[eclipse.org-architecture-council] FW: [Bug 307021] New: Host OpenGrok f

["Oberhuber, Martin" <Martin.Oberhuber@xxxxxxxxxxxxx>]

AC Fellows,

This may be very interesting for you.

Not sure if it makes sense to create a single search index to cover
Both the release train and non-train-projects. Maybe it does. Or
create separate databases for separate concerns.

Martin

-----Original Message-----
From: bugzilla-daemon@xxxxxxxxxxx [mailto:bugzilla-daemon@xxxxxxxxxxx] 
Sent: Wednesday, March 24, 2010 10:24 PM
To: Oberhuber, Martin
Subject: [Bug 307021] New: Host OpenGrok for Eclipse source code

https://bugs.eclipse.org/bugs/show_bug.cgi?id=307021
Product/Component: Community / Cross-Project

           Summary: Host OpenGrok for Eclipse source code
    Classification: Eclipse Foundation
           Product: Community
           Version: unspecified
          Platform: PC
        OS/Version: Mac OS X
            Status: NEW
          Severity: normal
          Priority: P3
         Component: Cross-Project
        AssignedTo: cross-project.inbox@xxxxxxxxxxx
        ReportedBy: gnormington@xxxxxxxxxx


At the recent EclipseRT BoF at EclipseCon, we discussed the advantages
of having OpenGrok search available to search the whole EclipseRT
codebase. We in the Virgo project have found this extremely valuable for
quickly searching the
24 git repositories for usages of types or for keywords (sometimes the
only way in when you are yet to learn a particular area of code).

OpenGrok efficiently indexes large volumes of code and provides search
results to its web UI in a fraction of a second. This is much more
efficient than pulling massive numbers of projects into Eclipse and then
attempting to do a search across the workspace, especially when doing a
text search for likely keywords.

OpenGrok has been successfully used to index/search the whole of Open
Solaris as well as OpenJDK. Clearly the larger the scope of code that
can be included the better, so it seemed reasonable to include all the
source code checked in to Eclipse repositories rather than just limiting
this to RT.

OpenGrok enables more focussed searches where necessary based on the
grouping of source directories which are indexed.

The basic implementation approach is to run a cron job periodically to
incrementally update the OpenGrok index and then host a web application
which serves the OpenGrok web UI pages. The set-up typically takes
somewhere of the order of half a day to comprehend and install OpenGrok
plus more time for crafting a cron job to cvs/svn update or git pull and
then run the OpenGrok indexing program.

For more information on OpenGrok see
http://hub.opensolaris.org/bin/view/Project+opengrok/. To try it out on
the OpenSolaris codebase, see: http://src.opensolaris.org/source/.

--
Configure bugmail: https://bugs.eclipse.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: ------- You are watching
the assignee of the bug.