Combining images to PDF in folder based on prefix

Question

I have an arbitary folder with images in them, that are named like this:
00_01.tiff
00_02.tiff
00_03.tiff
00_03.tiff
01.tiff
02-1_01.tiff
02-1_02.tiff
02-2_01.tiff
02-2_02.tiff
I want to combine all the files starting with 00 into a new pdf 00.pdf, all starting 01 to 01.pdf,  all starting with 02-1 to 02-1.pdf and all starting with 02-2 to 02-2.pdf
In short the rule would be:
if text 3 in fileName = "_" then
     set newFileName to text 1 thru 2 of fileName
else
     set newFileName to text 1 thru 4 of fileName
end if
It is easy to combine files to PDF's with Automator, but introducing this logic and doing the combining in Applescript is out of my league.
Does anybody have some suggestions on how I should go about solving this?
Best regards,
Knut

Answer 1

Jacques Rioux wrote:
Does the script should search in the subfolders?
No, I did not intend it to search through subfolders. I have thought about maybe first putting each group of files in to subfolders, and then iterate through the folders and do the PDF conversion then. That would be quite close to what the script you wrote is doing. More about that at the end of this reply.
Do you have multiple image formats (the name extension) to find in the source folder or in subfolders?
No, there will only ever be .tiff images in the folder. The contents of the folder is created by another script (as a part of a bigger workflow).
Do you have any PDF documents or PDF images  to find in the source folder or in subfolders that you want to combine?
No, there will only be images in .tiff format. I mistyped in my previous post - it should have said .tiff there too, and not .pdf. Sorry about that.
xx.tif and xx_y.tif : Is it the number in xx can be the same or always different?
If the answer is not always different, (11.tiff and 11_2.tiff) Does the script should create one PDF or two PDF?
If the answer is two PDF, how do you rename the new PDF file? --> (11.pdf and 11_2.pdf) both must be renamed "11.pdf"
I will remove the xx.tiff as a possibility. We will only have xx_yy.tiff or xx-z_yy.tiff to care about, and not xx.tiff.
I will give you some background information, maybe it is easier when understanding the use case. You can skip to the table below if this is not relevant. The reason for this naming convention, is so that yet another script can finalize the workflow later, when doing renaming based on the output name of the created PDF-file. My workflow is for digital archiving of sheet music.
The entire workflow is like this:
Scanning the sheet music.
Converting pdf-files to tiff.
Cleaning, straigthening and convert from grayscale to b/w with ScanTailor
Rename output files to the naming scheme I describe.
Convert the tiff images to pdf, per part.  (this is the part I'm asking about here)
Renaming all the files to fit the archiving scheme.
Grouping and sorting the files, and creating zip-archives.
Archiving the entire digital copy of the set in our digital archive.
For our case, we're only interested in item 5 in the list. Another series of scripts and Automator workflows take care of the rest.
Each music piece consists of several parts.
In the filing system I have created, each part has a prefix with the xx-number, like this:
00 = Score
01 = Soprano cornet
02 = Solo cornet
17 = BbBass
18 = Percussion
The final files will have a name like (only an example):
02_SoloCn_31_SilentNight_Gordon.pdf
For some pieces there are more than one solo cornet part. That's why we get the z-component, to differentiate between the parts:
02-1 = Solo cornet 1 --> will get filename 02_SoloCn1_31_SilentNight_Gordon.pdf
02-2 = Solo cornet 2 --> will get filename 02_SoloCn2_31_SilentNight_Gordon.pdf
Many pieces has more than one page per part. That is where the yy component of the file name comes in, the only reason for the yy component is to keep all the pages in the correct order to make it easy to combine them programatically:
02-1_10.tiff = Solo cornet 1, page 1
02-1_11.tiff = Solo cornet 1, page 2
02-1_12.tiff = Solo cornet 1, page 3
Here is a table with the pattern, example of a possible filename and what the output pdf should be called:
Pattern
Example file name
Output pdf
xx_yy.tiff
01_10.tiff
01_11.tiff
01_12.tiff
01_13.tiff
01.pdf
xx_yy.tiff
12_10.tiff
12_11.tiff
12_12.tiff
12.pdf
xx-z_yy.tiff
02-1_10.tiff
02-1_11.tiff
02-1_12.tiff
02-1.pdf
xx-z_yy.tiff
02-2_10.tiff
02-2_11.tiff
02-2_12.tiff
02-2.pdf
Is it possible that there would be other characters after the pattern? --> example: 33_2test.tiff, 40-4_99ball.tiff
No, all the files will have just the components described above in the file name. No need for exception handling, since this is a tightly controlled workflow.
As i mentioned earlier in this reply, I have played with the thought of first running a script to sort all the files for one part into a subfolder with the correct name.
This I can do with the following script (I guess it would be more efficient with shell script or using System Event to do the moving):
          tell application "Finder"
set thisfolder to (choose folder)
set mylist to every file in thisfolder as alias list
repeat with eachfile in mylist
set filename to name of eachfile
if (character 3 of filename is "-") then
set prefix to characters 1 thru 4 of filename as text
else
set prefix to characters 1 thru 2 of filename as text
end if
try
set newfolder to ((thisfolder as text) & prefix) as alias
on error -- file not found
set newfolder to make new folder at thisfolder with properties {name:prefix}
end try
move eachfile to newfolder
end repeat
   end tell
After this, all the files have moved to folders with the folder name corresponding to the part number, like this:
01/01_10.tiff
01/01_11.tiff
02-1/02-1_10.tiff
02-1/02-1_11.tiff
and so on...
It would perhaps be easier to extend on this script, or create a new one, that will iterate each subfolder in a folder, and run the Automator action for combining images to files with each iteration?
What do you think? I'm sorry if this was too much information. Do not hesitate to ask if something is unclear (I bet it is!).
I really appreciate your help, Jacques!
Knut