PDF doc search II

Introduction

I edited includes/SpecialSearch.php to make the saved PDFs searchable. This contribution is limited to a small amount of PDFs. When creating the PDFs don´t use compress text and images in e.g. acrobat distiller. SpecialSearch.php starts 2 AWK-Scripts. These scripts convert the PDFs into strings. These strings are searchable out of the wikisearchmask. Put the scripts searchAllPDFs.sh and searchPDF.sh in your mediawiki includes/ Folder. Tested under mediawiki 1.54 and LAMP.

see PDF doc search for the older not maintained code.

Authors

Bernd Flunger and Armin Lanzinger User:elai

Installation

1. In SpecialSearch.php copy and paste the contribution from Edit SpecialSearch.php.
2. You have to edit this line

   $output = shell_exec('includes/searchAllPDFs.sh -i '. $s . ' /home/xyz/mediawiki/images');
   (Note: change /home/xyz/mediawiki/images to your real server path.
   

3. Edit Spezial:Allmessages
4. In MediaWiki namespace you have to add 2 words:

   Pdfmatches, Noddfmatches
   http://xyz/mediawiki/index.php/MediaWiki:Nopdfmatches
   http://xyz/mediawiki/index.php/MediaWiki:Pdfmatches
   

   for English edit in Pdfmatches:
   Matches in PDFs
   in Nopdfmatches:
   No matches in PDFs
   

   for German edit in Pdfmatches:
   Übereinstimmungen in PDFs
   in Nopdfmatches:
   Keine Übereinstimmungen in PDFs
   

5. User rights

   The files searchAllPDFs.sh and searchPDF.sh have to be readable by your apache-user.
   

Edit SpecialSearch.php

Put the content after line 199:
(Note: Contribution starts with ### line)
(Note: replace www.xyz.com with with your domain and replace /home/xyz/ with your mediawiki path)

                if( $titleMatches ) {
                        if( $titleMatches->numRows() ) {
                                $wgOut->addWikiText( '==' . wfMsg( 'titlematches' ) . "==\n" );
                                $wgOut->addHTML( $this->showMatches( $titleMatches ) );



                        } else {
                                $wgOut->addWikiText( '==' . wfMsg( 'notitlematches' ) . "==\n" );
                        }
                }


######################################
######################################
#####
#####   contribution PDF doc search II
#####


                $s = $term;

                $output = shell_exec('includes/searchAllPDFs.sh -i '. $s . ' /home/xyz/mediawiki/images');
                $output_array = preg_split('/[\n\r]+/', $output);

                $n =  count($output_array);
                $n = $n - 1;

                $PDFi = 0;
                $realn = 0;


                while($PDFi<=$n) {
                        $output_array_short =  split("\.", $output_array[$i]);
                        $output_array_short[1] = $output_array_short[1].".".$output_array_short[2];



                        $filename = basename($output_array[$PDFi]);
                        $dirname = dirname($output_array[$PDFi]);


                        $archive = strpos($dirname, 'archiv');
                        $temp = strpos($dirname, 'temp');


                        if ( ($temp===false) && ($archive===false) ) {
                                $dirname = str_replace("/home/xyz/mediawiki/images", "", $dirname);
                                $wgOut->addHTML( "$ausgabe \n" );
#                               $ausgabe =  "<li>(<a href=\"http://www.xyz.com/mediawiki/index.php/Bild:$filename\">Beschreibung</a>) <a href=\"http://www.xyz.com/mediawiki/images$dirname/$filename\">$filename</a></li>";
                                $ausgabe_array[$realn] =  "<li>(<a href=\"http://www.xyz.com/mediawiki/index.php/Bild:$filename\">Beschreibung</a>) <a href=\"http://www.xyz.com/mediawiki/images$dirname/$filename\">$filename</a></li>";
                                $realn++;
                        }
                        $PDFi++;
                }

                $PDFi = 0;

                $realn--;
                $PDFi = 0;

                if ($realn>0) {
                        $wgOut->addWikiText( '==' . wfMsg( 'pdfmatches' ) . "==\n" );
                        $PDFsuchabfrage = "<p></p>\n Ihr Suchbegriff wurde <b>$realn</b> mal gefunden. Ihr Suchbegriff: <b>$s</b> ";
                        $wgOut->addHTML( "$PDFsuchabfrage<br /> " );
                } else {
                        $wgOut->addWikiText( '==' . wfMsg( 'nopdfmatches' ) . "==\n" );
                }

                while($PDFi<$realn) {
                        if($ol!=1) {
                                $wgOut->addHTML( " <ol start='1' style='none'>" );
                                $ol=1;
                        }
                        $wgOut->addHTML( "$ausgabe_array[$PDFi]\n" );
                        $PDFi++;
                }

                $wgOut->addHTML( "</ol>" );


#               $wgOut->addHTML( "<b>$PDFsearchstring</b><br /> " );


###############################
###############################


 

searchAllPDFs.sh

 
#!/bin/bash
#---------------------------------------------------
bindir=includes
#---------------------------------------------------
if [ "$1" = "-i" ]
then
        ic=$1
        shift
fi
#---------------------------------------------------
what=$1
if [ "$what" = "" ]
then
        echo ""
        echo "usage: $0 [-i] <what> [<where>]"
        echo "        -i      ... ignore case"
        echo "        <what>  ... what to search for"
        echo "        <where> ... where (directory) to search for"
        echo ""

        exit
fi
shift
#---------------------------------------------------
#curdir=$(pwd)
dir=$1
if [ "$dir" = "" ]
then
        dir="./"
fi
#---------------------------------------------------
find $dir -name "*pdf" -exec ${bindir}/searchPDF.sh $ic $what \{\} \; -print
#---------------------------------------------------
#cd $curdir
#---------------------------------------------------
 

searchPDF.sh

#!/bin/bash
#----------------------------------------------------------
if [ "$1" = "-i" ]
then
        ic=$1
        shift
fi
#----------------------------------------------------------
what=$1
file=$2
#----------------------------------------------------------
#strings $file | grep $ic $what |grep $ic -v "/$what" 1>/dev/null 2>&1
gawk 'BEGIN{
        n=0
}
{
        if($0 ~ ")Tj$"){
                n++
                txt=substr($0,2,length($0)-4)
                print txt
                next
        }
        if($0 ~ ")]TJ$"){
                n++
                #printf ">>>"$0"<<<"
                txt=substr($0,3,length($0)-6)
                ka="@klammer-auf@"
                kz="@klammer-zu@"
                kb="@backslash@"
                gsub(/\\\(/,ka,txt)
                gsub(/\\\)/,kz,txt)
                gsub(/\\\\/,kb,txt)
                gsub("[)][^(]+[(]","",txt)
                gsub(ka,"(",txt)
                gsub(kz,")",txt)
                gsub(kb,"\\",txt)
                print txt
                next
        }
}
END{
        if(n>0){
                exit 0
        }else{
                exit 1
        }
}' $file | grep $ic $what > /dev/null
#----------------------------------------------------------
if [ $? = 0 ]
then
        exit 0
else
        exit 1
fi
#----------------------------------------------------------