Recoll as an Alternative

Recoll [8] is an alternative to Clapgrep that also features a GUI. Unlike Clapgrep, Recoll relies on an index and therefore does not have to search the filesystem on subsequent calls once the index has been created. It is a good idea to create the index at night when your computer is not in use. Even on a system with AMD Ryzen 9 7900 and 64GB RAM, indexing my home directory took almost 20 minutes.

Recoll uses the Xapian [9] open source library as a storage and retrieval engine in the background. Xapian manages the index database, processes complex queries, and calculates relevance, while Recoll (Figure 6) provides mechanisms and interfaces for indexing document content and processing and presenting various file formats.

Figure 6: Unlike Clapgrep, Recoll relies on an index, which is faster when dealing with very large file collections.

According to information available on the Internet, Recoll's index gives it a speed advantage when searching in very large directories or frequently searched file collections. I was unable to verify this on state-of-the-art hardware where I used Clapgrep and Recoll to search for the term "linux" in the /usr directory, containing just south of 34,000 files. In Recoll, I entered the expression "linux and path:/usr/" in the search line. In both cases, the result was available in less than a blink of an eye. However, Recoll is likely to have an advantage in far larger directories with millions of files.

Recoll appears to be significantly more powerful when evaluating search results. The display (Figure 7) also appears more clear-cut, delivering the results page by page. In addition to Linux, Recoll is currently available for macOS, Windows, and Android. It searches files, email attachments, and archives, which it decompresses transparently.

Figure 7: Thanks to its powerful query language and the way it processes results, Recoll is more suitable for complex searches; this is further accelerated by the use of an index.

Besides Recoll's command line or GUI, you can also use a web interface. Recoll is recommended for search queries that require complex search functions and a detailed display and evaluation of the results. It supports a powerful query language with Boolean operators, phrase search, wildcards, and filtering by file attributes such as size, date, or MIME type.

Conclusion

Clapgrep and Recoll both have their merits; neither is better than the other. However, they are intended for different types of file system searches. For occasional quick searches, Clapgrep has the edge and offers several advantages over the veteran grep and its descendants Ripgrep and ripgrep-all. When it comes to complex or repetitive searches whose results need to be evaluated based on various criteria, Recoll is the clear winner. You will also find an overview of other desktop search engines in an earlier issue of Linux Magazine [10][11].