Feedback-ID: ie0a040ee:Fastmail
Message-ID: <1a89b3f0-e588-41ef-b712-aba766143cad@eisentraut.org>
Date: Fri, 7 Nov 2025 09:53:38 +0100
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Content-Language: en-US
From: Peter Eisentraut <peter@eisentraut.org>
Subject: translator comments for GUC descriptions useful?
To: "pgsql-translators@postgresql.org" <pgsql-translators@postgresql.org>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Archived-At: 
 <https://www.postgresql.org/message-id/1a89b3f0-e588-41ef-b712-aba766143cad%40eisentraut.org>
Precedence: bulk

Hi all,

I've done some work recently in the source code to make the GUC C 
structs automatically generated.  That now allows changing the output 
format more easily.

One random idea I had is that we could automatically generate 
"translator" comments on the short and long description strings of GUC 
parameters.  So the PO entries could then look like this:

#. translator: GUC huge_pages_status short description
#: utils/misc/guc_tables.c:5020
msgid "Indicates the status of huge pages."
msgstr ""

(and analogously for the long description).

This would be trivial to implement now (see below for an example).

Would this be useful?

I suppose the only possible disadvantage is that it would use more 
vertical space.

(Any related ideas while we're here?)


diff --git a/src/backend/utils/misc/gen_guc_tables.pl 
b/src/backend/utils/misc/gen_guc_tables.pl
index 601c34ec30b..25aa98debed 100644
--- a/src/backend/utils/misc/gen_guc_tables.pl
+++ b/src/backend/utils/misc/gen_guc_tables.pl
@@ -62,6 +62,7 @@ sub print_table
         printf $ofh "\t\t.name = %s,\n", dquote($entry->{name});
         printf $ofh "\t\t.context = %s,\n", $entry->{context};
         printf $ofh "\t\t.group = %s,\n", $entry->{group};
+       printf $ofh "\t\t/* translator: GUC %s short description */\n", 
$entry->{name};
         printf $ofh "\t\t.short_desc = gettext_noop(%s),\n",
           dquote($entry->{short_desc});
         printf $ofh "\t\t.long_desc = gettext_noop(%s),\n",