On 1/8/19 8:44 PM, Noah Misch wrote: > On Tue, Jan 08, 2019 at 08:17:43AM -0500, Andrew Dunstan wrote: >> On 1/3/19 12:53 AM, Noah Misch wrote: >>> If I run perltidy on 60d9979, then run perl-mode indent, the diff between >>> the >>> perltidy run and perl-mode indent run is: >>> 129 files changed, 8468 insertions(+), 8468 deletions(-) >>> If I add (perl-continued-brace-offset . -2): >>> 119 files changed, 3515 insertions(+), 3515 deletions(-) >>> If I add (perl-indent-continued-arguments . 4) as well: >>> 86 files changed, 2626 insertions(+), 2626 deletions(-) >>> If I add (perl-indent-parens-as-block . t) as well: >>> 65 files changed, 2373 insertions(+), 2373 deletions(-) >> Sounds good. What do the remaining diffs look like? > I've attached them. Most involve statement continuation in some form. For > example, src/backend/utils/mb/Unicode has numerous instances where perl-mode > indents hashref-constructor curly braces as though they were code blocks. > Other diff lines involve labels. Others are in string literals.
On a very quick glance I notice some things that looked just wrong, and some that were at best dubious. It's a pity that we can't get closer to what perltidy does, but +1 for applying your changes. cheers andrew -- Andrew Dunstan https://www.2ndQuadrant.com PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services
