Free

Free coded apps to suit all. No sign up, no email, usable offline, updates when I can, feedback welcome. Send feedback

How-to

How to search inside Wikipedia articles offline

Open an edition in Wonder by Zaweba, type in the search box and press Enter: Wonder searches inside articles using an index it builds itself.

How to search inside Wikipedia articles offline

Use Wonder by Zaweba. Type in its search box and title suggestions appear straight away; press Enter on words that are not an article's title and Wonder searches inside the articles' text instead. The first time you do that, Wonder builds its own full-text index in the background, and you can search while it grows.

Wonder is a free offline encyclopedia. You download an edition of Wikipedia once from the Kiwix project, and searching, reading and wandering all work with no internet from then on.

What you need

You need Wonder and an edition. Open File, then Library, press Show what Kiwix offers, and choose an edition. The full English Wikipedia with pictures is about 119 GB, and the Library also lists editions without pictures, introductions only, top articles and Simple English, each with its size, date and article count shown before you start. If you already have a .zim file from Kiwix, press Ctrl+O to open it instead.

You also need some free disk space for the search index, which is covered below.

Step by step

With the edition open, click in the search box and start typing. Suggestions come from the edition's title index as you type. If the article you want is in the list, pick it.

If you know the exact title, press Enter and it opens. Redirects are followed and capitalisation doesn't matter, so typing eiffel tower finds the right page.

If what you typed isn't a title, pressing Enter searches inside the articles. The first time, Wonder starts building its full-text index in the background. Results come from whatever has been indexed so far, and they fill out as the build carries on. Keep reading other articles while it works; nothing stops.

How long the index takes

For a small edition the index is ready in seconds. For the full English Wikipedia it will take hours and several GB of disk space. That sounds a lot, but it is a one-off for each edition, and Wonder saves progress every 500 articles, so closing the app or restarting the PC doesn't lose the work. It picks up where it left off next time.

Each edition gets its own index, a SQLite FTS5 database holding every article's title and the first 3,000 characters of prose. Title search never needs it, so suggestions always work instantly, whatever state the index is in.

Searching well

Because the index covers the opening prose of each article, search for words that would appear near the top of the page you have in mind: the subject's name, a place, a date, a defining phrase. Two or three distinctive words beat a long sentence. If a search finds nothing, try a different word for the same thing, or a broader one.

When you find what you were looking for, keep it. Ctrl+D bookmarks the article, Ctrl+Shift+D puts it on Read later and Ctrl+H brings back your history, all kept per edition.

Other ways in

Search is only one door into Wonder. Surprise me, on Ctrl+R or the big button on Home, opens a random good article. Follow a random link, on Ctrl+L, jumps to a random article the current one links to, with a trail bar recording each step. Subject shelves, Five a day and On this day are there too, for the times you don't know what you want to read.

So search when you have something in mind, and wander when you don't. Wonder by Zaweba does both without a connection.