Add a texinfo file for the download, build and install directions and

use this to replace README.Install and software-instructions.html.

Add the Makefile, filter and README files to support generating text
and html output from texinfo.

The advantage of texinfo is that it is possible to produce text, html
and latex output (almost anything) from a single source.

Remove spack/README.spack, superseded by top-level README.Install.

Author: mwkrentel

README.Install

@@ -31,6 +31,14 @@ directory used by latex. On Linux platfom, it's usually at
 
 /home/<user>/texmf/tex/latex/<package>
 
+---------------------------------
+Texinfo directory and html files
+---------------------------------
+
+The source for www/software-instructions.html is now in texinfo format
+in the texinfo directory.  See the README file in that directory to
+see how to edit that file.
+
 -----------------------------
 WWW directory and html files
 -----------------------------

doc/README

@@ -0,0 +1,46 @@
+#
+#  Make html and text files from texinfo.
+#
+#  make -- use makeinfo to convert .texi to html and text.
+#  make install -- copy .html files to ../www/
+#
+
+MAKEINFO = makeinfo
+
+BASE_FILES = software-instructions
+
+README = README.Install
+README_SOURCE = software-instructions.texi
+
+HTML_FILES = $(addsuffix .html,$(BASE_FILES))
+TEMP_FILES = $(addsuffix .temp,$(BASE_FILES))
+
+FILTER = ./filter-html.pl
+
+#------------------------------------------------------------
+
+.PHONY: all install clean
+
+# Don't automatically delete intermediate files.
+.SECONDARY: $(TEMP_FILES)
+
+all: $(HTML_FILES) $(README)
+
+install: all
+	for f in $(HTML_FILES) ; do  \
+		cp -f "$$f" "../www/$$f" ;  \
+	done
+	cp -f $(README) ../../$(README)
+
+$(README): $(README_SOURCE)
+	$(MAKEINFO) --plaintext  $< -o $@
+
+%.temp: %.texi
+	$(MAKEINFO) --html --no-split $< -o $@
+
+%.html: %.temp
+	$(FILTER) $< >$@
+
+clean:
+	rm -f $(TEMP_FILES) $(HTML_FILES) $(README)
+

doc/texinfo/Makefile

@@ -0,0 +1,36 @@
+This directory replaces some of the *.html.in files in ../www/ with
+texinfo files.  For now, only the build and install directions from
+software-instructions.html.  The advantage of texinfo is that it is
+possible to produce text, html and latex output (almost anything) from
+a single source.  This works well for directions that are mostly text.
+
+Requirements: texinfo (makeinfo) version 6.x or later, and perl
+version 5.x or later.  Texinfo is available as a spack package and
+builds simply and robustly with no prerequisites other than perl.
+
+Update the .html files in two phases.  First, outside of the normal
+build for hpctoolkit, edit the .texi files in this directory and use
+'make install' to copy them to the ../www/ directory.  Then, the
+normal 'make install' for hpctoolkit will copy them to the install
+prefix (in share/doc/hpctoolkit).
+
+  1. edit the .texi files with a text editor.
+
+  2. make (in this directory) -- this uses makeinfo to convert .texi
+  to html.
+
+  3. make install (this directory) -- this copies the .html files to
+  the ../www/ directory.
+
+  4. make install (for hpctoolkit) -- this copies the .html files to
+  the install prefix.
+
+Note: you don't need to rerun hpctoolkit configure after editing the
+texinfo files, just make install.
+
+Note: only developers editing the documentation ever need to run these
+steps.  The .html files are generated from texinfo and we keep the
+generated version in the repository.  This means that normal users who
+clone hpctoolkit already have the generated files and 'make install'
+for hpctoolkit will copy them to the install directory.
+

doc/texinfo/README

@@ -0,0 +1,88 @@
+#!/usr/bin/env perl
+#
+#  Read the raw html from texinfo on stdin and print the modified html
+#  for hpctoolkit on stdout.
+#
+#  ./filter-html.pl  <html-from-texinfo  >html-for-hpctoolkit
+#
+
+open(PART1, "<part1") || die "unable to open: part1";
+
+open(HEADER, "<../www/style/header-hpctoolkit.html")
+    || die "unable to open: header-hpctoolkit.html";
+
+open(FOOTER, "<../www/style/footer-hpctoolkit.html")
+    || die "unable to open: footer-hpctoolkit.html";
+
+#------------------------------------------------------------
+
+# 1 -- <?xml>, <!DOCTYPE>, <head> from part1 file.
+
+while ($_ = <PART1>) {
+    print $_;
+}
+
+#------------------------------------------------------------
+
+# 2 -- <title> from texinfo (stdin), skip lines before that.
+
+while ($_ = <>) {
+    if (m/<title/) {
+	print "  $_";
+	last;
+    }
+}
+
+#------------------------------------------------------------
+
+# 3 -- end </head>, begin <body>
+
+print "</head>\n\n<body>\n\n";
+
+#------------------------------------------------------------
+
+# 4 -- hpctoolkit header
+
+while ($_ = <HEADER>) {
+    print $_;
+}
+print "\n";
+
+#------------------------------------------------------------
+
+# 5 -- main body from texinfo, starting with <h1>, up to </body>.
+
+while ($_ = <>) {
+    if (m/<body/) {
+	last;
+    }
+}
+
+while ($_ = <>) {
+    if (m/<h/) {
+	print $_;
+	last;
+    }
+}
+
+while ($_ = <>) {
+    if (m|</body|) {
+	last;
+    }
+    print $_;
+}
+
+#------------------------------------------------------------
+
+# 6 -- hpctoolkit footer
+
+while ($_ = <FOOTER>) {
+    print $_;
+}
+
+#------------------------------------------------------------
+
+# 7 -- end </body> and </html>
+
+print "\n</body>\n</html>\n";
+

doc/texinfo/filter-html.pl

@@ -0,0 +1,12 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!-- -*- coding: utf-8 -*- -->
+
+<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"
+	  "http://www.w3.org/TR/html4/loose.dtd">
+
+<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
+
+<head>
+  <!-- meta http-equiv="Content-Type" content="text/html; charset=utf-8" -->
+  <link rel="stylesheet" type="text/css" href="style/style.css" />
+