Bug #3320: emacs ruby-mode.el font-lock fails on symboled string ending with ?

• Status: Third Party's Issue
• Assignee: nobu (Nobuyoshi Nakada)
• Target version: 1.9.4
• ruby -v: 1.9.2 (trunk and 1.8.7)+

[zev (Zev Blut)]

=begin
Fontification breaks when emacs sees a symbol like
:'this is a symbol?'

example code:

class EmacsExample
:symbol

'this is a test'
'is this a test?'
"Can this be a test"
:'this is an error?'

def bar
@help
end
end

---

I have a very hacked fix in ruby-font-lock-syntactic-keywords
for
;; the last $', $", $ in the respective string is not variable ;; the last ?', ?", ? in the respective string is not ascii code
("\(^\|[[ \t\n<+(,=:]\)\(['"]\\)\\(\\\\.\\|\\2\\|[^'\"\n\\]\)*?\\?[?$]\(\2\)"
(2 (7 . nil))
(4 (7 . nil)))

by adding : in the above matches with space tabs L etc...

See the attached patch

I am not sure this is the proper fix, but it fixes the above example.

Thanks,
Zev
=end

[zev (Zev Blut)]

• File ruby-mode.el.patch ruby-mode.el.patch added

=begin

=end

[zev (Zev Blut)]

=begin
I found another related parse error. Unfortunately, I am not sure how to fix this one.
This is a multi line string that starts on one line and ends on a different line with a ?
Such as:

x = 'What do you think about this long
text today ?'

def foo
end

---

=end

[zenspider (Ryan Davis)]

=begin

On May 20, 2010, at 06:18 , Zev Blut wrote:

Issue #3320 has been updated by Zev Blut.

I found another related parse error. Unfortunately, I am not sure how to fix this one.
This is a multi line string that starts on one line and ends on a different line with a ?
Such as:

x = 'What do you think about this long
text today ?'

def foo
end

In this case it is because it sees ?' or ?" and interprets that as the character notation (I have no idea what this is called). Putting a backslash in front of the ? fixes the problem locally.

There are a lot of different things that can trip up ruby-mode when inside a string. I tripped on one today with a multiline string with "def" in it:

eval "
def xxx
yyy
end
"

Notice that tab will indent up to the yyy as if it is actual code, not string content.

BTW, I'm using version 1.0 of ruby-mode.el as supplied in emacs 24.0.50. It may be more up to date in the ruby distro, but afaik, it should be shifting to emacs for distribution.

=end

[akr (Akira Tanaka)]

• Project changed from Ruby to Ruby master
• Category deleted (misc)
• Assignee set to nobu (Nobuyoshi Nakada)

[nahi (Hiroshi Nakamura)]

• Target version set to 1.9.3

[kosaki (Motohiro KOSAKI)]

• Status changed from Open to Assigned

[kosaki (Motohiro KOSAKI)]

• Target version changed from 1.9.3 to 1.9.4

ETIMEOUT. ruby-mode.el never be release stopper.

[naruse (Yui NARUSE)]

• Parent task set to #5142

[dgutov (Dmitry Gutov)]

All examples in this bug work fine for me with ruby-mode from the Emacs tree.
Not sure when they were fixed.

[nobu (Nobuyoshi Nakada)]

• Status changed from Assigned to Third Party's Issue