Dataview

sublimated renders Obsidian Dataview DQL queries as live tables and lists. no plugins needed — queries evaluate server-side when the page loads.

code block queries

wrap DQL in a dataview fenced code block:

```dataview
TABLE date, purpose
WHERE date
SORT date DESC
```

queries run against all files in the current folder. each file's YAML frontmatter fields are available as columns.

inline expressions

use `= expr` anywhere in text or markdown tables for computed values:

total files: `= length(rows)`
countries: `= length(unique(filter(flat(rows.tags), (t) => startswith(t, "country/"))))`

inline expressions have access to rows — all sibling files in the current folder.

query types

TABLE

renders as an HTML table. first column is always the file link.

LIST

renders as a bullet list with file links.

operations

operations apply in the order written:

• WHERE condition — filter rows

• SORT field ASC|DESC — sort rows (multiple fields supported)

• LIMIT n — cap number of results

• FLATTEN field AS alias — expand array fields into separate rows

• GROUP BY expr AS "Label" — group rows, access group items via rows variable

GROUP BY

groups all matching files by expression. each group becomes one row with rows containing the grouped items:

access grouped data: rows.file.name gives array of file names in the group, flat(rows.tags) flattens all tags.

FLATTEN + GROUP BY

combine FLATTEN and GROUP BY for disaggregation/reaggregation:

this expands each file's country tags into separate rows, then groups by country name.

expressions

fields

• file.name — filename without extension

• file.path — full path

• file.folder — parent folder

• file.size — size in bytes

• file.ctime — created time

• file.mtime — modified time

• file.link — link to the file

• any frontmatter field — date, tags, purpose, etc

operators

• comparison: =, !=, >, <, >=, <=

• arithmetic: +, -, *, /, %

• logical: AND, OR, !

lambdas

(x) => startswith(x, "country/")

used with filter, map, sort, any, all, none.

functions

array

• length(array) — count elements

• contains(array, value) — check membership

• filter(array, lambda) — keep matching elements

• map(array, lambda) — transform elements

• sort(array) — sort elements

• unique(array) — deduplicate

• flat(array) — flatten one level of nesting

• join(array, separator) — join to string

• reverse(array) — reverse order

• any(array, lambda) — true if any match

• all(array, lambda) — true if all match

• none(array, lambda) — true if none match

string

• contains(string, substring) — check substring

• startswith(string, prefix) — check prefix

• endswith(string, suffix) — check suffix

• lower(string), upper(string) — case conversion

• replace(string, from, to) — replace substring

• split(string, delimiter) — split to array

• regextest(pattern, string) — regex match test

date

• dateformat(date, format) — format date (e.g. "yyyy-MM-dd", "yyyy")

• date(string) — parse string to date

• striptime(date) — remove time component

math

• round(n), floor(n), ceil(n) — rounding

• min(array), max(array) — extremes

• sum(array), average(array) — aggregation

utility

• default(value, fallback) — null coalescing

• choice(condition, ifTrue, ifFalse) — ternary

• typeof(value) — type name

• string(value), number(value) — type conversion

• link(path, display) — create link

frontmatter

YAML frontmatter is automatically extracted and stored when files are pushed. string values that look like dates (YYYY-MM-DD) are parsed as date objects.

---
title: Japan 2024
date: 2024-03-15
purpose: tourism
tags:
  - country/japan
  - city/tokyo
  - city/kyoto
---

all fields are accessible in queries: WHERE date > date("2024-01-01"), SORT date DESC, etc.

limitations

• queries only access files in the current folder (no cross-folder queries yet)

• no FROM clause needed — defaults to current folder

• FROM "subfolder" resolves relative to the current folder

• TASK and CALENDAR query types not supported

• no custom JS views