[BRX]

I just noticed that the CVS import only accepts PDF as source files. That is in 99 % sufficient of course.

I wanted to import a CVS though with a (single) txt file as reference for multiple song titles with metadata.

(Placeholder with a list in the text file I can easily edit).

Is it difficult to extend the import funktion to txt, chord pro and other formats MSP can display?

(I'll try as a workaround for now to import a pdf file and use the swap feature to replace it with the text file. I guess that should work, right?)

[BRX]

Just came to me that I would have to swap the song for each single entry. That's not ideal for larger amounts of items.

Renaming the txt file to the pdf source didn't work either since MSP seems to check for headers or if it's a PDF. So my ideawas no real workaround.

[itsme]

What's the use case that makes you try this?
The CSV as we have it now links multiple songs to a certain page in one multi-page multi-song PDF of the same name and allows filling the metadata of several songs conveniently.
Are you trying to use the CSV just to supply metadata? For what?
A chordpro file can fill metadata with {title: , {key: and the like. Could that help?
Many songs in a single ChordPro file does not look very promising to me.

[BRX]

I admit it's a very specific usage of mine (and it's your fault, really, since you started me to care about alternative titles).

If there's a title like for instance "Black Orpheus" which is also known as "A Day In The Life Of A Fool" or "Manha Do Carnaval" and "Chanson D'Orphee" for which I already have 7 versions I decided not to have 28 entries but only 7 for the most used title (here Black Orpheus) and just make 3 Placeholders for the other names with a reference to the other titles and a text file in which I have a list of all the aliases in my song collection. And yes, for this file and the multiple placeholder entries I want to supply metadata by CVS.

Also the CVS import is more convenient even for adding metadata to single entries if you already have a database or an excel sheet with metadata you just have to copy and export.

[itsme]

OK, got it. It's about handling more than one title per song.
Generally MSP is designed around m:n relationships (ask your favorite search engine about "database cardinality m:n" to read more). The obvious ones are songs - composers (a song can have m composers, a composer might have written n songs) or songs - collections (a collection contains several songs, a song can be in several collections).
For song titles there is nothing like that, we just have "sort title" as a second option. And I doubt that CSV import is the best starting point, CSVs are mainly linear lists that link a number of songs to a certain file.
If I had the plan to organize a huge collection of sheet music in MSP I could imagine a way like this using only features that already exist:
- create a song in MSP for every song in your collection, write into MSP's "title" the title as it is printed on the sheet
- for songs with more than one title, decide which of the titles shall be your "main title", write this into "sort title" in MSP, so MSP will use this one in all its lists
- searching in MSP should find the song when searching for the title as well as searching for the sort title (I assume, need to check that)
- create a text file (or whatever is convenient) with the same title, maybe with a postfix "(song info)", import it as "local file" into MSP like any song
- write the list of all song titles for this specific song into its "song info" file
- add this "song info" file to MSP under all contained song titles, note: the file itself exists only once, but with several MSP database entries
To create the "song info" file, MSP can assist you:
- all these songs are listed close together with their sort title on the "Songs" tab, create a temporary setlist of them and export it with an adequate song list format containing both title and sort title
What you have achieved now:
If you can remember the "main title" you can find all your versions easily. If you can't remember the main title, any other title of the song will find the "song info" file where you can look up the main title.
The "song info" file could also be used to collect more details about a song in a central place.
Collecting and maintaining this is a time-consuming process, no doubt. But matching various titles for the same song is a manual process anyway (except some easy automatic removing of "the", "a" and so on that we already have).
Feel free to contact me by e-mail, if you have further questions, we can discuss that also in German.

[Skip]

I'm not sure but I think you may be able to use the custom fields in much the same way. Enter the alternate names in the custom field using commas to separate them. I tried it by entering alternate names in the custom field in the song editor then searching using the custom filter.

[Zubersoft]

I can pretty easily extend the CSV import to support those file types if I just completely ignore the page order field and assign the full range of pages to the new song(s). When it comes to matching the CSV with the file, I'll first search for a PDF with a matching name, then one of the other supported file types, and if I don't find any matches, I'll use the first line in the CSV as the filename. Even though I can add support for this, until I add a mechanism where you can import multiple files at once through CSV, I don't think it would be useful for most people.

[BRX]

@skip
Thanks, but I had that idea already. I'm using the first of the two custom fields for alternative titles which I also display with the song format.

And I use placeholders for the actual alt title songs (only one for each alt title). I wanted to assign (always the same) text file to these placeholders. It might be of use in other cases, too, if you want to add metadata to other files than pdf and prefer the CVS way. I admit it's not a priority at all. But if it's easy to add this function and it doesn't effect other functions and designs, I'd appreciate it.

@Mike
Independent from this request I think it would be useful if there would be a way to define different source files in the cvs for an import. Maybe an extra column for the source file name or something.

[sciurius]

Independent from this request I think it would be useful if there would be a way to define different source files in the cvs for an import. Maybe an extra column for the source file name or something.

Exactly. And not just for import, but also as a means to bulk-edit metadata after import.