image source: www.wpclipart.com

Tracker/recoll/docfetcher/locate/ deep file search + calibre database search

Ulauncher extension for deep searching the filesystem or the calibre database via the recoll / gnome tracker / docfetcher / locate index or calibre database file and handle the results in multiple ways (e.g. opening in default application).

If you like this extension then consider to star it on github so people can find it more easily (by sorting on github stars on the ulauncher extension page)

Requirements

• A running gnome-tracker daemon + tracker index (for gt and ts keywords)
• A recoll index and the recoll python api (for the rc keyword)
• A docfetcher daemon and index + setting the variable PythonApiEnabled = true in the 'program-conf.txt' file of the 'conf' subdirectory of your docfetcher directory (for the df keyword). The instructions about the PythonApiEnabled variable are given in the search.py file of your 'docfetcher' directory.
• A calibre sqlite database file (for the cb keyword)

A working daemon and index can be tested respectively in the terminal using the 'tracker search' command (gt and ts), with the recoll gui (rc) or with the docfetcher gui (df)

Description

This extension provides filesystem deep search functionality (i.e. full text search) via gnome tracker, recoll or docfetcher (respective keywords: gt, rc and df) and near instant full system file search via the locate command (keyword: lc + 2x g for grep -i (see image)). Additionally it provides functionality to search books in the calibre database. Searching with gnome tracker by default searches on exact match, so optionally each pattern in the query can be appended with as a wildcard (unfortunately I do not know how to use a wildcard in front of a SPARQL query). Additionally this extension has an option to deep search files using the tracker search command which returns text snippets (keyword: ts, also supports appending with as wilcard). The output can be opened with your prefered application.

Usage

Tracker/recoll/docfetcher and locate keywords

Type keyword followed by a search term. The locate keyword uses locate with the -i (case-insensitive) and the -r (regexp) flags to search for the query, which must be a single regexp without space (for historic reasons, but if a space should be included than simply adding a grep covers most cases. You are welcome to extend the functionality. I am using the really elegant and powerful Spacemacs now instead of this extension...). The results can be narrowed down twice with a grep using g keyword (see image). Tracker (default: gt and ts), recoll (default: rc) and docfetcher (default: df) by default search case-insensitive for matches with full words. However all these keywords allow to use wildcards (use symbol * ) at the and of a search term. As the tracker and recoll searches can return many results these searches can be narrowed down by an single grep using g followed by the keyword (see image) where the grep works only on the filename field.

Use wildcards by default

The extension has a setting option to use wildcards by default (there is no extended grep functionality for the tracker keyword here).

Documentation recoll query syntax

The rc keyword supports the very powerful recoll query syntax as documented here

Calibre keyword

The calibre keyword (default: cb) searches in the title and author_sort (can be different from te author field, see book's metadata) fields and can handle two query patterns separated by a space (e.g. "cb hello dan") where the search is case insensitive and includes wildcards before and after the patterns.

If you really want the optimal PDF and calibre experience (at least on linux and mac) then you should try out Spac(Emacs) with the PDF layer (or PDF-tools package) in addition with the calibre-layer (or calibredb package).

(For vim users... I find Spacemacs provides an even better (or more awesome) vim than vim itself... you should really check it out.)

Installation

Add the plugin via the extension menu in the ulauncher settings using the URL: https://github.com/dalanicolai/gnome-tracker-extension

Short note on development

• The py4j folder and the search.py file are part of docfetcher. They are required to use its search functionality from python. If a user wishes that the docfetcher keywords also gets a grep extension functionality (like the tracker, recoll and locate keywords), then someone who knows python can either implement this functionality easily himself or another user can simply open an issue on github or send me a request by mail (see below).

• The appchooser.py is a small script that launches a gtk application chooser window.

• This extension might be useful to use as a template for other seach extensions. Also, if you know some python, its (search) behavior can be easily adapted to your own preferences by adapting the code in main.py.

Contact

To report any bugs or if your have any requests or comments relating to this extension send an e-mail to dalanicolai@gmail.com, or just open an issue here on github.