Dataview sort on part of property value

[PatrickMarq]

What I’m trying to do

I have a dataview for displaying book, This work fine.
The data comes (if possible) from the book plugin.
Now I’m looking a way to sort the book on the last name of the author. As Lastname does not exists I need to extract the last woord from the property Schrijver.

---

tags:

• epub
  Titel: De boodschapper
  Schrijver: Kader Abdolah
  Achternaam: Abdolah
  Serie:
  Serienumber:
  Categorie: Fiction
  Uitgever: Singel Uitgeverijen
  Paginas: 244
  Taal: NL
  Bindwijze: Epub
  Cover: “[[De boodschapper - Kader Abdolah.jpeg]]”
  Gekocht: true
  Gelezen: true
  calibreURL:
  cssclasses:
• page-90

---

	Table 
	WITHOUT ID 
	link(file.name, Titel) AS Titel, embed(link(cover, "150")) AS Cover, 
	choice(gekocht , "✅", "✘") as Gekocht,	 Schrijver
	FROM "02 - Multimedia/02 - Boeken/Boeken"
	
	WHERE (Achternaam >= "A" and Achternaam <="F")
	Sort Achternaam, Titel ASC

Things I have tried

I have created an extra property called Achternaam, so the datives works. But this is a lot of work adding something extra.

I think it could be possible with Dataviewjs, but I don’t understand it. Tried it without luck

	// create table 
	dv.table(["Schrijver", "Achternaam"],
		// get the pages
		dv.pages('"02 - Multimedia/02 - Boeken/Boeken"')
		//.where(p )......
		//.sort(page => page.file.cday, "desc")
	);

Any advice would be helpful.

[cheezopath]

can you sort by SORT split(Schrijver," ")[1] ASC ?
https://blacksmithgu.github.io/obsidian-dataview/reference/functions/#splitstring-delimiter-limit

[holroy]

@PatrickMarq

As Lastname does not exists I need to extract the last woord from the property Schrijver.

I’d use the following to get the last name of the author: slice(split(Schrijver, " "), -1)

This will allow multiple forenames, but only use the last surname. So from John Ronald Reuel Tolkien you’d get Tolkien. But also only Beethoven from Ludwig van Beethoven.

[PatrickMarq]

Thank you all,
The solution works for the sort, but not on the WHERE.

I was under the impression that this code was only for dataviewjs.

[holroy]

If you want it in the WHERE clause, or other parts you can use it in combination with FLATTEN like siren in this variant:

```dataview
Table WITHOUT ID 
  LastName,
  link(file.name, Titel) AS Titel,
  embed(link(cover, "150")) AS Cover, 
  choice(gekocht , "✅", "✘") as Gekocht,
  Schrijver
FROM "02 - Multimedia/02 - Boeken/Boeken"
FLATTEN slice(split(Schrijver, " "), -1) as LastName
WHERE (LastName[0] >= "A" and LastName[0] <="F")
	Sort LastName, Titel ASC
```

[PatrickMarq]

Thanks, it works.
I still don’t understand the concept of Flatten.

Some books have a tag ‘book’ or ‘epub’ or both.
But with a table you see a line for each tag, so I tried flatten (to have then on one line) but then I got two books in the view one with each tag.
This seems logical because it are two things but still.

I’ll try to get a better understanding of this.

[holroy]

If you use FLATTEN on a list, it’ll split the current row into multiple rows where each new row has the unique values of that list in addition to all other available information.

If you use FLATTEN on a single value, it can be used to compute a new intermediate value for user later on in the query or for display.

If you want to filter/map a list, and store the intermediate result for later use, you’ll need to do something like: FLATTEN list( ... ) as newList where you replace the ... with the expression doing something with a list which returns a list.

Does this clarify a little on the usage of FLATTEN?