I have a need to strip all html code from a string (and thereby leaving the text intact). There are several functions out there

http://www.planet-source-code.com/vb/scripts/ShowCode.asp?txtCodeId=6269&lngWId=4, strip.tags (php) etc. that can do this.

What would be the best way to accomplish calling these external functions?

Here's how you would code the link you gave. I did not test this…

script = 'function StripHTML(asHTML)'

script = 'Dim loRegExp'

script = ''

script = 'Set loRegExp=New RegExp'

script = 'loRegExp.Pattern=*]"'

script = 'StripHTML=loRegExp.Replace(asHTML, "")'

script = 'Set loRegExp=Nothing'

script = 'End function'

SWAP @FM WITH CHAR(13):CHAR(10) IN script

oWscript=OleCreateInstance('ScriptControl')

OlePutProperty(oWscript, 'Language', 'VbScript')

zz=OleCallMethod(oWscript, "AddCode", script)

HTML=HTML text'

strippedHTML =OleCallMethod(oWscript, "Run", "StripHTML", HTML)

oWscript='

The regular expression engine is itself an OLE object that you can use inside OI. This offers interesting possibilites for indexed symbolics.

the progid is

oRegEx=OleCreateInstance("VBScript.RegExp")

For a code example see

OleregExp

Thanks Hank.

I tested this out and it only stripped the front part.

I passed it "Hello" and it returned

"Hello".

Hey Bob.

I'm not sure I follow what you saying in this post. Can you expand?

Thanks.

I ran it as vbscript outside of OI and it did the same thing. I suspect it has more to do with the vbscript code.

Hi Bruce

I was trying to say that you can call the regular expression engine from OI. In the example below I run the HTML through several filters.

Note - I just googled filters without understanding them, the ouptut is not exactly what you need. More time with google and a regular expression tutorial will get you a better result.

The general concept is that regular expressions do sophisticated searching and matching on any text. One could create calculated fields that are very smart - like finding repeated words, parsing email or addresses from raw text, find multiline comments, etc. Note that regular expressions are difficult to master and OI has good string handling, so we can usually avoid using the expressions. But it is nice to know they are available.

function StripHTML_Eqample(HTML)

/*

Whipped together from a bunch of googled regulars expressions

*/

Equ true$ To 1

Declare Function Ole_GetWebPage

html=Ole_GetWebPage('http://www.revelation.com')

If Assigned(html) Else html='

oRegEx=OleCreateInstance("VBScript.RegExp")

oRegEx-]Global=true$

oRegEx-]IgnoreCase=true$

* Kill Scripts

* See http://regexblogs.com/blogs/dneimke/archive/2004/01/17/200.aspx

oRegEx-]Pattern=*].*?*]"

work= oRegEx-]Replace(HTML, "")

* Kill css

oRegEx-]Pattern=^\t*a-zA-Z0-9\.# -_:@+\t*\{.*\t*$"

work=oRegEx-]Replace(work, "")

* Kill Comments

* http://www.webmasterworld.com/forum88/11584.htm

oRegEx-]pattern=/(\/\*(?!()).*?\*\/¦\/\/.*?\r\n+¦\r\n+)/s'

work=oRegEx-]Replace(work, "")

* Kill HTML tags

oRegEx-]Pattern=*]"

ans=oRegEx-]Replace(work, "")

Return ans

Bob, I gotcha, that's great info. Thanks.

Bob,

A continuation of the question…

I can take and open an xxxx.html file in IE for example.

I can then do a "File -] Save As" plain text (txt) and all the

html code is stripped.

Is there a way to do that in OI using the ole Shell.Explorer?

Just a thought.

Bruce

Oh, you meant that way

Make a form with an ole control named browser, a textbox named RESULT

add a button with this script:

url=http://www.google.com"

Call Send_Message(@window:'.BROWSER', 'Navigate2', url)

loop

call yield()

readystate=get_property(@window:'.BROWSER', 'ReadyState')

until readystate=4

repeat

text=Get_Property(@window:'.BROWSER', 'Document.Body.OuterText')

x=Set_Property(@window:'.RESULT', 'TEXT', text)

That is sweet.

I modified it so that in a function I write my string out to "c:\temp\xxx.html" then start the window invisible, then put the "c:\temp\xxx.html" in for the url and do a send_event click for the code and grab the .RESULTS. Works for me for now! Thanks Bob!

If I wanted to convert your example into a function would it be…

Function StripHtml(datastring)

*

oleObj=OleCreateInstance('Shell.Explorer')

status=OleStatus()

if status then return 'OLE Error code: ':status

* It doesn't work here on the method, not sure what is the correct way

ReturnValue=OleCallMethod(oleObj, "Navigate2","www.google.com")

status=OleStatus()

if status then return 'OLE Error code: ':status

loop

 call yield()

 readystate=OleGetProperty(oleObj, 'ReadyState')

until readystate=4

repeat

strippedstring=OleGetProperty(oleObj, 'Document.Body.OuterText')

status=OleStatus()

if status then return 'OLE Error code: ':status

Return strippedstring

If I wanted to convert your example into a function would it be…

Function StripHtml(datastring)

*

oleObj=OleCreateInstance('Shell.Explorer')

status=OleStatus()

if status then return 'OLE Error code: ':status

* It doesn't work here on the method, not sure what is the correct way

ReturnValue=OleCallMethod(oleObj, "Navigate2","www.google.com")

status=OleStatus()

if status then return 'OLE Error code: ':status

loop

 call yield()

 readystate=OleGetProperty(oleObj, 'ReadyState')

until readystate=4

repeat

strippedstring=OleGetProperty(oleObj, 'Document.Body.OuterText')

status=OleStatus()

if status then return 'OLE Error code: ':status

Return strippedstring

Bruce,

Shell.Explorer wants an owner window to display the HTML.

It will not work as a plain OLE object. You need to use InternetExplorer.Application instead

For Example

Subroutine Ole_GetWebText(void)

/*

** Pull up a web page, strip out the text

*/

IE=OleCreateInstance('InternetExplorer.Application')

status=OleStatus()

if status then

return 'OLE Error code: ':status

End

IE-]Visible=0

x=IE-]Navigate("http://www.google.com")

Loop

Call Yield()

while IE-]Busy

repeat

doc=IE-]Document

body=doc-]Body

result=body-]InnerText

x=IE-]Quit()

Return result

Bob,

Perfect! I changed it to a function and added an arg for datalocation so you can put in a url or pathfilename.

Thank you, again.

bc

View this thread on the Works forum...