Skip to content

Obsidian Dataview

Is an Obsidian plugin that allows you to you to database everything in a Vault.

Usage

Metadata

Are data about the data, like context

To add then in Obsidian - key:: value right in the text - In the Front Matter, in YAML

Types of metadata

  • Explicit (user-defined in YAML): List of all data types
  • Implicit related to the page
    • file.name
    • file.ctime
    • file.mtime
    • file.day
    • file.tags an array of all tags in note, with subtags broken down: #tag/1/A will be [#tag, #tag/1, #tag/1/A]
  • Implicit related to the task or list

Queries

Are done in a thing similar to SQL

In note

  • Do a Markdown with dataview as language
    • TABLE key FROM XXX
      • key is the key defined in the annotation
      • xxx is to sort where it search
    • WHERE type = "person"
    • SORT file.name ASC

Queries type

  • /```dataview to start the query - without the /
  • list: Create a list of all specified notes
    • list "File path: " + file.path to add the file path after the note name
    • list list_indented to add indented [[YAML#Lists]] sub-list defined as metadata
  • task: searches for all checkboxes -[ ]
  • table shows a table of various metadata field linked to a note
    • table fieldA, fieldB for multiple metadata
  • from: Define where you get the notes
    • from #tag
    • from "folder/sub"
    • from [noteA](../noteA "noteA") all notes with links coming into noteA
      • from [[]] is the self note
    • from outgoing([noteA](../noteA "noteA")) all notes with links coming out of (leaving) noteA
    • Operators
      • And: from #a and #b
      • Or: from "Uni" or "Work"
      • Not: from -#tag
  • where: to narrow down the list further using comparison operators
    • operators: >, >=, <, <=, =, !=
    • where file.size > 1000 to have all big files
    • where !complete to show all notes without the complete metadata field
  • sort to define which order to list the result
    • sort field asc/desc
    • sort field1 asc, field2 desc for multiple sorts
  • flatten to unroll lists
    • flatten list_indented to have one row for each item in list_indented, in every notes
  • group by to gather together results based on a field value
    • group by fieldA will only show the different value of fieldA, not the content inside each note
    • the rows object allow you to access each notes
    • table rows.file.name from #tag group by intensity gives a table of each intensity with the name of each notes that have the corresponding intensity
    • Group by is difficult, so try to concatenate several groups with +
  • /``` to start the query - without the/`

Functions

Goal is to evaluate whether something is true or false - Reference page - contains(key, value) will search in key for value - regexmatch(pattern, string) with pattern the regex and string the string to be tested against (example: file.name)

Examples: - list where contains(file.name, "2022") will return all notes with 2022 in the name - GROUP BY split(tags[0],"/")[1] will extract from the 1st tag the 1st thing after /

Group by the 2nd sub-item of the #area tag (for example #area/a and #area/b)

GROUP BY "Task: " +
     split(
         filter(
             tags, 
             (x) => 
                 startswith(x, "#area")
         )[0],                           // extract only the 1st match
         "/",
         2                               // stop at the 2nd split occurence
     )[1]                                // extract only the 1st tag

Ressources

Examples