UNIX for Dummies Questions & Answers

I am trying to grab some text from an HTML file.

I have something similar to this:

Code:

<tr>
<td><a  href="somepath.html">Some Text<HR></HR></a></td>
</tr>
<tr>
<td><a  href="somepath.html" >Some More Text<HR></HR></a></td>
</tr>

And just want the bolded text. Any suggestions?

You can try this.

cat a.txt |awk -F">" '{print $3}' | awk -F"<" '{print $1}'

Code:

awk  'NF!=0'  RS="<[^<>]+>" filename

Thanks Rubin, that worked great! Now I need to figure out how to store each line as a new line. When I just plain run the script and look at the output in the command prompt, it looks fine. If I try and pipe it to a text file, its all mashed together.

Any suggestions on how to separate it?

If you have lynx:

Code:

lynx --force-html --nolist --dump file

Or if you have html2text:

Code:

html2text file

When I try and run this through sed to remove all numbers that are in my file, I lose my line breaks. Any idea on why or how to keep them?

sed 's/[0-9]*//g'

is the command I am using.

Can you post the input file ( the processed output file ) before running the sed command, and your expected / final output ?

I tested the command with a bigger sample file, redirected the output to another file, saved it as a text file, and it opened just fine either in Linux or Windows, with their respective editors and with no line breaks. Note though that the command works fine if a record has one part to extract as shown in your sample, if there are more, then each extract will be on a newline. But the command can be changed to handle that too. Post a more complete sample if needed.