Match mixed alphanumeric text with spaces between tags, no grouping, not a number, no garbage chars

[a-zA-Z0-9]+

basically any mix of lower and uppercase letters and digits, but it can't be only a digit (must have letters).
Does that answer your question?

No, it is not one regex. You have to strip all unwanted text first. Here is Perl:

$tmp = $input;
$tmp =~ s/<[^>]+>/\s/g; # strip HTML
$tmp =~ s/\&\w+;/\s/g; # strip special characters
$tmp =~ s/\s[\d\.]+\s/\s/g; # strip raw decimal digits
$output = $tmp; #Done

I think it can be done with one regex?

This gets pretty close:
(?<![\w])[a-zA-Z0-9]+(?>)

See how it matches each individual string of characters? For example: "1st" then "third" then "word" then "Quarter" (total of 4 matches). I want it to match the whole thing, for example: "1st third word Quarter" would be matched (for a total of 1 match).

Anyone know how to do that?

this almost does it:

(?!>| )(\b[0-9]?[a-zA-Z]+ ?)+<

but it matches "<" at the end...how do i get rid of this?

[Contents of the file dinesh.txt]
<tr>
         <td width="140" class="datahl2">&nbsp;</td>
         <td width="90" class="datahl2b">Offense</td>
         <td width="90" class="datahl2b">Defense</td>
         <td class="datahl4b" width="90">&nbsp;</td>
         <td width="90" class="datahl2b">Offense</td>
         <td width="90" class="datahl2b">Defense</td>
       </tr>
       <tr>
         <td class="datahl2">
           <strong>Points</strong><br>
             >1st third word Quarter<br>
             2nd Quarter<br>
             3rd Quarter<br>
             4th Quarter
         <td class="datacell">
           <strong>96.4</strong><br>
           24.7<br>
           23.2<br>
           21.9<br>
           25.1</td>
[Contents of the file dinesh.txt]


To show the expected output at shell on linux oriented platforms or at at "Cygwin bash shell" on Windows :
              sed 's/<.[^>]*>//g;s/&nbsp;//g;s/[0-9]*\.[0-9]*//g;s/>//;s/ *//;' dinesh.txt | egrep -v "^$"

This can also be done with the input given at the terminal itself Command:
sed 's/<.[^>]*>//g;s/&nbsp;//g;s/[0-9]*\.[0-9]*//g;s/>//;s/ *//;' | egrep -v "^$"
Press enter
Give the input.
Press enter
Press Control D
This will show the output.

Murug, Sorry, I am using C# and your regex engine and syntax are not familiar to me.

It is starting to look like no one will answer this question. However, I will try out eRescuers method (although I wanted a single regex). Perhaps it will work and I can give him credit.

Re:
(?!>| )(\b[0-9]?[a-zA-Z]+ ?)+<

but it matches "<" at the end...how do i get rid of this?

Essentially if you use a non-capturing match (positive look-ahead) on the <-symbol that would fix your regexp?
(?!>| )(\b[0-9]?[a-zA-Z]+ ?)+(?=<)

Hope this helps ;)

 Martin

mgh,

That's just about it! If you can show me one more thing I will award you the points.

sample input:
           <strong>Points</strong><br>
             1st Quarter<br>
             2nd Quarter<br>
             3rd Quarter<br>
             4th Quarter
         <td class="datacell">

Regex I tried: (mgh's)
(?!>| )(\b[0-9]?[a-zA-Z]+ ?)+(?:<)

Sample output:
Points<
1st Quarter<
2nd Quarter<
3rd Quarter<

See how it doesn't match 4th Quarter (because '<' is on the next line)?

Is there a way to match 4th Quarter?

Try:

(?!>| )(\b[0-9]?[a-zA-Z]+ ?)+(?:\n?<)

or

you can you /s modifier at the end like

/(?!>| )(\b[0-9]?[a-z]+ ?)+(?:<)/igsm

I've been wondering about how to use those options. However, you are using what looks like PERL. I am using C# and do not know the notation to use /igsm. Your second option does not work for me because of this.

Your first option doesn't match "4th Quarter" so it also doesn't work.

If your solution does work I need to figure out how to use /igsm with C#. Anyone know?

You've to set Multiline, Global, IgnoreCase and SingleLine options to be true in C#