Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!rutgers!mailrus!cornell!uw-beaver!uw-june!uw-entropy!dataio!bright
From: bright@Data-IO.COM (Walter Bright)
Newsgroups: comp.binaries.ibm.pc.d
Subject: Re: WILDCARD
Keywords: wildcard
Message-ID: <1623@dataio.Data-IO.COM>
Date: 1 Aug 88 18:42:04 GMT
References: <11675@steinmetz.ge.com> <1622@dataio.Data-IO.COM> <546@utacs.UTA.FI>
Reply-To: bright@Data-IO.COM.Data-IO.COM (Walter Bright)
Organization: Data-IO Corporation; Redmond, WA
Lines: 16

In article <546@utacs.UTA.FI> av@utacs.UUCP (Arto Viitanen) writes:
<When I used my program (called do) to unpack ARC-files taken from net using
<command line
<	do *.arc pkxarc ^1
<first archivefile was handled ok, but after that system hang up. Finally I
<found the reason: ``do'' used DOS's FINDFIRST-function to parse wildcard
<and executed pxarc with first matching filename. But pkxarc created
<several new files, so directory no more matched the parsed wildcard, so
<next FINDNEXT-function searched for file, which wasn't in same place in
<directory, as it was with FINDFIRST.
<With ``for'' command, there is no such problems.

The WILDCARD program I wrote doesn't suffer from those problems either.
The reason is that *all* the wildcard expansion is done *prior* to executing
the command, so if the command changes the files or the directory, the
wildcard expansion is not affected.