Bug #14500: Missing Regexp documentation and clarification on behavior of \K for edge case

• Status: Closed
• Assignee: -
• Target version: -
• Backport: 2.6: UNKNOWN, 2.7: UNKNOWN, 3.0: UNKNOWN

[Sundeep (Sundeep Agarwal)]

Capturing section (https://ruby-doc.org/core-2.5.0/Regexp.html#class-Regexp-label-Capturing)

• formatting issue: n<sup>th</sup>
• suggestion to add a note on \0 similar to the note mentioned in Special global variables (https://ruby-doc.org/core-2.5.0/Regexp.html#class-Regexp-label-Special+global+variables) section about $&
• suggestion to add that numbered capturing groups is limited to 9

Anchors section (https://ruby-doc.org/core-2.5.0/Regexp.html#class-Regexp-label-Anchors)

• suggestion to add documentation on \K
• need clarification if the below behavior seen is expected and mention it while adding documentation

$ echo 'aaa' | ruby -pe 'gsub(/a\K/, ":")'
a:aa:

$ # what I expected
$ echo 'aaa' | ruby -pe 'gsub(/(a)/, "\\1:")'
a:a:a:

[znz (Kazuhiro NISHIYAMA)]

Sundeep (Sundeep Agarwal) wrote:

• suggestion to add that numbered capturing groups is limited to 9

I don't think so.
Why did you think so?

irb -r irb/completion --simple-prompt
>> /()()()()()()()()()(a)/ =~ "a"
=> 0
>> $10
=> "a"

[Sundeep (Sundeep Agarwal)]

oh, I didn't check with $10. I'd tried with backreference. Any idea how to use \10?

$ echo 'abcdefghij' | ruby -pe 'sub(/(.)(.)(.)(.)(.)(.)(.)(.)(.)(.)/, ":\\9:")'
:i:
$ echo 'abcdefghij' | ruby -pe 'sub(/(.)(.)(.)(.)(.)(.)(.)(.)(.)(.)/, ":\\10:")'
:a0:
$ echo 'abcdefghij' | ruby -pe 'sub(/(.)(.)(.)(.)(.)(.)(.)(.)(.)(.)/){":#{$10}:"}'
:j:

[hsbt (Hiroshi SHIBATA)]

• Tracker changed from Misc to Bug
• Backport set to 2.6: UNKNOWN, 2.7: UNKNOWN, 3.0: UNKNOWN

[jeremyevans0 (Jeremy Evans)]

Sundeep (Sundeep Agarwal) wrote:

Anchors section (https://ruby-doc.org/core-2.5.0/Regexp.html#class-Regexp-label-Anchors)

• suggestion to add documentation on \K
• need clarification if the below behavior seen is expected and mention it while adding documentation

$ echo 'aaa' | ruby -pe 'gsub(/a\K/, ":")'
a:aa:

$ # what I expected
$ echo 'aaa' | ruby -pe 'gsub(/(a)/, "\\1:")'
a:a:a:

The \K behavior at the end of a regexp is a bug that has already been filed upstream in Onigmo: https://github.com/k-takata/Onigmo/issues/152

I agree about the other documentation issues, and I'll push a commit shortly to address them.

• Status changed from Open to Closed

Applied in changeset git|4fc9ddd7b6af54abf88d702c2e11e97ca7750ce3.

---

Update Capturing and Anchors sections of regexp documention

Document that only first 9 numbered capture groups can use the \n
backreference syntax. Document \0 backreference. Document \K anchor.

Fixes [Bug #14500]