From: Johannes Sixt Date: Fri, 13 Dec 2019 19:24:47 GMT Subject: Re: [PATCH v2] userdiff: remove empty subexpression from elixir regex Message-ID: <0c9d891e-382f-03d1-bcbd-d652f1d58f4d@kdbg.org> In-Reply-To: <20191213175535.87725-1-emaste@FreeBSD.org> Am 13.12.19 um 18:55 schrieb Ed Maste: > The regex failed to compile on FreeBSD. > > Fixes: a807200f67588f6e Having a references is this form is unusual for our codebase. (Not that I mind a lot, though.) I expect that Junio will commit the fix on top of the commit that introduced the bogus regex anyway (branch ln/userdiff-elixir), and then it will be easy find. > Signed-off-by: Ed Maste > --- > Add /* -- */ to make things more clear and be consistent with other > patterns. This text would be nice to have in the commit message. > > userdiff.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/userdiff.c b/userdiff.c > index 324916f20f..efbe05e5a5 100644 > --- a/userdiff.c > +++ b/userdiff.c > @@ -34,8 +34,9 @@ PATTERNS("dts", > "|[-+*/%&^|!~]|>>|<<|&&|\\|\\|"), > PATTERNS("elixir", > "^[ \t]*((def(macro|module|impl|protocol|p)?|test)[ \t].*)$", > + /* -- */ > /* Atoms, names, and module attributes */ > - "|[@:]?[a-zA-Z0-9@_?!]+" > + "[@:]?[a-zA-Z0-9@_?!]+" > /* Numbers with specific base */ > "|[-+]?0[xob][0-9a-fA-F]+" > /* Numbers */ > Good catch! Tested-by: Johannes Sixt Thanks! -- Hannes