Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.
Bug 367635 - support decomposed start.ini
Summary: support decomposed start.ini
Status: RESOLVED FIXED
Alias: None
Product: Jetty
Classification: RT
Component: server (show other bugs)
Version: 7.6.0.RC0   Edit
Hardware: All All
: P3 enhancement (vote)
Target Milestone: 7.5.x   Edit
Assignee: Thomas Becker CLA
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2011-12-29 17:23 EST by Greg Wilkins CLA
Modified: 2012-01-08 19:27 EST (History)
2 users (show)

See Also:


Attachments
POC patch provided by tbecker (30.63 KB, patch)
2011-12-29 17:23 EST, Greg Wilkins CLA
no flags Details | Diff
proposed patch (6.57 KB, patch)
2012-01-02 11:56 EST, Thomas Becker CLA
no flags Details | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Greg Wilkins CLA 2011-12-29 17:23:11 EST
Support a start.d directory of start.ini fragments
Comment 1 Greg Wilkins CLA 2011-12-29 17:23:57 EST
Created attachment 208856 [details]
POC patch provided by tbecker
Comment 2 Greg Wilkins CLA 2011-12-29 17:43:56 EST
Thomas,

thanks for the patch.    I've refactored it a little bit and made sure that there is good ordering of the start.d fragments.  I also have moved it all back to Main.java, so that jetty_home processing can be centralised.  I also broke up the parseCommandLine into better testable methods.... but more tests are needed.

I have committed this for 7.6.0.RC3 - but not refactored start.ini yet

Can you review (and usage text probably needs more work) - thanks
Comment 3 Thomas Becker CLA 2012-01-02 04:50:57 EST
Greg, thx for the work. Will review your changes now. Would have been good if you had commited my patch and then applied your changes in another commit. Then it's easier for me to track what you've changed compared to my version.
Comment 4 Thomas Becker CLA 2012-01-02 11:56:04 EST
Created attachment 208922 [details]
proposed patch

Like your changes. Changed the test and refactored a bit to extend the test to parse files in start.d directory again. 
Also changed usage.txt a bit. 

In Main.java I've found this in Main.loadStartIni(...):

catch (IOException e)
        {
            // usageExit(e,ERR_UNKNOWN);
        }

What's the intention in commenting the usageExit call here? Any IOException will silently be ignored when reading start.ini files. Is this intended?

Patch attached. Would be good if you can commit my patch even if you change it also to get me some commit karma.
Comment 5 Greg Wilkins CLA 2012-01-08 19:27:42 EST
Applied (and exit usage uncommented)