Richard Mix wrote:Max, could you point me to that discussion?
The discussion is
here, however it is only visible to admins (so you can read it).
In a nutshell, I created a spider that visits ChoralWiki and creates a database structure where logical relationships between various items are represented. In a nutshell:
- each file (e.g., pdf, MIDI, source) relevant to a certain CPDL edition is identified with an individual file ID (format xxxxx.yy) where the integer part is the CPDL edition number and the decimal part is a specific file progressive number within the edition file set. For example, files pursuant to edition number 12345 are identified as 12345.1, 12345.2, and so on;
- for each file, the file "class" is identified (score, sound, or source), as well as the file type (pdf, midi, mp3, Finale, ...), and also the associated icon to be displayed;
- each edition is associated to the corresponding work name and composer name;
- other edition information (editor, submission date, copyright, etc.) are also captured and stored in the database.
The database is currently being used (mostly by Claude) for maintenance purposes, as one can extract information from it much more easily than looking for keywords in the wiki page text.
Richard Mix wrote:If to obtain means to click, then this is just counting files, something that could probably be accomplished without requiring contributors to use the Add Work form for every midi, pdf, letter-size formatting...
You're right, and this is actually what I've done already by assigning a file ID to each file. File IDs may be easily converted to ISMN numbers. Then, creating and maintaining the logical relationships between all items requires the Add Work form and all the rest, and this is a completely different subject than just numbering editions and files.
Max