[comp.binaries.ibm.pc.d] Getting flex to compile on the PC

cs411134@umbc5.umbc.edu (Peter Johansson) (05/21/89)

Hello netfolk, I have a question, but first a bit of introduction...

Over the past several months, I have seen several requests for an
implementation of "lex" for the PC.  These requests have gone by
without notice, until recently.  (public notice, at least)

A response to a recent query (I'm editing out of rn, sorry this isn't
in the message chain) suggested a patch (posted earlier to c.b.i.p.)
to the flex sources posted to comp.sources.unix.  Though I didn't
happen to have that patch, I remembered seeing it on some anonymous
ftp, but since any informational material had been removed (i.e.  what
the patch was for) I trashed it.

Well, now that I know what it's for, I ftp it (from chyde.uwasa.fi,
btw) along with the origonal sources from uunet.uu.net.  Ok, I'm all
set to apply the patches and compile, right?  Wrong!

First off, there are two short patches to the origonal flex.
Unfortunately, our version of patch doesn't like them.  Here's what
happens: 

fubar> patch < patch.1
Hmm...  Looks like a new-style context diff to me...
The text leading up to this was:
--------------------------
|Greg Lee mentioned a bug regarding "\^" not correctly matching a caret
|in the input.  This is due to a throw-back to the dark ages, and the
|fix follows.
|
|revision 1.3 
|*** scan.l      Thu May  5 14:32:53 1988
|--- /tmp/da1563 Thu May  5 14:36:11 1988
--------------------------
Patching file scan.l using Plan A...
Hunk #1 failed at 48.
1 out of 1 hunks failed--saving rejects to scan.l.rej
Hmm...  Ignoring the trailing garbage.
done

Ok, the two patches (something similar happens on patch.2) are pretty
simple, so maybe I can just make the diffs manualy.  The second patch
(patch.2) looks fairly ovbious, but I'm not sure about the first.
Best not to edit indiscriminatly...

At this point I figure those patches weren't that important (for now,
at least) so now it's time to apply the MS-DOS patches.  The
"flex.fix" patch is actually a shell script that wants to run a
program called "patch2".  Of course, our patch isn't good enough for
patch2.  It appears that whatever patch program would work on
"flex.fix" would also work for the first set of patches, though at
this point, I wouldn't assume anything.

BTW, I'm doing all this on a VAX 3200 running Ultrix 3.0.  I assume
I'm using the distribution patch binary.

So, can anybody who has compiled flex on their PC offer any
suggestions?  It appears as if another patch program will do the
trick.  Is this so?

Any help graciously appreciated!  Post or Email your suggestions and
I'll summarize to the net.  If you too are looking for a FLEX.EXE, as
one nice packaged executable, just drop me a line and I'll mail it off
to you when I get it working (I'm optimistic!)

--
cs411134@umbc5.umbc.edu (for another week or so)
peter@umbc2.umbc.edu    (permanent I-net)
peter@umbc2.bitnet      (permanent Bitnet)