[rfc][icedtea-web] moving rust launchers to cargo

[Jiri Vanek]

On 03/16/2018 10:19 PM, Alex Kashchenko wrote:
> Hi Jiri,
> 
> On 03/16/2018 04:56 PM, Jiri Vanek wrote:
>>
>> [...]
>>
>>>
>>> Hello!
>>>
>>> One more attempt to make default build pass, and still protect the launcher from being built 
>>> without vars set.
>>>
>>> Using the original approach (no build.rs), havinf  option_env! to private fields, and having 
>>> unwrapping gettets in hardcoded.rs.  With difference that the getter will return default instead 
>>> of failing.
>>>
>>>
>>> In addition I would keep the test, ensuring that the necessary values are not empty nor default.
>>> By that, the default build you require will be kept, and  my fear that default-vars build will 
>>> leak removed.
>>>
>>>
>>
>> Patch with above approach attached.
> 
> Patch looks good to me. Compile-time checks may be better, but runtime ones will work too.

With compile checks, the default build would be broken. And I had serious fars about duplicating 
configure checks in the build.rs. Thank you for the dsicussion about the patch and I'm happ we found 
the common way.
> 
> Minor problems:
> 
> 1. redundant mod declaration in hardcoded_paths.rs
> 
> 2. excessive trims in tests
> 
> 3. variables_non_em*o*pty typo
> 

All three should be fixed. I'm a bit paranoid about dangling spaces, but yah.. you are right :)
> 


-- 
Jiri Vanek
Senior QE engineer, OpenJDK QE lead, Mgr.
Red Hat Czech
jvanek at redhat.com    M: +420775390109
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cargoisation_testdrivenFail2.patch
Type: text/x-patch
Size: 9052 bytes
Desc: not available
URL: <http://mail.openjdk.java.net/pipermail/distro-pkg-dev/attachments/20180322/522f133c/cargoisation_testdrivenFail2.patch>