parser: support on conflict do select & sync regression suite

crates/squawk_parser/src/grammar.rs

@@ -12420,6 +12420,10 @@ fn conflict_action(p: &mut Parser<'_>) {
     p.expect(DO_KW);
     if p.eat(NOTHING_KW) {
         m.complete(p, CONFLICT_DO_NOTHING);
+    } else if p.eat(SELECT_KW) {
+        opt_locking_clause(p);
+        opt_where_clause(p);
+        m.complete(p, CONFLICT_DO_SELECT);
     } else {
         p.expect(UPDATE_KW);
         set_clause(p);

crates/squawk_parser/tests/data/ok/insert_pg18.sql

@@ -0,0 +1,3 @@
+-- on conflict do select
+INSERT INTO t (a) VALUES (1)
+    ON CONFLICT (a) DO SELECT FOR UPDATE WHERE a > 0 RETURNING *;

crates/squawk_parser/tests/snapshots/tests__insert_pg18_ok.snap

@@ -0,0 +1,77 @@
+---
+source: crates/squawk_parser/tests/tests.rs
+input_file: crates/squawk_parser/tests/data/ok/insert_pg18.sql
+---
+SOURCE_FILE
+  COMMENT "-- on conflict do select"
+  WHITESPACE "\n"
+  INSERT
+    INSERT_KW "INSERT"
+    WHITESPACE " "
+    INTO_KW "INTO"
+    WHITESPACE " "
+    PATH
+      PATH_SEGMENT
+        NAME_REF
+          IDENT "t"
+    WHITESPACE " "
+    COLUMN_LIST
+      L_PAREN "("
+      COLUMN
+        NAME_REF
+          IDENT "a"
+      R_PAREN ")"
+    WHITESPACE " "
+    VALUES
+      VALUES_KW "VALUES"
+      WHITESPACE " "
+      ROW_LIST
+        ROW
+          L_PAREN "("
+          LITERAL
+            INT_NUMBER "1"
+          R_PAREN ")"
+    WHITESPACE "\n    "
+    ON_CONFLICT_CLAUSE
+      ON_KW "ON"
+      WHITESPACE " "
+      CONFLICT_KW "CONFLICT"
+      WHITESPACE " "
+      CONFLICT_ON_INDEX
+        CONFLICT_INDEX_ITEM_LIST
+          L_PAREN "("
+          CONFLICT_INDEX_ITEM
+            NAME_REF
+              IDENT "a"
+          R_PAREN ")"
+      WHITESPACE " "
+      CONFLICT_DO_SELECT
+        DO_KW "DO"
+        WHITESPACE " "
+        SELECT_KW "SELECT"
+        WHITESPACE " "
+        LOCKING_CLAUSE
+          FOR_KW "FOR"
+          WHITESPACE " "
+          UPDATE_KW "UPDATE"
+        WHITESPACE " "
+        WHERE_CLAUSE
+          WHERE_KW "WHERE"
+          WHITESPACE " "
+          BIN_EXPR
+            NAME_REF
+              IDENT "a"
+            WHITESPACE " "
+            R_ANGLE ">"
+            WHITESPACE " "
+            LITERAL
+              INT_NUMBER "0"
+    WHITESPACE " "
+    RETURNING_CLAUSE
+      RETURNING_KW "RETURNING"
+      WHITESPACE " "
+      TARGET_LIST
+        TARGET
+          STAR "*"
+  SEMICOLON ";"
+  WHITESPACE "\n"

crates/squawk_parser/tests/snapshots/tests__regression_encoding.snap

@@ -0,0 +1,5 @@
+---
+source: crates/squawk_parser/tests/tests.rs
+input_file: postgres/regression_suite/encoding.sql
+---
+

crates/squawk_parser/tests/snapshots/tests__regression_euc_kr.snap

@@ -0,0 +1,5 @@
+---
+source: crates/squawk_parser/tests/tests.rs
+input_file: postgres/regression_suite/euc_kr.sql
+---
+

crates/squawk_parser/tests/snapshots/tests__regression_oid8.snap

@@ -0,0 +1,5 @@
+---
+source: crates/squawk_parser/tests/tests.rs
+input_file: postgres/regression_suite/oid8.sql
+---
+

crates/squawk_syntax/src/postgresql.ungram

@@ -1305,13 +1305,17 @@ ConflictIndexItem =
 ConflictAction =
   ConflictDoNothing
 | ConflictDoUpdateSet
+| ConflictDoSelect
 
 ConflictDoUpdateSet =
   'do' 'update' SetClause WhereClause?
 
 ConflictDoNothing =
   'do' 'nothing'
 
+ConflictDoSelect =
+  'do' 'select' LockingClause? WhereClause?
+
 SetClause =
   'set' SetColumnList
 

crates/xtask/src/sync_regression_suite.rs

@@ -282,7 +282,9 @@ pub(crate) fn preprocess_sql<R: BufRead, W: Write>(source: R, mut dest: W) -> Re
                 in_copy_stdin = false;
             }
             should_comment = true;
-        } else if (line.trim_start().starts_with('\\') && !line.contains("\\gset"))
+        } else if (line.trim_start().starts_with('\\')
+            && !line.contains("\\gset")
+            && !line.contains("\\gx"))
             || line.starts_with("'show_data'")
             || line.starts_with(':')
         {
@@ -318,8 +320,8 @@ pub(crate) fn preprocess_sql<R: BufRead, W: Write>(source: R, mut dest: W) -> Re
             line = line.replace(from, to);
         }
 
-        if line.contains("\\gset") {
-            if let Some(start) = line.find("\\gset") {
+        for pattern in ["\\gx", "\\gset"] {
+            if let Some(start) = line.find(pattern) {
                 let end = line[start..]
                     .find('\n')
                     .map(|i| start + i)

postgres/kwlist.h

@@ -1,7 +1,7 @@
 // synced from: 
-//   commit: e5f3839af685c303d8ebcc1ea0d407c124372931
-//   committed at: 2025-12-22T22:41:34Z
-//   file: https://github.com/postgres/postgres/blob/e5f3839af685c303d8ebcc1ea0d407c124372931/src/include/parser/kwlist.h
+//   commit: ef3c3cf6d021ff9884c513afd850a9fe85cad736
+//   committed at: 2026-02-14T06:50:06Z
+//   file: https://github.com/postgres/postgres/blob/ef3c3cf6d021ff9884c513afd850a9fe85cad736/src/include/parser/kwlist.h
 //
 // update via:
 //   cargo xtask sync-kwlist
@@ -15,7 +15,7 @@
  * by the PG_KEYWORD macro, which is not defined in this file; it can
  * be defined by the caller for special purposes.
  *
- * Portions Copyright (c) 1996-2025, PostgreSQL Global Development Group
+ * Portions Copyright (c) 1996-2026, PostgreSQL Global Development Group
  * Portions Copyright (c) 1994, Regents of the University of California
  *
  * IDENTIFICATION

postgres/regression_suite/arrays.sql

@@ -528,6 +528,10 @@ select '[2147483646:2147483646]={1}'::int[];
 select '[-2147483648:-2147483647]={1,2}'::int[];
 -- all of the above should be accepted
 
+-- some day we might allow these cases, but for now they're errors:
+select array[]::oidvector;
+select array[]::int2vector;
+
 -- tests for array aggregates
 CREATE TEMP TABLE arraggtest ( f1 INT[], f2 TEXT[][], f3 FLOAT[]);
 
@@ -555,19 +559,22 @@ SELECT max(f1), min(f1), max(f2), min(f2), max(f3), min(f3) FROM arraggtest;
 
 -- A few simple tests for arrays of composite types
 
-create type comptype as (f1 int, f2 text);
+create type comptype as (f1 int, f2 text, f3 int[]);
 
 create table comptable (c1 comptype, c2 comptype[]);
 
 -- XXX would like to not have to specify row() construct types here ...
 insert into comptable
-  values (row(1,'foo'), array[row(2,'bar')::comptype, row(3,'baz')::comptype]);
+  values (row(1,'foo',array[10,20]), array[row(2,'bar',array[30,40])::comptype, row(3,'baz',array[50,60])::comptype]);
 
 -- check that implicitly named array type _comptype isn't a problem
 create type _comptype as enum('fooey');
 
 select * from comptable;
 select c2[2].f2 from comptable;
+select c2[2].f3 from comptable;
+select c2[2].f3[1:2] from comptable;
+select c2[1:2].f3[1:2] from comptable;
 
 drop type _comptype;
 drop table comptable;

postgres/regression_suite/constraints.sql

@@ -568,6 +568,9 @@ INSERT INTO circles VALUES('<(20,20), 10>', '<(0,0), 4>')
 -- fail, because DO UPDATE variant requires unique index
 INSERT INTO circles VALUES('<(20,20), 10>', '<(0,0), 4>')
   ON CONFLICT ON CONSTRAINT circles_c1_c2_excl DO UPDATE SET c2 = EXCLUDED.c2;
+-- fail, because DO SELECT variant requires unique index
+INSERT INTO circles VALUES('<(20,20), 10>', '<(0,0), 4>')
+  ON CONFLICT ON CONSTRAINT circles_c1_c2_excl DO SELECT RETURNING *;
 -- succeed because c1 doesn't overlap
 INSERT INTO circles VALUES('<(20,20), 1>', '<(0,0), 5>');
 -- succeed because c2 doesn't overlap
@@ -623,7 +626,9 @@ DROP TABLE deferred_excl;
 -- verify constraints created for NOT NULL clauses
 CREATE TABLE notnull_tbl1 (a INTEGER NOT NULL NOT NULL);
 -- \d+ notnull_tbl1
--- no-op
+-- specifying an existing constraint is a no-op
+ALTER TABLE notnull_tbl1 ADD CONSTRAINT notnull_tbl1_a_not_null NOT NULL a;
+-- but using a different constraint name is not allowed
 ALTER TABLE notnull_tbl1 ADD CONSTRAINT nn NOT NULL a;
 -- \d+ notnull_tbl1
 -- duplicate name

postgres/regression_suite/copy.sql

@@ -124,6 +124,21 @@ copy copytest5 from stdin (format csv, header 5);
 -- \.
 select count(*) from copytest5;
 
+-- test header line feature (given as strings)
+truncate copytest5;
+copy copytest5 from stdin (format csv, header '0');
+-- 1
+-- 2
+-- \.
+select * from copytest5 order by c1;
+
+truncate copytest5;
+copy copytest5 from stdin (format csv, header '1');
+-- 1
+-- 2
+-- \.
+select * from copytest5 order by c1;
+
 -- test copy from with a partitioned table
 create table parted_copytest (
 	a int,

postgres/regression_suite/copy2.sql

@@ -93,6 +93,9 @@ COPY x from stdin with (on_error ignore, reject_limit 0);
 COPY x from stdin with (header -1);
 COPY x from stdin with (header 2.5);
 -- COPY x to stdout with (header 2);
+COPY x to stdout with (header '-1');
+COPY x from stdin with (header '2.5');
+-- COPY x to stdout with (header '2');
 
 -- too many columns in column list: should fail
 COPY x (a, b, c, d, e, d, c) from stdin;

postgres/regression_suite/copyencoding.sql

@@ -23,6 +23,13 @@ CREATE TABLE copy_encoding_tab (t text);
 COPY (SELECT E'\u3042') TO 'utf8_csv' WITH (FORMAT csv, ENCODING 'UTF8');
 -- Read UTF8 data as LATIN1: no error
 COPY copy_encoding_tab FROM 'utf8_csv' WITH (FORMAT csv, ENCODING 'LATIN1');
+-- Non-server encodings have distinct code paths.
+-- \set fname :abs_builddir '/results/copyencoding_gb18030.csv'
+COPY (SELECT E'\u3042,') TO 'fname' WITH (FORMAT csv, ENCODING 'GB18030');
+COPY copy_encoding_tab FROM 'fname' WITH (FORMAT csv, ENCODING 'GB18030');
+-- \set fname :abs_builddir '/results/copyencoding_gb18030.data'
+COPY (SELECT E'\u3042,') TO 'fname' WITH (FORMAT text, ENCODING 'GB18030');
+COPY copy_encoding_tab FROM 'fname' WITH (FORMAT text, ENCODING 'GB18030');
 
 -- Use client_encoding
 SET client_encoding TO UTF8;