doc: \copy can get data values \. and end-of-input confused

Reported-by: Svante Richter

Discussion: https://postgr.es/m/fcd57e4-8f23-4c3e-a5db-2571d09208e2@beta.fastmail.com

Backpatch-through: 11

doc/src/sgml/manage-ag.sgml

@@ -207,6 +207,12 @@ createdb -O <replaceable>rolename</replaceable> <replaceable>dbname</replaceable
    action being taken when those databases are created.
   </para>
 
+  <para>
+   However, <command>CREATE DATABASE</command> does not copy database-level
+   <command>GRANT</command> permissions attached to the source database.
+   The new database has default database-level permissions.
+  </para>
+
   <para>
    There is a second standard system database named
    <literal>template0</literal>.<indexterm><primary>template0</primary></indexterm> This

doc/src/sgml/ref/alter_default_privileges.sgml

@@ -137,7 +137,11 @@ REVOKE [ GRANT OPTION FOR ]
     <listitem>
      <para>
       The name of an existing role of which the current role is a member.
-      If <literal>FOR ROLE</literal> is omitted, the current role is assumed.
+      Default access privileges are not inherited, so member roles
+      must use <command>SET ROLE</command> to access these privileges,
+      or <command>ALTER DEFAULT PRIVILEGES</command> must be run for
+      each member role.  If <literal>FOR ROLE</literal> is omitted,
+      the current role is assumed.
      </para>
     </listitem>
    </varlistentry>

doc/src/sgml/ref/create_rule.sgml

@@ -59,15 +59,17 @@ CREATE [ OR REPLACE ] RULE <replaceable class="parameter">name</replaceable> AS
   </para>
 
   <para>
-   Presently, <literal>ON SELECT</literal> rules must be unconditional
-   <literal>INSTEAD</literal> rules and must have actions that consist
-   of a single <command>SELECT</command> command.  Thus, an
-   <literal>ON SELECT</literal> rule effectively turns the table into
-   a view, whose visible contents are the rows returned by the rule's
-   <command>SELECT</command> command rather than whatever had been
-   stored in the table (if anything).  It is considered better style
-   to write a <command>CREATE VIEW</command> command than to create a
-   real table and define an <literal>ON SELECT</literal> rule for it.
+   Presently, <literal>ON SELECT</literal> rules can only be attached
+   to views.  (Attaching one to a table converts the table into a view.)
+   Such a rule must be named <literal>"_RETURN"</literal>,
+   must be an unconditional <literal>INSTEAD</literal> rule, and must have
+   an action that consists of a single <command>SELECT</command> command.
+   This command defines the visible contents of the view.  (The view
+   itself is basically a dummy table with no storage.)  It's best to
+   regard such a rule as an implementation detail.  While a view can be
+   redefined via <literal>CREATE OR REPLACE RULE "_RETURN" AS
+   ...</literal>, it's better style to use <literal>CREATE OR REPLACE
+   VIEW</literal>.
   </para>
 
   <para>

doc/src/sgml/ref/psql-ref.sgml

@@ -1060,6 +1060,10 @@ testdb=>
         destination, because all data must pass through the client/server
         connection.  For large amounts of data the <acronym>SQL</acronym>
         command might be preferable.
+        Also, because of this pass-through method, <literal>\copy
+        ... from</literal> in <acronym>CSV</acronym> mode will erroneously
+        treat a <literal>\.</literal> data value alone on a line as an
+        end-of-input marker.
         </para>
         </tip>
 

src/bin/psql/copy.c

@@ -633,6 +633,7 @@ handleCopyIn(PGconn *conn, FILE *copystream, bool isbinary, PGresult **res)
 					 * This code erroneously assumes '\.' on a line alone
 					 * inside a quoted CSV string terminates the \copy.
 					 * http://www.postgresql.org/message-id/E1TdNVQ-0001ju-GO@wrigleys.postgresql.org
+					 * https://www.postgresql.org/message-id/bfcd57e4-8f23-4c3e-a5db-2571d09208e2@beta.fastmail.com
 					 */
 					if (strcmp(buf, "\\.\n") == 0 ||
 						strcmp(buf, "\\.\r\n") == 0)