|
How to decode encoded/escaped URLs ?
Apache JMeter FAQ
(Continued from previous question...)
How to decode encoded/escaped URLs ?
URLs that are present in HTML pages are usually encoded. For example:
href="/search.cgi?search1=abc&search2=xyx"
If the URL is extracted using a Regular Expression Post-Processor and used directly in another HTTP Sampler, then the sample may well fail, as the server is expecting to see:
/search.cgi?search1=abc&search2=xyx
As of JMeter 2.3.1 there is no automatic way of handling this, however there are some work-rounds. Assume that the URL is stored in the variable ESCAPED_URL.
Using Jexl:
${__jexl(vars.get("ESCAPED_URL").replace("&"\,"&"))}
Using BeanShell:
Set the JMeter property beanshell.function.init to point to a file; add the following definition to the file:
// Fix ampersands in a string
String fixAmps(s) {
return s.replaceAll("&","&");
}
[the current release of JMeter already contains the fixAmps() function in the file BeanShellFunction.bshrc]
The function can then be called as follows:
${__BeanShell(fixAmps(vars.get("ESCAPED_URL"))}
or
${__BeanShell(fixAmps(vars.get("ESCAPED_URL"),PLAIN_URL)}
The latter function call will store the result in the variable PLAIN_URL The work-rounds above can be extended to handle other transformations such as %20 -> space, %2F -> / and %3A -> :
(Continued on next question...)
Other Interview Questions
| 