text parsing in a field

[ken]

RE: text parsing in a field

Hi,

There is also a particular case when there is no space at the beginning of email and the semicolon is concatened with the email like in the foll example

E-MAIL:[email protected]

so the whole phrase is parsed like an email address.

Is there any possibility to consider the semicolon the left boundary of the email?

regards,
Corneliu

[ken]

RE: text parsing in a field

Hi Stan,

Is there any possibility to add a check box on the form, when checked to disable the "on_change" event of the field?

I mean that there are some cases when I want to disable the parsing facility of the field.

Regards,
Corneliu

[Steve Andrews]

RE: text parsing in a field

Corneliu,

As for the emails, it now picks up the appropriate substring (not always the last one). That was accomplished using filter_string() as mentioned above. Likewise, it now accounts for when substrings are preceded by a colon rather than a space. I simply used a strtran() within a *for_each() to place a space after every colon.

As for the web addresses, however, they still pick the last space-delimited word on the line. Since I would have to pick the word which contains "hppt:", "www:" or ".net" it gets rather complicated again. Maybe you can make do and do manual changes for web addresses that are not on the end of a line.

Steve

[ken]

RE: text parsing in a field

Hi Steve,
Thanks a lot for the new version.
generaly it works correctly but however I encountered some strange situations like follows

-----

ZEINEB B.Y F / MOB: 216 98 370 777
E.Mail : [email protected]
E.Mail: [email protected]
please visit : www.mmuagroup.net

-----

where an email address is parsed as www2

-----

Sale & Purchase : [email protected]

-----

is parsed both as an email and an web address

Maybe you can figure out what happens

Regards,
Corneliu

[Steve Andrews]

RE: text parsing in a field

Corneliu,

I cannot reproduce the "www2" that you got parsed out. As for the items that come back as both emails and web addresses, maybe it would be wise to remove the search for ".net" on the web part of the script. That means that web addresses would be identified solely by the substrings of "hppt:" or "www:".

Steve

[ken]

RE: text parsing in a field

Hi Steve,

Also the foll are not parsed correctly

email: [email protected] or [email protected]

-------

Email: [email protected] or [email protected]

Regards,
Corneliu

[ken]

RE: text parsing in a field

Yes, you are right
after removing the search for ".net" on the web part
the emails ending in .net were parsed correctly.

Thanks and regards,
Corneliu

[Steve Andrews]

RE: text parsing in a field

Corneliu,

With all of the different formats you are encountering, you need to first to establish them all. Deciding on how to approach a problem depends on assumptions made as to how the input data can be laid out.

First we assumed that the last word on a given line would be what you wanted and then you came up with examples where that was not the case. First we assumed that the desired substring would be space-delimited and then you showed a case ("Email:[email protected]") to the contrary. First I assumed at most one entity per line and now you show some double entities.

This could go on and on but I am not taking any more looks at this unless every possibility is fully documented. If someone else wants to take a look at this they are more than welcome.

Steve

[ken]

RE: text parsing in a field

Hi Steve,

You are absolutely right but unfortunatelly I can not control the way the messages I receive are conceived.
Unfortunatelly everyone finds a new way to write a message that was never encountered before.
However you did a great job and your routine is close to perfection. I am sure that the error percentage is a way under 1% which is absolutely amazing.
The slight errors that may occur can be correctly manually.
For example I use 3 different intelligent antispam engines working in a chain and even so there are occassions when some spam messages find their way into my mailbox.
Thank you very much for all your hard efforts and your patience.

One more thing I kindly ask you. Can you show me how to parse the first line of the message to the field COMPANY.
In this case it is easy as I will take care
that the first line is always the company name and it will always end with a space or CRLF.
Just to check that this line is not empty, and does not contain any numbers, or @, or http or www.

Regards,
Corneliu

[ken]

RE: text parsing in a field

Hi Steve,

Just asking an explanation about the code.

in the web searching routine you kept
"line_count(emails)". This is the way to be or should it be changed to "webs". The fact is that it is working correctly but I just wish to understand myself.

Regards,
Corneliu


if line_count(emails)"0
company_input:website.value = word(word(webs,1,crlf()),-1," ")
end if

[Steve Andrews]

RE: text parsing in a field

You're right. It should be:

if line_count(webs)"0
&
if line_count(webs)"1

Steve

[ken]

RE: text parsing in a field

OK, thanks again for the beautiful piece of code.
Regards,
Corneliu