Re: Parameter expansion, patterns and fnmatch

[Harald van Dijk <harald@gigawatt.nl>]

[-- Attachment #1: Type: text/plain, Size: 1534 bytes --]

On 09/08/16 11:28, Olof Johansson wrote:
> Hello,
>
> I'm seeing some discrepancies between dash built with --enable-fnmatch
> and dash built without that got me curious. Maybe you could help shed
> some light?
>
>  foo='[abc]'
>  echo ${foo#[}
>  echo ${foo#\[}
>
> With dash built with --enable-fnmatch:
>
>  abc]
>  abc]
>
> With dash built without --enable-fnmatch:
>
>  [abc]
>  abc]

Nice find.

> I was able to reproduce this behavior on latest git master
> (17a5f24e0). The dash manual states:
>
>> The end of the character class is indicated by a ]; if the ] is
>> missing then the [ matches a [ rather than introducing a character
>> class.
>
> Which to me suggests that the non-fnmatch case is not the expected
> behavior. Is this interpretation correct?

Yes, this looks like a bug in dash. With the default --disable-fnmatch 
code, when dash encounters [ in a pattern, it immediately treats the 
following characters as part of the set. If it then encounters the end 
of the pattern without having seen a matching ], it attempts to reset 
the state and continue as if [ was treated as a literal character right 
from the start. The attempt to reset the state doesn't look right, and 
has been like this since at least the initial Git commit in 2005.

This also affects

     case [a in [?) echo ok ;; *) echo bad ;; esac

which should print ok.

Attached is a patch that attempts to reset the state correctly. It 
passes your test and mine, but I have not yet tested it extensively.

> Thanks,
>

Cheers,
Harald van Dijk

[-- Attachment #2: lit-lbrack.patch --]
[-- Type: text/x-patch, Size: 234 bytes --]

--- a/src/expand.c
+++ b/src/expand.c
@@ -1594,7 +1594,8 @@ pmatch(const char *pattern, const char *string)
 			do {
 				if (!c) {
 					p = startp;
-					c = *p;
+					q--;
+					c = '[';
 					goto dft;
 				}
 				if (c == '[') {