mprove tab completion for ALTER EXTENSION ADD/DROP
Hi,
Tab completion for ALTER EXTENSION ADD and DROP was missing, this
patch adds the tab completion for the same.
Regards,
Vignesh
Attachments:
v1-0001-Missing-tab-completion-for-ALTER-EXTENSION-ADD-DR.patchtext/x-patch; charset=US-ASCII; name=v1-0001-Missing-tab-completion-for-ALTER-EXTENSION-ADD-DR.patchDownload
From 35e9e8bd22ae8767baa95a04c41ee8f68a9c5338 Mon Sep 17 00:00:00 2001
From: Vignesh C <vignesh21@gmail.com>
Date: Sat, 26 Nov 2022 20:18:58 +0530
Subject: [PATCH v1] Missing tab completion for ALTER EXTENSION ADD/DROP
Missing tab completion for ALTER EXTENSION ADD/DROP
---
src/bin/psql/tab-complete.c | 21 +++++++++++++++++++++
1 file changed, 21 insertions(+)
diff --git a/src/bin/psql/tab-complete.c b/src/bin/psql/tab-complete.c
index 13014f074f..9c3bc751dd 100644
--- a/src/bin/psql/tab-complete.c
+++ b/src/bin/psql/tab-complete.c
@@ -1937,6 +1937,27 @@ psql_completion(const char *text, int start, int end)
else if (Matches("ALTER", "EXTENSION", MatchAny))
COMPLETE_WITH("ADD", "DROP", "UPDATE", "SET SCHEMA");
+ /* ALTER EXTENSION <name> ADD|DROP */
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD|DROP"))
+ COMPLETE_WITH("ACCESS METHOD", "AGGREGATE", "CAST", "COLLATION",
+ "CONVERSION", "DOMAIN", "EVENT TRIGGER", "FOREIGN",
+ "FUNCTION", "MATERIALIZED VIEW", "OPERATOR",
+ "PROCEDURAL LANGUAGE", "PROCEDURE", "LANGUAGE",
+ "ROUTINE", "SCHEMA", "SEQUENCE", "SERVER", "TABLE",
+ "TEXT SEARCH", "TRANSFORM FOR", "TYPE", "VIEW");
+
+ /* ALTER EXTENSION <name> ADD|DROP FOREIGN*/
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD|DROP", "FOREIGN"))
+ COMPLETE_WITH("DATA WRAPPER", "TABLE");
+
+ /* ALTER EXTENSION <name> ADD|DROP OPERATOR*/
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD|DROP", "OPERATOR"))
+ COMPLETE_WITH("CLASS", "FAMILY");
+
+ /* ALTER EXTENSION <name> ADD|DROP TEXT SEARCH*/
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD|DROP", "TEXT", "SEARCH"))
+ COMPLETE_WITH("CONFIGURATION", "DICTIONARY", "PARSER", "TEMPLATE");
+
/* ALTER EXTENSION <name> UPDATE */
else if (Matches("ALTER", "EXTENSION", MatchAny, "UPDATE"))
COMPLETE_WITH("TO");
--
2.34.1
------- Original Message -------
On Sunday, November 27th, 2022 at 10:24, vignesh C <vignesh21@gmail.com> wrote:
Hi,
Tab completion for ALTER EXTENSION ADD and DROP was missing, this
patch adds the tab completion for the same.Regards,
Vignesh
Hi Vignesh
I've tested your patched on current master and seems to be working properly.
I'm starting reviewing some patches here, let's see what more experience hackers
has to say about this, but as far I can tell is that is working as expected.
--
Matheus Alcantara
On Sat, Dec 03, 2022 at 05:34:57PM +0000, Matheus Alcantara wrote:
I've tested your patched on current master and seems to be working properly.
I'm starting reviewing some patches here, let's see what more experience hackers
has to say about this, but as far I can tell is that is working as expected.
+ /* ALTER EXTENSION <name> ADD|DROP */
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD|DROP"))
+ COMPLETE_WITH("ACCESS METHOD", "AGGREGATE", "CAST", "COLLATION",
+ "CONVERSION", "DOMAIN", "EVENT TRIGGER", "FOREIGN",
+ "FUNCTION", "MATERIALIZED VIEW", "OPERATOR",
+ "PROCEDURAL LANGUAGE", "PROCEDURE", "LANGUAGE",
+ "ROUTINE", "SCHEMA", "SEQUENCE", "SERVER", "TABLE",
+ "TEXT SEARCH", "TRANSFORM FOR", "TYPE", "VIEW");
+
+ /* ALTER EXTENSION <name> ADD|DROP FOREIGN*/
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD|DROP", "FOREIGN"))
+ COMPLETE_WITH("DATA WRAPPER", "TABLE");The DROP could be matched with the objects that are actually part of
the so-said extension?
--
Michael
On Mon, 5 Dec 2022 at 06:53, Michael Paquier <michael@paquier.xyz> wrote:
On Sat, Dec 03, 2022 at 05:34:57PM +0000, Matheus Alcantara wrote:
I've tested your patched on current master and seems to be working properly.
I'm starting reviewing some patches here, let's see what more experience hackers
has to say about this, but as far I can tell is that is working as expected.+ /* ALTER EXTENSION <name> ADD|DROP */ + else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD|DROP")) + COMPLETE_WITH("ACCESS METHOD", "AGGREGATE", "CAST", "COLLATION", + "CONVERSION", "DOMAIN", "EVENT TRIGGER", "FOREIGN", + "FUNCTION", "MATERIALIZED VIEW", "OPERATOR", + "PROCEDURAL LANGUAGE", "PROCEDURE", "LANGUAGE", + "ROUTINE", "SCHEMA", "SEQUENCE", "SERVER", "TABLE", + "TEXT SEARCH", "TRANSFORM FOR", "TYPE", "VIEW"); + + /* ALTER EXTENSION <name> ADD|DROP FOREIGN*/ + else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD|DROP", "FOREIGN")) + COMPLETE_WITH("DATA WRAPPER", "TABLE");The DROP could be matched with the objects that are actually part of
the so-said extension?
The modified v2 version has the changes to handle the same. Sorry for
the delay as I was working on another project.
Regards,
Vignesh
Attachments:
v2-0002-Missing-tab-completion-for-ALTER-EXTENSION-DROP.patchtext/x-patch; charset=US-ASCII; name=v2-0002-Missing-tab-completion-for-ALTER-EXTENSION-DROP.patchDownload
From c2aeeadac88b709ce823a01c80b0b8455c62d3d9 Mon Sep 17 00:00:00 2001
From: Vignesh C <vignesh21@gmail.com>
Date: Mon, 2 Jan 2023 08:01:45 +0530
Subject: [PATCH v2 2/2] Missing tab completion for ALTER EXTENSION DROP
Missing tab completion for ALTER EXTENSION DROP
---
src/bin/psql/tab-complete.c | 25 ++++++++++++++++++++++---
1 file changed, 22 insertions(+), 3 deletions(-)
diff --git a/src/bin/psql/tab-complete.c b/src/bin/psql/tab-complete.c
index 4d1a664bf7..a6f27ed9b8 100644
--- a/src/bin/psql/tab-complete.c
+++ b/src/bin/psql/tab-complete.c
@@ -224,6 +224,7 @@ static char *completion_ref_schema; /* schema name of reference object */
static bool completion_case_sensitive; /* completion is case sensitive */
static bool completion_verbatim; /* completion is verbatim */
static bool completion_force_quote; /* true to force-quote filenames */
+static bool completion_dont_quote; /* true to not quote the result */
/*
* A few macros to ease typing. You can use these to complete the given
@@ -1011,6 +1012,15 @@ static const SchemaQuery Query_for_trigger_of_table = {
"SELECT nspname FROM pg_catalog.pg_namespace "\
" WHERE nspname LIKE '%s'"
+#define Query_for_extension_objs \
+"SELECT distinct CASE WHEN p.type = 'foreign-data wrapper' "\
+" THEN 'FOREIGN DATA WRAPPER'"\
+" ELSE trim ('\"' from pg_catalog.upper(p.type::text)) END "\
+" FROM pg_depend, pg_identify_object(classid, objid, objsubid) AS p"\
+" WHERE refclassid = 'pg_extension'::regclass AND"\
+" pg_catalog.lower(p.type) LIKE pg_catalog.lower('%s') AND"\
+" refobjid = (SELECT oid FROM pg_extension WHERE extname = '%s')"
+
/* Use COMPLETE_WITH_QUERY_VERBATIM with these queries for GUC names: */
#define Query_for_list_of_alter_system_set_vars \
"SELECT pg_catalog.lower(name) FROM pg_catalog.pg_settings "\
@@ -1946,6 +1956,15 @@ psql_completion(const char *text, int start, int end)
"ROUTINE", "SCHEMA", "SEQUENCE", "SERVER", "TABLE",
"TEXT SEARCH", "TRANSFORM FOR", "TYPE", "VIEW");
+ /* ALTER EXTENSION <name> DROP */
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "DROP"))
+ {
+ set_completion_reference(prev2_wd);
+ completion_dont_quote = true;
+ COMPLETE_WITH_QUERY(Query_for_extension_objs);
+ completion_dont_quote = false;
+ }
+
/* ALTER EXTENSION <name> ADD FOREIGN*/
else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD", "FOREIGN"))
COMPLETE_WITH("DATA WRAPPER", "TABLE");
@@ -5313,7 +5332,7 @@ _complete_from_query(const char *simple_query,
* surprising. This restriction also dodges some odd behaviors of
* some versions of readline/libedit.
*/
- if (non_empty_object)
+ if (non_empty_object && !completion_dont_quote)
{
if (item && !objectquoted && identifier_needs_quotes(item))
continue;
@@ -5910,7 +5929,7 @@ requote_identifier(const char *schemaname, const char *objectname,
if (schemaname)
{
buflen += strlen(schemaname) + 1; /* +1 for the dot */
- if (!quote_schema)
+ if (!quote_schema && !completion_dont_quote)
quote_schema = identifier_needs_quotes(schemaname);
if (quote_schema)
{
@@ -5925,7 +5944,7 @@ requote_identifier(const char *schemaname, const char *objectname,
if (objectname)
{
buflen += strlen(objectname);
- if (!quote_object)
+ if (!quote_object && !completion_dont_quote)
quote_object = identifier_needs_quotes(objectname);
if (quote_object)
{
--
2.34.1
v2-0001-Missing-tab-completion-for-ALTER-EXTENSION-ADD.patchtext/x-patch; charset=US-ASCII; name=v2-0001-Missing-tab-completion-for-ALTER-EXTENSION-ADD.patchDownload
From 328509d60c1ec5489f5e29efbabc4c32466d1bac Mon Sep 17 00:00:00 2001
From: Vignesh C <vignesh21@gmail.com>
Date: Sat, 26 Nov 2022 20:18:58 +0530
Subject: [PATCH v2 1/2] Missing tab completion for ALTER EXTENSION ADD
Missing tab completion for ALTER EXTENSION ADD
---
src/bin/psql/tab-complete.c | 21 +++++++++++++++++++++
1 file changed, 21 insertions(+)
diff --git a/src/bin/psql/tab-complete.c b/src/bin/psql/tab-complete.c
index 2a3921937c..4d1a664bf7 100644
--- a/src/bin/psql/tab-complete.c
+++ b/src/bin/psql/tab-complete.c
@@ -1937,6 +1937,27 @@ psql_completion(const char *text, int start, int end)
else if (Matches("ALTER", "EXTENSION", MatchAny))
COMPLETE_WITH("ADD", "DROP", "UPDATE", "SET SCHEMA");
+ /* ALTER EXTENSION <name> ADD */
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD"))
+ COMPLETE_WITH("ACCESS METHOD", "AGGREGATE", "CAST", "COLLATION",
+ "CONVERSION", "DOMAIN", "EVENT TRIGGER", "FOREIGN",
+ "FUNCTION", "MATERIALIZED VIEW", "OPERATOR",
+ "PROCEDURAL LANGUAGE", "PROCEDURE", "LANGUAGE",
+ "ROUTINE", "SCHEMA", "SEQUENCE", "SERVER", "TABLE",
+ "TEXT SEARCH", "TRANSFORM FOR", "TYPE", "VIEW");
+
+ /* ALTER EXTENSION <name> ADD FOREIGN*/
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD", "FOREIGN"))
+ COMPLETE_WITH("DATA WRAPPER", "TABLE");
+
+ /* ALTER EXTENSION <name> ADD OPERATOR*/
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD", "OPERATOR"))
+ COMPLETE_WITH("CLASS", "FAMILY");
+
+ /* ALTER EXTENSION <name> ADD TEXT SEARCH*/
+ else if (Matches("ALTER", "EXTENSION", MatchAny, "ADD|DROP", "TEXT", "SEARCH"))
+ COMPLETE_WITH("CONFIGURATION", "DICTIONARY", "PARSER", "TEMPLATE");
+
/* ALTER EXTENSION <name> UPDATE */
else if (Matches("ALTER", "EXTENSION", MatchAny, "UPDATE"))
COMPLETE_WITH("TO");
--
2.34.1
At Mon, 2 Jan 2023 13:19:50 +0530, vignesh C <vignesh21@gmail.com> wrote in
On Mon, 5 Dec 2022 at 06:53, Michael Paquier <michael@paquier.xyz> wrote:
The DROP could be matched with the objects that are actually part of
the so-said extension?The modified v2 version has the changes to handle the same. Sorry for
the delay as I was working on another project.
It suggests the *kinds* of objects that are part of the extension, but
lists the objects of that kind regardless of dependency. I read
Michael suggested (and I agree) to restrict the objects (not kinds) to
actually be a part of the extension. (And not for object kinds.)
However I'm not sure it is useful to restrict object kinds since the
operator already knows what to drop, if you still want to do that, the
use of completion_dont_quote looks ugly since the function
(requote_identifier) is assuming an identifier as input. I didn't
looked closer but maybe we need another way to do that.
regards.
--
Kyotaro Horiguchi
NTT Open Source Software Center
On Wed, Jan 11, 2023 at 12:10:33PM +0900, Kyotaro Horiguchi wrote:
It suggests the *kinds* of objects that are part of the extension, but
lists the objects of that kind regardless of dependency. I read
Michael suggested (and I agree) to restrict the objects (not kinds) to
actually be a part of the extension. (And not for object kinds.)
Yeah, that's what I meant. Now, if Vignesh does not want to extend
that, that's fine for me as well at the end on second thought, as this
involves much more code for each DROP path depending on the object
type involved.
Adding the object names after DROP/ADD is useful on its own, and we
already have some completion once the object type is specified, so
simpler is perhaps just better here.
--
Michael
On Wed, 11 Jan 2023 at 12:19, Michael Paquier <michael@paquier.xyz> wrote:
On Wed, Jan 11, 2023 at 12:10:33PM +0900, Kyotaro Horiguchi wrote:
It suggests the *kinds* of objects that are part of the extension, but
lists the objects of that kind regardless of dependency. I read
Michael suggested (and I agree) to restrict the objects (not kinds) to
actually be a part of the extension. (And not for object kinds.)Yeah, that's what I meant. Now, if Vignesh does not want to extend
that, that's fine for me as well at the end on second thought, as this
involves much more code for each DROP path depending on the object
type involved.Adding the object names after DROP/ADD is useful on its own, and we
already have some completion once the object type is specified, so
simpler is perhaps just better here.
I too felt keeping it simpler is better. How about using the simple
first version of patch itself?
Regards,
Vignesh
On Wed, Jan 11, 2023 at 10:29:25PM +0530, vignesh C wrote:
I too felt keeping it simpler is better. How about using the simple
first version of patch itself?
Okay, I have just done that, then, after checking that all the object
types were covered (28). Note that PROCEDURAL LANGUAGE has been
removed for simplicity.
--
Michael
On Thu, 12 Jan 2023 at 05:22, Michael Paquier <michael@paquier.xyz> wrote:
On Wed, Jan 11, 2023 at 10:29:25PM +0530, vignesh C wrote:
I too felt keeping it simpler is better. How about using the simple
first version of patch itself?Okay, I have just done that, then, after checking that all the object
types were covered (28). Note that PROCEDURAL LANGUAGE has been
removed for simplicity.
Thanks for pushing this.
Regards,
Vignesh