Thread: Re: [COMMITTERS] pgsql: Remove more SGML tabs.
On Wed, Feb 09, 2011 at 01:17:06PM +0000, Bruce Momjian wrote: > Remove more SGML tabs. Perhaps we should see about putting something in .git/hooks/pre-commit so people can focus on more substantive matters. Is there some kind of cross-platform way to do this? I'm thinking that given the fact that our build system already requires Perl, there should be, but I'm not quite sure how this would be accomplished. Cheers, David. -- David Fetter <david@fetter.org> http://fetter.org/ Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter Skype: davidfetter XMPP: david.fetter@gmail.com iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics Remember to vote! Consider donating to Postgres: http://www.postgresql.org/about/donate
<p><br /> On Feb 9, 2011 5:01 PM, "David Fetter" <<a href="mailto:david@fetter.org">david@fetter.org</a>> wrote:<br/> ><br /> > On Wed, Feb 09, 2011 at 01:17:06PM +0000, Bruce Momjian wrote:<br /> > > Remove more SGMLtabs.<br /> ><br /> > Perhaps we should see about putting something in .git/hooks/pre-commit<br /> > so peoplecan focus on more substantive matters.<br /> ><br /> > Is there some kind of cross-platform way to do this? I'mthinking<br /> > that given the fact that our build system already requires Perl, there<br /> > should be, butI'm not quite sure how this would be accomplished.<p>If you can write a reliable perl script for this, we can run it likepgindent. No need to complicate it with hooks. <p>/Magnus
On ons, 2011-02-09 at 08:00 -0800, David Fetter wrote: > On Wed, Feb 09, 2011 at 01:17:06PM +0000, Bruce Momjian wrote: > > Remove more SGML tabs. > > Perhaps we should see about putting something in .git/hooks/pre-commit > so people can focus on more substantive matters. > > Is there some kind of cross-platform way to do this? I'm thinking > that given the fact that our build system already requires Perl, there > should be, but I'm not quite sure how this would be accomplished. There is make check target in doc/src/sgml/ that is supposed to catch this. But it's probably hard to remember to run that. One thing I was thinking of was that we could add a global make maintainer-check target (a name I picked up from other projects) which would run various source code sanity checks. Besides the SGML tabs issue, my favourite would be duplicate_oids. Maybe if we could find a third use case, we'd have a quorum for implementing this.
Excerpts from Peter Eisentraut's message of jue feb 10 07:58:16 -0300 2011: > One thing I was thinking of was that we could add a global make > maintainer-check target (a name I picked up from other projects) which > would run various source code sanity checks. Besides the SGML tabs > issue, my favourite would be duplicate_oids. Maybe if we could find a > third use case, we'd have a quorum for implementing this. The c++ headers thing? Or the compiles-in-isolation test for headers? -- Álvaro Herrera <alvherre@commandprompt.com> The PostgreSQL Company - Command Prompt, Inc. PostgreSQL Replication, Consulting, Custom Development, 24x7 support
On Thu, Feb 10, 2011 at 12:58:16PM +0200, Peter Eisentraut wrote: > On ons, 2011-02-09 at 08:00 -0800, David Fetter wrote: > > On Wed, Feb 09, 2011 at 01:17:06PM +0000, Bruce Momjian wrote: > > > Remove more SGML tabs. > > > > Perhaps we should see about putting something in .git/hooks/pre-commit > > so people can focus on more substantive matters. > > > > Is there some kind of cross-platform way to do this? I'm thinking > > that given the fact that our build system already requires Perl, there > > should be, but I'm not quite sure how this would be accomplished. > > There is make check target in doc/src/sgml/ that is supposed to catch > this. But it's probably hard to remember to run that. It's "check-tabs", and if you hadn't mentioned it, I'd never have seen it. > One thing I was thinking of was that we could add a global make > maintainer-check target (a name I picked up from other projects) which > would run various source code sanity checks. Besides the SGML tabs > issue, my favourite would be duplicate_oids. Maybe if we could find a > third use case, we'd have a quorum for implementing this. I think all such checks belong in .git/hooks/pre-commit, and need to be as cross-platform as needed for committers. Would a *n*x-based version do for a start? Cheers, David. -- David Fetter <david@fetter.org> http://fetter.org/ Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter Skype: davidfetter XMPP: david.fetter@gmail.com iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics Remember to vote! Consider donating to Postgres: http://www.postgresql.org/about/donate
On tor, 2011-02-10 at 10:40 -0800, David Fetter wrote: > I think all such checks belong in .git/hooks/pre-commit, and need to > be as cross-platform as needed for committers. Would a *n*x-based > version do for a start? I think as a matter of principle, the only things that belongs into git hooks are things that relate to the repository itself, such as commit metadata (author checks, commit message spell checks, etc.). Anything that relates to the source belongs into the source.
On Thu, Feb 10, 2011 at 09:16:09PM +0200, Peter Eisentraut wrote: > On tor, 2011-02-10 at 10:40 -0800, David Fetter wrote: > > I think all such checks belong in .git/hooks/pre-commit, and need > > to be as cross-platform as needed for committers. Would a > > *n*x-based version do for a start? > > I think as a matter of principle, the only things that belongs into > git hooks are things that relate to the repository itself, such as > commit metadata (author checks, commit message spell checks, etc.). > Anything that relates to the source belongs into the source. So on that principle, should "make check" do these checks? Just to put another viewpoint out there, the reason that such hooks exist is precisely to enable people to do deeper checks than commit metadata. Cheers, David. -- David Fetter <david@fetter.org> http://fetter.org/ Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter Skype: davidfetter XMPP: david.fetter@gmail.com iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics Remember to vote! Consider donating to Postgres: http://www.postgresql.org/about/donate
On Thu, Feb 10, 2011 at 11:29:43AM -0300, Alvaro Herrera wrote: > Excerpts from Peter Eisentraut's message of jue feb 10 07:58:16 -0300 2011: > > > One thing I was thinking of was that we could add a global make > > maintainer-check target (a name I picked up from other projects) > > which would run various source code sanity checks. Besides the > > SGML tabs issue, my favourite would be duplicate_oids. Maybe if > > we could find a third use case, we'd have a quorum for > > implementing this. > > The c++ headers thing? Or the compiles-in-isolation test for > headers? Both sound reasonable, given the number of times they come up and the ease of checking them mechanically. Cheers, David. -- David Fetter <david@fetter.org> http://fetter.org/ Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter Skype: davidfetter XMPP: david.fetter@gmail.com iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics Remember to vote! Consider donating to Postgres: http://www.postgresql.org/about/donate
On tor, 2011-02-10 at 12:58 +0200, Peter Eisentraut wrote: > There is make check target in doc/src/sgml/ that is supposed to catch > this. But it's probably hard to remember to run that. > > One thing I was thinking of was that we could add a global make > maintainer-check target (a name I picked up from other projects) which > would run various source code sanity checks. Besides the SGML tabs > issue, my favourite would be duplicate_oids. Maybe if we could find a > third use case, we'd have a quorum for implementing this. Hah, I found one. There's already a hidden maintainer-check target that checks the NLS files syntax. Anyway, here is a small patch that changes the duplicate_oids script to return a nonzero exit status in case of a problem, and then creates a global maintainer-check target that checks that and the SGML syntax and the NLS thing. Other things could be added in the future. Documentation should be added. Useful?
Attachment
On Thu, Mar 24, 2011 at 4:05 PM, Peter Eisentraut <peter_e@gmx.net> wrote: > On tor, 2011-02-10 at 12:58 +0200, Peter Eisentraut wrote: >> There is make check target in doc/src/sgml/ that is supposed to catch >> this. But it's probably hard to remember to run that. >> >> One thing I was thinking of was that we could add a global make >> maintainer-check target (a name I picked up from other projects) which >> would run various source code sanity checks. Besides the SGML tabs >> issue, my favourite would be duplicate_oids. Maybe if we could find a >> third use case, we'd have a quorum for implementing this. > > Hah, I found one. There's already a hidden maintainer-check target that > checks the NLS files syntax. > > Anyway, here is a small patch that changes the duplicate_oids script to > return a nonzero exit status in case of a problem, and then creates a > global maintainer-check target that checks that and the SGML syntax and > the NLS thing. Other things could be added in the future. > Documentation should be added. > > Useful? +1 from me. -- Robert Haas EnterpriseDB: http://www.enterprisedb.com The Enterprise PostgreSQL Company
On tor, 2011-03-24 at 16:05 -0400, Peter Eisentraut wrote: > Anyway, here is a small patch that changes the duplicate_oids script > to return a nonzero exit status in case of a problem, and then creates > a global maintainer-check target that checks that and the SGML syntax > and the NLS thing. Other things could be added in the future. > Documentation should be added. Committed, but didn't find a good place to document it.
On Mon, Mar 28, 2011 at 10:56:39PM +0300, Peter Eisentraut wrote: > On tor, 2011-03-24 at 16:05 -0400, Peter Eisentraut wrote: > > Anyway, here is a small patch that changes the duplicate_oids script > > to return a nonzero exit status in case of a problem, and then creates > > a global maintainer-check target that checks that and the SGML syntax > > and the NLS thing. Other things could be added in the future. > > Documentation should be added. > > Committed, but didn't find a good place to document it. Is there a section for any of the maintainer- stuff in make? If not, should there be? Cheers, David. -- David Fetter <david@fetter.org> http://fetter.org/ Phone: +1 415 235 3778 AIM: dfetter666 Yahoo!: dfetter Skype: davidfetter XMPP: david.fetter@gmail.com iCal: webcal://www.tripit.com/feed/ical/people/david74/tripit.ics Remember to vote! Consider donating to Postgres: http://www.postgresql.org/about/donate