Sorry for the thread title.

Here's what I'm trying to do. I need some arbitrary names for a database project I'm
working on. I got some stuff off the net but I need to get rid of the html code.
So I have a file with names like this.

<tr bgcolor="white"><td>MARY</td><td>2.629</td><td> 3,991,060 </td><td>1</td></tr>
<tr bgcolor="#f5fdd9"><td>PATRICIA</td><td>1.073</td><td> 1,628,911 </td><td>2</td></tr>
<tr bgcolor="white"><td>LINDA</td><td>1.035</td><td> 1,571,224 </td><td>3</td></tr>
<tr bgcolor="#f5fdd9"><td>BARBARA</td><td>0.98</td><td> 1,487,729 </td><td>4</td></tr>

So what I want to do is use sed, awk, grep.... to search for "<td>" and delete
it and everything before it and then search for "</td>" and delete it and everything
after it leaving only the name. I hope that makes sense.


Thanks for your time and help.

Zackarya

I suggest using the "-dump" option of lynx or w3m, rather than doing this yourself.

ataraxia, thanks for the reply.
I've used lynx -dump before but the site has extra information on the same line as the
names that I want so I'm left with still having to go and delete those fields out.

I know of a couple ways I could hack this to make it work but from time to time I've run
across situations where I need to do something that's similar to this problem and I've
always had to basically hack around it because I don't know of a "good" way to do it.

Zackarya

There's few variations on stripping out HTML tags in Perl Cookbook eg:

Use HTML::Parser;
use HTML::FormatText;
$plain_text = HTML::FormatText->new->format(parse_html($html_text));
Then you can just grab the fields you want.
This might be a good starting point.

Thanks chrism01.

I ended up do what I usually do and just use a bad hack.
I'm tyring to find a way to do what I outlined in my first post but in a very
generic way. It's not always html I'm trying to get stuff out of.

For this case what I did was:

sed -e 's/<tr bgcolor="#f5fdd9"><td>//' -e 's/<tr bgcolor="white"><td>//' <most_common_surnames.txt >surnamesfirst.txt

Which strips out everything before the name.

Then
sed 's/</\n/' <surnamesfirst.txt >surnamessecond.txt

Which finds the character after the name and replaces it with a newline.
NOTE: I could have done this with the -i option of sed but I wanted to make sure
I didn't make any mistakes.


Then
awk '(NR%2==0) {print $0; }' <surnamessecond.txt >surnamesthird

Which prints out every second line and outputs to a file. (There's a space as the first line.)


That's basically how I did it this time but it's such a specific solution.
If anyone knows of a "cleaner" way to do this please let me know because I
run across this kind of thing fairly often and would like a more generic
approach.