FAQ Guide for Obsidian Dataview

Who this guide is for?

Any person looking to get started but feels that the documentation is too intimidating. SkepticMystic’s guide is actually a better guide to help you get started writing queries. This should be viewed as a supplement to it. That said, if you’re a programmer or database expert who’s looking to get started quickly, this guide probably isn’t it. Check out the docs instead.

What is dataview?

It’s a plugin by blacksmithgu. It lets you query data from your Obsidian vault.

What’s a query?

Think of a query as a question you want to ask your Obsidian vault. For example,

• What notes did I modify in the last 5 days?
• Which movies are tagged #towatch?
• What tasks have I tagged #critical?
• List my favorite books in descending order of rating.

These are just a few examples of what is possible with dataview. Note that you can’t change your metadata with dataview. Look into Christian’s MetaEdit for that instead.

Great, how do I get started?

It’s actually pretty easy to get started. Just install the plugin from the “Community Plugins” list after turning off safe mode.

But wait, how do I query?

Use a dataview code block like so:

```dataview
list
from "Folder"
```

The syntax is rather extensive so refer to the docs copiously. I still do so even after having used dataview for 6 months at the time of writing.

What can I query?

Dataview indexesJust a fancy way of saying “collects”.

“metadata” from your notes. Metadata is meta information about your notes and files. Some of the common metadata fields you might be familiar with are things like “last modified date”, “file creation date”, “file extension”, etc.

There are two types of metadata, for the purposes of dataview. The one we looked at above was “implicit metadata”, which exists for every file in your vault. The list of fields can be found in the docs. Note some of the more useful ones:

• file.name
• file.path
• file.mtime

There’s a second type of (arguably) more useful metadata — the one we define. This lets you create your own indexes to query. Lets look at some examples.

Book.md

---
title: "Iliad"
author: "Homer"
---

Movie.md

---
title: "Citizen Kane"
year: 1941
director: "Orson Welles"
---

Note that for Book.md and Movie.md, the metadata defined there is actually YAML frontmatter. The Obsidian Help docs have a useful section on what that is. Dataview supports YAML but also has support for a custom variant called “inline YAML”, specific only to Dataview.

Inline YAML is a non-standard YAML-like way of specifying additional metadata. This is what is depicted in the example below. It follows the key:: value style. Note that it currently doesn’t support nested fields the same way standard YAML does.

Example.md

This is a sample note. Today I met person:: Adam.

Now that we have some files and metadata defined, let’s talk about indexes, which is what you’re querying against. It’s helpful to think of an index as one big ’ol table that you’re trying to narrow down to precisely what you need. So, in this case, say we had these three metadata examples in our vault, our index would look like this:

Querying our three files for title like:

```dataview
table
    title
from
""
```

would give us:

We queried the title from our index but Dataview also gives us the File column by default.

The rows in your index will depend on what source you’re using. If you’re using "" as a source, the rows will be all the files in the vault. If you’re using #tag as a source, the rows will be all the files with that tag in them. The columns in your index will depend on the metadata that you’ve defined in your files. Task indexes (- [ ] elements that you use for TODOs) are different and they’ll be covered in a separate guide.

To narrow this table down, we typically use filters or WHERE conditions. For layering complication logic, we use functions. For more on what you can do with this index, check SkepticMystic’s guide linked earlier.

What are some gotchas about YAML?

A very important caveat is that this is all you can query. You cannot query the contents of your files. File content is not part of metadata. Remember, metadata is either explicity or implicitly defined as we discussed earlier.

Another gotcha is that wikilinks in Obsidian Frontmatter are not clickable, nor do they show up on your graph. This is a current Obsidian limitation.

Finally, metadata for each file is independent of the other. To rephrase, it means that each file has its own metadata.

How should I define my YAML?

This is a difficult question. You can add nearly any kind of metadata you want to your notes. The open ended nature of this makes it difficult to design “schemas”.Schemas are the database equivalent term for your “metadata structure”.

One solid piece of advice when it comes to “designing metadata” is to think of your end goal first. It’s counterintuitive, I know, but it’ll give you a good idea of what fields you need to use. You can look at the forums for inspiration:

• SlRvb’s media db setup
• Arminta’s Daily + Weekly Reviews in Dataview
• Arminta’s Dataview Task and Project Examples

How to debug your queries?

It’s a little difficult to write a correct query on your first try. I still find it tricky even after having used Dataview for more than 8 months at the time of writing. A general rule of thumb is to start with the simpler version of a query and then slowly progress to the actual query that you want.

Another helpful tip is to actually inspect the metadata that’s being queried. You can do this by enabling inline DVJS in the plugin settings and then typing `dv.list([dv.current()])` to inspect that note’s metadata. Remember to use that in a page that actually has the metadata since you’re likely writing the query in a separate file. You’ll note that everything you can query about that page shows up in a bulleted list.

Finally, for debugging more complex queries, these often error out because of some assumptions you’ve made about the data you’re querying or the functions you’re using. Common ones include syntax errors and type mismatches. Test those assumptions by checking the docs.

Let’s also do a quick rundown of some of the common errors/issues.

Feedback

If you feel there are some other FAQs you want covered in this guide, feel free to reach out to me on Discord or via email.