Recoll

Recoll is a desktop search tool that provides full-text search in a GUI with a few mandatory external dependencies. It runs on many Unix-like operating systems and is mostly independent of the desktop environment.

Installing

This Page was written with Ubuntu 24.04 LTS in mind, and may not work correctly with other versions or distributions.

sudo apt install recoll recollcmd libimage-exiftool-perl antiword unrtf poppler-utils libwpd-tools untex

Configuring

First start by creating the wanted recoll index directory and copy the example config to it, as shown below:

mkdir -p /var/www/recoll
cp -R /usr/share/recoll/examples/* /var/www/recoll/

Open the recoll.conf file in it's new location and change the following two values.

topdirs is the location the files to be index and in turn searched.

topdirs = /var/www/html

cachedir is the location the indexes are actually stored on the system.

cachedir = /var/www/recoll

Indexing

Now index the system, depending on the size of the site, this may take a long time.

recollindex -c /var/www//recoll -z

The full usage list for recollindex my be found by using the -h option, as seen below.

recollindex: Usage:
recollindex [-h]
    Print help
recollindex [-z|-Z] [-k]
    Index everything according to configuration file
    -z : reset database before starting indexing
    -Z : in place reset: consider all documents as changed. Can also
         be combined with -i or -r but not -m
    -k : retry files on which we previously failed
    --diagsfile <outputpath> : list skipped or otherwise not indexed documents to <outputpath>
       <outputpath> will be truncated
recollindex -m [-w <secs>] -x [-D] [-C]
    Perform real time indexing. Don't become a daemon if -D is set.
    -w sets number of seconds to wait before starting.
    -C disables monitoring config for changes/reexecuting.
    -n disables initial incremental indexing (!and purge!).
    -x disables exit on end of x11 session
recollindex -e [<filepath [path ...]>]
    Purge data for individual files. No stem database updates.
    Reads paths on stdin if none is given as argument.
recollindex -i [-f] [-Z] [<filepath [path ...]>]
    Index individual files. No database purge or stem database updates
    Will read paths on stdin if none is given as argument
    -f : ignore skippedPaths and skippedNames while doing this
recollindex -r [-K] [-f] [-Z] [-p pattern] <top>
   Recursive partial reindex.
     -p : filter file names, multiple instances are allowed, e.g.:
        -p *.odt -p *.pdf
     -K : skip previously failed files (they are retried by default)
recollindex -l
    List available stemming languages
recollindex -s <lang>
    Build stem database for additional language <lang>
recollindex -E
    Check configuration file for topdirs and other paths existence
recollindex --webcache-compact : recover wasted space from the Web cache
recollindex --webcache-burst <targetdir> : extract entries from the Web cache to the target
recollindex --notindexed [filepath [filepath ...]] : check if the file arguments are indexed will read file paths from stdin if there are no arguments
recollindex -S
    Build aspell spelling dictionary.>
Common options:
    -c <configdir> : specify config directory, overriding $RECOLL_CONFDIR
    -d : call fadvise() with the POSIX_FADV_DONTNEED flag on indexed files
          (avoids trashing the page cache)

WebUI

Run standalone

Installed required dependence.

apt install -y git python3-waitress

Then download WebUI to the directory that is not accessible via the web server.

git clone https://framagit.org/medoc92/recollwebui.git /var/www/search

Run webui-standalone.py and connect to http://localhost:8080.

/var/www/search/webui-standalone.py -c /var/www/recoll

There's some optional command-line arguments available:

-h, --help            show this help message and exit
-a ADDR, --addr ADDR  address to bind to [127.0.0.1]
-p PORT, --port PORT  port to listen on [8080]
-c CONFDIR, --config CONFDIR Recoll configuration directory to use

Nginx WSGI/CGI Setup

apt install -y pwgen git nginx xxd apache2-utils

git clone https://framagit.org/medoc92/recollwebui.git /var/www/html/search

sed -i 's/127.0.0.1/0.0.0.0/g' /var/www/html/search/webui-standalone.py