Use mode "r" for popen() in psql's evaluate_backtick().

tglsfdc · tglsfdc · commit 504f963f76f0 · 2020-10-28T14:35:53.000-04:00
In almost all other places, we use plain "r" or "w" mode in popen() calls (the exceptions being for COPY data). This one has been overlooked (possibly because it's buried in a ".l" flex file?), but it's using PG_BINARY_R. Kensuke Okamura complained in bug #16688 that we fail to strip \r when stripping the trailing newline from a backtick result string. That's true enough, but we'd also fail to convert embedded \r\n cleanly, which also seems undesirable. Fixing the popen() mode seems like the best way to deal with this. It's been like this for a long time, so back-patch to all supported branches. Discussion: https://postgr.es/m/16688-c649c7b69cd7e6f8@postgresql.org
diff --git a/src/bin/psql/psqlscanslash.l b/src/bin/psql/psqlscanslash.l
@@ -754,7 +754,7 @@ evaluate_backtick(PsqlScanState state)
 
 	initPQExpBuffer(&cmd_output);
 
-	fd = popen(cmd, PG_BINARY_R);
+	fd = popen(cmd, "r");
 	if (!fd)
 	{
 		state->callbacks->write_error("%s: %s\n", cmd, strerror(errno));
@@ -795,7 +795,7 @@ evaluate_backtick(PsqlScanState state)
 	/* If no error, transfer result to output_buf */
 	if (!error)
 	{
-		/* strip any trailing newline */
+		/* strip any trailing newline (but only one) */
 		if (cmd_output.len > 0 &&
 			cmd_output.data[cmd_output.len - 1] == '\n')
 			cmd_output.len--;
