Re: PATCH: hiding of specials and namerefs

X-seq: zsh-workers 55060

From: Philippe Altherr <philippe.altherr@xxxxxxxxx>

To: Bart Schaefer <schaefer@xxxxxxxxxxxxxxxx>

Cc: Zsh hackers list <zsh-workers@xxxxxxx>

Subject: Re: PATCH: hiding of specials and namerefs

Date: Tue, 4 Aug 2026 12:23:32 +0200

Arc-authentication-results: i=1; mx.google.com; arc=none

Arc-message-signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=arc-20260327; h=cc:to:subject:message-id:date:from:in-reply-to:references :mime-version:dkim-signature; bh=RWvlReBK3r+ey9GG+uNIhG7ZmCafaIU6YrDbUbS36WY=; fh=vuDAGjptCPc/P1/HEZ3j98yPJEW1XjuEoEAmmwDS/+U=; b=HyHN0Z1Q7VXepxPy3X+0vptrcqyAk3JvdPQPYIFeK3MyINBx8uHIRxb4tKT9Oa4nHM KoHvfvF0ILcGfcbsD5Uwe06QDm3rYU2E8dRZgBsNW0j3tdCVN8W1oJzJ+6PHUdtiw47Y 1oXs3tsrCz31IXd5QEuDJ9QQfwkbfdzwWU5VcdvTxLTF7NrtWyTnOurGyRXI1bi5RGlb 8KMxGqFhHCtkdsHQdQif8oboBRUxMlii8t/dEB0mvyr4Q0/yLiQpJ1WvpecGgqiN39fA fAki+/ZXssty3PEHaFyuezT2+A/qVwV6wNepU7OQb2+YV2+pbLf8P1lLNHYvxd2yKB/z Gsag==; darn=zsh.org

Arc-seal: i=1; a=rsa-sha256; t=1785839024; cv=none; d=google.com; s=arc-20260327; b=b5iHLj8Bd93Z6buyA5TtGhHymQ3BtT1tHcDlsm9LWeezHQL5S/RUXltmudhcv5pAtN sbZiVJJIYPpucpObMC2ADdZoJoxdHRvVaGUtnPTwc5rpUXP4DcO9dcLlEFSOgPSXiC83 xnH2Bq8jGPui7Xb4MR+0WQR/Zfdbyg10PM0Mvf4U02L4VxHjpJF2tJP9fwlRExtm2l4j x+5c7LK6mI64e6ixiOm2jB0Ty+Df6Wx8Pn4Mxa74eadxmni/aTOWKBYCTwx0UINt+Cqn D1TcbQ7LSFoF+Q0zEhn4CCtbg1K9t07BN59PXV9Tl0bzuYEn1q7uwAJyzjQkNgucgeUD S2sA==

Archived-at: <https://zsh.org/workers/55060>

In-reply-to: <CAH+w=7b1ABbVRAhtxW3iDtnh1g4Ea2mvjndy-aY2qAF-gv=cVA@mail.gmail.com>

List-id: <zsh-workers.zsh.org>

References: <CAH+w=7b1ABbVRAhtxW3iDtnh1g4Ea2mvjndy-aY2qAF-gv=cVA@mail.gmail.com>

+ if (usepm && (on & PM_HIDE) &&
+ (pm->node.flags & (PM_SPECIAL|PM_AUTOLOAD)) &&
+ pm->level <= locallevel) {

Since my first review, I was disturbed by the last condition. First I thought that it was wrong and instead should be "pm->level == locallevel" but no, we need the "<" in case one attempts a "typeset -gh" from a local scope. Now, it occurred to me that the condition isn't needed at all; it's always true because the parameter table is never supposed to contain a parameter whose level is higher than the locallevel. The condition should be dropped.

Philippe

diff --git a/Src/builtin.c b/Src/builtin.c index e35d7fe2c..55fa7ffb2 100644 --- a/Src/builtin.c +++ b/Src/builtin.c @@ -2210,8 +2210,7 @@ typeset_single(char *cname, char *pname, Param pm, int func, */ } if (usepm && (on & PM_HIDE) && - (pm->node.flags & (PM_SPECIAL|PM_AUTOLOAD)) && - pm->level <= locallevel) { + (pm->node.flags & (PM_SPECIAL|PM_AUTOLOAD))) { #if 0 zwarnnam(cname, "%s: can't change parameter attribute", pname); /* return NULL; */ /* this has always been a no-op */

Reverse Date

Date

Thread

Author