HttpSend Activity with Url Encode

I've been using HttpSend Activity and the querystring contains some special characters. I've found the querystring is not encode automatically by the HttpSend activity. I was wondering if there's any activity or way to UrlEncode the querystring? I've tried
BuildSPUri activity but it doesn't help.
Thanks,
Sohel Rana
http://ranaictiu-technicalblog.blogspot.com

Hi,
It's probably late for your problem but it could help some others, as I were 10mn ago :)
I had a similar problem. Use a HttpSend activity with the accountname of the current user (such as i:0#.w|domain\user).
With the BuildUri activity, I was able to obtain an encoded URL.
I set a web service URL as "URI.Source" parameter and set the "Query" parameter with
"accountname="+AccountName
Then, my URL was encoded and I could use it in the HttpSend activity.

Similar Messages

Problem with URL encoding conversion

Hi all,
I am working on an I18N application and in my application one component sends the request to another component and then this component fetch that requet and extract the query-parameters from the request (HTTP request).
Now the problem is that the input to first component can be given in one of the 5 character encodings:-
UTF-8
Shift_JIS
EUC_JP
Windows-31J
ISO-2022-JP
I have created a test program that convert the encoded URL from one character encoding to another character encoding. It is working successfully for the above 4 encodings but for the last encoding that is "ISO-2022-JP" this fails. The test programs is: -
import java.io.*;
import java.util.*;
import java.net.URLDecoder;
import java.net.URLEncoder;
class JPtoUTF8{
     public static void main(String[] args){
          try{
              String shift_jis = "%82%C8%82%A4%82%8B%82%E8";          // This is Shift_JIS encoded URL
              String iso2022jp = "%1B%24B%24J%24%26%23k%24j%1B%28B"; // This is ISO-2022-JP encoded URL
              String utf8 = "%E3%81%AA%E3%81%86%EF%BD%8B%E3%82%8A";   // This is the result that should be obtained
              String decodedShift_jis = URLDecoder.decode(shift_jis,"Shift_JIS");
              String decodedIso2022jp = URLDecoder.decode(iso2022jp,"ISO-2022-JP");
              String encodedShift_JIS = URLEncoder.encode(decodedShift_jis,"UTF-8");
              String encodedIso2022jp = URLEncoder.encode(decodedIso2022jp,"UTF-8");
               System.out.println("shift_jis        = "+shift_jis);
               System.out.println("encodedShift_JIS = "+encodedShift_JIS);
               System.out.println("iso2022jp        = "+iso2022jp);
               System.out.println("encodedIso2022jp = "+encodedIso2022jp);
          }catch(Exception e){
               e.printStackTrace();
}I am using jdk5 for this application.
Please give your valuable suggestions.
Thanks in advance.

Could the cause be that ISO-2022-JP is not just ISO-2022-JP:
http://www.w3.org/TR/japanese-xml/#AEN28427904
Maybe what you are getting is one of the flavors, while the java urldecoder uses another flavor? Or maybe the string you are getting is incorrectly encoded to being with (might have been incorrectly converted from shift-jis)?
With the shift-in shift-out design it is a difficult encoding to deal with under the best of circumstances, so you have my sympathy.

H:outputlink stop url encoding

Hi, I have a simple outputlink like this:
<h:outputLink value="#{myBean.myUrlFile}"/>the problem is that myUrl file can contains characters like é, à, I know this is not very a good idea but I had to deal with it.
When the JSF is parsed a link like this is created:
<a href="/R%E9glement.pdf"></a>I'd like that jsf don't encode the link beacause like it is not valid.
Like this it is valid:
<a href="/Réglement.pdf"></a>any idea?
Thx

But this link is used file, like a pdf file for exemple.So?
This file can be uploaded and so can contains characters like é à ù etcThat has nothing to do with how the URL is encoded.
What I don't understand is when the link is encoded, file is not found => 404 error.It means Not Found. Again that has nothing to do with URL-encoding.
when the link is not encoded (so as you've said with a bad format) the file is found.Then you have a co-operating bug. Your server should be URL-decoding the URL. What is your server exactly?

Encoding character � to send with url

I need to encode the character � to send it as an url. I do this with URLEncoder.encode(str, "UTF-8");
But i get %C3%9F instead of %DF.
What could i do to get %DF?

Since the W3C recommends "UTF-8" as character encoding, I think you're doing right. Remember that different character sets encode certain characters differently. What character set is %DF of? Have you tried whether the UTF-8-encoded URL (with %C3%9F) works?
Two Bears.

XML to URL Encoded (XSLT?)

I have some data in a Document object and need to convert it to URL encoded format so, for example this:
<share>
<code>GSK</code>
<price>13.34</price>
</share>
becomes something like...
&code=GSK&price=13.34& ...
I've been looking into doing this with XSLT but have run into difficulties. Do you think this is the best way of solving this problem, or is there another easier way?
Thanks.

I did something similar, basically takes a incoming XML file and output a HTML with certain link containing XML element value as query string. One of the things you have to be careful is, like Dr. Clap mentioned, to properly encode the value. I accompolished by using java extension in my XSL.
For example:
<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform"
          xmlns:encode="java.net.URLEncoder"
          extension-element-prefixes="encode"
          version="1.0">
     <xsl:template match="/">
          <xsl:variable name="myURL" select="concat('?blah=', encode:encode($myValueFromXML))" >
     </xls:template>
</xsl:stylesheet>

URL Encoder not working in Netscape

Hi All,
I am using url encoding in my servlet for redirecting a request. The query string values has some spaces in them. The code I have used is shown below
String mname = "Jason Perry"
response.sendRedirect
("http://194.216.8.238:8081/merchant/merch.merchant?msgid=RP&mctid=1111242&totcp="+price.toString()+"&mname="+URLEncoder.encode(mname)+"&mtxnid=UY675432");
Please note that 194.216.8.238:8081 is an IP of another server located outside our network.
In Netscape the url is shown as http://194.216.8.238:8081/merchant/merch.merchant?msgid=RP&mctid=1111242&totcp=15
&mname=Jason Perry&mtxnid=UY675432
The space is not getting replaced with the plus.
But I have observed that if the redirection is to the same server on which the servlet is running, it works fine.
I am at loss as to why url encode is not working when routing the request to another server. I am using Netscape 4.77 .
This is works in IE fine.
Your inputs are appreciated.
Thanks
Malini

Yeah had this problem before, use the java.net.URLEncoder class:
<%@ page language="java" import="java.net.URLEncoder"%>
<%
String url = "http://194.216.8.238:8081/merchant/merch.merchant?msgid=RP&mctid=1111242&totcp=15
&mname=Jason Perry&mtxnid=UY675432"
String encodedUrl = URLEncoder.encode(url);
response.sendRedirect(encodedUrl);
%>

FF5 error parsing CSS font-face with url inline base64 data

Firefox 5 refuses to parse CSS @font-face with url inline base64 data.
I use the declaration:
<style type="text/css">
@font-face {
font-family: 'MyFont';
src: url(data:font/truetype;charset=utf-8;base64,[base64data]);
</style>
then used this way:
<div style="font-family:'MyFont'; font-size:12.0pt">Test text</div>
But Firefox is not using the font and in the error console, there is always the message:
''Error parsing the "src" value. Skipped to next declaration.''
(more or less, I actually have this message in Czech)
Tried with different mime types (font/ttf,font/otf,font/opentype,application/x-font-ttf etc.), with or without charset specification, with or without quoting the font family name, with different specifications:
<style type="text/css">
@font-face {
font-family: 'MyFont';
src: url(data:font/truetype;charset=utf-8;base64,[base64data]) format(truetype);
</style>
(tried also with opentype format, etc.)
<style type="text/css">
@font-face {
font-family: 'MyFont';
src: url('myfont-webfont.eot?');
src: local('☺'), url(data:font/truetype;charset=utf-8;base64,[base64data]);
</style>
If I provide the font path:
<style type="text/css">
@font-face {
font-family: 'MyFont';
src: url('Arial.ttf');
</style>
(the font actually is Arial, for testing), it works (but I need to embed the font in the HTML for specific reason, so having the font externally is not the option).

Finally I got it work! Thanks, cor-el, you pointed me the right way to solve this problem.
There was problem with the encoding too (there was part of the font missing at the end, because of the bug in the program - I forgot to flush the buffered output stream), after then I was able to download the same copy of the TTF. - I didn't know about the possibility to put the entire url data to the location bar and try to download it, thanks cor-el.
But it still didn't solve the problem ... the problem was, that the base64 stream was divided to multiple lines, like
data:font/truetype;charset=utf-8;base64,
AAEAAAAYAQAABACARFNJRwMaCRYAC8m8AAAXfEdERUaJ+Y1JAAr/JAAAAsJHUE9T
e1arnwALAegAAKwaR1NVQt5CYFEAC64EAAAbmEpTVEZtKmkGAAvJnAAAAB5MVFNI
RExjrAAAN8wAAA1dT1MvMhAyXXMAAAIIAAAAYFBDTFT9ez5DAAr+7AAAADZWRE1Y
After I removed the line breaks, it works now! (the line is quite long then, because the base64 string is about 1MB, but it works)
Strange that I do the same for images (jpeg, png) and there is no problem with base64 string divided to multiple lines.
But anyway, I'm fine with that.

Apache plugin url encoding

the following is on HP Apache-based Web Server/1.3.26
with the mod_wl.so from 6.1 sp3 proxying to a 6.1 sp3 backend and I tried setting
up QueryFromRequest ON to no avail.
I am having the following problem with the BEA proxy plugin for Apache. I have
a very simply scenario to demonstrate it. Lets say that you have the following
directory/file structure:
/applications
     /DefaultWebApp
          spaceLink.htm
          /space demo
               resource.htm
NOTE: There is a "space" in the "space demo" directory name.
Now, spaceLink.htm has a link to the file resource.htm in the "space demo" directory.
Here is the html:
<html> <body> click here </body> </html>
Now assume that apache is sitting in front of this BEA box and proxying requests
to the BEA box.
If the Apache-BEA proxy plugin is proxying the request to BEA, then the request
is made correctly from the browser to the proxy as:
GET /space%20demo/resource.htm HTTP/1.1
Notic the browser is correctly URL encoding the request.
BUT, the apache-BEA proxy plugin is incorrectly proxying the request to the BEA
box:
GET /space demo/resource.htm
Notice the Apache-BEA proxy is not URL encoding the request, so when BEA gets
it he thinks this is a request for "/space" and "demo/resource.htm" is the http
version.
This problem DOES NOT happen if I use the generic http proxy module provided with
Apache.
Has anyone seen this problem before?
Please get back to me if you have some information.
Thanks,
Cole Crawford
Paymentech

the following is on HP Apache-based Web Server/1.3.26
with the mod_wl.so from 6.1 sp3 proxying to a 6.1 sp3 backend and I tried setting
up QueryFromRequest ON to no avail.
I am having the following problem with the BEA proxy plugin for Apache. I have
a very simply scenario to demonstrate it. Lets say that you have the following
directory/file structure:
/applications
     /DefaultWebApp
          spaceLink.htm
          /space demo
               resource.htm
NOTE: There is a "space" in the "space demo" directory name.
Now, spaceLink.htm has a link to the file resource.htm in the "space demo" directory.
Here is the html:
<html> <body> click here </body> </html>
Now assume that apache is sitting in front of this BEA box and proxying requests
to the BEA box.
If the Apache-BEA proxy plugin is proxying the request to BEA, then the request
is made correctly from the browser to the proxy as:
GET /space%20demo/resource.htm HTTP/1.1
Notic the browser is correctly URL encoding the request.
BUT, the apache-BEA proxy plugin is incorrectly proxying the request to the BEA
box:
GET /space demo/resource.htm
Notice the Apache-BEA proxy is not URL encoding the request, so when BEA gets
it he thinks this is a request for "/space" and "demo/resource.htm" is the http
version.
This problem DOES NOT happen if I use the generic http proxy module provided with
Apache.
Has anyone seen this problem before?
Please get back to me if you have some information.
Thanks,
Cole Crawford
Paymentech

WCF-WebHttp Adapter Substituting Nonstandard Characters In Address URI With Web Encoded Equivalent

I’m using the WCF-WebHttp adapter to send an HTTP Get request to a REST API. The request has to be formatted as shown…
http://xxx.xxx.xxx.xxx/HPRMServiceAPI/Record?q=title:test*&format=json
Essentially this is searching for any records with “test” in their title and returning a JSON formatted response containing information on any records that match the search criteria. I don’t have any flexibility in how this request is formatted. It has to
confirm to the syntax above.
If I hardcode the above uri in the Address URI field in the WCF-WebHttp adapter configuration and put a request through my BizTalk application, BizTalk correctly sends that request and I get my expected response.
So far so good.
The problem is the query portion of the request string (“?q=title:test*” in the example above) needs to be created dynamically based on what is specified in an incoming request to my BizTalk application. It may contain many search criteria and is not limited
to just record titles as in my example.
This, I thought would be straight forward: I construct the query portion of the request string in a custom send pipeline and insert that value into a promoted property which is then used by the URL mapping mechanic of the WCF-WebHttp adapter.
So now the uri in the Address URI field in the WCF-WebHttp adapter configuration simply contains…
http://xxx.xxx.xxx.xxx/HPRMServiceAPI
And the remaining part of the request is defined using a URL mapping…
<BtsHttpUrlMapping>
<Operation Name="DocumentSearchRequest" Method="GET" Url="/Record{searchquery}" />
</BtsHttpUrlMapping>
Where the {searchquery} variable is mapped to a promoted property.
I can see that my custom send pipeline is correctly promoting the property and that it contains the correct value (e.g “?q=title:test*&format=json”). The problem is that the WCF-WebHttp adapter is now substituting the “?” with its web encoded equivalent
(%3F). When I view the outgoing HTTP request in Fiddler it looks like this…
http://xxx.xxx.xxx.xxx/HPRMServiceAPI/Record%3Fq=title:test*&format=json
I have tried numerous ways of constructing the HTTP request, including hardcoding the offending “?q=” section into the URL mapping rather than having it in the promoted property…
<Operation Name="DocumentSearchRequest" Method="GET" Url="/Record?q={searchquery}" />
But this results in other characters (“:”, “&”, “=”) being web encoded in the outgoing HTTP request. Fiddler shows this…
/HPRMServiceAPI/Record/?q=title%3atest*%26format%3djson
Similarly, hardcoding the “/Record?q=” section in the Address URI field in the WCF-WebHttp adapter configuration like this…
http://xxx.xxx.xxx.xxx/HPRMServiceAPI/Record?q=
…and having the URL mapping contain just…
<Operation Name="DocumentSearchRequest" Method="GET" Url="{searchquery}" />
…results in the following outgoing HTTP request (Fiddler shows “?q=” portion now being placed at the end of the request string)…
http://xxx.xxx.xxx.xxx/HPRMServiceAPI/Record/title:test*&format=json?q=
So in summary it seems the WCF-WebHttp adapter is struggling to correctly assemble the HTTP request string when a portion of it is supplied in a promoted property and used by the URL mapping mechanic. Specifically when that portion contains nonstandard characters
(“?”, “:”, “&”, “=” etc), the WCF-WebHttp adapter substitutes them for their web encoded equivalent.
Any suggestions?

Thanks Ravindar,
Unfortunately, I don’t have control of the ‘receive side’. I can’t write a custom pipeline component at the receive side as it is not a BizTalk application, it’s a 3rd party web site exposing REST-like behaviour.
I have however made progress. It occurred to me the WCF-WebHttp adapter isn't the problem. It's doing exactly what it should by URL encoding any potentially 'unsafe' characters that have been 'injected' into the URL via a promoted property. Potentially the
promoted property could contain malicious content, especially if that content has been sourced directly from text in an incoming message to BizTalk.
So the WCF-WebHttp adapter is URL encoding ‘?’ to ‘%3F’ which is probably good behaviour from a security point of view.
http://xxx.xxx.xxx.xxx/HPRMServiceAPI/Record%3Fq=title:test*&format=json is a perfectly valid address.
So I looked again at the actual error I was receiving from the 3rd party site. It was…
HTTP/1.1 400 Bad Request
System.Web.HttpException: A potentially dangerous Request.Path value was detected from the client (?).
So, after some research, I edited the web.config of the 3rd party site so it no longer excludes any specific characters from the request.
<system.web>
<httpRuntime requestPathInvalidCharacters="" requestValidationMode="2.0" />
</system.web>
The 400 Bad Request has gone away and the 3rd party site is processing my request now. Unfortunately it still seems to have an issue with ‘%3F’ being used in place of ‘?’, but I think that will have to be resolved in IIS/web config on the receive side (maybe
by using request filtering
http://www.iis.net/configreference/system.webserver/security/requestfiltering) or by discussing with the sites developers.
In summary, I now believe the WCF-WebHttp adapter is behaving correctly by substituting URL encoded values in place of potentially unsafe characters.
Thanks for your help.

URL encoding special characters in episode name

Hello,
Configuration :
1 capture agent on a MAC MINI under MAC OS X 10.6.5
1 PCP2 server on an XServe under MAC OS X 10.6.5
I encounter this problem since 10.6.5 update :
When I stop a Podcast capture, if episode name is for example abcd#, '#' character is "URL encoded" to '%23' sequence. So the final file name of the podcast is abcd%23.m4v instead of abcd#.m4v.
This was not the case under 10.6.4.
I tried this solution : renaming the file adding a task in the workflow ; this task launches a bash script which renames all files replacing '%23' sequences by a '#' character.
Files are well renamed, but the task don't give hand back, and XGrid task never ends.
Got this error in /var/log/system.log :
sandbox [nnn] : bash(12146) : deny file-ioctl /path/to/bash-script
Thanks for help

Problem solved !
I had a problem with rights on my podcasts directory.
So, for those who are interested, you can add a task to your workflow, for example a bash batch file which renames all files in your podcast directory whose name contains a special character :
for fic in `ls`
do
mv $fic `echo $fic | sed -e s/%23/#/g`
done
the additional workflow task must contain 2 items :
--- an item "command" of type String with the path of your batch file as value
--- an item "dependsOnTasks" of type array with an element of type string with "postflight" as value

URL encoding in web.show_document

Hi.
I have to pass some parameteres with spaces and such in them. Is there any available pl/sql functions to url encode/decode ?

Thanks.
I had some trouble though, as the compiler didn't recognize it. It turned out that this package did not exist in the v8 db we're using for development, but the recent 9i production db had it.
We decided to put the parameter in question into the db, and extract it from the report instead of passing it on the url...

URL Encoding and Decoding in OSB

Hi,
Do we have out-of-box solution in OSB to decode XML which is URL encoded and encode XML to URL encoding. If so how can I do it.
This might be basic question. I am unable to figure out. new to OSB.
Thanks
Sham.

Hi Sham,
Probably these are what you looking for...
fn-bea:inlinedXML()
fn-bea:serialize()
fn-bea:serialize() to represent an XML document as a string
fn-bea:inlinedXML() to parses textual XML and returns an instance of the XQuery 1.0 Data Model.
Supported Function Extensions from Oracle
http://docs.oracle.com/cd/E23943_01/admin.1111/e15867/xquery.htm#i1101407
Oracle’s XQuery Implementation
http://docs.oracle.com/cd/E13162_01/odsi/docs10gr3/xquery/extensions.html
Cheers,
Vlad
It is considered good etiquette to reward answerers with points (as "helpful" - 5 pts - or "correct" - 10pts)
https://forums.oracle.com/forums/ann.jspa?annID=893

BSP Url encoding

Hi all!
I'd like to use URL encoding (I guess base64) to mask url parameters in our URL (instead of default.htm?param1=value&param2=value etc.) to prevent URL manipulation. In most parts of our application we use PopUp-Windows without the location bar of the browser, but we have email links to workitems which launch BSPs and these new opened browser windows do have a location bar and a URL which can be manipulated.
How do you set up URL encoding in your BSP application? (entries in the BSP service of /sicf for example or BSP Application Attributes ?). Is there a way to encode the only parts of the application? How to use and when the if_http_utility->en/decode_base64 methods?
best regards,
Christoph

did you try the following?
ex. you have page a.htm and from there you are calling page b.htm with params param1 and param2.
use
CALL METHOD cl_http_utility=>if_http_utility~encode_base64
                 EXPORTING
                   unencoded = param1value
                 receiving
                   encoded   = param1value .
and pass this as the value for param1 and call the URL for page b.htm.
and within page b.htm decode it and use .
CALL METHOD cl_http_utility=>IF_HTTP_UTILITY~DECODE_BASE64                 EXPORTING
                   encoded = param1value
                 receiving
                   decoded   = param1value .
<b>Sample application:</b>
Create a BSP application withe two pages a.htm and b.htm
and have a page attribute called name in both the pages.
<u><b>a.htm layout code</b></u>
<%@page language="abap"%>
<%@extension name="htmlb" prefix="htmlb"%>
<htmlb:content design="design2003">
<htmlb:page title = " ">
    <htmlb:form>
<%
if name is initial .
name = sy-uname .
endif .
%>
      <htmlb:textView     text          = "<%= name %>"
                          design        = "EMPHASIZED" />
<%
CALL METHOD cl_http_utility=>if_http_utility~encode_base64
                 EXPORTING
                   unencoded = name
                 receiving
                   encoded   = name .
%>
      <htmlb:button       text          = "Press Me"
                          onClientClick       = "javascript:window.open('b.htm?name=<%= name %>');" />
    </htmlb:form>
</htmlb:page>
</htmlb:content>
<u><b>b.htm layout code</b></u>
<%@page language="abap"%>
<%@extension name="htmlb" prefix="htmlb"%>
<htmlb:content design="design2003">
<htmlb:page title = " ">
    <htmlb:form>
<%
CALL METHOD cl_http_utility=>if_http_utility~DECODE_BASE64
                 EXPORTING
                   encoded = name
                 receiving
                   decoded   = name .
%>
      <htmlb:textView     text          = "<%= name %>"
                          design        = "EMPHASIZED" />
    </htmlb:form>
</htmlb:page>
</htmlb:content>
Hope this helps.
Regards
Raja

URL ENCODING

hi,
I am devoloping using JSP pages and servlets on JRUN web server. I found that when the browser passes the URL (along with my custom parameters) to the servlet, it encodes any spaces in my parameter values with '%20'. Now am I correct that this is done using Base64encoding?? However, I am not using any encoding mechanism and this automatic conversion seems to cause a problem in that the web server thinks that it needs to authenticate this URL and is unable to process it. Does anyone know how I can turn this encoding off and pass any spaces in my parameter values as they are, or is there any other known solution to this.
Any help will be much appreciated.
Thanks

hi,
I am devoloping using JSP pages and servlets on JRUN
web server. I found that when the browser passes the
URL (along with my custom parameters) to the servlet,
it encodes any spaces in my parameter values with
'%20'. Now am I correct that this is done using
Base64encoding?? All browsers will do this, and it's called "URL encoding". It's a little trick to use other characters than simple letters and digits in URLs (including parameters). This has nothing to do with Base64 encoding.
However, I am not using any encoding
mechanism The browser uses this encoding, and you (the server) cannot do anything about it. It shouldn't actually be a problem at all.
and this automatic conversion seems to cause
a problem in that the web server thinks that it needs
to authenticate this URL and is unable to process it.I think you're making a mistake here. URL encoding has nothing to do with URL authentication. I guess you mean that the webserver is requesting username/password ? Then you have configured the requested URL to be authenticated at yuor webserver.
Does anyone know how I can turn this encoding off and
pass any spaces in my parameter values as they are, or
is there any other known solution to this.You really don't want to turn URL encoding off. Next, it can't be 'turned off'. Jrun will automatically decode the %20 to a space again; you won't notice it has been encoded at all.
Your problem must be somewhere else. Have a close look at your http authentication settings...
>
Any help will be much appreciated.
ThanksNo problem
Cheers,
Tom

Converting to PDF issue with URLs

I've edited a file for a client that was originally created in Adobe Illustrator and added phone numbers and URLs to each coupon. When I open the PDF of the coupons and click on any of the links on the second row of Page 2 with florists, the "fl" in florist in each URL is converted to "%CB%9C". I realize these are encoded characters, but they are the wrong ones and the links don't open up correctly as a result. All other links in all 3 pages of coupons open up correctly except for these three....HELP!!
http://gallery.mailchimp.com/9948dc9b0f4637762c82e0a7f/files/Coupons.pdf

Try turning off Standard Ligatures in the Opentype panel. I don’t usually use Illustrator to make PDFs with URLs, but I think what’s happening is Acrobat is just reading the text as a URL and assuming it is a link. I don’t think you can make a URL link from selected text in Illustrator. Really. Fracking ridiculous.
You can open the PDF in Acrobat Pro and add he links there, without worrying about the actual text on the screen.

HttpSend Activity with Url Encode

Similar Messages

Maybe you are looking for