Horton

Re: WIP: a way forward on bootstrap data

John Naylor <jcnaylor@gmail.com>

View thread View on postgr.es
From: John Naylor <jcnaylor@gmail.com>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2018-04-06T10:17:51Z
Lists: pgsql-hackers 
On 4/5/18, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> Something that bothered me a bit while writing the warning-producing code
> is that showing %bki_values isn't actually that great a way of identifying
> the trouble spot.  By this point we've expanded out defaults and possibly
> replaced some other macros, so it doesn't look that much like what was
> in the .dat file.  I think what would be ideal, both here and in some
> other places like AddDefaultValues, is to be able to finger the location
> of the bad tuple by filename and line number, but I have no idea whether
> it's practical to annotate the tuples with that while reading the .dat
> files.  Any thoughts?

We could use the $. variable to save the line number, which is what
the old code had. AddDefaultValues will report the line number on
failure, so I left out explicit line number reporting. If memory
serves, Perl is sensitive to how you format the "die" message. If I
delete a default value from the header, I get this, reporting line 16:

Failed to form full tuple for pg_opfamily
Missing values for: opfnamespace
Showing other values for context:
oid => 421, opfmethod => 403, opfowner => PGUID, opfname =>
abstime_ops,  at ../../../src/backend/catalog/Catalog.pm line 259,
<$ifd> line 16.
Makefile:23: recipe for target 'reformat-dat-files' failed
make: *** [reformat-dat-files] Error 25

I think the context is good for pg_attribute, because there is no file
to read from.
I'll think about the lookup code.

-John Naylor

Commits

  1. Clarify handling of special-case values in bootstrap catalog data.

  2. Replace our traditional initial-catalog-data format with a better design.

  3. Faster partition pruning

  4. Minor cleanup in genbki.pl.

  5. Trivial adjustments in preparation for bootstrap data conversion.

  6. Remove hard-coded schema knowledge about pg_attribute from genbki.pl

  7. Minor edits to catalog files and scripts

  8. Hide most variable-length fields from Form_pg_* structs