Special query recommendations

[Feedback1000]

Hello.

I'm looking for a way to list book, film or series recommendations.
How can I query these things using a TABLE?

Following: File clippings:
File 1

---
tags: 
  - daily
---
# 2024-10-02
recommendation/book:: Matrix

File 2

---
tags: 
  - daily
---
# 2024-10-01
recommendations/artist:: The Police

Can you help?

Which does not work:

TABLE recommendation/movie AS Recommendation
FROM ""
WHERE recommendation/movie != null
SORT file.name desc

[rbennett00]

I don't believe you can next properties like you're trying to do with recommendation/foo syntax. What about keeping it simple and just using a generic recommendation:: inline field and tags to identify the type of media. e.g.,

---
tags: 
  - daily
---
# 2024-10-02
recommendation:: The Matrix #movie
recommendation:: The Police #artist 
recommendation:: Smartless #podcast 

Then you just need a dataview table like:

TABLE recommendation AS Recommendation
WHERE contains(recommendation,"")
SORT file.name desc

[holroy]

If you add the following query, you can see which fields are defined in the current file:

`$= const p = dv.current(); delete p.file; dv.span(p) `

And in the case of recommending the artist Police it displays this:

So the answer is one either needs to do row["recommendation/artist"] or recommendationartist to access the nested inline fields.

[holroy]

The nested inline fields are allowed, but currently there is no easy way in a normal query to list all of those starting with recommendation/ for an arbitrary set of nested categories. If you're able to switch to dataviewjs, you can use Object.keys() (or Object.entries()) to manipulate the lists of inline fields.

For starters try the following in a document of its own:

recommendation/artist:: The Police
recommendation/book:: Matrix

Local fields: `$= const p = dv.current(); delete p.file; dv.span(p) `

```dataviewjs

const allFields = dv.current()
delete allFields.page
const rec = Object.entries(allFields)
  .filter(e =>  e[0].startsWith('recommendation/') )
  .map(e => [e[0].split("/")[1], e[1]])

dv.table(["Type", "Value"], rec)
```

It should produce a table like:

---

If you're just using a limited set of nested variants, you could do similar stuff where you hardcode the various alternatives, and check for their existence. But in general it would then be easier to use a simpler recommendation:: type of inline fields. On a side note, if the field value is a reference to existing documents, you could put the document type in that and do something like to get all the book recommendations in your vault:

```dataview
TABLE WITHOUT ID recommendation
WHERE recommendation AND recommendation.type == "book"
```

(Or something similar where you lookup the type in the link. This would require a field definition like: recommendation:: [[Matrix]])