast_alloc()` is to create VALUE vast outside ruby_parser.c - `iseq.c` and `vm_core.h` - Amend the first parameter of `rb_iseq_new_XXXX()` functions from `rb_ast_body_t *` to `VALUE` - This keeps the VALUE of AST on the machine stack to prevent being removed by GC - `ast.c` - Almost all change is replacement `rb_ast_t *ast` with `VALUE vast` (sorry for the big diff) - Fix `node_memsize()` - Now it includes `rb_ast_local_table_link`, `tokens` and script_lines - `compile.c`, `load.c`, `node.c`, `parse.y`, `proc.c`, `ruby.c`, `template/prelude.c.tmpl`, `vm.c` and `vm_eval.c` - Follow-up due to the above changes - `imemo.{c|h}` - If an object with `imemo_ast` appears, considers it a bug Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
2024-04-16 12:42:42 +03:00
static VALUE node_children(VALUE, const NODE*);
static VALUE
node_find(VALUE self, const int node_id)
{
VALUE ary;
long i;
struct ASTNodeData *data;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
if (nd_node_id(data->node) == node_id) return self;
ary = node_children(data->ast_value, data->node);
for (i = 0; i < RARRAY_LEN(ary); i++) {
VALUE child = RARRAY_AREF(ary, i);
if (CLASS_OF(child) == rb_cNode) {
VALUE result = node_find(child, node_id);
if (RTEST(result)) return result;
}
}
return Qnil;
}
extern VALUE rb_e_script;
static VALUE
node_id_for_backtrace_location(rb_execution_context_t *ec, VALUE module, VALUE location)
{
int node_id;
if (!rb_frame_info_p(location)) {
rb_raise(rb_eTypeError, "Thread::Backtrace::Location object expected");
}
node_id = rb_get_node_id_from_frame_info(location);
if (node_id == -1) {
return Qnil;
}
return INT2NUM(node_id);
}
static VALUE
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
ast_s_of(rb_execution_context_t *ec, VALUE module, VALUE body, VALUE keep_script_lines, VALUE error_tolerant, VALUE keep_tokens)
{
VALUE node, lines = Qnil;
const rb_iseq_t *iseq;
int node_id;
if (rb_frame_info_p(body)) {
iseq = rb_get_iseq_from_frame_info(body);
node_id = rb_get_node_id_from_frame_info(body);
}
else {
iseq = NULL;
if (rb_obj_is_proc(body)) {
iseq = vm_proc_iseq(body);
if (!rb_obj_is_iseq((VALUE)iseq)) return Qnil;
}
else {
iseq = rb_method_iseq(body);
}
if (iseq) {
node_id = ISEQ_BODY(iseq)->location.node_id;
}
}
if (!iseq) {
return Qnil;
}
2024-02-14 18:58:20 +03:00
if (ISEQ_BODY(iseq)->prism) {
rb_raise(rb_eRuntimeError, "cannot get AST for ISEQ compiled by prism");
}
lines = ISEQ_BODY(iseq)->variable.script_lines;
VALUE path = rb_iseq_path(iseq);
int e_option = RSTRING_LEN(path) == 2 && memcmp(RSTRING_PTR(path), "-e", 2) == 0;
if (NIL_P(lines) && rb_iseq_from_eval_p(iseq) && !e_option) {
rb_raise(rb_eArgError, "cannot get AST for method defined in eval");
}
Separate SCRIPT_LINES__ from ast.c This patch suggests relocating the code dealing with `SCRIPT_LINES__` from ast.c to ruby_parser.c. ## Background - I guess `AbstractSyntaxTree.of` method used to use `SCRIPT_LINES__` internally for some reason before - However, now it appears `SCRIPT_LINES__` is no longer used meaningfully by the method - As evidence of this, (and as my patch shows,) removing the function call of `rb_script_lines_for()` from `ast_s_of()` does not affect the result of `test/ruby/test_ast.rb` Given the above, I think two possibilities can be considered: - (A) `AbstractSyntaxTree.of` has not needed `SCRIPT_LINES__` already (I pick this) - (B) We lack a test case of `AbstractSyntaxTree.of` that needs to use `SCRIPT_LINES__` ## Besides, The current implementation causes strange behavior: ```console ruby -e"SCRIPT_LINES__ = {__FILE__ => []}; puts RubyVM::AbstractSyntaxTree.of(->{ 1 + 2 }, keep_script_lines: true).script_lines" => `-e:1:in '<main>': undefined method 'script_lines' for nil (NoMethodError)` ``` I think this is a bug because `AbstractSyntaxTree.of` is not supposed to return `nil` even in this case. This happens due to the ast.c's dependence on `SCRIPT_LINES__`. And at the end of the `ast_s_of()`, `node_find()` can not find the target child node obviously because it doesn't make sense to look for a corresponding node made from the parameter of `AbstractSyntaxTree.of` in the AST tree made from the value of `{__FILE__ => []}` ## Solution Since I think it's good enough `SCRIPT_LINES__` to be only referred by ruby.c, I chose the possibility "(A)" and wrote this patch which moves `rb_script_lines_for()` from ast.c to ruby_parser.c. So as the result: - `ast_s_of()` function no longer look up `SCRIPT_LINES__` - Even so, this patched code passes the existing tests - The strange behavior above no longer happens (I also added a test for it) Please correct me if I miss something🙏
2024-03-29 10:18:14 +03:00
if (!NIL_P(lines)) {
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
node = rb_ast_parse_array(lines, keep_script_lines, error_tolerant, keep_tokens);
}
else if (e_option) {
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
node = rb_ast_parse_str(rb_e_script, keep_script_lines, error_tolerant, keep_tokens);
}
else {
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
node = rb_ast_parse_file(path, keep_script_lines, error_tolerant, keep_tokens);
}
return node_find(node, node_id);
}
static VALUE
rb_ast_node_alloc(VALUE klass)
{
struct ASTNodeData *data;
VALUE obj = TypedData_Make_Struct(klass, struct ASTNodeData, &rb_node_type, data);
return obj;
}
static const char*
node_type_to_str(const NODE *node)
{
return (ruby_node_name(nd_type(node)) + rb_strlen_lit("NODE_"));
}
static VALUE
ast_node_type(rb_execution_context_t *ec, VALUE self)
{
struct ASTNodeData *data;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
return rb_sym_intern_ascii_cstr(node_type_to_str(data->node));
}
static VALUE
ast_node_node_id(rb_execution_context_t *ec, VALUE self)
{
struct ASTNodeData *data;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
return INT2FIX(nd_node_id(data->node));
}
#define NEW_CHILD(ast_value, node) (node ? ast_new_internal(ast_value, node) : Qnil)
static VALUE
rb_ary_new_from_node_args(VALUE ast_value, long n, ...)
{
va_list ar;
VALUE ary;
long i;
ary = rb_ary_new2(n);
va_start(ar, n);
for (i=0; i<n; i++) {
NODE *node;
node = va_arg(ar, NODE *);
rb_ary_push(ary, NEW_CHILD(ast_value, node));
}
va_end(ar);
return ary;
}
static VALUE
dump_block(VALUE ast_value, const struct RNode_BLOCK *node)
{
VALUE ary = rb_ary_new();
do {
rb_ary_push(ary, NEW_CHILD(ast_value, node->nd_head));
} while (node->nd_next &&
2021-12-03 18:01:24 +03:00
nd_type_p(node->nd_next, NODE_BLOCK) &&
(node = RNODE_BLOCK(node->nd_next), 1));
if (node->nd_next) {
rb_ary_push(ary, NEW_CHILD(ast_value, node->nd_next));
}
return ary;
}
static VALUE
dump_array(VALUE ast_value, const struct RNode_LIST *node)
{
VALUE ary = rb_ary_new();
rb_ary_push(ary, NEW_CHILD(ast_value, node->nd_head));
2021-12-03 18:01:24 +03:00
while (node->nd_next && nd_type_p(node->nd_next, NODE_LIST)) {
node = RNODE_LIST(node->nd_next);
rb_ary_push(ary, NEW_CHILD(ast_value, node->nd_head));
}
rb_ary_push(ary, NEW_CHILD(ast_value, node->nd_next));
return ary;
}
static VALUE
dump_parser_array(VALUE ast_value, rb_parser_ary_t *p_ary)
{
VALUE ary;
if (p_ary->data_type != PARSER_ARY_DATA_NODE) {
rb_bug("unexpected rb_parser_ary_data_type: %d", p_ary->data_type);
}
ary = rb_ary_new();
for (long i = 0; i < p_ary->len; i++) {
rb_ary_push(ary, NEW_CHILD(ast_value, p_ary->data[i]));
}
return ary;
}
static VALUE
var_name(ID id)
{
if (!id) return Qnil;
if (!rb_id2str(id)) return Qnil;
return ID2SYM(id);
}
static VALUE
no_name_rest(void)
{
ID rest;
CONST_ID(rest, "NODE_SPECIAL_NO_NAME_REST");
return ID2SYM(rest);
}
static VALUE
rest_arg(VALUE ast_value, const NODE *rest_arg)
{
return NODE_NAMED_REST_P(rest_arg) ? NEW_CHILD(ast_value, rest_arg) : no_name_rest();
}
static VALUE
node_children(VALUE ast_value, const NODE *node)
{
char name[sizeof("$") + DECIMAL_SIZE_OF(long)];
enum node_type type = nd_type(node);
switch (type) {
case NODE_BLOCK:
return dump_block(ast_value, RNODE_BLOCK(node));
case NODE_IF:
return rb_ary_new_from_node_args(ast_value, 3, RNODE_IF(node)->nd_cond, RNODE_IF(node)->nd_body, RNODE_IF(node)->nd_else);
case NODE_UNLESS:
return rb_ary_new_from_node_args(ast_value, 3, RNODE_UNLESS(node)->nd_cond, RNODE_UNLESS(node)->nd_body, RNODE_UNLESS(node)->nd_else);
case NODE_CASE:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_CASE(node)->nd_head, RNODE_CASE(node)->nd_body);
case NODE_CASE2:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_CASE2(node)->nd_head, RNODE_CASE2(node)->nd_body);
case NODE_CASE3:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_CASE3(node)->nd_head, RNODE_CASE3(node)->nd_body);
case NODE_WHEN:
return rb_ary_new_from_node_args(ast_value, 3, RNODE_WHEN(node)->nd_head, RNODE_WHEN(node)->nd_body, RNODE_WHEN(node)->nd_next);
case NODE_IN:
return rb_ary_new_from_node_args(ast_value, 3, RNODE_IN(node)->nd_head, RNODE_IN(node)->nd_body, RNODE_IN(node)->nd_next);
case NODE_WHILE:
case NODE_UNTIL:
return rb_ary_push(rb_ary_new_from_node_args(ast_value, 2, RNODE_WHILE(node)->nd_cond, RNODE_WHILE(node)->nd_body),
RBOOL(RNODE_WHILE(node)->nd_state));
case NODE_ITER:
case NODE_FOR:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_ITER(node)->nd_iter, RNODE_ITER(node)->nd_body);
case NODE_FOR_MASGN:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_FOR_MASGN(node)->nd_var);
case NODE_BREAK:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_BREAK(node)->nd_stts);
case NODE_NEXT:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_NEXT(node)->nd_stts);
case NODE_RETURN:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_RETURN(node)->nd_stts);
case NODE_REDO:
return rb_ary_new_from_node_args(ast_value, 0);
case NODE_RETRY:
return rb_ary_new_from_node_args(ast_value, 0);
case NODE_BEGIN:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_BEGIN(node)->nd_body);
case NODE_RESCUE:
return rb_ary_new_from_node_args(ast_value, 3, RNODE_RESCUE(node)->nd_head, RNODE_RESCUE(node)->nd_resq, RNODE_RESCUE(node)->nd_else);
case NODE_RESBODY:
return rb_ary_new_from_node_args(ast_value, 4, RNODE_RESBODY(node)->nd_args, RNODE_RESBODY(node)->nd_exc_var, RNODE_RESBODY(node)->nd_body, RNODE_RESBODY(node)->nd_next);
case NODE_ENSURE:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_ENSURE(node)->nd_head, RNODE_ENSURE(node)->nd_ensr);
case NODE_AND:
case NODE_OR:
{
VALUE ary = rb_ary_new();
while (1) {
rb_ary_push(ary, NEW_CHILD(ast_value, RNODE_AND(node)->nd_1st));
if (!RNODE_AND(node)->nd_2nd || !nd_type_p(RNODE_AND(node)->nd_2nd, type))
break;
node = RNODE_AND(node)->nd_2nd;
}
rb_ary_push(ary, NEW_CHILD(ast_value, RNODE_AND(node)->nd_2nd));
return ary;
}
case NODE_MASGN:
if (NODE_NAMED_REST_P(RNODE_MASGN(node)->nd_args)) {
return rb_ary_new_from_node_args(ast_value, 3, RNODE_MASGN(node)->nd_value, RNODE_MASGN(node)->nd_head, RNODE_MASGN(node)->nd_args);
}
else {
return rb_ary_new_from_args(3, NEW_CHILD(ast_value, RNODE_MASGN(node)->nd_value),
NEW_CHILD(ast_value, RNODE_MASGN(node)->nd_head),
no_name_rest());
}
case NODE_LASGN:
if (NODE_REQUIRED_KEYWORD_P(RNODE_LASGN(node)->nd_value)) {
return rb_ary_new_from_args(2, var_name(RNODE_LASGN(node)->nd_vid), ID2SYM(rb_intern("NODE_SPECIAL_REQUIRED_KEYWORD")));
}
return rb_ary_new_from_args(2, var_name(RNODE_LASGN(node)->nd_vid), NEW_CHILD(ast_value, RNODE_LASGN(node)->nd_value));
2023-10-07 04:37:33 +03:00
case NODE_DASGN:
if (NODE_REQUIRED_KEYWORD_P(RNODE_DASGN(node)->nd_value)) {
return rb_ary_new_from_args(2, var_name(RNODE_DASGN(node)->nd_vid), ID2SYM(rb_intern("NODE_SPECIAL_REQUIRED_KEYWORD")));
}
return rb_ary_new_from_args(2, var_name(RNODE_DASGN(node)->nd_vid), NEW_CHILD(ast_value, RNODE_DASGN(node)->nd_value));
2023-10-07 04:37:33 +03:00
case NODE_IASGN:
return rb_ary_new_from_args(2, var_name(RNODE_IASGN(node)->nd_vid), NEW_CHILD(ast_value, RNODE_IASGN(node)->nd_value));
2023-10-07 04:37:33 +03:00
case NODE_CVASGN:
return rb_ary_new_from_args(2, var_name(RNODE_CVASGN(node)->nd_vid), NEW_CHILD(ast_value, RNODE_CVASGN(node)->nd_value));
2023-10-07 04:37:33 +03:00
case NODE_GASGN:
return rb_ary_new_from_args(2, var_name(RNODE_GASGN(node)->nd_vid), NEW_CHILD(ast_value, RNODE_GASGN(node)->nd_value));
case NODE_CDECL:
if (RNODE_CDECL(node)->nd_vid) {
return rb_ary_new_from_args(2, ID2SYM(RNODE_CDECL(node)->nd_vid), NEW_CHILD(ast_value, RNODE_CDECL(node)->nd_value));
}
return rb_ary_new_from_args(3, NEW_CHILD(ast_value, RNODE_CDECL(node)->nd_else), ID2SYM(RNODE_COLON2(RNODE_CDECL(node)->nd_else)->nd_mid), NEW_CHILD(ast_value, RNODE_CDECL(node)->nd_value));
case NODE_OP_ASGN1:
return rb_ary_new_from_args(4, NEW_CHILD(ast_value, RNODE_OP_ASGN1(node)->nd_recv),
ID2SYM(RNODE_OP_ASGN1(node)->nd_mid),
NEW_CHILD(ast_value, RNODE_OP_ASGN1(node)->nd_index),
NEW_CHILD(ast_value, RNODE_OP_ASGN1(node)->nd_rvalue));
case NODE_OP_ASGN2:
return rb_ary_new_from_args(5, NEW_CHILD(ast_value, RNODE_OP_ASGN2(node)->nd_recv),
RBOOL(RNODE_OP_ASGN2(node)->nd_aid),
ID2SYM(RNODE_OP_ASGN2(node)->nd_vid),
ID2SYM(RNODE_OP_ASGN2(node)->nd_mid),
NEW_CHILD(ast_value, RNODE_OP_ASGN2(node)->nd_value));
case NODE_OP_ASGN_AND:
return rb_ary_new_from_args(3, NEW_CHILD(ast_value, RNODE_OP_ASGN_AND(node)->nd_head), ID2SYM(idANDOP),
NEW_CHILD(ast_value, RNODE_OP_ASGN_AND(node)->nd_value));
case NODE_OP_ASGN_OR:
return rb_ary_new_from_args(3, NEW_CHILD(ast_value, RNODE_OP_ASGN_OR(node)->nd_head), ID2SYM(idOROP),
NEW_CHILD(ast_value, RNODE_OP_ASGN_OR(node)->nd_value));
case NODE_OP_CDECL:
return rb_ary_new_from_args(3, NEW_CHILD(ast_value, RNODE_OP_CDECL(node)->nd_head),
ID2SYM(RNODE_OP_CDECL(node)->nd_aid),
NEW_CHILD(ast_value, RNODE_OP_CDECL(node)->nd_value));
case NODE_CALL:
return rb_ary_new_from_args(3, NEW_CHILD(ast_value, RNODE_CALL(node)->nd_recv),
ID2SYM(RNODE_CALL(node)->nd_mid),
NEW_CHILD(ast_value, RNODE_CALL(node)->nd_args));
2023-10-09 03:20:53 +03:00
case NODE_OPCALL:
return rb_ary_new_from_args(3, NEW_CHILD(ast_value, RNODE_OPCALL(node)->nd_recv),
2023-10-09 03:20:53 +03:00
ID2SYM(RNODE_OPCALL(node)->nd_mid),
NEW_CHILD(ast_value, RNODE_OPCALL(node)->nd_args));
2023-10-09 03:20:53 +03:00
case NODE_QCALL:
return rb_ary_new_from_args(3, NEW_CHILD(ast_value, RNODE_QCALL(node)->nd_recv),
2023-10-09 03:20:53 +03:00
ID2SYM(RNODE_QCALL(node)->nd_mid),
NEW_CHILD(ast_value, RNODE_QCALL(node)->nd_args));
case NODE_FCALL:
return rb_ary_new_from_args(2, ID2SYM(RNODE_FCALL(node)->nd_mid),
NEW_CHILD(ast_value, RNODE_FCALL(node)->nd_args));
case NODE_VCALL:
return rb_ary_new_from_args(1, ID2SYM(RNODE_VCALL(node)->nd_mid));
case NODE_SUPER:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_SUPER(node)->nd_args);
case NODE_ZSUPER:
return rb_ary_new_from_node_args(ast_value, 0);
case NODE_LIST:
return dump_array(ast_value, RNODE_LIST(node));
case NODE_ZLIST:
return rb_ary_new_from_node_args(ast_value, 0);
case NODE_HASH:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_HASH(node)->nd_head);
case NODE_YIELD:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_YIELD(node)->nd_head);
case NODE_LVAR:
return rb_ary_new_from_args(1, var_name(RNODE_LVAR(node)->nd_vid));
2023-10-09 03:27:24 +03:00
case NODE_DVAR:
return rb_ary_new_from_args(1, var_name(RNODE_DVAR(node)->nd_vid));
case NODE_IVAR:
2023-10-09 03:27:24 +03:00
return rb_ary_new_from_args(1, ID2SYM(RNODE_IVAR(node)->nd_vid));
case NODE_CONST:
2023-10-09 03:27:24 +03:00
return rb_ary_new_from_args(1, ID2SYM(RNODE_CONST(node)->nd_vid));
case NODE_CVAR:
2023-10-09 03:27:24 +03:00
return rb_ary_new_from_args(1, ID2SYM(RNODE_CVAR(node)->nd_vid));
case NODE_GVAR:
2023-10-09 03:27:24 +03:00
return rb_ary_new_from_args(1, ID2SYM(RNODE_GVAR(node)->nd_vid));
case NODE_NTH_REF:
snprintf(name, sizeof(name), "$%ld", RNODE_NTH_REF(node)->nd_nth);
return rb_ary_new_from_args(1, ID2SYM(rb_intern(name)));
case NODE_BACK_REF:
name[0] = '$';
name[1] = (char)RNODE_BACK_REF(node)->nd_nth;
name[2] = '\0';
return rb_ary_new_from_args(1, ID2SYM(rb_intern(name)));
2024-04-04 01:46:28 +03:00
case NODE_MATCH:
return rb_ary_new_from_args(1, rb_node_regx_string_val(node));
case NODE_MATCH2:
if (RNODE_MATCH2(node)->nd_args) {
return rb_ary_new_from_node_args(ast_value, 3, RNODE_MATCH2(node)->nd_recv, RNODE_MATCH2(node)->nd_value, RNODE_MATCH2(node)->nd_args);
}
return rb_ary_new_from_node_args(ast_value, 2, RNODE_MATCH2(node)->nd_recv, RNODE_MATCH2(node)->nd_value);
case NODE_MATCH3:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_MATCH3(node)->nd_recv, RNODE_MATCH3(node)->nd_value);
case NODE_STR:
case NODE_XSTR:
return rb_ary_new_from_args(1, rb_node_str_string_val(node));
2024-01-05 06:03:11 +03:00
case NODE_INTEGER:
return rb_ary_new_from_args(1, rb_node_integer_literal_val(node));
2024-01-05 06:03:11 +03:00
case NODE_FLOAT:
return rb_ary_new_from_args(1, rb_node_float_literal_val(node));
2024-01-05 06:03:11 +03:00
case NODE_RATIONAL:
return rb_ary_new_from_args(1, rb_node_rational_literal_val(node));
2024-01-05 06:03:11 +03:00
case NODE_IMAGINARY:
return rb_ary_new_from_args(1, rb_node_imaginary_literal_val(node));
case NODE_REGX:
return rb_ary_new_from_args(1, rb_node_regx_string_val(node));
case NODE_ONCE:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_ONCE(node)->nd_body);
case NODE_DSTR:
case NODE_DXSTR:
case NODE_DREGX:
case NODE_DSYM:
{
struct RNode_LIST *n = RNODE_DSTR(node)->nd_next;
VALUE head = Qnil, next = Qnil;
if (n) {
head = NEW_CHILD(ast_value, n->nd_head);
next = NEW_CHILD(ast_value, n->nd_next);
}
return rb_ary_new_from_args(3, rb_node_dstr_string_val(node), head, next);
}
case NODE_SYM:
return rb_ary_new_from_args(1, rb_node_sym_string_val(node));
case NODE_EVSTR:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_EVSTR(node)->nd_body);
case NODE_ARGSCAT:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_ARGSCAT(node)->nd_head, RNODE_ARGSCAT(node)->nd_body);
case NODE_ARGSPUSH:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_ARGSPUSH(node)->nd_head, RNODE_ARGSPUSH(node)->nd_body);
case NODE_SPLAT:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_SPLAT(node)->nd_head);
case NODE_BLOCK_PASS:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_BLOCK_PASS(node)->nd_head, RNODE_BLOCK_PASS(node)->nd_body);
case NODE_DEFN:
return rb_ary_new_from_args(2, ID2SYM(RNODE_DEFN(node)->nd_mid), NEW_CHILD(ast_value, RNODE_DEFN(node)->nd_defn));
case NODE_DEFS:
return rb_ary_new_from_args(3, NEW_CHILD(ast_value, RNODE_DEFS(node)->nd_recv), ID2SYM(RNODE_DEFS(node)->nd_mid), NEW_CHILD(ast_value, RNODE_DEFS(node)->nd_defn));
case NODE_ALIAS:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_ALIAS(node)->nd_1st, RNODE_ALIAS(node)->nd_2nd);
case NODE_VALIAS:
return rb_ary_new_from_args(2, ID2SYM(RNODE_VALIAS(node)->nd_alias), ID2SYM(RNODE_VALIAS(node)->nd_orig));
case NODE_UNDEF:
return rb_ary_new_from_args(1, dump_parser_array(ast_value, RNODE_UNDEF(node)->nd_undefs));
case NODE_CLASS:
return rb_ary_new_from_node_args(ast_value, 3, RNODE_CLASS(node)->nd_cpath, RNODE_CLASS(node)->nd_super, RNODE_CLASS(node)->nd_body);
case NODE_MODULE:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_MODULE(node)->nd_cpath, RNODE_MODULE(node)->nd_body);
case NODE_SCLASS:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_SCLASS(node)->nd_recv, RNODE_SCLASS(node)->nd_body);
case NODE_COLON2:
return rb_ary_new_from_args(2, NEW_CHILD(ast_value, RNODE_COLON2(node)->nd_head), ID2SYM(RNODE_COLON2(node)->nd_mid));
case NODE_COLON3:
return rb_ary_new_from_args(1, ID2SYM(RNODE_COLON3(node)->nd_mid));
case NODE_DOT2:
case NODE_DOT3:
case NODE_FLIP2:
case NODE_FLIP3:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_DOT2(node)->nd_beg, RNODE_DOT2(node)->nd_end);
case NODE_SELF:
return rb_ary_new_from_node_args(ast_value, 0);
case NODE_NIL:
return rb_ary_new_from_node_args(ast_value, 0);
case NODE_TRUE:
return rb_ary_new_from_node_args(ast_value, 0);
case NODE_FALSE:
return rb_ary_new_from_node_args(ast_value, 0);
case NODE_ERRINFO:
return rb_ary_new_from_node_args(ast_value, 0);
case NODE_DEFINED:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_DEFINED(node)->nd_head);
case NODE_POSTEXE:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_POSTEXE(node)->nd_body);
case NODE_ATTRASGN:
return rb_ary_new_from_args(3, NEW_CHILD(ast_value, RNODE_ATTRASGN(node)->nd_recv), ID2SYM(RNODE_ATTRASGN(node)->nd_mid), NEW_CHILD(ast_value, RNODE_ATTRASGN(node)->nd_args));
case NODE_LAMBDA:
return rb_ary_new_from_node_args(ast_value, 1, RNODE_LAMBDA(node)->nd_body);
case NODE_OPT_ARG:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_OPT_ARG(node)->nd_body, RNODE_OPT_ARG(node)->nd_next);
case NODE_KW_ARG:
return rb_ary_new_from_node_args(ast_value, 2, RNODE_KW_ARG(node)->nd_body, RNODE_KW_ARG(node)->nd_next);
case NODE_POSTARG:
if (NODE_NAMED_REST_P(RNODE_POSTARG(node)->nd_1st)) {
return rb_ary_new_from_node_args(ast_value, 2, RNODE_POSTARG(node)->nd_1st, RNODE_POSTARG(node)->nd_2nd);
}
return rb_ary_new_from_args(2, no_name_rest(),
NEW_CHILD(ast_value, RNODE_POSTARG(node)->nd_2nd));
case NODE_ARGS:
{
struct rb_args_info *ainfo = &RNODE_ARGS(node)->nd_ainfo;
return rb_ary_new_from_args(10,
INT2NUM(ainfo->pre_args_num),
NEW_CHILD(ast_value, ainfo->pre_init),
NEW_CHILD(ast_value, (NODE *)ainfo->opt_args),
var_name(ainfo->first_post_arg),
INT2NUM(ainfo->post_args_num),
NEW_CHILD(ast_value, ainfo->post_init),
(ainfo->rest_arg == NODE_SPECIAL_EXCESSIVE_COMMA
? ID2SYM(rb_intern("NODE_SPECIAL_EXCESSIVE_COMMA"))
: var_name(ainfo->rest_arg)),
(ainfo->no_kwarg ? Qfalse : NEW_CHILD(ast_value, (NODE *)ainfo->kw_args)),
(ainfo->no_kwarg ? Qfalse : NEW_CHILD(ast_value, ainfo->kw_rest_arg)),
var_name(ainfo->block_arg));
}
case NODE_SCOPE:
{
rb_ast_id_table_t *tbl = RNODE_SCOPE(node)->nd_tbl;
int i, size = tbl ? tbl->size : 0;
VALUE locals = rb_ary_new_capa(size);
for (i = 0; i < size; i++) {
rb_ary_push(locals, var_name(tbl->ids[i]));
}
return rb_ary_new_from_args(3, locals, NEW_CHILD(ast_value, (NODE *)RNODE_SCOPE(node)->nd_args), NEW_CHILD(ast_value, RNODE_SCOPE(node)->nd_body));
}
case NODE_ARYPTN:
{
VALUE rest = rest_arg(ast_value, RNODE_ARYPTN(node)->rest_arg);
return rb_ary_new_from_args(4,
NEW_CHILD(ast_value, RNODE_ARYPTN(node)->nd_pconst),
NEW_CHILD(ast_value, RNODE_ARYPTN(node)->pre_args),
rest,
NEW_CHILD(ast_value, RNODE_ARYPTN(node)->post_args));
}
case NODE_FNDPTN:
{
VALUE pre_rest = rest_arg(ast_value, RNODE_FNDPTN(node)->pre_rest_arg);
VALUE post_rest = rest_arg(ast_value, RNODE_FNDPTN(node)->post_rest_arg);
return rb_ary_new_from_args(4,
NEW_CHILD(ast_value, RNODE_FNDPTN(node)->nd_pconst),
pre_rest,
NEW_CHILD(ast_value, RNODE_FNDPTN(node)->args),
post_rest);
}
case NODE_HSHPTN:
{
VALUE kwrest = RNODE_HSHPTN(node)->nd_pkwrestarg == NODE_SPECIAL_NO_REST_KEYWORD ? ID2SYM(rb_intern("NODE_SPECIAL_NO_REST_KEYWORD")) :
NEW_CHILD(ast_value, RNODE_HSHPTN(node)->nd_pkwrestarg);
return rb_ary_new_from_args(3,
NEW_CHILD(ast_value, RNODE_HSHPTN(node)->nd_pconst),
NEW_CHILD(ast_value, RNODE_HSHPTN(node)->nd_pkwargs),
kwrest);
}
case NODE_LINE:
return rb_ary_new_from_args(1, rb_node_line_lineno_val(node));
case NODE_FILE:
return rb_ary_new_from_args(1, rb_node_file_path_val(node));
case NODE_ENCODING:
return rb_ary_new_from_args(1, rb_node_encoding_val(node));
case NODE_ERROR:
return rb_ary_new_from_node_args(ast_value, 0);
case NODE_ARGS_AUX:
case NODE_LAST:
break;
}
rb_bug("node_children: unknown node: %s", ruby_node_name(type));
}
static VALUE
ast_node_children(rb_execution_context_t *ec, VALUE self)
{
struct ASTNodeData *data;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
return node_children(data->ast_value, data->node);
}
static int
null_loc_p(rb_code_location_t *loc)
{
return (loc->beg_pos.lineno == 0 && loc->beg_pos.column == -1 && loc->end_pos.lineno == 0 && loc->end_pos.column == -1);
}
static VALUE
location_new(rb_code_location_t *loc)
{
VALUE obj;
struct ASTLocationData *data;
if (null_loc_p(loc)) return Qnil;
obj = TypedData_Make_Struct(rb_cLocation, struct ASTLocationData, &rb_location_type, data);
data->first_lineno = loc->beg_pos.lineno;
data->first_column = loc->beg_pos.column;
data->last_lineno = loc->end_pos.lineno;
data->last_column = loc->end_pos.column;
return obj;
}
static VALUE
node_locations(VALUE ast_value, const NODE *node)
{
enum node_type type = nd_type(node);
switch (type) {
2024-09-03 10:12:30 +03:00
case NODE_ALIAS:
return rb_ary_new_from_args(2,
location_new(nd_code_loc(node)),
location_new(&RNODE_ALIAS(node)->keyword_loc));
case NODE_AND:
return rb_ary_new_from_args(2,
location_new(nd_code_loc(node)),
location_new(&RNODE_AND(node)->operator_loc));
case NODE_BREAK:
return rb_ary_new_from_args(2,
location_new(nd_code_loc(node)),
location_new(&RNODE_BREAK(node)->keyword_loc));
case NODE_NEXT:
return rb_ary_new_from_args(2,
location_new(nd_code_loc(node)),
location_new(&RNODE_NEXT(node)->keyword_loc));
case NODE_OR:
return rb_ary_new_from_args(2,
location_new(nd_code_loc(node)),
location_new(&RNODE_OR(node)->operator_loc));
case NODE_REDO:
return rb_ary_new_from_args(2,
location_new(nd_code_loc(node)),
location_new(&RNODE_REDO(node)->keyword_loc));
2024-09-03 11:53:55 +03:00
case NODE_UNDEF:
return rb_ary_new_from_args(2,
location_new(nd_code_loc(node)),
location_new(&RNODE_UNDEF(node)->keyword_loc));
case NODE_UNLESS:
return rb_ary_new_from_args(4,
location_new(nd_code_loc(node)),
location_new(&RNODE_UNLESS(node)->keyword_loc),
location_new(&RNODE_UNLESS(node)->then_keyword_loc),
location_new(&RNODE_UNLESS(node)->end_keyword_loc));
case NODE_VALIAS:
return rb_ary_new_from_args(2,
location_new(nd_code_loc(node)),
location_new(&RNODE_VALIAS(node)->keyword_loc));
2024-09-05 12:41:10 +03:00
case NODE_WHEN:
return rb_ary_new_from_args(3,
location_new(nd_code_loc(node)),
location_new(&RNODE_WHEN(node)->keyword_loc),
location_new(&RNODE_WHEN(node)->then_keyword_loc));
case NODE_WHILE:
return rb_ary_new_from_args(3,
location_new(nd_code_loc(node)),
location_new(&RNODE_WHILE(node)->keyword_loc),
location_new(&RNODE_WHILE(node)->closing_loc));
case NODE_UNTIL:
return rb_ary_new_from_args(3,
location_new(nd_code_loc(node)),
location_new(&RNODE_UNTIL(node)->keyword_loc),
location_new(&RNODE_UNTIL(node)->closing_loc));
case NODE_ARGS_AUX:
case NODE_LAST:
break;
default:
return rb_ary_new_from_args(1, location_new(nd_code_loc(node)));
}
rb_bug("node_locations: unknown node: %s", ruby_node_name(type));
}
static VALUE
ast_node_locations(rb_execution_context_t *ec, VALUE self)
{
struct ASTNodeData *data;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
return node_locations(data->ast_value, data->node);
}
static VALUE
ast_node_first_lineno(rb_execution_context_t *ec, VALUE self)
{
struct ASTNodeData *data;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
return INT2NUM(nd_first_lineno(data->node));
}
static VALUE
ast_node_first_column(rb_execution_context_t *ec, VALUE self)
{
struct ASTNodeData *data;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
return INT2NUM(nd_first_column(data->node));
}
static VALUE
ast_node_last_lineno(rb_execution_context_t *ec, VALUE self)
{
struct ASTNodeData *data;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
return INT2NUM(nd_last_lineno(data->node));
}
static VALUE
ast_node_last_column(rb_execution_context_t *ec, VALUE self)
{
struct ASTNodeData *data;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
return INT2NUM(nd_last_column(data->node));
}
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
static VALUE
ast_node_all_tokens(rb_execution_context_t *ec, VALUE self)
{
long i;
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
struct ASTNodeData *data;
[Universal parser] Decouple IMEMO from rb_ast_t This patch removes the `VALUE flags` member from the `rb_ast_t` structure making `rb_ast_t` no longer an IMEMO object. ## Background We are trying to make the Ruby parser generated from parse.y a universal parser that can be used by other implementations such as mruby. To achieve this, it is necessary to exclude VALUE and IMEMO from parse.y, AST, and NODE. ## Summary (file by file) - `rubyparser.h` - Remove the `VALUE flags` member from `rb_ast_t` - `ruby_parser.c` and `internal/ruby_parser.h` - Use TypedData_Make_Struct VALUE which wraps `rb_ast_t` `in ast_alloc()` so that GC can manage it - You can retrieve `rb_ast_t` from the VALUE by `rb_ruby_ast_data_get()` - Change the return type of `rb_parser_compile_XXXX()` functions from `rb_ast_t *` to `VALUE` - rb_ruby_ast_new() which internally `calls ast_alloc()` is to create VALUE vast outside ruby_parser.c - `iseq.c` and `vm_core.h` - Amend the first parameter of `rb_iseq_new_XXXX()` functions from `rb_ast_body_t *` to `VALUE` - This keeps the VALUE of AST on the machine stack to prevent being removed by GC - `ast.c` - Almost all change is replacement `rb_ast_t *ast` with `VALUE vast` (sorry for the big diff) - Fix `node_memsize()` - Now it includes `rb_ast_local_table_link`, `tokens` and script_lines - `compile.c`, `load.c`, `node.c`, `parse.y`, `proc.c`, `ruby.c`, `template/prelude.c.tmpl`, `vm.c` and `vm_eval.c` - Follow-up due to the above changes - `imemo.{c|h}` - If an object with `imemo_ast` appears, considers it a bug Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
2024-04-16 12:42:42 +03:00
rb_ast_t *ast;
rb_parser_ary_t *parser_tokens;
rb_parser_ast_token_t *parser_token;
VALUE str, loc, token, all_tokens;
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
ast = rb_ruby_ast_data_get(data->ast_value);
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
[Universal parser] Decouple IMEMO from rb_ast_t This patch removes the `VALUE flags` member from the `rb_ast_t` structure making `rb_ast_t` no longer an IMEMO object. ## Background We are trying to make the Ruby parser generated from parse.y a universal parser that can be used by other implementations such as mruby. To achieve this, it is necessary to exclude VALUE and IMEMO from parse.y, AST, and NODE. ## Summary (file by file) - `rubyparser.h` - Remove the `VALUE flags` member from `rb_ast_t` - `ruby_parser.c` and `internal/ruby_parser.h` - Use TypedData_Make_Struct VALUE which wraps `rb_ast_t` `in ast_alloc()` so that GC can manage it - You can retrieve `rb_ast_t` from the VALUE by `rb_ruby_ast_data_get()` - Change the return type of `rb_parser_compile_XXXX()` functions from `rb_ast_t *` to `VALUE` - rb_ruby_ast_new() which internally `calls ast_alloc()` is to create VALUE vast outside ruby_parser.c - `iseq.c` and `vm_core.h` - Amend the first parameter of `rb_iseq_new_XXXX()` functions from `rb_ast_body_t *` to `VALUE` - This keeps the VALUE of AST on the machine stack to prevent being removed by GC - `ast.c` - Almost all change is replacement `rb_ast_t *ast` with `VALUE vast` (sorry for the big diff) - Fix `node_memsize()` - Now it includes `rb_ast_local_table_link`, `tokens` and script_lines - `compile.c`, `load.c`, `node.c`, `parse.y`, `proc.c`, `ruby.c`, `template/prelude.c.tmpl`, `vm.c` and `vm_eval.c` - Follow-up due to the above changes - `imemo.{c|h}` - If an object with `imemo_ast` appears, considers it a bug Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
2024-04-16 12:42:42 +03:00
parser_tokens = ast->node_buffer->tokens;
if (parser_tokens == NULL) {
return Qnil;
}
all_tokens = rb_ary_new2(parser_tokens->len);
for (i = 0; i < parser_tokens->len; i++) {
parser_token = parser_tokens->data[i];
str = rb_str_new(parser_token->str->ptr, parser_token->str->len);
loc = rb_ary_new_from_args(4,
INT2FIX(parser_token->loc.beg_pos.lineno),
INT2FIX(parser_token->loc.beg_pos.column),
INT2FIX(parser_token->loc.end_pos.lineno),
INT2FIX(parser_token->loc.end_pos.column)
);
token = rb_ary_new_from_args(4, INT2FIX(parser_token->id), ID2SYM(rb_intern(parser_token->type_name)), str, loc);
rb_ary_push(all_tokens, token);
}
Resize arrays in `rb_ary_freeze` and use it for freezing arrays While working on a separate issue we found that in some cases `ary_heap_realloc` was being called on frozen arrays. To fix this, this change does the following: 1) Updates `rb_ary_freeze` to assert the type is an array, return if already frozen, and shrink the capacity if it is not embedded, shared or a shared root. 2) Replaces `rb_obj_freeze` with `rb_ary_freeze` when the object is always an array. 3) In `ary_heap_realloc`, ensure the new capa is set with `ARY_SET_CAPA`. Previously the change in capa was not set. 4) Adds an assertion to `ary_heap_realloc` that the array is not frozen. Some of this work was originally done in https://github.com/ruby/ruby/pull/2640, referencing this issue https://bugs.ruby-lang.org/issues/16291. There didn't appear to be any objections to this PR, it appears to have simply lost traction. The original PR made changes to arrays and strings at the same time, this PR only does arrays. Also it was old enough that rather than revive that branch I've made a new one. I added Lourens as co-author in addtion to Aaron who helped me with this patch. The original PR made this change for performance reasons, and while that's still true for this PR, the goal of this PR is to avoid calling `ary_heap_realloc` on frozen arrays. The capacity should be shrunk _before_ the array is frozen, not after. Co-authored-by: Aaron Patterson <tenderlove@ruby-lang.org> Co-Authored-By: methodmissing <lourens@methodmissing.com>
2024-06-18 21:52:18 +03:00
rb_ary_freeze(all_tokens);
return all_tokens;
Enhance keep_tokens option for RubyVM::AbstractSyntaxTree parsing methods Implementation for Language Server Protocol (LSP) sometimes needs token information. For example both `m(1)` and `m(1, )` has same AST structure other than node locations then it's impossible to check the existence of `,` from AST. However in later case, it might be better to suggest variables list for the second argument. Token information is important for such case. This commit adds these methods. * Add `keep_tokens` option for `RubyVM::AbstractSyntaxTree.parse`, `.parse_file` and `.of` * Add `RubyVM::AbstractSyntaxTree::Node#tokens` which returns tokens for the node including tokens for descendants nodes. * Add `RubyVM::AbstractSyntaxTree::Node#all_tokens` which returns all tokens for the input script regardless the receiver node. [Feature #19070] Impacts on memory usage and performance are below: Memory usage: ``` $ cat test.rb root = RubyVM::AbstractSyntaxTree.parse_file(File.expand_path('../test/ruby/test_keyword.rb', __FILE__), keep_tokens: true) $ /usr/bin/time -f %Mkb /usr/local/bin/ruby -v ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] 11408kb # keep_tokens :false $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 17508kb # keep_tokens :true $ /usr/bin/time -f %Mkb /usr/local/bin/ruby test.rb 30960kb ``` Performance: ``` $ cat ../ast_keep_tokens.yml prelude: | src = <<~SRC module M class C def m1(a, b) 1 + a + b end end end SRC benchmark: without_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: false) with_keep_tokens: | RubyVM::AbstractSyntaxTree.parse(src, keep_tokens: true) $ make benchmark COMPARE_RUBY="./ruby" ARGS=../ast_keep_tokens.yml /home/kaneko.y/.rbenv/shims/ruby --disable=gems -rrubygems -I../benchmark/lib ../benchmark/benchmark-driver/exe/benchmark-driver \ --executables="compare-ruby::./ruby -I.ext/common --disable-gem" \ --executables="built-ruby::./miniruby -I../lib -I. -I.ext/common ../tool/runruby.rb --extout=.ext -- --disable-gems --disable-gem" \ --output=markdown --output-compare -v ../ast_keep_tokens.yml compare-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] built-ruby: ruby 3.2.0dev (2022-11-19T09:41:54Z 19070-keep_tokens d3af1b8057) [x86_64-linux] warming up.. | |compare-ruby|built-ruby| |:--------------------|-----------:|---------:| |without_keep_tokens | 21.659k| 21.303k| | | 1.02x| -| |with_keep_tokens | 6.220k| 5.691k| | | 1.09x| -| ```
2022-09-23 16:40:02 +03:00
}
static VALUE
ast_node_inspect(rb_execution_context_t *ec, VALUE self)
{
VALUE str;
VALUE cname;
struct ASTNodeData *data;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
cname = rb_class_path(rb_obj_class(self));
str = rb_str_new2("#<");
rb_str_append(str, cname);
rb_str_catf(str, ":%s@%d:%d-%d:%d>",
node_type_to_str(data->node),
nd_first_lineno(data->node), nd_first_column(data->node),
nd_last_lineno(data->node), nd_last_column(data->node));
return str;
}
static VALUE
ast_node_script_lines(rb_execution_context_t *ec, VALUE self)
{
struct ASTNodeData *data;
[Universal parser] Decouple IMEMO from rb_ast_t This patch removes the `VALUE flags` member from the `rb_ast_t` structure making `rb_ast_t` no longer an IMEMO object. ## Background We are trying to make the Ruby parser generated from parse.y a universal parser that can be used by other implementations such as mruby. To achieve this, it is necessary to exclude VALUE and IMEMO from parse.y, AST, and NODE. ## Summary (file by file) - `rubyparser.h` - Remove the `VALUE flags` member from `rb_ast_t` - `ruby_parser.c` and `internal/ruby_parser.h` - Use TypedData_Make_Struct VALUE which wraps `rb_ast_t` `in ast_alloc()` so that GC can manage it - You can retrieve `rb_ast_t` from the VALUE by `rb_ruby_ast_data_get()` - Change the return type of `rb_parser_compile_XXXX()` functions from `rb_ast_t *` to `VALUE` - rb_ruby_ast_new() which internally `calls ast_alloc()` is to create VALUE vast outside ruby_parser.c - `iseq.c` and `vm_core.h` - Amend the first parameter of `rb_iseq_new_XXXX()` functions from `rb_ast_body_t *` to `VALUE` - This keeps the VALUE of AST on the machine stack to prevent being removed by GC - `ast.c` - Almost all change is replacement `rb_ast_t *ast` with `VALUE vast` (sorry for the big diff) - Fix `node_memsize()` - Now it includes `rb_ast_local_table_link`, `tokens` and script_lines - `compile.c`, `load.c`, `node.c`, `parse.y`, `proc.c`, `ruby.c`, `template/prelude.c.tmpl`, `vm.c` and `vm_eval.c` - Follow-up due to the above changes - `imemo.{c|h}` - If an object with `imemo_ast` appears, considers it a bug Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
2024-04-16 12:42:42 +03:00
rb_ast_t *ast;
TypedData_Get_Struct(self, struct ASTNodeData, &rb_node_type, data);
ast = rb_ruby_ast_data_get(data->ast_value);
[Universal parser] Decouple IMEMO from rb_ast_t This patch removes the `VALUE flags` member from the `rb_ast_t` structure making `rb_ast_t` no longer an IMEMO object. ## Background We are trying to make the Ruby parser generated from parse.y a universal parser that can be used by other implementations such as mruby. To achieve this, it is necessary to exclude VALUE and IMEMO from parse.y, AST, and NODE. ## Summary (file by file) - `rubyparser.h` - Remove the `VALUE flags` member from `rb_ast_t` - `ruby_parser.c` and `internal/ruby_parser.h` - Use TypedData_Make_Struct VALUE which wraps `rb_ast_t` `in ast_alloc()` so that GC can manage it - You can retrieve `rb_ast_t` from the VALUE by `rb_ruby_ast_data_get()` - Change the return type of `rb_parser_compile_XXXX()` functions from `rb_ast_t *` to `VALUE` - rb_ruby_ast_new() which internally `calls ast_alloc()` is to create VALUE vast outside ruby_parser.c - `iseq.c` and `vm_core.h` - Amend the first parameter of `rb_iseq_new_XXXX()` functions from `rb_ast_body_t *` to `VALUE` - This keeps the VALUE of AST on the machine stack to prevent being removed by GC - `ast.c` - Almost all change is replacement `rb_ast_t *ast` with `VALUE vast` (sorry for the big diff) - Fix `node_memsize()` - Now it includes `rb_ast_local_table_link`, `tokens` and script_lines - `compile.c`, `load.c`, `node.c`, `parse.y`, `proc.c`, `ruby.c`, `template/prelude.c.tmpl`, `vm.c` and `vm_eval.c` - Follow-up due to the above changes - `imemo.{c|h}` - If an object with `imemo_ast` appears, considers it a bug Co-authored-by: Nobuyoshi Nakada <nobu@ruby-lang.org>
2024-04-16 12:42:42 +03:00
rb_parser_ary_t *ret = ast->body.script_lines;
[Universal parser] DeVALUE of p->debug_lines and ast->body.script_lines This patch is part of universal parser work. ## Summary - Decouple VALUE from members below: - `(struct parser_params *)->debug_lines` - `(rb_ast_t *)->body.script_lines` - Instead, they are now `rb_parser_ary_t *` - They can also be a `(VALUE)FIXNUM` as before to hold line count - `ISEQ_BODY(iseq)->variable.script_lines` remains VALUE - In order to do this, - Add `VALUE script_lines` param to `rb_iseq_new_with_opt()` - Introduce `rb_parser_build_script_lines_from()` to convert `rb_parser_ary_t *` into `VALUE` ## Other details - Extend `rb_parser_ary_t *`. It previously could only store `rb_parser_ast_token *`, now can store script_lines, too - Change tactics of building the top-level `SCRIPT_LINES__` in `yycompile0()` - Before: While parsing, each line of the script is added to `SCRIPT_LINES__[path]` - After: After `yyparse(p)`, `SCRIPT_LINES__[path]` will be built from `p->debug_lines` - Remove the second parameter of `rb_parser_set_script_lines()` to make it simple - Introduce `script_lines_free()` to be called from `rb_ast_free()` because the GC no longer takes care of the script_lines - Introduce `rb_parser_string_deep_copy()` in parse.y to maintain script_lines when `rb_ruby_parser_free()` called - With regard to this, please see *Future tasks* below ## Future tasks - Decouple IMEMO from `rb_ast_t *` - This lifts the five-members-restriction of Ruby object, - So we will be able to move the ownership of the `lex.string_buffer` from parser to AST - Then we remove `rb_parser_string_deep_copy()` to make the whole thing simple
2024-03-28 04:26:42 +03:00
return rb_parser_build_script_lines_from(ret);
}
static VALUE
ast_location_first_lineno(rb_execution_context_t *ec, VALUE self)
{
struct ASTLocationData *data;
TypedData_Get_Struct(self, struct ASTLocationData, &rb_location_type, data);
return INT2NUM(data->first_lineno);
}
static VALUE
ast_location_first_column(rb_execution_context_t *ec, VALUE self)
{
struct ASTLocationData *data;
TypedData_Get_Struct(self, struct ASTLocationData, &rb_location_type, data);
return INT2NUM(data->first_column);
}
static VALUE
ast_location_last_lineno(rb_execution_context_t *ec, VALUE self)
{
struct ASTLocationData *data;
TypedData_Get_Struct(self, struct ASTLocationData, &rb_location_type, data);
return INT2NUM(data->last_lineno);
}
static VALUE
ast_location_last_column(rb_execution_context_t *ec, VALUE self)
{
struct ASTLocationData *data;
TypedData_Get_Struct(self, struct ASTLocationData, &rb_location_type, data);
return INT2NUM(data->last_column);
}
static VALUE
ast_location_inspect(rb_execution_context_t *ec, VALUE self)
{
VALUE str;
VALUE cname;
struct ASTLocationData *data;
TypedData_Get_Struct(self, struct ASTLocationData, &rb_location_type, data);
cname = rb_class_path(rb_obj_class(self));
str = rb_str_new2("#<");
rb_str_append(str, cname);
rb_str_catf(str, ":@%d:%d-%d:%d>",
data->first_lineno, data->first_column,
data->last_lineno, data->last_column);
return str;
}
#include "ast.rbinc"
void
Init_ast(void)
{
rb_mAST = rb_define_module_under(rb_cRubyVM, "AbstractSyntaxTree");
rb_cNode = rb_define_class_under(rb_mAST, "Node", rb_cObject);
rb_cLocation = rb_define_class_under(rb_mAST, "Location", rb_cObject);
rb_undef_alloc_func(rb_cNode);
rb_undef_alloc_func(rb_cLocation);
}