Shell Programming and Scripting

I am trying to setup to automatically import a series of mysql database files. I am doing manually now and its a royal pain.

All the sql files are sequentially numbered in a format of 4 numbers underscore text with spaces replaced by underscores.

example:
There are 3 databases each setup in the same fashion. The main directory has the database structure file higher sequence replacing lower sequence. Each database has its own subdirectory with sequenced updates the files with numbers higher than the structured files need to be loaded in numerical sequenced. There are duplicate sequence numbers in the update directories those all have to be loaded before the next number.

I figured I can do this to gain the proper structure file as they do not have duplicate numbers.

I tried the below code getting rid of the directory name with cut but they are in the wrong order, they are in alphabetic not numeric order.
The variable h has the number alone and g has the full file name .
I lack a lot of basic knowledge under linux that I had in a dos/windows environment.

I have been looking at this code so long I can't see the forest for the trees, if you know what I mean. So any suggestions to get me on the right train of thought would be greatly appreciated

Thanks,
dlm1065

If you only use the output from find once, you don't really need to put it in a variable.

My suggestion would be to extract the sequence number and the database to separate fields, and sort numerically on the sequence number.

Code:

find logon_updates -name "*.sql" -print |
while read f
do
  g=$(echo $f | cut -b15-)
  h=$(echo $f | cut -b15-18)
  echo $h:$g
done | 
sort -t : -k1n

Once you are confident that this works correctly (I don't have the data to test on), you can continue the pipeline:

Code:

... sort -t : -k1n |
cut -d: -f2- |
while read f; do
  sql PERFORM ACTS OF horror WITH "$f" USING BIG STICK
done

If the output from find is fairly regular, probably you could find some clever options which would allow you to pass that directly to sort without the while look. Maybe pass the output from find through sed to temporarily regularize it?

Maybe something like this would work already?

Code:

find -name "*logon_structure.sql" -print |
sort -t / -k2n