Fix libpq's code for searching .pgpass; rationalize empty-list-item cases.

Before v10, we always searched ~/.pgpass using the host parameter,
and nothing else, to match to the "hostname" field of ~/.pgpass.
(However, null host or host matching DEFAULT_PGSOCKET_DIR was replaced by
"localhost".)  In v10, this got broken by commit 274bb2b38, repaired by
commit bdac9836d, and broken again by commit 7b02ba62e; in the code
actually shipped, we'd search with hostaddr if both that and host were
specified --- though oddly, *not* if only hostaddr were specified.
Since this is directly contrary to the documentation, and not
backwards-compatible, it's clearly a bug.

However, the change wasn't totally without justification, even though it
wasn't done quite right, because the pre-v10 behavior has arguably been
buggy since we added hostaddr.  If hostaddr is specified and host isn't,
the pre-v10 code will search ~/.pgpass for "localhost", and ship that
password off to a server that most likely isn't local at all.  That's
unhelpful at best, and could be a security breach at worst.

Therefore, rather than just revert to that old behavior, let's define
the behavior as "search with host if provided, else with hostaddr if
provided, else search for localhost".  (As before, a host name matching
DEFAULT_PGSOCKET_DIR is replaced by localhost.)  This matches the
behavior of the actual connection code, so that we don't pick up an
inappropriate password; and it allows useful searches to happen when
only hostaddr is given.

While we're messing around here, ensure that empty elements within a
host or hostaddr list select the same behavior as a totally-empty
field would; for instance "host=a,,b" is equivalent to "host=a,/tmp,b"
if DEFAULT_PGSOCKET_DIR is /tmp.  Things worked that way in some cases
already, but not consistently so, which contributed to the confusion
about what key ~/.pgpass would get searched with.

Update documentation accordingly, and also clarify some nearby text.

Back-patch to v10 where the host/hostaddr list functionality was
introduced.

Discussion: https://postgr.es/m/30805.1532749137@sss.pgh.pa.us

diff --git a/doc/src/sgml/libpq.sgml b/doc/src/sgml/libpq.sgml
index d67212b8311d249663cf87762dc0c2f0e0746703..caab9700b86e8b67d0c9bde774d34a0a1714c749 100644 (file)
--- a/doc/src/sgml/libpq.sgml
+++ b/doc/src/sgml/libpq.sgml
@@ -938,8 +938,8 @@ postgresql://%2Fvar%2Flib%2Fpostgresql/dbname
      <para>
        If a password file is used, you can have different passwords for
        different hosts. All the other connection options are the same for every
-       host, it is not possible to e.g. specify a different username for
-       different hosts.
+       host in the list; it is not possible to e.g. specify different
+       usernames for different hosts.
      </para>
    </sect3>
   </sect2>
@@ -961,7 +961,7 @@ postgresql://%2Fvar%2Flib%2Fpostgresql/dbname
         name of the directory in which the socket file is stored.  If
         multiple host names are specified, each will be tried in turn in
         the order given.  The default behavior when <literal>host</literal> is
-        not specified is to connect to a Unix-domain
+        not specified, or is empty, is to connect to a Unix-domain
         socket<indexterm><primary>Unix domain socket</primary></indexterm> in
         <filename>/tmp</filename> (or whatever socket directory was specified
         when <productname>PostgreSQL</productname> was built). On machines without
@@ -969,7 +969,8 @@ postgresql://%2Fvar%2Flib%2Fpostgresql/dbname
        </para>
        <para>
         A comma-separated list of host names is also accepted, in which case
-        each host name in the list is tried in order. See
+        each host name in the list is tried in order; an empty item in the
+        list selects the default behavior as explained above. See
         <xref linkend="libpq-multiple-hosts"/> for details.
        </para>
       </listitem>
@@ -1020,14 +1021,17 @@ postgresql://%2Fvar%2Flib%2Fpostgresql/dbname
         </itemizedlist>
         Note that authentication is likely to fail if <literal>host</literal>
         is not the name of the server at network address <literal>hostaddr</literal>.
-        Also, note that <literal>host</literal> rather than <literal>hostaddr</literal>
+        Also, when both <literal>host</literal> and <literal>hostaddr</literal>
+        are specified, <literal>host</literal>
         is used to identify the connection in a password file (see
         <xref linkend="libpq-pgpass"/>).
        </para>
 
        <para>
         A comma-separated list of <literal>hostaddr</literal> values is also
-        accepted, in which case each host in the list is tried in order. See
+        accepted, in which case each host in the list is tried in order.
+        An empty item in the list causes the corresponding host name to be
+        used, or the default host name if that is empty as well. See
         <xref linkend="libpq-multiple-hosts"/> for details.
        </para>
        <para>
@@ -1047,9 +1051,12 @@ postgresql://%2Fvar%2Flib%2Fpostgresql/dbname
         name extension for Unix-domain
         connections.<indexterm><primary>port</primary></indexterm>
         If multiple hosts were given in the <literal>host</literal> or
-        <literal>hostaddr</literal> parameters, this parameter may specify a list
-        of ports of equal length, or it may specify a single port number to
-        be used for all hosts.
+        <literal>hostaddr</literal> parameters, this parameter may specify a
+        comma-separated list of ports of the same length as the host list, or
+        it may specify a single port number to be used for all hosts.
+        An empty string, or an empty item in a comma-separated list,
+        specifies the default port number established
+        when <productname>PostgreSQL</productname> was built.
        </para>
       </listitem>
      </varlistentry>
@@ -1683,6 +1690,17 @@ char *PQuser(const PGconn *conn);
 char *PQpass(const PGconn *conn);
 </synopsis>
       </para>
+
+      <para>
+       <function>PQpass</function> will return either the password specified
+       in the connection parameters, or if there was none and the password
+       was obtained from the <link linkend="libpq-pgpass">password
+       file</link>, it will return that.  In the latter case,
+       if multiple hosts were specified in the connection parameters, it is
+       not possible to rely on the result of <function>PQpass</function> until
+       the connection is established.  The status of the connection can be
+       checked using the function <function>PQstatus</function>.
+      </para>
      </listitem>
     </varlistentry>
 
@@ -7521,13 +7539,18 @@ myEventProc(PGEventId evtId, void *evtInfo, void *passThrough)
    used.  (Therefore, put more-specific entries first when you are using
    wildcards.) If an entry needs to contain <literal>:</literal> or
    <literal>\</literal>, escape this character with <literal>\</literal>.
-   A host name of <literal>localhost</literal> matches both TCP (host name
-   <literal>localhost</literal>) and Unix domain socket (<literal>pghost</literal> empty
-   or the default socket directory) connections coming from the local
-   machine. In a standby server, a database name of <literal>replication</literal>
+   The host name field is matched to the <literal>host</literal> connection
+   parameter if that is specified, otherwise to
+   the <literal>hostaddr</literal> parameter if that is specified; if neither
+   are given then the host name <literal>localhost</literal> is searched for.
+   The host name <literal>localhost</literal> is also searched for when
+   the connection is a Unix-domain socket connection and
+   the <literal>host</literal> parameter
+   matches <application>libpq</application>'s default socket directory path.
+   In a standby server, a database field of <literal>replication</literal>
    matches streaming replication connections made to the master server.
-   The <literal>database</literal> field is of limited usefulness because
-   users have the same password for all databases in the same cluster.
+   The database field is of limited usefulness otherwise, because users have
+   the same password for all databases in the same cluster.
   </para>
 
   <para>

diff --git a/src/interfaces/libpq/fe-connect.c b/src/interfaces/libpq/fe-connect.c
index bd7dac120d1fc44202a542afa6ddadd332defd12..4b35994394a426983b897f36209320a748d18e29 100644 (file)
--- a/src/interfaces/libpq/fe-connect.c
+++ b/src/interfaces/libpq/fe-connect.c
@@ -901,6 +901,8 @@ parse_comma_separated_list(char **startptr, bool *more)
 static bool
 connectOptions2(PGconn *conn)
 {
+       int                     i;
+
        /*
         * Allocate memory for details about each host to which we might possibly
         * try to connect.  For that, count the number of elements in the hostaddr
@@ -920,11 +922,10 @@ connectOptions2(PGconn *conn)
 
        /*
         * We now have one pg_conn_host structure per possible host.  Fill in the
-        * host details for each one.
+        * host and hostaddr fields for each, by splitting the parameter strings.
         */
        if (conn->pghostaddr != NULL && conn->pghostaddr[0] != '\0')
        {
-               int                     i;
                char       *s = conn->pghostaddr;
                bool            more = true;
 
@@ -933,8 +934,6 @@ connectOptions2(PGconn *conn)
                        conn->connhost[i].hostaddr = parse_comma_separated_list(&s, &more);
                        if (conn->connhost[i].hostaddr == NULL)
                                goto oom_error;
-
-                       conn->connhost[i].type = CHT_HOST_ADDRESS;
                }
 
                /*
@@ -948,7 +947,6 @@ connectOptions2(PGconn *conn)
 
        if (conn->pghost != NULL && conn->pghost[0] != '\0')
        {
-               int                     i;
                char       *s = conn->pghost;
                bool            more = true;
 
@@ -957,17 +955,9 @@ connectOptions2(PGconn *conn)
                        conn->connhost[i].host = parse_comma_separated_list(&s, &more);
                        if (conn->connhost[i].host == NULL)
                                goto oom_error;
-
-                       /* Identify the type of host. */
-                       if (conn->pghostaddr == NULL || conn->pghostaddr[0] == '\0')
-                       {
-                               conn->connhost[i].type = CHT_HOST_NAME;
-#ifdef HAVE_UNIX_SOCKETS
-                               if (is_absolute_path(conn->connhost[i].host))
-                                       conn->connhost[i].type = CHT_UNIX_SOCKET;
-#endif
-                       }
                }
+
+               /* Check for wrong number of host items. */
                if (more || i != conn->nconnhost)
                {
                        conn->status = CONNECTION_BAD;
@@ -979,29 +969,48 @@ connectOptions2(PGconn *conn)
        }
 
        /*
-        * If neither host or hostaddr options was given, connect to default host.
+        * Now, for each host slot, identify the type of address spec, and fill in
+        * the default address if nothing was given.
         */
-       if ((conn->pghostaddr == NULL || conn->pghostaddr[0] == '\0') &&
-               (conn->pghost == NULL || conn->pghost[0] == '\0'))
+       for (i = 0; i < conn->nconnhost; i++)
        {
-               Assert(conn->nconnhost == 1);
+               pg_conn_host *ch = &conn->connhost[i];
+
+               if (ch->hostaddr != NULL && ch->hostaddr[0] != '\0')
+                       ch->type = CHT_HOST_ADDRESS;
+               else if (ch->host != NULL && ch->host[0] != '\0')
+               {
+                       ch->type = CHT_HOST_NAME;
 #ifdef HAVE_UNIX_SOCKETS
-               conn->connhost[0].host = strdup(DEFAULT_PGSOCKET_DIR);
-               conn->connhost[0].type = CHT_UNIX_SOCKET;
+                       if (is_absolute_path(ch->host))
+                               ch->type = CHT_UNIX_SOCKET;
+#endif
+               }
+               else
+               {
+                       if (ch->host)
+                               free(ch->host);
+#ifdef HAVE_UNIX_SOCKETS
+                       ch->host = strdup(DEFAULT_PGSOCKET_DIR);
+                       ch->type = CHT_UNIX_SOCKET;
 #else
-               conn->connhost[0].host = strdup(DefaultHost);
-               conn->connhost[0].type = CHT_HOST_NAME;
+                       ch->host = strdup(DefaultHost);
+                       ch->type = CHT_HOST_NAME;
 #endif
-               if (conn->connhost[0].host == NULL)
-                       goto oom_error;
+                       if (ch->host == NULL)
+                               goto oom_error;
+               }
        }
 
        /*
         * Next, work out the port number corresponding to each host name.
+        *
+        * Note: unlike the above for host names, this could leave the port fields
+        * as null or empty strings.  We will substitute DEF_PGPORT whenever we
+        * read such a port field.
         */
        if (conn->pgport != NULL && conn->pgport[0] != '\0')
        {
-               int                     i;
                char       *s = conn->pgport;
                bool            more = true;
 
@@ -1065,8 +1074,8 @@ connectOptions2(PGconn *conn)
        }
 
        /*
-        * Supply default password if none given.  Note that the password might be
-        * different for each host/port pair.
+        * If password was not given, try to look it up in password file.  Note
+        * that the result might be different for each host/port pair.
         */
        if (conn->pgpass == NULL || conn->pgpass[0] == '\0')
        {
@@ -1089,20 +1098,16 @@ connectOptions2(PGconn *conn)
 
                if (conn->pgpassfile != NULL && conn->pgpassfile[0] != '\0')
                {
-                       int                     i;
-
                        for (i = 0; i < conn->nconnhost; i++)
                        {
                                /*
-                                * Try to get a password for this host from pgpassfile. We use
-                                * host name rather than host address in the same manner as
-                                * PQhost().
+                                * Try to get a password for this host from file.  We use host
+                                * for the hostname search key if given, else hostaddr (at
+                                * least one of them is guaranteed nonempty by now).
                                 */
-                               char       *pwhost = conn->connhost[i].host;
+                               const char *pwhost = conn->connhost[i].host;
 
-                               if (conn->connhost[i].type == CHT_HOST_ADDRESS &&
-                                       conn->connhost[i].host != NULL &&
-                                       conn->connhost[i].host[0] != '\0')
+                               if (pwhost == NULL || pwhost[0] == '\0')
                                        pwhost = conn->connhost[i].hostaddr;
 
                                conn->connhost[i].password =
@@ -6385,14 +6390,14 @@ passwordFromFile(const char *hostname, const char *port, const char *dbname,
 #define LINELEN NAMEDATALEN*5
        char            buf[LINELEN];
 
-       if (dbname == NULL || strlen(dbname) == 0)
+       if (dbname == NULL || dbname[0] == '\0')
                return NULL;
 
-       if (username == NULL || strlen(username) == 0)
+       if (username == NULL || username[0] == '\0')
                return NULL;
 
        /* 'localhost' matches pghost of '' or the default socket directory */
-       if (hostname == NULL)
+       if (hostname == NULL || hostname[0] == '\0')
                hostname = DefaultHost;
        else if (is_absolute_path(hostname))
 
@@ -6403,7 +6408,7 @@ passwordFromFile(const char *hostname, const char *port, const char *dbname,
                if (strcmp(hostname, DEFAULT_PGSOCKET_DIR) == 0)
                        hostname = DefaultHost;
 
-       if (port == NULL)
+       if (port == NULL || port[0] == '\0')
                port = DEF_PGPORT_STR;
 
        /* If password file cannot be opened, ignore it. */