Re: Flexible configuration for full-text search - Mailing list pgsql-hackers
| From | Alexander Korotkov |
|---|---|
| Subject | Re: Flexible configuration for full-text search |
| Date | |
| Msg-id | CAPpHfdvjTM7R+GFvLEUsNLdn=qFaDJN7ya7zQwe5EoXG+uvgVg@mail.gmail.com Whole thread Raw |
| In response to | Re: Flexible configuration for full-text search (Aleksandr Parfenov <a.parfenov@postgrespro.ru>) |
| Responses |
Re: Flexible configuration for full-text search
|
| List | pgsql-hackers |
Hi, Aleksandr!
On Mon, Jul 9, 2018 at 10:26 AM Aleksandr Parfenov <a.parfenov@postgrespro.ru> wrote:
A new version of the patch in the attachment. There are no changes since
the last version except refreshing it to current HEAD.
I took a look at this patch. It applied cleanly, but didn't pass regression tests.
*** /Users/smagen/projects/postgresql/env/master/src/src/test/regress/expected/misc_sanity.out 2018-07-20 13:44:54.000000000 +0300
--- /Users/smagen/projects/postgresql/env/master/src/src/test/regress/results/misc_sanity.out 2018-07-20 13:47:00.000000000 +0300
***************
*** 105,109 ****
pg_index | indpred | pg_node_tree
pg_largeobject | data | bytea
pg_largeobject_metadata | lomacl | aclitem[]
! (11 rows)
--- 105,110 ----
pg_index | indpred | pg_node_tree
pg_largeobject | data | bytea
pg_largeobject_metadata | lomacl | aclitem[]
! pg_ts_config_map | mapdicts | jsonb
! (12 rows)
It seems to be related to recent patches which adds toast tables to majority of system tables with varlena column. Regression diff indicates that mapdicts field of pg_ts_config_map can't be toasted. I think we should add toast table to pg_ts_config_map table.
Also, I see that you add extra grammar rules for ALTER TEXT SEARCH CONFIGURATION to the documentation, which allows specifying config instead of dictionary_name.
+ALTER TEXT SEARCH CONFIGURATION <replaceable>name</replaceable>
+ ADD MAPPING FOR <replaceable class="parameter">token_type</replaceable> [, ... ] WITH <replaceable class="parameter">config</replaceable>
ALTER TEXT SEARCH CONFIGURATION <replaceable>name</replaceable>
ADD MAPPING FOR <replaceable class="parameter">token_type</replaceable> [, ... ] WITH <replaceable class="parameter">dictionary_name</replaceable> [, ... ]
+ALTER TEXT SEARCH CONFIGURATION <replaceable>name</replaceable>
+ ALTER MAPPING FOR <replaceable class="parameter">token_type</replaceable> [, ... ] WITH <replaceable class="parameter">config</replaceable>
ALTER TEXT SEARCH CONFIGURATION <replaceable>name</replaceable>
ALTER MAPPING FOR <replaceable class="parameter">token_type</replaceable> [, ... ] WITH <replaceable class="parameter">dictionary_name</replaceable> [, ... ]
ALTER TEXT SEARCH CONFIGURATION <replaceable>name</replaceable>
In the same time you declare config as following.
+ <para>
+ Formally <replaceable class="parameter">config</replaceable> is one of:
+ </para>
+ <programlisting>
+ * dictionary_name
+
+ * config { UNION | INTERSECT | EXCEPT | MAP } config
+
+ * CASE config
+ WHEN [ NO ] MATCH THEN { KEEP | config }
+ [ ELSE config ]
+ END
+ </programlisting>
That is config itself could be a dictionary_name. I think this makes grammar rules for ALTER TEXT SEARCH CONFIGURATION redundant. We can specify those rules to always expect config, assuming that it can be actually a dictionary nay.
+ if (fout->remoteVersion >= 110000)
PostgreSQL 11 already passed feature freeze. Thus, we should be aimed to PostgreSQL 12.
That's all for now, but I'm going to do more detailed code review.
------
Alexander Korotkov
Postgres Professional: http://www.postgrespro.com
The Russian Postgres Company
Alexander Korotkov
Postgres Professional: http://www.postgrespro.com
The Russian Postgres Company
pgsql-hackers by date: