Extract words from JSP into text file

[Michelle]

Hi,

I have a big problem:

I want to extract selected words from a jsp file.
Following are found in one of my jsp file:

E.g.
1) <td width="217"><font face="verdana,arial" size=2>ORGANISATION UNIT NAME<font
color="#FF0000"> <font
id="fontMandatory">*</font></font>

I want to retrieve "ORGANISATION UNIT NAME"

2)errPrompt(frm.txtDesc, "Maximum length is just 100 character only");

I want to retrieve "Maximum length is just 100 character only"

I try String tokenising, then check that is the token ends with ">", then
the next token is the one
I want, then loop until "<" is found. BUT this would not work as there is
no spacing between some of
the tags and the words.(e.g. <p>abc)--> This cauese the whole token to be
<p>abc, so the "abc" will
not be extracted as it does not have a "ends with >" in front of it.
Even with using ">" as a checking does not work for pop up messages, as it
does not have tags(refer
to e.g 2).

Please reply a.s.a.p...
Really urgent!!
Thank You

Michelle

[Paul Clapham]

If you set up your StringTokenizer correctly, you should be using "<>" for
your delimiters. You should also be using the constructor that allows you
to get the delimiters back as tokens. Then, when you get ">" as a token,
you know the next token is one of the things you want.

PC2

"Michelle" <michellelimmx@hotmail.com> wrote in message
news:3c746530$1@10.1.10.29...
>
> Hi,
>
> I have a big problem:
>
> I want to extract selected words from a jsp file.
> Following are found in one of my jsp file:
>
> E.g.
> 1) <td width="217"><font face="verdana,arial" size=2>ORGANISATION UNIT
NAME<font
> color="#FF0000"> <font
> id="fontMandatory">*</font></font>
>
> I want to retrieve "ORGANISATION UNIT NAME"
>
> 2)errPrompt(frm.txtDesc, "Maximum length is just 100 character only");
>
> I want to retrieve "Maximum length is just 100 character only"
>
> I try String tokenising, then check that is the token ends with ">", then
> the next token is the one
> I want, then loop until "<" is found. BUT this would not work as there is
> no spacing between some of
> the tags and the words.(e.g. <p>abc)--> This cauese the whole token to be
> <p>abc, so the "abc" will
> not be extracted as it does not have a "ends with >" in front of it.
> Even with using ">" as a checking does not work for pop up messages, as it
> does not have tags(refer
> to e.g 2).
>
> Please reply a.s.a.p...
> Really urgent!!
> Thank You
>
> Michelle

[foo]

Michelle,

If this is a 1-time thing (say, as an attempt at internationalization), you
may be better off using Perl rather than Java as your extraction language.

For example:

while (<>) {
@matches = />([^<]+)<)/g;
}

Will read in a passed-in file line by line and for each line return an array
of >...< matches.

IMO what you're going to come up against is that you'll be hard pressed to
come up with parsing rules that are specific enough to cull out just the
text you want.

--A