Make psql handle EOF during COPY FROM STDIN properly on all platforms.

When stdin is a terminal, it's possible to end a COPY FROM STDIN with
a keyboard EOF signal (typically control-D), and then keep on issuing
SQL commands.  One would expect another COPY FROM STDIN to work as well,
but on some platforms it did not.  This turns out to be because we were
not resetting the stream's feof() flag, and BSD-ish versions of fread()
and fgets() won't attempt to read more data if that's set.

The misbehavior is observed on BSDen (including macOS), but not Linux,
Windows, or SysV-ish Unixen, which makes this a portability bug not
just a missing feature.

Add a clearerr() call to fix the behavior, and improve the prompt that's
issued when copying from a TTY to mention that EOF signals work.

It's been like this forever, so back-patch to all supported branches.

Thomas Munro

Discussion: https://postgr.es/m/CAEepm=0MCGfYf=JAMiYhO6JPtv9-3ZfBo8fcGeCZ8oMzaw+Z+Q@mail.gmail.com

Author: tglsfdc

src/bin/psql/copy.c

@@ -452,7 +452,7 @@ handleCopyIn(PGconn *conn, FILE *copystream, bool isbinary)
 	{
 		if (!pset.quiet)
 			puts(_("Enter data to be copied followed by a newline.\n"
-				   "End with a backslash and a period on a line by itself."));
+				   "End with a backslash and a period on a line by itself, or an EOF signal."));
 		prompt = get_prompt(PROMPT_COPY);
 	}
 	else
@@ -573,6 +573,16 @@ handleCopyIn(PGconn *conn, FILE *copystream, bool isbinary)
 
 copyin_cleanup:
 
+	/*
+	 * Clear the EOF flag on the stream, in case copying ended due to an EOF
+	 * signal.  This allows an interactive TTY session to perform another COPY
+	 * FROM STDIN later.  (In non-STDIN cases, we're about to close the file
+	 * anyway, so it doesn't matter.)  Although we don't ever test the flag
+	 * with feof(), some fread() implementations won't read more data if it's
+	 * set.  This also clears the error flag, but we already checked that.
+	 */
+	clearerr(copystream);
+
 	/*
 	 * Check command status and return to normal libpq state.
 	 *