Issue #13949 has been updated by nagachika (Tomoyuki Chikanaga).

Backport changed from 2.3: REQUIRED, 2.4: REQUIRED to 2.3: REQUIRED, 2.4: DONE

ruby_2_4 r61454 merged revision(s) 60059.

----------------------------------------
Bug #13949: String#unpack with 'M' directive can create strings with wrong code range
https://bugs.ruby-lang.org/issues/13949#change-68626

* Author: nirvdrum (Kevin Menard)
* Status: Closed
* Priority: Normal
* Assignee: 
* Target version: 
* ruby -v: ruby 2.4.2p198 (2017-09-14 revision 59899) [x86_64-linux]
* Backport: 2.3: REQUIRED, 2.4: DONE
----------------------------------------
I've noticed that `String#unpack` with the `'M'` directive can create strings that should be `CR_7BIT` as `CR_VALID`. The issue appears to have been introduced in r30542, which assumes that all `ASCII-8BIT` strings must be `CR_VALID`. It's possible this was correct back during Ruby 1.9.3 development and just wasn't updated. I'm not familiar enough with the history to tell.

A simple reproduction showing the issue is:

```
res = '0123456789=\n'.unpack('M').first
p res
p res.encoding
p res.bytes
p res.ascii_only?

puts

packed = res.bytes.pack('c*')
p packed
p packed.encoding
p packed.bytes
p packed.ascii_only?

```

This yields the following output:

```
"0123456789=\\n"
#<Encoding:ASCII-8BIT>
[48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 61, 92, 110]
false

"0123456789=\\n"
#<Encoding:ASCII-8BIT>
[48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 61, 92, 110]
true
```

Both strings have exactly the same contents with the same encoding. But, depending on how you construct them, one is consider to be `CR_7BIT` value (indicated by the `String#ascii_only?` output), and one is considered to be `CR_VALID`. I believe `CR_7BIT` is the correct code range value in this situation.



-- 
https://bugs.ruby-lang.org/

Unsubscribe: <mailto:ruby-core-request / ruby-lang.org?subject=unsubscribe>
<http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>