Managing Application Data

As I said at the beginning, several programs you already use on your desktop run SQLite under the hood. Knowing how to manage SQLite databases can help you get more from those programs than they offer you from their own interfaces. As just two of many examples, I'll look at Firefox and digiKam. If you look at the hidden folder in which Firefox stores its data, you will find a number of SQLite databases, which all have the .sqlite file name extension. With the following simple shell script, you can periodically clean and compact all of these files to help Firefox run a bit faster:

cd $HOME/.mozilla/firefox/*.default
for i in *.sqlite
do
  echo "VACUUM;" | sqlite3 $i
done

digiKam, on the other hand, stores location and all other metadata for each of the pictures it manages in one SQLite database called digikam4.db (Figure 4).

Figure 4: DB Browser is quite useful whenever you need to study and understand the structure of an existing SQLite database that you did not create yourself. This screenshot shows some of the tables from the digiKam database.

Assume, for example, that you have hundreds of scanned photographs from the past 30 years scattered among many different digiKam albums. What do you do if your grandparents ask for a copy of all pictures that were taken at your parents' house? Should you find all the corresponding albums manually? If they are properly geotagged (which is much easier than you might think, but that is a topic for another tutorial), you can tell SQLite to find those pictures for you.

Looking at the tables in the digiKam database with the metacommands .tables and .schema shows that the location of each album and the photograph file names are kept in the tables Albums and Images, respectively; another table, ImagePositions, stores the latitude and longitude of each image. In these conditions, asking SQLite to list the locations of all the pictures whose latitude is the same as that of your parents' home (43,33 in this example) is relatively simple (Listing 2).

sqlite> .separator "/"
sqlite> .mode list
sqlite> .output photolist.txt
sqlite> SELECT relativePath, name from ImagePositions as P JOIN
        (Albums as A JOIN Images AS I ON A.id = I.album) ON P.imageid =
        I.id WHERE latitude = '43,33' ORDER BY relativePath, name;

The SELECT command concatenates the three tables with the JOIN statement, producing the photolist.txt file with lines like those shown in Listing 3: a file, for example, that you can easily use in a tar command or in a script that copies the photographs onto a DVD. On the other hand, if you use .mode html, you can export your data as an HTML file (Figure 5).

[marco@localhost ~]$ head -2 photolist.txt
/marco/1978/family/birthdays/20080815-grandma-birthday/200808151951.jpg
/marco/2002/holidays/20021225-christmas-dinner/200212251505.jpg

Figure 5: An HTML gallery of all (and only) the pictures from digiKam albums from a certain location.

Graphical SQLite Interfaces

Command-line tools are the most efficient, but not the best for every situation. The most convenient graphical interfaces for managing SQLite databases on Linux may be LibreOffice Base and Kexi [2]. Several drivers can connect a SQLite database with the LibreOffice Suite; the easiest to set up at the time of writing was the ODBC Driver Extension [3]. Kexi, on the other hand, can deal with SQLite out of the box.

If you want a graphical interface, however, I recommend you at least test drive DB Browser for SQLite [4], as shown in the figures for this article. Like its command-line counterpart sqlite3, DB Browser is a multiplatform application and is available as a binary package for all the most common Linux distributions. For complex, raw queries, as well as heavy data processing tasks, I continue to prefer the command-line tool, but if you just need to tweak your SQLite configuration or study the structure of an existing database, DB Browser is a better choice.

Conclusion

If you need a database that is simple to set up or want better access to the data that your software is already handling with SQLite, you now know how to proceed. The next step is to look at the official SQLite documentation, especially the "SQL Syntax" and "Core SQL Functions" sections [5], and bookmark some handy cheat sheets that will help you create queries, starting with the resources listed in the Info section [6]-[9]. Enjoy your portable databases!