Manki is a library and commandline tool for producing bare-bones flashcards to be imported in Anki.

This should be useful to anyone set on learning a foreign language using flashcards. If you are like me, you prefer working with text files rather than excel files, the Anki GUI or whatnot. Manki's purpose is to generate flashcards (via Anki) out of a text file.

Manki converts text files to a csv format ready for import in the Anki app. You'll write a lightly marked up text like:

Here is a cell/field with @words to be converted to sound@.
---
And this is another cell/field with ~a cloze deletion~1
---
Third cell, with a reference to the audio string in first cell: |>

This will be converted to a csv string (or file) marked up for Anki use. So any @blah@ markup disappears, and the insert marker |> will be replaced by a [sound:blah.mp3] string (the Anki's markup for sound files). The Manki cloze markup ~foo~1 becomes the Anki string {{c1:foo}}. And three dashes on their own lines delimit what will become the Anki fields. In the csv string they will be replaced by commas.

Note: Manki does not produce the sound files for you. It generates a reference to sound files which can be added to the Anki collection independently, for instance, using Google Cloud's Text-to-Speech or similar services. A list of audio strings picked up from the input file can be obtained with the --audios flag on the command line or using the functions provided by the library.

Manki is more constrained than other light markup languages in the sense that not every stream of characters is a valid Manki document.

• cloze deletion: ~check~1 this ~out~1 converts to the Anki string {{c1:check}} this {{c1:out}}
• *bold* and *italic* convert to html <b>bold</b> and <i>italic</i> (recognised as such by Anki)
• @audio string@3 converts to a bare string with no markup (audio string) but can be collected and inserted in the same block; if you don't want audio string to occur in the output file (yet need the string in order to produce audio) use one or two dashes before the starting marker: -@audio string.
• the 3|> inserts a coindexed audio string, and converts to something like [[sound:audio_strings]] (assuming the audio string just above)
• indices (on cloze-deletion markers, audio markers and insert markers) are optional and should be used for cross-referencing.
• two or more dashes on a new line (\n---\n) separate one field (csv cell) from another; no start or end delimiters are allowed; so \n--\ncell one\n--\ncell two\n--\n is wrong; use instead: cell one\n--\ncell two
• blocks in the input file (which will become csv rows, which in turn will become flash cards after import) are separated from one another by two or more newlines. Example: row one, cell one\n--\nrow two, cell two\n\nrow two, cell one\n--\nrowtwo, cell two

Note: at the moment the Manki tags can't overlap. For instance, *_bold italic_* is invalid. One exception is the audio marker @ which can wrap around any other tags @*bold*, _italic_, etc.@. This is particularly useful if you plan to convert cloze-deleted words to a sound file: @~hidden but pronounceable~@.

Check out /test/test_files/example.txt to get a input file sample using the Manki markup. example.csv, in the same directory, is the output that Manki's commandline tool produces out of the example.txt input file.

You'll find the executable in Releases.

To convert file.txt to file.csv, run:

$ manki-exe -f file.txt -o file.csv

You can then import file.csv in Anki. There is a thread on Stack Exchange that explains how to import csv files. And the Anki Manual provides more detail.

Run manki --help for more information about the commandline options.

Note: Manki has not yet been tested on Windows and macOS.