head	1.13;
access;
symbols;
locks
	ids:1.13; strict;
comment	@# @;


1.13
date	2000.07.08.18.25.25;	author ids;	state Exp;
branches;
next	1.12;

1.12
date	99.06.12.20.48.32;	author ids;	state Exp;
branches;
next	1.11;

1.11
date	99.06.12.20.25.44;	author ids;	state Exp;
branches;
next	1.10;

1.10
date	98.10.01.14.01.32;	author ids;	state Exp;
branches;
next	1.9;

1.9
date	98.08.28.18.59.10;	author ids;	state Exp;
branches;
next	1.8;

1.8
date	98.06.21.18.19.47;	author ids;	state Exp;
branches;
next	1.7;

1.7
date	96.01.19.16.02.54;	author ids;	state Exp;
branches;
next	1.6;

1.6
date	94.08.01.12.08.50;	author ids;	state Exp;
branches;
next	1.5;

1.5
date	94.07.31.19.07.17;	author ids;	state Exp;
branches;
next	1.4;

1.4
date	94.07.29.21.15.25;	author ids;	state Exp;
branches;
next	1.3;

1.3
date	94.07.29.17.23.02;	author ids;	state Exp;
branches;
next	1.2;

1.2
date	94.07.29.13.37.54;	author ids;	state Exp;
branches;
next	1.1;

1.1
date	94.07.28.14.24.53;	author ids;	state Exp;
branches;
next	;


desc
@@


1.13
log
@better choices of logical vs. physical HTML styles.
@
text
@#!/bin/sh -
#
#	wais_queries.cgi: given a query string of the form "whatever.src"
#	specifying a WAIS server, generates
#	an HTML page describing (and offering search facilities on) that server.
#
#	NOTICE!!! This does not work any more. The local listing of WAIS servers
#	is gone. But maybe the local listing of WAIS servers will be back
#	someday...or maybe this script will be useful to cannibalize for
#	other similar applications!
#
PATH=/usr/local/bin:/bin:/usr/bin export PATH
progname=`basename $0`
waisdir="/usr/local/lib/wais/wais-sources"
default_port_no=210	#WAIS default port

echo "Content-type: text/html"
echo ""

exec 2>&1		#we want to see any errors in the output!


cd $waisdir
#
#	Amazingly, on our Linux and FreeBSD machines, the Bourne shell
#	"cd" command doesn't cause exit on failure - it just barges on!
#	Naturally, all hell can break loose in a shell script which is
#	in a different directory from where it thinks it is.
#	So, to avoid such chaos, we check explicitly for failure here.
#
if [ $? -ne 0 ]; then
	echo "$progname: cannot find local listing of WAIS servers, aborting" >&2
	exit 1
fi

#
#	check query string has a meaningful syntax
#
if [ -z "`echo $QUERY_STRING | grep '\.src$'`" ]; then
	echo "$progname: query string must end in \".src\", aborting" >&2
	exit 1
fi

server_file="$QUERY_STRING"

#
#	Next, a security alert! (Exciting innit?)
#
#	CGI-scripts in DoC home directory areas now run as that user.
#	(So this one runs as me!) This rather dubious practice means
#	we have to be much more careful about exactly what's being requested.
#	So, we now check the purported server file pathname we've just
#	collected from the query string doesn't try to "wander" out of the
#	little bubble-world of the WAIS directory.
#
grepcount=`echo "$server_file" | egrep -c '/|\.\.'`
status=$?

if [ $status -ne 1 -o "x$grepcount" != "x0" ]; then
	echo "$progname: dodgy WAIS server file pathname detected, aborting" >&2
	exit 1
fi

#
#	now check the requested server is actually locally registered!
#
if [ ! -f "$server_file" ]; then
	echo "$progname: WAIS server $server_file not locally registered, aborting" >&2
	exit 1
fi

#
#	now attempt to collect access data from the server file.
#
wais_host=`grep '^[ 	]*:ip-name[ 	][ 	]*"..*"[ 	]*$' "$server_file" | sed -e 's/^[ 	]*:ip-name[ 	][ 	]*"//' -e 's/"[ 	]*$//'`
if [ -z "$wais_host" ]; then
	#try making do with ip-address field instead of ip-name field
	wais_host=`grep '^[ 	]*:ip-address[ 	][ 	]*"..*"[ 	]*$' "$server_file" | sed -e 's/^[ 	]*:ip-address[ 	][ 	]*"//' -e 's/"[ 	]*$//'`
fi
if [ -z "$wais_host" ]; then
	#still no luck---give up!
	echo "$progname: cannot find IP host address for WAIS server $server_file, aborting" >&2
	exit 1
fi

wais_port=`grep '^[ 	]*:tcp-port[ 	][ 	]*..*[ 	]*$' "$server_file" | sed -e 's/^[ 	]*:tcp-port[ 	][ 	]*//' -e 's/[ 	]*$//'`
if [ -z "$wais_port" ]; then
	#try making do with the default WAIS port number
	wais_port=$default_port_no
fi

database_name=`grep '^[ 	]*:database-name[ 	][ 	]*"..*"[ 	]*$' "$server_file" | sed -e 's/^[ 	]*:database-name[ 	][ 	]*"//' -e 's/"[ 	]*$//' -e 's|/|%2F|g'`
if [ -z "$database_name" ]; then
	echo "$progname: cannot find database name for WAIS server $server_file, aborting" >&2
	exit 1
fi


#
# HTML output (at last!)
# Note: the "isindex" in the form below has to be in small letters.
#
sed 's/^	//' << here
	<HTML>
	<HEAD>
	<TITLE>WAIS server $server_file</TITLE>
	</HEAD>

	<BODY BGCOLOR="#FFFFFF">

	<H1>WAIS server $server_file</H1>

	<STRONG>General warning:</STRONG> WAIS servers can be very slow.
	Even <EM>interrupting</EM> a WAIS server can be very slow!<P>

	<HR>

	<FORM ACTION="wais://$wais_host:$wais_port/$database_name">
	<B>This is a searchable index. Enter search keywords:</B>
	<INPUT NAME="isindex">
	</FORM>
here


#
# the host and description, from the server file
#
echo "<PRE>"
echo "Host is $wais_host"
sed -n -e 's/^[ 	]*:description/:description/' -e '/^:description/,$p' "$server_file"
echo "</PRE>"


#
# final courtesy link to WAIS welcome page
#
sed 's/^	//' << here
	<HR>

	<A HREF="wais_menu.cgi">
	<I>Back to the WAIS welcome page</I></A>

	</BODY>
	</HTML>
here
@


1.12
log
@whoops! missed one case of never having strings that could have dodgy meanings
 used as they stand as the first argument of "test" (alias "[").

(This is in fact a case of what I thought I'd dealt with in all outstanding
 contexts earlier, but I forgot CGI-scripts. The revision message from my
 earlier sweep is reproduced below.)

bloody hell! we need to change things like

        if [ "$<VAR>" = "<WHATEVER>" ]; then

 to:

        if [ "x$<VAR>" = "x<WHATEVER>" ]; then

 (or analogous changes which achieve the same protective goal)
 now that our current Linux version of "test", alias "[", has become
 too clever by half and allows (quote from manual page)
 "the special expression `-l STRING', which evaluates to the length of STRING".
 (If $<VAR> in the unprotected example above happens to be "-l",
  the next argument ("=" above) is taken to be a string whose length is to
  be measured, and the parsing gets totally screwed up.)

[Note: in some files being dealt with in this checking-in, this revision message
 is essentially the same as in one or more previous revision(s).
 That's simply because I missed a few cases first time(s) round!]
@
text
@d113 1
a113 1
	<B>General warning:</B> WAIS servers can be very slow.
d141 1
a141 1
	<EM>Back to the WAIS welcome page</EM></A>
@


1.11
log
@careful scrutiny of purported server file pathname:

> #
> #     Next, a security alert! (Exciting innit?)
> #
> #     CGI-scripts in DoC home directory areas now run as that user.
> #     (So this one runs as me!) This rather dubious practice means
> #     we have to be much more careful about exactly what's being requested.
> #     So, we now check the purported server file pathname we've just
> #     collected from the query string doesn't try to "wander" out of the
> #     little bubble-world of the WAIS directory.
> #

Plus similarly careful handling of strings that could have dodgy meanings
 in various contexts: we take care to protect against this by things like
 having them in quotes, never having them as they stand as the first
 argument of "test" (alias "["), and so on.
@
text
@d76 1
a76 1
if [ "$wais_host" = "" ]; then
@


1.10
log
@simpler, more minimalist replacement for the now-defunct
 "lab basic environment": we now just set the path.
@
text
@d39 1
a39 1
if [ "`echo $QUERY_STRING | grep '\.src$'`" = "" ]; then
d47 18
d67 1
a67 1
if [ ! -f $server_file ]; then
d75 1
a75 1
wais_host=`grep '^[ 	]*:ip-name[ 	][ 	]*"..*"[ 	]*$' $server_file | sed -e 's/^[ 	]*:ip-name[ 	][ 	]*"//' -e 's/"[ 	]*$//'`
d78 1
a78 1
	wais_host=`grep '^[ 	]*:ip-address[ 	][ 	]*"..*"[ 	]*$' $server_file | sed -e 's/^[ 	]*:ip-address[ 	][ 	]*"//' -e 's/"[ 	]*$//'`
d80 1
a80 1
if [ "$wais_host" = "" ]; then
d86 2
a87 2
wais_port=`grep '^[ 	]*:tcp-port[ 	][ 	]*..*[ 	]*$' $server_file | sed -e 's/^[ 	]*:tcp-port[ 	][ 	]*//' -e 's/[ 	]*$//'`
if [ "$wais_port" = "" ]; then
d92 2
a93 2
database_name=`grep '^[ 	]*:database-name[ 	][ 	]*"..*"[ 	]*$' $server_file | sed -e 's/^[ 	]*:database-name[ 	][ 	]*"//' -e 's/"[ 	]*$//' -e 's|/|%2F|g'`
if [ "$database_name" = "" ]; then
d130 1
a130 1
sed -n -e 's/^[ 	]*:description/:description/' -e '/^:description/,$p' $server_file
@


1.9
log
@added explicit exit status check on cd:

#
#       Amazingly, on our Linux and FreeBSD machines, the Bourne shell
#       "cd" command doesn't cause exit on failure - it just barges on!
#       Naturally, all hell can break loose in a shell script which is
#       in a different directory from where it thinks it is.
#       So, to avoid such chaos, we check explicitly for failure here.
#
@
text
@d12 1
a20 1
. /vol/lab/basicenv.sh	#use the standard lab environment.
@


1.8
log
@layout, comment, and HTML standardization improvements.
@
text
@d23 9
a31 1
if [ ! -d $waisdir ]; then
a34 2

cd $waisdir
@


1.7
log
@added comment saying it doesn't work any more!
@
text
@d80 2
d83 3
d101 1
d103 1
a103 1

d110 1
d112 1
a112 1

d118 3
@


1.6
log
@slightly more readable syntax in the form.
@
text
@d7 5
@


1.5
log
@much nicer arrangement, using a form with an action and an "isindex"
printed host as well as description, now that you can't see the host by
 "sitting on the link" (which no longer exists, explicitly anyway!)
corrected a comment
less verbose, less cluttered layout.
@
text
@d86 1
a86 1
	<INPUT NAME=isindex>
@


1.4
log
@proper handling of WAIS servers with slashes in their database names
nicer delimiting of description in layout.
@
text
@d5 1
a5 1
#	an HTML page describing (and offering a link to) that server.
d70 1
d72 2
a73 1

d79 2
d82 1
a82 1
	<UL>
d84 4
a87 17
	<LI>
	<B>Remember:</B> the server
	will simply expect a search string from you; it will <EM>not</EM>
	offer you a home page!

	<LI>
	<B>GENERAL WARNING:</B> WAIS servers can be very slow.
	Even <EM>interrupting</EM> a WAIS server can be very slow!

	<LI>
	<A HREF="wais://$wais_host:$wais_port/$database_name">
	<B>Here's the server.</B></A>

	</UL>


	<HR>
d91 1
a91 1
# the contents of the server file
d94 1
@


1.3
log
@further shortening and simplifying of whole layout.
@
text
@d63 1
a63 1
database_name=`grep '^[ 	]*:database-name[ 	][ 	]*"..*"[ 	]*$' $server_file | sed -e 's/^[ 	]*:database-name[ 	][ 	]*"//' -e 's/"[ 	]*$//'`
d94 3
@


1.2
log
@cleared away technical crap in description section
new cleaner, less cluttered layout.
@
text
@d82 1
a82 1
	itself will simply expect a search string from you; it will <EM>not</EM>
d89 1
a89 6
	</UL>


	<HR>

	<B>The server itself:</B>
d91 1
a91 2
	WAIS server $server_file.</A>
	<EM>Here it is!</EM>
d93 1
a93 7
	<HR>


	<UL>

	<LI>
	<B>Description:</B>
d100 1
a100 1
sed -n '/:description/,$p' $server_file
a106 3
	</UL>


@


1.1
log
@Initial revision
@
text
@d77 4
a83 1
	<P>
d85 3
d89 1
a89 1
	<HR>
a90 1
	<H2>The server itself</H2>
d92 1
a92 1
	<UL>
d94 1
a94 1
	<LI>
a98 7
	<LI>
	<B>GENERAL WARNING:</B> WAIS servers can be very slow.
	Even <EM>interrupting</EM> a WAIS server can be very slow!

	</UL>


a100 1
	<H2>A description of the server</H2>
d105 1
a105 3
	The stuff from the label ":description" onwards is usually the most
	useful part.
	The earlier stuff is technical details, included for completeness.
d112 1
a112 1
cat $server_file
@