Shell Programming and Scripting

Here is I want to do:
find lines in file filteredtrace.txt which is not continued as
a multiply of 4 and strip them from file original_trace_framno.

problem is awk used the ' symbol
so pipe of getline has to use the " symbol
while agument of sed has to use the " symbol,
it doesn't recongize the ' symbol.

Code:

awk 'BEGIN{old=0}
{
    if(NR>5) 
    {   
        while($9>old&&$9-old!=4) 
        {   
            a=old*2+6; 
            print a;
            "'sed -n "${a}pi" original_trace_framno'"|getline al; 
            print al; 
            a++;
            "sed -n '${a}p' original_trace_framno"|getline al; 
            print al; 
            old +=4;
            exit
        }   
        if($9<=old)
            print $0; 
        else if($9>old&&$9-old==4) 
        {   
            old=$9;
            print $0
        }   
    }   
    else 
        print $0
}' 
filteredtrace.txt

part of original_trace_framno (vim with 'set nu' )
78 0x00009f7c 9 0 0 0 SliceData No No 40 0
79 0x00009f85 2976 0 0 0 SliceData No No 40 0
80 0x0000ab25 9 0 1 0 SliceData Yes No 38 0
81 0x0000ab2e 713 0 1 0 SliceData Yes No 38 0
82 0x0000adf7 8 0 2 0 SliceData Yes No 37 0
83 0x0000adff 383 0 2 0 SliceData Yes No 37 0
84 0x0000af7e 8 0 2 0 SliceData Yes No 39 0
85 0x0000af86 400 0 2 0 SliceData Yes No 39 0
86 0x0000b116 9 0 0 0 SliceData No No 44 0
87 0x0000b11f 2972 0 0 0 SliceData No No 44 0
88 0x0000bcbb 9 0 1 0 SliceData Yes No 42 0
89 0x0000bcc4 638 0 1 0 SliceData Yes No 42 0
90 0x0000bf42 8 0 2 0 SliceData Yes No 41 0
91 0x0000bf4a 415 0 2 0 SliceData Yes No 41 0
92 0x0000c0e9 8 0 2 0 SliceData Yes No 43 0
93 0x0000c0f1 325 0 2 0 SliceData Yes No 43 0
94 0x0000c236 9 0 0 0 SliceData No No 48 0
95 0x0000c23f 2568 0 0 0 SliceData No No 48 0

part of filteredtrace.txt (vim with 'set nu' )
38 0x00009f7c 9 0 0 0 SliceData No No 40 1317630108771
39 0x00009f85 2976 0 0 0 SliceData No No 40 1317630108771
40 0x0003f006 9 0 0 0 SliceData No No 352 1317630119121
41 0x0003f00f 977 0 0 0 SliceData No No 352 1317630119121
42 0x0003f537 9 0 0 0 SliceData No No 356 1317630119205
43 0x0003f540 1138 0 0 0 SliceData No No 356 1317630119205
44 0x0003fba4 9 0 0 0 SliceData No No 360 1317630119300
45 0x0003fbad 1300 0 0 0 SliceData No No 360 1317630119300
46 0x000402f0 9 0 0 0 SliceData No No 364 1317630119375

need to add lines which has $8 of 44 48 ... 348 from original_trace_framno to file filteredtrace.txt.

What are those seds supposed to be doing? You know sed will only get run once, before awk even runs, not once per loop right?

filteredtrace.txt
I have to modify the suffix to .txt
originaltrace-frameno.txt

I have to print 2 lines at one time for data processing reason.

You answered me twice and still didn't really answer my questions... Modify what suffix? what is "a"? Is this supposed to change every loop or not?

Try getline <"filename.txt" instead of "something | getline".

Or better:

Code:

awk -v FILE="filename" '{ ...  getline < FILE }'

instead of embedding shell statements inside your awk statement.

I modify the original_trace_framno to original_trace_framno.txt in order to upload the attachment.
"a" is described the relation between line number and $8(frame number) in file original_trace_framno.txt.
the change happened only when frame number is not contiued as a multiply of 4.
file filteredtrace.txt is genarated from original_trace_framno, with some of lines missing. During these missing lines, some are very important and need to be recovered from original_trace_framno. for example,
In file filteredtrace.txt line with $8==44 is missing, so I need to strip it from original_trace_framno.

I don't think that crossed the language barrier.

You can't call sed inside awk.

You can't pipe inside awk either.

---------- Post updated at 11:08 AM ---------- Previous update was at 11:06 AM ----------

Could you show the actual output you want too, not just your input files?