libtool
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: I think binary-branch is almost ready for play time


From: Bruce Korb
Subject: Re: I think binary-branch is almost ready for play time
Date: Sat, 12 May 2001 11:41:58 -0700

Alexandre Oliva wrote:
> 
> I've taken a quick look at the code.  I can't build ltmain: `autogen
> ltopts.def' fails saying it can't map data file `options'.  Did you
> forget to check something in?  Or is `options' generated?  Or part of
> some AutoGen release I don't have (I've got 5.0.9).

Run this command:

   autogen -L ${prefix}/share/autogen --save=$HOME/.autogenrc

where "${prefix}" is whereever you installed it.
5.0.9 should be recent enough.

> > Here is my hope and wish:
> > that this be used for all further development on the mlb.
> 
> I still have to be convinced it will be advantageous to use this
> instead of say m4 or sic.  I can't see how we're going to factor out
> multiple occurrences of certain constructs ($echo "Xsomething" |

The method is to put that stuff into the lt_mumble.def files.
>From there, it gets copied into the ltmain.in file and sucked
into the C program as well.  There is only one place to fix it,
unless you update ltmain.in.  Then, I would have to move fixes
into one of the lt_*.def files.  Ick.  At client-package-build time,
"configure" is run.  "configure" incorporates ltconfig.in.  I
modified that file to produce ltconfig.x thus:

Index: ltconfig.in
===================================================================
RCS file: /cvsroot/libtool/libtool/Attic/ltconfig.in,v
retrieving revision 1.246.2.58
diff -u -r1.246.2.58 ltconfig.in
--- ltconfig.in 2001/04/24 22:20:08     1.246.2.58
+++ ltconfig.in 2001/05/12 18:26:30
@@ -2489,6 +2489,33 @@

 if test -z "$tagname"; then
   echo '### END LIBTOOL CONFIG' >> "$ofile"
+  ( cat <<EOF
+/*
+ *  DO NOT EDIT THIS FILE - it was generated by
+ *  version $LTCONFIG_VERSION of the $PACKAGE configurator.
+ *
+ *  The shell script source for creating this is in ltconfig.in.
+ *  It gets incorporated into the `configure' script.
+ */
+#ifndef LIBTOOL_CONFIG_TEXT
+#define LIBTOOL_CONFIG_TEXT 1
+
+static const char z_ltconfig[] = "\\n\\
+EOF
+  sed -n '/### BEGIN LIBTOOL CONFIG/,/### END LIBTOOL CONFIG/p' "$ofile" |
\
+  egrep -v '^[         ]*($|#)' | \
+  sed 's/"/\\"/g;s/\\/\\\\/g;s/$/\\n\\/;$s/n\\$/n";/'
+  cat <<EOF
+
+const char* pz_ltconfig  = z_ltconfig;
+const char* pz_shell     = "$SHELL";
+const char* pz_vers      = "$VERSION";
+const char* pz_package   = "$PACKAGE";
+const char* pz_timestamp = " $TIMESTAMP";
+
+#endif /* LIBTOOL_CONFIG_TEXT */
+EOF
+  ) > ltconfig.x
 else
   echo "### END LIBTOOL TAG CONFIG: $tagname" >> "$ofile"
 fi

If I need more specific pieces of these assignments, I will
either emit it directly as with the pz_* variables at the bottom,
or write code to actually hunt through the "z_ltconfig" string.

> Please prove me wrong, Bruce.  Please, please! :-D

I hope this does so.  :-D

> Make sure you tag the binary-branch and the MLB so that we can easily
> identify and merge any changes, should they occur.

I don't know what this means.  I tagged the "binary-branch".
What more do I need do?

Thanks!
 - Bruce



reply via email to

[Prev in Thread] Current Thread [Next in Thread]