Re: proposal: possibility to read dumped table's name from file
Justin Pryzby <pryzby@telsasoft.com>
From: Justin Pryzby <pryzby@telsasoft.com>
To: Pavel Stehule <pavel.stehule@gmail.com>
Cc: Julien Rouhaud <rjuju123@gmail.com>, Daniel Gustafsson <daniel@yesql.se>, "Gregory Stark (as CFM)" <stark.cfm@gmail.com>, Andres Freund <andres@anarazel.de>, Andrew Dunstan <andrew@dunslane.net>, John Naylor <john.naylor@enterprisedb.com>, Erik Rijkers <er@xs4all.nl>, Tom Lane <tgl@sss.pgh.pa.us>, Alvaro Herrera <alvherre@2ndquadrant.com>, Tomas Vondra <tomas.vondra@enterprisedb.com>, Dean Rasheed <dean.a.rasheed@gmail.com>, Stephen Frost <sfrost@snowman.net>, Surafel Temesgen <surafel3000@gmail.com>, vignesh C <vignesh21@gmail.com>, pgsql-hackers@postgresql.org
Date: 2023-03-19T14:01:51Z
Lists: pgsql-hackers
Attachments
- 0001-possibility-to-read-options-for-dump-from-file.txt (text/x-diff)
On Thu, Mar 16, 2023 at 01:05:41PM +0100, Pavel Stehule wrote: > rebase + enhancing about related option from a563c24 Thanks. It looks like this doesn't currently handle extensions, which were added at 6568cef26e. > + <literal>table_and_children</literal>: tables, works like > + <option>-t</option>/<option>--table</option>, except that > + it also includes any partitions or inheritance child > + tables of the table(s) matching the > + <replaceable class="parameter">pattern</replaceable>. Why doesn't this just say "works like --table-and-children" ? I think as you wrote log_invalid_filter_format(), the messages wouldn't be translated, because they're printed via %s. One option is to call _() on the message. > +ok($dump !=~ qr/^CREATE TABLE public\.bootab/m, "exclude dumped children table"); !=~ is being interpretted as as numeric "!=" and throwing warnings. It should be a !~ b, right ? It'd be nice if perl warnings during the tests were less easy to miss. > + * char is not alpha. The char '_' is allowed too (exclude first position). Why is it treated specially? Could it be treated the same as alpha? > + log_invalid_filter_format(&fstate, > + "\"include\" table data filter is not allowed"); > + log_invalid_filter_format(&fstate, > + "\"include\" table data and children filter is not allowed"); For these, it might be better to write the literal option: > + "include filter for \"table_data_and_children\" is not allowed"); Because the option is a literal and shouldn't be translated. And it's probably better to write that using %s, like: > + "include filter for \"%s\" is not allowed"); That makes shorter and fewer strings. Find attached a bunch of other corrections as 0002.txt I also dug up what I'd started in november, trying to reduce the code duplication betwen pg_restore/dump/all. This isn't done, but I might never finish it, so I'll at least show what I have in case you think it's a good idea. This passes tests on CI, except for autoconf, due to using exit_nicely() differently. -- Justin
Commits
Same data as JSON:
GET /api/v1/messages/:b64id/commits
the thread's linked commits as JSON, with link sources.
API reference →
-
Fix array subscript warnings
- 17935e1fdf0a 17.0 landed
-
Read include/exclude commands for dump/restore from file
- a5cf808be55b 17.0 landed
-
Allow records to span multiple lines in pg_hba.conf and pg_ident.conf.
- 8f8154a503c7 14.0 cited