Project

General

Profile

Bug #17015

RubyVM::AbstractSyntaxTree.parse has the same result on `proc { |a| }` and `proc { |a,| }`

Added by osyo (manga osyo) 4 months ago. Updated about 2 months ago.

Status:
Closed
Priority:
Normal
Assignee:
-
Target version:
-
ruby -v:
ruby 2.8.0dev (2020-07-05T16:15:47Z master efe851a0df) [x86_64-linux]
[ruby-core:99077]

Description

Summary

RubyVM::AbstractSyntaxTree.parse("proc { |a| }") and RubyVM::AbstractSyntaxTree.parse("proc { |a,| }") have the same result.
Since proc { |a| } and proc { |a,| } have different meanings, I think we need a way to distinguish them.

node1 = RubyVM::AbstractSyntaxTree.parse("proc { |a| }")
body1 = node1.children.last
pp body1
# => (ITER@1:0-1:12 (FCALL@1:0-1:4 :proc nil)
#       (SCOPE@1:5-1:12
#        tbl: [:a]
#        args:
#          (ARGS@1:8-1:9
#           pre_num: 1
#           pre_init: nil
#           opt: nil
#           first_post: nil
#           post_num: 0
#           post_init: nil
#           rest: nil
#           kw: nil
#           kwrest: nil
#           block: nil)
#        body: (BEGIN@1:10-1:10 nil)))


# returns the same result as proc { |a| }
node2 = RubyVM::AbstractSyntaxTree.parse("proc { |a,| }")
body2 = node2.children.last
pp body2
# => (ITER@1:0-1:13 (FCALL@1:0-1:4 :proc nil)
#       (SCOPE@1:5-1:13
#        tbl: [:a]
#        args:
#          (ARGS@1:8-1:10
#           pre_num: 1
#           pre_init: nil
#           opt: nil
#           first_post: nil
#           post_num: 0
#           post_init: nil
#           rest: nil
#           kw: nil
#           kwrest: nil
#           block: nil)
#        body: (BEGIN@1:11-1:11 nil)))

Proposal

In case of proc { |a,| }, rest: :NODE_SPECIAL_EXCESSIVE_COMMA is added.

# no change
node1 = RubyVM::AbstractSyntaxTree.parse("proc { |a| }")
body1 = node1.children.last
pp body1
# => (ITER@1:0-1:12 (FCALL@1:0-1:4 :proc nil)
#       (SCOPE@1:5-1:12
#        tbl: [:a]
#        args:
#          (ARGS@1:8-1:9
#           pre_num: 1
#           pre_init: nil
#           opt: nil
#           first_post: nil
#           post_num: 0
#           post_init: nil
#           rest: nil
#           kw: nil
#           kwrest: nil
#           block: nil)
#        body: (BEGIN@1:10-1:10 nil)))


# rest: :NODE_SPECIAL_EXCESSIVE_COMMA is added
node2 = RubyVM::AbstractSyntaxTree.parse("proc { |a,| }")
body2 = node2.children.last
pp body2
# => (ITER@1:0-1:13 (FCALL@1:0-1:4 :proc nil)
#       (SCOPE@1:5-1:13
#        tbl: [:a]
#        args:
#          (ARGS@1:8-1:10
#           pre_num: 1
#           pre_init: nil
#           opt: nil
#           first_post: nil
#           post_num: 0
#           post_init: nil
#           rest: :NODE_SPECIAL_EXCESSIVE_COMMA
#           kw: nil
#           kwrest: nil
#           block: nil)
#        body: (BEGIN@1:11-1:11 nil)))

This correspondence is similar to :NODE_SPECIAL_NO_NAME_REST for MASGN.

node = RubyVM::AbstractSyntaxTree.parse("*a = foo")
body = node.children.last
pp body
# => (MASGN@1:0-1:8 (VCALL@1:5-1:8 :foo) nil (LASGN@1:1-1:2 :a nil))

# :NODE_SPECIAL_NO_NAME_REST is added if the variable name is not defined
node = RubyVM::AbstractSyntaxTree.parse("* = foo")
body = node.children.last
pp body
# => (MASGN@1:0-1:7 (VCALL@1:4-1:7 :foo) nil :NODE_SPECIAL_NO_NAME_REST)

NOTE

NODE_SPECIAL_EXCESSIVE_COMMA was implemented in 0872ea5 and renamed in f258137.
So backport to 2.6 can be difficult.

Versions

  • ruby 2.6.6p146 (2020-03-31 revision 67876) [x86_64-linux]
  • ruby 2.7.1p83 (2020-03-31 revision a0c7c23c9c) [x86_64-linux]
  • ruby 2.8.0dev (2020-07-05T16:15:47Z master efe851a0df) [x86_64-linux]

Patch

Also available in: Atom PDF