------------------------------------------------------------------------------
 HTMLsearch 0.09   ---   (C) Dec  6 2002 by ReallySoft
 ------------------------------------------------------------------------------
 Purpose: Search local HTML-trees and dump results as HTML to stdout
 Usage  : HTMLsearch <opts> <search> <index-dir> <rootDirList>

 <search>     ::= string to search for
 <index-dir>  ::= directory for search indices
 <rootDirList>::= Textfile containing directories (one per line) where search
                  should start. Prefixing a directory with - disables recursive
                  search into this directory

 <opts>::= -e<extensions> List of extensions treated as HTML files
                          [Default=.htm;.html;.sht;.shtm;.shtml;.html.en;.html.d
                          e]
           -v             be verboose.
           -i             ignore case.
           -x             never throw exception (write error to HTML code
                          instead)
           -L<len>        Maximum length of text per result [Default=300]
           -H<fontdef>    attribute for results ($$$ will be replaced by
                          searchstring) [Default=<b>$$$</b>]
           -T             highlight results in page titles?
           -S<size>       max. size of searched files (kb) [Default=150]
           -m<num>        max. number of results in output [Default=500]
           -n<id>[;<id>]* search only files which have <id> in their full
                          filename
           -D             Exchanges DOS-drive letters by small drive letters as
                          top-dir (done before -P). Example:
                          'C:/Test/Example.html' -> '/c/Test/Example.html'.
           -P<exchange>   Prefix to exchange at beginning of each result URL
                          (done before -p). Format of exchange: "old = new".
           -p<prefix>     Prefix to add to each result URL (you may i.e. use -p
                          "file://")
           -t<template>   How to display results ($xxx$ expressions get
                          replaced) [Default=<tr><td colspan=2><a
                          href="$url$">$title$</a> <font
                          size="-1">[$url$]</font></td></tr><tr><td
                          width=20%>&nbsp;</td><td>$text$<br>Score:
                          $score$</td></tr>]
           -d<title>      What to display, if no title exists in linked HTML
                          file [Default=Untitled]
           If you want to use HTMLsearch as cgi-tool you should experiment with
           -P and -p to create correct URLs
           -u             force index update.
           -b<kb>         do not add files bigger than <kb> kilobytes to index.
           -I<prefix>     Prefix to add to global index files. Use this to
                          create multiple indices in same index-directory. The
                          <rootDir>s used for different indices may overlap. In
                          this case word extraction is only done once. 

           Indexing mechanism:
           1. Create list of all HTML files.
           2. Create wordlist for each file (if not existing or HTML newer)
           3. Create global index file (describes occurances of words)
           
 ------------------------------------------------------------------------------
 Contact: HTMLsearch@ReallySoft.de
 ------------------------------------------------------------------------------

HTMLsearch: Help requested