On Sat, 18 Aug 2018, at 08:54, ToddAndMargo wrote:
> Hi All,
>
> I want to download
> jre-8u181-windows-i586.exe
>
> from
> http://www.oracle.com/technetwork/java/javase/downloads/jre8-downloads-2133155.html
>
> According to Firefox's cliget, the address is:
>
> https://download.oracle.com/otn-pub/java/jdk/8u181-b13/96a7b8442fe848ef90c96a2fad6ed6d1/jre-8u181-windows-i586.exe?AuthParam=1534577721_2ae0bad94b6341c2907ebf82aa395171
>
> But how do I get the "AuthParam"?

How did FF get it for the session in which you used Firefox's 'cliget' feature?

Is it what happens when (as a user) you acccept the software licence? If so, your curl process will
have to mimic that stage first.

In my limited experience of doing something complicated with curl, you need to start at the
beginning - so maybe going to the oracle website pretending you've never been there before,
then see what happens at each necessary interaction, then find a way to replicate that in
curl. In my case that's meant using a scripting language to build each successive curl
request, and also to check that what comes back from each one has the expected return
code, no unexpected headers (for example no redirects), the expected set of cookies (ie
their names if not their values), and also do to some sanity checking on the contents of
any html you get, before (perhaps) parsing values out of that to plug in to the next
request.

It might be impossible, if the website uses a lot of javascript and requires a human being
to interact with it. You might get on better using an automation scripting package (the
sort of thing used to test that GUI applications still work) which will simulate a person
pointing and clicking on various things on a series of pages being displayed in their
browser - something like AutoHotkey or AutoIt.

-- 
Jeremy Nicoll - my opinions are my own.
-----------------------------------------------------------
Unsubscribe: https://cool.haxx.se/list/listinfo/curl-users
Etiquette:   https://curl.haxx.se/mail/etiquette.html