<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<title>RECOLL: a personal text search system for
Unix/Linux</title>
<meta name="generator" content="HTML Tidy, see www.w3.org">
<meta name="Author" content="Jean-Francois Dockes">
<meta name="Description" content=
"recoll is a simple full-text search system for unix and linux based on the powerful and mature xapian engine">
<meta name="Keywords" content=
"full text search,fulltext,desktop search,unix,linux,solaris,open source,free">
<meta http-equiv="Content-language" content="en">
<meta http-equiv="content-type" content=
"text/html; charset=iso-8859-1">
<meta name="robots" content="All,Index,Follow">
<link type="text/css" rel="stylesheet" href="styles/style.css">
</head>
<body>
<div class="rightlinks">
<ul>
<li><a href="index.html">Home</a></li>
<li><a href="pics/index.html">Screenshots</a></li>
<li><a href="download.html">Downloads</a></li>
<li><a href="usermanual/index.html">User manual</a></li>
<li><a href="index.html#support">Support</a></li>
<li><a href="devel.html">Development</a></li>
</ul>
</div>
<div class="content">
<h1 class="intro">Recoll features</h1>
<table width=100%>
<tbody>
<tr>
<td><a href="#systems">Supported systems</a></td>
<td><a href="#doctypes">Document types</a></td>
<td><a href="#other">Other features</a></td>
<td><a href="#integration">Desktop and web integration</a></td>
<td><a href="#stemming">Stemming</a></td>
</tr>
</tbody>
</table>
<h2><a name="systems">Supported systems</a></h2>
<p><span class="application">Recoll</span> has been compiled
and tested on FreeBSD, Linux, Darwin and Solaris (initial
versions FreeBSD 5, Redhat 7, Fedora Core 5, Suse 10, Gentoo,
Debian 3.1, Solaris 8). It should compile and run on all
subsequent releases of these systems and probably a few
others too.</p>
<p>Qt versions from 3.1 to 4.7</p>
<h2><a name="doctypes">Document types</a></h2>
<p>Recoll can index many document types (along with their
compressed versions). Some types are handled internally (no
external application needed). Other types need a separate
application to be installed to extract the text. Types that
only need very common utilities (awk/sed/groff etc.) are
listed in the native section.</p>
<h4>File types indexed natively</h4>
<ul>
<li><span class="literal">text</span>.</li>
<li><span class="literal">html</span>.</li>
<li><span class="literal">maildir</span> and <span class=
"literal">mailbox</span> (<span class=
"literal">Mozilla</span>, <span class=
"literal">Thunderbird</span> and <span class=
"literal">Evolution</span> mail ok).</li>
<li><span class="literal">gaim</span> and <span class=
"literal">purple</span> log files.</li>
<li><span class="literal">Lyx</span> files (needs <span
class="literal">Lyx</span> to be installed).</li>
<li><span class="literal">Scribus</span> files.</li>
<li><span class="literal">Man pages</span> (need <span
class="command">groff</span>).</li>
</ul>
<h4>File types indexed with external helpers</h4>
<p>Many document types need the <span class="command">iconv</span>
command in addition to the applications specifically listed.</p>
<h5>The XML ones</h5>
<p>The following types need <span class=
"command">xsltproc</span> from the <b>libxslt</b> package.
Quite a few also need <span class="command">unzip</span>:</p>
<ul>
<li><span class="literal">Abiword</span> files.</li>
<li><span class="literal">Fb2</span> ebooks.</li>
<li><span class="literal">Kword</span> files.</li>
<li><span class="literal">Microsoft Office Open XML</span>
files.</li>
<li><span class="literal">OpenOffice</span> files.</li>
<li><span class="literal">SVG</span> files.</li>
</ul>
<h5>Other formats</h5>
<ul>
<li><span class="literal">pdf</span> with the <span class=
"command">pdftotext</span> command, which can be installed
as part of <a href="http://www.foolabs.com/xpdf/">xpdf</a>
or <a href="http://poppler.freedesktop.org/">poppler</a>,
depending on your distribution.</li>
<li><span class="literal">msword</span> with <a href=
"http://www.winfield.demon.nl/">antiword</a>. It is also useful to
have <a href="http://wvware.sourceforge.net/">wvWare</a> installed
as it may be be used as a fallback for some files which antiword
does not handle.</li>
<li><span class="literal">Powerpoint</span> and <span
class="literal">Excel</span> with the <a href=
"http://catdoc.klik.atekon.de">catdoc</a> utilities.</li>
<li><span class="literal">CHM (Microsoft help)</span> files
with <span class="command">Python, <a href="http://gnochm.sourceforge.net/pychm.html">pychm</a>
and <a href="http://www.jedrea.com/chmlib/">chmlib</a></span>.</li>
<li><span class="literal">GNU info</span> files
with <span class="command">Python</span> and the
<span class="command">info</span> command.</li>
<li><span class="literal">Zip</span> archives (needs <span
class="command">Python</span>).</li>
<li><span class="literal">Rar</span> archives (needs <span
class="command">Python</span>), the
<a href="http://pypi.python.org/pypi/rarfile/">rarfile</a> Python
module and the <a
href="http://www.rarlab.com/rar_add.htm">unrar</a> utility.</li>
<li><span class="literal">iCalendar</span>(.ics) files
(needs <span class="command">Python, <a href=
"http://pypi.python.org/pypi/icalendar/2.1">icalendar</a></span>).</li>
<li><span class="literal">Mozilla calendar data</span> See
<a href=
"http://bitbucket.org/medoc/recoll/wiki/IndexMozillaCalendari">
the wiki</a> about this.</li>
<li><span class="literal">Wordperfect</span> with the
<span class="command">wpd2html</span> command from <a href=
"http://libwpd.sourceforge.net">libwpd</a>. On some distributions,
the command may come with an package named <span
class="literal">libwpd-tools</span> or such, not the base <a
span="literal">libwpd</a> package.</li>
<li><span class="literal">postscript</span> with <a href=
"http://www.gnu.org/software/ghostscript/ghostscript.html">
ghostscript</a> and <a href=
"http://www.cs.wisc.edu/~ghost/doc/pstotext.htm">pstotext</a>.
Actually the pstotext 1.9 found at the latter link has a
problem with file names using special shell characters, and
you should either use the version packaged for your system
which is probably patched, or apply the Debian patch which
is stored <a href=
"files/pstotext-1.9_4-debian.patch">here</a> for
convenience. See
http://packages.debian.org/squeeze/pstotext and
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=356988 for
references/explanations.</li>
<li><span class="literal">RTF</span> files with <a href=
"http://www.gnu.org/software/unrtf/unrtf.html">unrtf</a>. Please
note that up to version
0.21, <span class="command">unrtf</span> mostly does not work
with non western-european character sets. If you have a need
for indexing, ie, russian or chinese RTF files, I have
produced a modified version which works much better (as
indicated by my tests and a few external ones). You can
download the <a href="unrtf/unrtf-0.22.2beta.tar.gz">source
here</a>. The development is hosted
on <a href="http://www.bitbucket.org/medoc/unrtf-int">
bitbucket.org</a>.</li>
<li><span class="literal">TeX</span> with <span class=
"command">untex</span>. If there is no untex package for
your distribution, <a href="untex/untex-1.3.jf.tar.gz">a
source package is stored on this site</a> (as untex has no
obvious home). Will also work with <a href=
"http://www.cs.purdue.edu/homes/trinkle/detex/">detex</a>
if this is installed.</li>
<li><span class="literal">dvi</span> with <a href=
"http://www.radicaleye.com/dvips.html">dvips</a>.</li>
<li><span class="literal">djvu</span> with <a href=
"http://djvu.sourceforge.net">DjVuLibre</a>.</li>
<li>Audio file tags: Recoll releases 1.13 and older use <a
href="http://id3lib.sourceforge.net/">id3info (id3lib)</a>
(compiling id3lib on recent systems may need a small patch,
see <a href="id3lib.html">here.</a>) or the ogg and flac
tools.<br>
Recoll releases 1.14 and later use a Python filter based
on <a href="http://code.google.com/p/mutagen/">mutagen</a>
for all audio types.</li>
<li>Image file tags with <a href=
"http://www.sno.phy.queensu.ca/~phil/exiftool/">exiftool</a>.
This is a perl program, so you also need perl on the
system. This works with about any possible image file and
tag format (jpg, png, tiff, gif etc.).</li>
<li>Midi karaoke files with Python, the
<a href="http://pypi.python.org/pypi/midi/0.2.1">
midi module</a>, and some help
from <a href="http://chardet.feedparser.org/">chardet</a>. There
is probably a <tt>chardet</tt> package for your distribution,
but you will quite probably need to build the midi
package. This is easy but see the
to <a href="helpernotes.html#midi">notes here</a>.
</li>
<li>Konqueror webarchive format with Python (uses the tarfile
module).</li>
<li>mimehtml web archive format (support based on the mail
filter, which introduces some mild weirdness, but still
usable).</li>
</ul>
<h2><a name="other">Other features</a></h2>
<ul>
<li>Can use <b>Beagle</b> browser plug-ins to index web
history. See the <a href=
"http://bitbucket.org/medoc/recoll/wiki/IndexBeagleWeb">the
Wiki</a> for more detail.</li>
<li>Processes all email attachments, and more generally any
realistic level of container imbrication (the "msword attachment to
a message inside a mailbox in a zip" thingy...) .</li>
<li>Multiple selectable databases.</li>
<li>Powerful query facilities, with boolean searches,
phrases, filter on file types and directory tree.</li>
<li>Xesam-compatible query language.</li>
<li>Wildcard searches (with a specific and faster function
for file names).</li>
<li>Support for multiple charsets. Internal processing and
storage uses Unicode UTF-8.</li>
<li><a href="#Stemming">Stemming</a> performed at query
time (can switch stemming language after indexing).</li>
<li>Easy installation. No database daemon, web server or
exotic language necessary.</li>
<li>An indexer which runs either as a thread inside the
GUI, as an external, batch, cron'able program, or as a
real-time indexing daemon.</li>
</ul>
<h2><a name="integration">Desktop and web integration</a></h2>
<p>The <span class="application">Recoll</span> GUI has many
features that help to specify an efficient search and to manage
the results. However it maybe sometimes preferable to use a
simpler tool with a better integration with your desktop
interfaces. Several solutions exist, at the moment mostly for
the KDE desktop:</p>
<ul>
<li>The <span class="application">Recoll</span> KIO module
allows starting queries and viewing results from the
Konqueror browser or KDE applications <em>Open</em> dialogs.</li>
<li>The <a href="http://kde-apps.org">recollrunner</a> krunner
module allows integrating Recoll search results into a
krunner query.</li>
</ul>
<p>Recoll also has
<a href="usermanual/rcl.program.api.html#RCL.PROGRAM.API.PYTHON">
<span class="application">Python</span></a> and
<span class="application">PHP</span> modules which can allow
easy integration with web or other applications.</p>
<h2><a name="stemming"></a>Stemming</h2>
<p>Stemming is a process which transforms inflected words
into their most basic form. For example, <i>flooring</i>,
<i>floors</i>, <i>floored</i> would probably all be
transformed to <i>floor</i> by a stemmer for the English
language.</p>
<p>In many search engines, the stemming process occurs during
indexing. The index will only contain the stemmed form of
words, with exceptions for terms which are detected as being
probably proper nouns (ie: capitalized). At query time, the
terms entered by the user are stemmed, then matched against
the index.</p>
<p>This process results into a smaller index, but it has the
grave inconvenient of irrevocably losing information during
indexing.</p>
<p>Recoll works in a different way. No stemming is performed
at query time, so that all information gets into the index.
The resulting index is bigger, but most people probably don't
care much about this nowadays, because they have a 100Gb disk
95% full of binary data <em>which does not get
indexed</em>.</p>
<p>At the end of an indexing pass, Recoll builds one or
several stemming dictionaries, where all word stems are
listed in correspondence to the list of their
derivatives.</p>
<p>At query time, by default, user-entered terms are stemmed,
then matched against the stem database, and the query is
expanded to include all derivatives. This will yield search
results analogous to those obtained by a classical engine.
The benefits of this approach is that stem expansion can be
controlled instantly at query time in several ways:</p>
<ul>
<li>It can be selectively turned-off for any query term by
capitalizing it (<i>Floor</i>).</li>
<li>The stemming language (ie: english, french...) can be
selected (this supposes that several stemming databases
have been built, which can be configured as part of the
indexing, or done later, in a reasonably fast way).</li>
</ul>
</div>
</body>
</html>