Shell Programming and Scripting

I've never written scripts (just switched from Ada to C++). I have a book that's over my head and a few examples, other then that I'm floundering. Everything here at work is being done in C Shell. None of the C++ programmers are experienced in shell scripting.

I have a data file with the following format, "key" words in bold:

-------------------------------------------
{header info: 20 lines}

DATA_POINT 99
DATA_HEADER
TYPE: AppendNewRecordStruc
NAME: AppendNewRecordStruc
TITLE: AppendNewRecordStruc
SOURCE: N/A
END_DATA_HEADER
DEFINITIION
SOURCE_FILE: K:/home/source_code/my_code.C
DATA: append_new_struc
END_DEFINITION
END_DATA_POINT

DATA_POINT 100
DATA_HEADER
TYPE: DeleteRecordStruc
NAME: DeleteRecordStruc
TITLE: DeleteRecordStruc
SOURCE: N/A
END_DATA_HEADER
DEFINITIION
SOURCE_FILE: K:/home/source_code/my_code.C
DATA: del_struc
END_DEFINITION
END_DATA_POINT

END_FILE
-----------------------------------------------

After I run this through an in-house tool, I get a log file listing the errors. However the log is thousands of lines, and only some of these lines are relevent to me. For example:
-----------------------------------------------

{header}

The entry for Data_Point #99 could not be processed.
The entry for Data_Point #100 could not be processed.
The entry for Data_Point #101 processed successfully.
{irrelevent error messages}
The entry for Data_Point #212 processed successfully.
The entry for Data_Point #211 could not be processed.
{irrelevent error messages}
-----------------------------------------------

I'd like to take the log file, cross reference it back to the original data file and come up with something useful. A new file listing only the failed points with the format

POINT..........NAME....................................DATA
99...............AppendNewRecordStruc............append_new_struc
100..............DeleteRecordStruc..................del_struc

...'s are blank spaces


Any help would be greatly appreciated.

Steve

need a little more info... but here's how you might grab points of interest and then grep them from another file:

Code:

#!/bin/ksh

for point in `grep "The entry for Data_Point.*process.*" log | sed -e 's/^.*#//' -e 's/ .*$//'` ; do

print "POINT NAME $point DATA"
grep "^$point " point_data.file

done

Thanks. That will give me a starting point at least, if not more. I modified the original output to be clearer. It had formatted oddly, the confusion was my fault.