Re: [dpdk-dev] [PATCH 2/2] eal: add fnmatch implementation on Windows

[Thomas Monjalon <thomas@monjalon.net>]

01/05/2020 03:08, Ranjit Menon:
> On 4/30/2020 12:30 AM, Dmitry Kozlyuk wrote:
> > On 2020-04-30 08:52 GMT+0200 Thomas Monjalon wrote:
> >> 30/04/2020 01:24, Pallavi Kadam:
> >>> Added fnmatch implementation on Windows to support
> >>> log level arguments.
> >>> The source file is with BSD-3-Clause license.
> >>> https://github.com/lattera/freebsd/blob/master/usr.bin/csup/fnmatch.c
> >>
> >> Sorry for the naive question, I don't know Windows programming.
> >>
> >> Do we really need this external code?
> >> Why RtlIsNameInExpression from Windows cannot be used?
> >> https://docs.microsoft.com/en-us/windows/win32/devnotes/rtlisnameinexpression
> > 
> > The general reason not to use Win32 API for globbing is poorly documented
> > contract: what are the exact matching rules? They're definitely incompatible
> > with fnmatch(3). IMO small external code is better than unknown behavior.
> > 
> > RtlIsNameInExpression is an internal call for drivers and services with a
> > cumbersome API. PathMatchSpecA is the user-mode interface, but see above.
> > 
> > https://docs.microsoft.com/en-us/windows/win32/api/shlwapi/nf-shlwapi-pathmatchspeca
> > 
> 
> I will agree with Dmitry here. The fnmatch external code is shipping, 
> production quality code. So it's better to use it than writing something 
> new.

OK thank you for the explanations and opinions.