Ok, I have some experience with awk, but I haven't touched it for years ... back again and I have one problem.

I have large char file, it has hundreds of major text blocks. First, I need to grab and process those, and ignore everything else. That bit is simple.

Within those major text blocks, there are 3 to 32 minor text blocks. Second, I need to process those. Here's what the data looks like this:

Code:

major_start
major data

minor_start
minor data 1
minor_end

minor_start
minor data 2
minor_end

...
minor_start
minor data 99
minor_end

major_end

This is what I would like to code (of course it doesn't work). Note, I need bits from the major_block while processing the minor_block, such as var1:

Code:

/major_start/,/major_end/ {

    if ( $1 == "fat" ) { var1=$7 }

    /minor_start/,/minor_end/ {

        / 0:/ { array1[0]=var1; array2[0]=$3; var3=$3 }
        if ( $1 == "thin" ) { var2=$9 }

        }

    }

I was certain I have done this before (without leaving awk), but I cannot find evidence of that. I looked at my old awk scripts, they seem to process the major_block; then use if-then-else. But the need was not quite the same.

I searched this forum, and the closest question is "parsing-file-extracting-useful-data-block". (close but not quite, because that particular bit did not get answered). I an seeking the following:

1. Instructions to get awk to do what I want.

2. Alternately, can I capture everything in major_block and write that to a file_B; then process it with "system(awk ... file_B)"

3. Alternately, can someone who is an awk geek confirm to me that awk is a single-pass design, with no iteration.

TIA