BibGlimpse.README

# *************************************************************************** #
#  BibGlimpse, the Web Repository                                             #
# *************************************************************************** #
#
#
# If you want to use WebGlimpse as a scientific reprints repository
# follow these instructions installing Glimpse and WebGlimpse.
# Apart from two flags that need to be set during Glimpse installation,
# everything else is done after WebGlimpse has been successfully installed.
#
# (c) Boku Bioinformatics, Vienna, Austria
#     Thomas Tuechler 2006



# 0) Glimpse installation
# -----------------------------------------------------------------------------
# make sure that Glimpse is configured with the following two flags:

./configure --with-file-end-mark='\t' --enable-structured-queries;



# *************************************************************************** #
#  BibGlimpse specific modifications                                          #
# *************************************************************************** #
#
# you may do the BibGlimpse installation manually, by goint through the
# points below, or you may simply call the BibGlimpse.SETUP script to
# take you through the procedure:

./BibGlimpse.SETUP





xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Manual installation instructions:
xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx


# 1) some abbreviations
# -----------------------------------------------------------------------------
# We will refer to some WebGlimpse installation paths as follows:
#
# $WG  = WebGlimpse installation directory (eg. /usr/lib/wg2)
# $ARC = your archive directory (eg. /usr/lib/wg2/archives)
# $CGI = your cgi-bin directory (eg. /usr/lib/apache/cgi-bin/wg2)
#
# If you set these as variables in your shell, you can directly copy,
# paste the commands given here.



# 2) $WG/lib/(wr)usexpdf.sh
# -----------------------------------------------------------------------------
# Use the advanced pdf filtering tool wrusexpdf.sh instead of the old one:

mv $V $WG/lib/usexpdf.sh $WG/lib/old.usexpdf.sh;
mv $V $WG/lib/wrusexpdf.sh $WG/lib/usexpdf.sh;

# Adjust bin $PATH in usexpdf.sh to include the directory where the
# binaries for pdftotext, md5sum and so on can be found on your system.
# (here we change it to /bi/arch/bin):

# old:   export PATH=/usr/local/bin:$PATH;
# new:   export PATH=/bi/arch/bin:$PATH;

emacs $WG/lib/usexpdf.sh &



# 3) $WG/templates/.glimpse_filters
# -----------------------------------------------------------------------------
# Add the paths to usexpdf.sh to the .glimpse_filters template. This
# will make the program index pdf files as well.
# 

# new:   *.pdf	|WGHOME|/lib/usexpdf.sh
# new:   *.PDF	|WGHOME|/lib/usexpdf.sh

emacs $WG/templates/.glimpse_filters &



# 4) $WG/dist/wgfilter-index
# -----------------------------------------------------------------------------
# Adapt the filters such that it accepts *.pdf's by replacing and adding

# old:   Deny \.pdf$
# new:   Allow \.pdf$
# new:   Allow \.PDF$

# old:   Allow \..?html?$
# new:   Deny \.s?html?$

# new:   Deny \.pl$      
# new:   Deny \.sh$      
# new:   Deny \.tmp$     
# new:   Deny \.ag\.     
# new:   Deny \.bib$     
# new:   Deny \.medline$ 
# new:   Deny \.medl$    
# new:   Deny \.suppl$   
# new:   Deny \.pubmed   
# new:   Deny \.anno$    
# new:   Deny \.htaccess 
# new:   Deny README\.txt

emacs $WG/dist/wgfilter-index &



# 5) $WG/templates/wgreindex
# -----------------------------------------------------------------------------
# Enable structured queries with -s and -z flag in glimpseindex commands:

# old:   |CAT| |INDEXDIR|/.wg_toindex | |GLIMPSEIDX_LOC| -H |INDEXDIR| -t -o -h -n -X -U -C -F > /dev/null
# new:   |CAT| |INDEXDIR|/.wg_toindex | |GLIMPSEIDX_LOC| -H |INDEXDIR| -s -z -t -o -h -n -X -U -C -F > /dev/null

# old:   |CAT| |INDEXDIR|/.wg_toindex | |GLIMPSEIDX_LOC| -H |INDEXDIR| -t -o -h -n -X -U -C -F
# new:   |CAT| |INDEXDIR|/.wg_toindex | |GLIMPSEIDX_LOC| -H |INDEXDIR| -s -z -t -o -h -n -X -U -C -F

# ATTENTION: Take care that emacs does not introduced a new newline
#            in the middle of the glimpseindex command line!

emacs $WG/templates/wgreindex &



# 6) $WG/lib/wgHeader.pm
# -----------------------------------------------------------------------------
# Change the $FILE_END_MARK=" " definition in the $WG/lib/wgHeader.pm file
# to "\t". in accordance with the --file-end-mark='\t' flag used in the
# Glimpse installation. This will allow filenames with spaces.

# old:   $FILE_END_MARK=" ";
# new:   $FILE_END_MARK="\t";

# set a flag that we use WebGlimpse as a Web Repository:

# old:   $WRREPOS = 0;    # to use WebGlimpse as a scientific reprints repository [TT]
# new:   $WRREPOS = 1;    # to use WebGlimpse as a scientific reprints repository [TT]

emacs $WG/lib/wgHeader.pm &



# 7) $CGI/wrrepos.cgi
#    $CGI/wrsearch.cgi
# -----------------------------------------------------------------------------
# Specify path to webglimpse libraries (eg. to '/bi/arch/lib/wg2'):

# old:   my $WEBGLIMPSE_LIB='';
# new:   my $WEBGLIMPSE_LIB='/bi/arch/lib/wg2/lib';

emacs $CGI/wrrepos.cgi &
emacs $CGI/wrsearch.cgi &



# *************************************************************************** #
#  Begin of OPTIONAL modifications 8) - 10)                                   #
# *************************************************************************** #


# OPTIONAL 8) $CGI/logout/logout.cgi
# -----------------------------------------------------------------------------
# If you want to restrict access to your repository by means of .htaccess,
# create a logout directory within $CGI, that has a separate .htaccess
# file containing:
#
#  AuthName "Logout"
#  AuthUserFile /your/path/to/.htpasswd 
#  Require user logout
#
# which allows for the user 'logout' only. Do not forget to provide
# a user 'logout' with password logout in your .htpasswd file!
# All this provides a basic logout option using .htaccess user control.

mkdir $CGI/logout;
cat > "$CGI"/logout/logout.cgi << EOF
#!/usr/bin/perl -w
use strict;
use CGI;
use CGI::Carp qw(fatalsToBrowser);
my \$q = new CGI;
print \$q->header();
my $user = \$ENV\{REMOTE_USER\};
print
    "You are now logged out!<br />",
    "Current REMOTE_USER is set to \$user.<br />",
    "<br />",
    "(On some browsers you still need to close the browser,<br />",
    " to ensure that the previous pages can not be reloaded!)";
EOF
chmod +x "$CGI"/logout/logout.cgi


# OPTIONAL 9) $WG/templates/wgindex.html
#             $WG/templates/wgall.html	
#             $WG/templates/wgany.html
# ------------------------------------------------------------------------------
# Changes in the template html files allow you to customize the standard
# settings for the search results page upfront. We suggest to first
# adjust the matches per file, but feel free to make changes as you like:

#        <br>Maximum number of matches per file returned:
#        <SELECT NAME=maxlines>
# old:   <OPTION selected>10
# new:   <OPTION selected>1

# Modesty also for the maximum characteroutput per file:

#        <br>Maximum number of characters output per file:
# old:   <INPUT NAME="maxchars" VALUE=2000>
# new:   <INPUT NAME="maxchars" VALUE=250>

# Another make-it-look-nicer feature (for wgindex.html only):

# old:   <input type=checkbox name="sentence"> Try to print only sentences ...
# new:   <input type=checkbox name="sentence" value="on"> Try to print ...

emacs $WG/templates/wgindex.html &
emacs $WG/templates/wgall.html &
emacs $WG/templates/wgany.html &



# OPTIONAL 10) $WG/templates/wgoutput.cfg
# -----------------------------------------------------------------------------
# This is also an optional modification. If you want to modify the output,
# remove the limit lines reached line. Again, its up to you to choose the
# setting you like best. This should just point you to the user configurable 
# parts of the code:

# old:   lines_exceeded		<i>...limit of [MAXLINES] lines reached, additional matching lines are not shown...</i>
# new:   lines_exceeded		<i></i>

emacs $WG/templates/wgoutput.cfg &


# *************************************************************************** #
#  End of optional modifications, the following is required again...          #
# *************************************************************************** #



# 11) configure archive via webinterface
# -----------------------------------------------------------------------------
# configure archive with wgarcmin.cgi via the web
# with two archives as using YOUR settings
# 
# NOTE: one archive is the one to be accessed from the web all the time,
#       the other one is for stealthly reindexing in background.
#       check for correct permissions! Webuser must have full access
#       the directories to index!
# 
# Edit domain configureation:
#   Server Name: your.server.com
#   Document Root: /your/servers/htdocs
#   Port: xxxx
#   Protocol: http
#   User Dir: public_html
#   Directory Index Files: index.html index.htm index.shtml index.php
#   ScriptAlias & Extensions: \.php.?(\?|$) \/cgi
# 
# Add new Archive
#   Title: Your_prints_repository_main
#   Description: Prints repository
#   Index by Directory
#     Directory Path: /your/servers/htdocs/prints
#     Equivalent URL: http://your.server.com:xxxx/prints
#     Max #Pages: 99999
# 
# Add new Archive
#   Title: Your_prints_repository_stealth
#   Description: Prints repository
#   Index by Directory
#     Directory Path: /your/servers/htdocs/prints
#     Equivalent URL: http://your.server.com:xxxx/prints
#     Max #Pages: 99999
# 
# Build Indices for both archives



# 12) Reindexing in background
# -----------------------------------------------------------------------------
# From the above you should now have two directories in $ARC,
# namely $ARC/1 and $ARC/2.
# To keep the working archive $ARC/1 searchable while reindexing
# wrwgreindex will mirror $ARC/1 to $ARC/2, build the index there
# and then port the resulting index back to $ARC/1.
# To do so, we need to replace the standard wgreindex script in $ARC/1.
# Set $ARC to your full archive path and execute the commands below:
# Note that reindexing can also be called from Crontab, as described
# in the WebGlimpse documentation!

ARC_MAIN="$ARC/1";
ARC_STEALTH="$ARC/2";

cat > "$ARC_MAIN"/wgreindex << EOF
#!/bin/sh

# Since this is the visible archive we will not do any
# reindexing here. instead we will move the reindexing
# to the stealth twin of this archive.
# 
# (c) Boku Bioinformatics
#     Thomas Tuechler, David Kreil 2006
#

rm -f /tmp/.wg*;
"$ARC_MAIN"/wrwgreindex \$1 "$ARC_STEALTH";
EOF

chmod +x $ARC_MAIN/wgreindex;

# ATTENTION: Take care that sym-links are not interpolated:
#            /your/archives/1 must not become
#	     /your/somewhere/else/because/of/symlink/arch/...
#            or the 'sed' command in wrwgreindex will fail
#            and the abra retrieval in wrrepos.cgi will
#            silently produce crap!



# 13) final test
# -----------------------------------------------------------------------------
# Build Index from command line:

$ARC/1/wgreindex

# Test if archive is searchable from webinterface. Enjoy!