I have mrtg and want to dump the current in/out to a text file.
Is there anyone knows how to grep certain row (like ROW3 and ROW4 only)?
awk can print certain column, is there a binary that can print certain rows? Like the ff, when I do "cat 202.84.22.161_5.html | egrep "kb/s|b/s" it shows the ff. How can I print ROW3 and ROW4 only?

Anyone suggest

<td>231.7 kb/s (9.7%)</td>
<td>25.0 kb/s (1.0%) </td>
<td>21.7 kb/s (0.9%) </td> --> ROW3
<td>180.1 kb/s (7.5%) </td>
<td>6160.0 b/s (0.3%) </td>
<td>7768.0 b/s (0.3%) </td> --> ROW 6
<td>241.8 kb/s (10.1%)</td>
<td>9952.0 b/s (0.4%) </td>
<td>24.1 kb/s (1.0%) </td>
<td>68.8 kb/s (2.9%) </td>
<td>2520.0 b/s (0.1%) </td>
<td>6848.0 b/s (0.3%) </td>
<td>183.2 kb/s (7.6%)</td>
<td>4608.0 b/s (0.2%) </td>
<td>23.7 kb/s (1.0%) </td>
<td>28.5 kb/s (1.2%) </td>
<td>1480.0 b/s (0.1%) </td>
<td>4384.0 b/s (0.2%) </td>
<td>28.2 kb/s (1.2%)</td>
<td>3432.0 b/s (0.1%) </td>
<td>14.4 kb/s (0.6%) </td>
<td>5288.0 b/s (0.2%) </td>
<td>1216.0 b/s (0.1%) </td>
<td>3552.0 b/s (0.1%) </td>

awk and sed can ...





Cheers,
Tink

This one works

This one didn't. I tried to show certain 3 and 6 but it print 4 lines.
The command I did is "cat test.txt | sed -n '3,6p'" and it print the ff:

<td>28.5 kb/s (1.2%) </td> --> ROW 3
<td>180.1 kb/s (7.5%) </td> --> ROW4
<td>6248.0 b/s (0.3%) </td>
<td>5936.0 b/s (0.2%) </td>

I tried this:

cat test.txt | awk ' NR == 1 || NR == 9 {print }' # 1 and 9 are the column

<td>231.7 kb/s (9.7%)</td>
<td>16.8 kb/s (0.7%) </td>

It works. The limitation here is how to get rid of <td> and show only 231.7 kb/s and 16.8 kb/s?
If only there is a space after the > , there could be no problem since awk/cut can remove this but since not all has no space, is there another way to remove them?

In sed & awk, you can precede a command to select just the lines you want, and then in sed use the substitute command, or in awk print out just the fields you need.

The example given before, "sed -n '3,6p' file", will actually print the lines between 3 and 6 inclusive.

This might be a better version:
Every line has the tags removed but only lines 3 & 6 are printed. When it gets to line 7, sed quits. This can save time if you have a very long file.

Note that there is a slight irregularity in your sample file. On the 19th line there isn't a space before the end tag. If this wasn't a mistake, then remove the space before the endtag in the sed program to handle such a case:
example:
Otherwise the tags will be in the output. This way however, there will be a space at the end of the line.

Or you could add another sed command to remove any trailing spaces:

Stop the cat'ing, will you? :}

If you find yourself using cat with only one argument you're
abusing it.

And 1 and 9 are the rows.



Cheers,
Tink