Ritt database format

The Ritt database (.ritt files) uses a simple and efficient format to store information. Essentially, it is a compressed plain text file that is written in JSON format. Here, you can learn more about the data structure used by Ritt, which will allow you to write your own code to read and modify the Ritt database file.

Ritt data structure

Ritt uses a directed graph structure that consists of vertices and edges to store data.


There are three types of vertices.

  1. Space. A root vertex. There is only one space vertex per Ritt database.
  2. Link. A vertex that links to user data such as files, folders and tasks.
  3. Tag. A vertex that stores information of a tag.


There are five types of directed edges.

  1. Parent. Together with the Child edge, this defines a directed acyclic graph between vertices of the same type. Parent edges of a particular vertex point to the parents of the said vertex.
  2. Child. Similar to above, child edges of a particular vertex point to the children of the said vertex.
  3. Space. Points to the associated Space.
  4. Tag. Points to the associated Tags.
  5. Link. Points to the associated Links.

Information on the edges is stored with the originating vertices as lists.

Edges are always bidirectional. For example, if vertex A is the parent of vertex B, then A will have a Child edge pointing toward B, and B will have a Parent edge pointing toward A.

Similarly, if link A has a tag B, then A will have a tag edge pointing toward B, and B will have a link edge pointing toward A.


An example graph of a small Ritt database is shown below. The Space, Link and Tag vertices and internal Parent/Child edges (directed arrows) are shown in green, orange and blue respectively. Edges connecting two vertices of two different types are shown as non-directed lines in gray.

Small Example Graph

The full database file (in plain-text) that generated the above graph can be downloaded here.

File format and operations


To reduce file size, .ritt files are compressed using Gzip. To decompress, use any supported software such as 7zip. The result is a text file that can be opened with any text editor.

Interpreting the JSON file

Line 1

Contains favorites and history. See below for an example and the comments for the description of the field.

    "i": [ // Favorite emojis used a tag icons
    "s": [ // Search history

Line 2

Contains metadata about the graph.

    "id": "4817f99e-9940-4fb5-94c9-c9c18bf858b0", // Graph unique ID
    "v": "0.13", // Graph specification version
    "l": 324, // Number of vertices in graph
    "s": { // Dictionary of indices of special vertices, typically the space root
        "root_space": 0

Line 3 to end

The rest of the file contains information on the vertices, with one vertex per line.

Example Vertex

    "p": [ // Indices of Parent vertices
    "c": [ // Indices of Children vertices
    "s": [], // Indices of Space vertices
    "t": [], // Indices of Tag vertices
    "l": [], // Indices of Link vertices
    "m": { // Metadata of this vertex
        "t": 2, // Vertex type, see below for details
        "n": "Work", // Name
        "c": { // Content
            "t": 2, // Content type, see below for details
            "id": "373963cd-9d3f-4305-b577-4959888b9a10" // Content ID. Used to identify and track Sources
        "i": "", // Icon for tags
        "a": {} // Attribute dictionary, see below for details
    "i": 14 // Index of this vertex

Vertex Type

Type of vertex. See the previous section for details.

Value Name
0 Space
1 Tag
2 Link

Content Type

Used in Link vertices to indicate the type of content they hold.

Value Name
0 None
1 File
2 Folder
3 Task
4 Task Folder
5 StorageItemPlaceholder (For internal use)


Some vertices have special attributes, which are stored in key-value pairs in an attribute dictionary.

Key Attribute Name Description
2848 Hidden For future implementation of hidden tags.
3217 Dashboard Dashboard tags
4626 Badge Badge tags
35528 ToDoState State (completed or not) of a task
74140 Mirror Mirror tags

Opening a modified .ritt file

After editing the database file in plain text, there is no need to zip it. Simply ensure that it has the β€œ.ritt” extension and open it with Ritt. Note that Ritt will recompress the file on save.