Re: capturing output of !! not working

[Peter Stephenson <p.stephenson@samsung.com>]

On Wed, 18 Mar 2015 21:49:37 -0400
Vin Shelton <acs@alumni.princeton.edu> wrote:
> $ ls -l $(!!)
> ls -l $()

This is due to special code for handling aliases inside command
substitutions that got applied too widely because the flag that
indicates aliases also indicates history expansion; however, there's a
discriminator in another bit.  I suspect restricting application of
INP_ALIAS is a better fix overall, but it needs a lot more code tracking
down plus safety checks.

Another case where we need a framework for testing interactive shells.

pws

diff --git a/Src/hist.c b/Src/hist.c
index aa07ce8..b7ef522 100644
--- a/Src/hist.c
+++ b/Src/hist.c
@@ -338,7 +338,8 @@ ihwaddc(int c)
 	 * fashion as we never need the expansion in the history
 	 * line, only in the lexer and above.
 	 */
-	!((histactive & HA_INWORD) && (inbufflags & INP_ALIAS))) {
+	!((histactive & HA_INWORD) &&
+	  (inbufflags & (INP_ALIAS|INP_HIST)) == INP_ALIAS)) {
 	/* Quote un-expanded bangs in the history line. */
 	if (c == bangchar && stophist < 2 && qbang)
 	    /* If qbang is not set, we do not escape this bangchar as it's *
@@ -901,7 +902,8 @@ ihungetc(int c)
 	    zlemetall--;
 	    exlast++;
 	}
-	if (!(histactive & HA_INWORD) || !(inbufflags & INP_ALIAS)) {
+	if (!(histactive & HA_INWORD) ||
+	    (inbufflags & (INP_ALIAS|INP_HIST)) != INP_ALIAS) {
 	    DPUTS(hptr <= chline, "BUG: hungetc attempted at buffer start");
 	    hptr--;
 	    DPUTS(*hptr != (char) c, "BUG: wrong character in hungetc() ");