On 5/22/22 15:22, Paul Eggert wrote:

> We've already uncovered one area, where \a doesn't work as expected and 
> where a warning diagnostic would be helpful.

I installed the attached patch to cause 'grep' to warn about these. 
Comments welcome. Most of the changes were in Gnulib's dfa module, which 
see.


> Here's another one, where an oddly-placed '*' doesn't work as one would expect:
> 
> $ printf '*\na\n*a\n' | grep '\(*\)'
> *
> *a
> $ printf '*\na\n*a\n' | grep -E '(*)'
> grep: Unmatched ( or \(
> $ printf '*\na\n*a\n' | grep '\(*a\)'
> *a
> $ printf '*\na\n*a\n' | grep -E '(*a)'
> a
> *a 

I plan to look at this next. We shouldn't warn about BREs like '\(*\)' 
and '\(*a\)' as these conform to POSIX and work fine. But it makes sense 
to warn about EREs like '(*)', '(*a)', '(+)', '(+a)', '({1})', '({1}a)' 
as POSIX does not specify their behavior, their semantics are 
unpredictable with GNU grep, and it's plausible that people are making 
mistakes in this area.