Welcome to LQ!
Please note that "Urgent" does not make us move faster---some people will not even open the thread with such a title.
You have several constructs in here:
variable substitution (eg $ALLMOD)
sed addressing (eg sed "/address1/,/address2/d"---deletes everything from address 1 to 2)
escaping to force literal interpretation (eg \"Core Rect Sources")
To decipher a particular expression, look for the basic construct, and then take it apart. I'll do one example:
sed "/^# Begin Group \"Core Rect Sources\"/,/^# End Group/d;/^# Begin Group \"Core Rect Headers\"/,/^# End Group/d" $TMPLIST > $OUT_FILE
This has the basic form:
sed "/addr1/,/addr2/d" $var1 > $var2 (there are two instances of the (address range/d) construct, separated by the semicolon.)
The above decodes to:
read from the location stored in var1
In the range of addr1 to addr2, delete all lines
write to the location stored in var2
Now decode the first address:
/^# Begin Group \"Core Rect Sources\"/
At the beginning of the line (^), find this phrase: # Begin Group "Core Rect Sources"
The escaped quotes are required to force them to be interpreted correctly in this context. (Read more on quoting in a good BASH manual)
For a good SED tutorial:http://www.grymoire.com/Unix/Sed.html