[jira] [Resolved] (COMDEV-167) parsepmcs.py regex for committer groups also matches -pmc groups

["Sebb (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/COMDEV-167?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sebb resolved COMDEV-167.
-------------------------
    Resolution: Fixed

URL: http://svn.apache.org/viewvc?rev=1706165&view=rev
Log:
COMDEV-164 parsepmcs.py does not handle new groups correctly
COMDEV-167 parsepmcs.py regex for committer groups also matches -pmc groups

Modified:
    comdev/reporter.apache.org/trunk/data/parsepmcs.py


> parsepmcs.py regex for committer groups also matches -pmc groups
> ----------------------------------------------------------------
>
>                 Key: COMDEV-167
>                 URL: https://issues.apache.org/jira/browse/COMDEV-167
>             Project: Community Development
>          Issue Type: Bug
>          Components: Reporter Tool
>            Reporter: Sebb
>            Priority: Minor
>
> The regex for selecting committer (non-pmc) lists does not work.
> The regex is currently:
> r"#([-a-z0-9._]+)(?!-pmc)"
> This is applied to data of the form
> committers-by-project.html#abdera'
> and
> committers-by-project.html#abdera-pmc'
> However the greedy match at the start of the RE allows it to grab everything up to the trailing single-quote; this is not followed by -pmc so the negative look-ahead has no effect.
> The effect of this is that the -pmc groups are stored in the projects.json file.
> AFAICT these are not used elsewhere; they just increase the filesize that needs to be downloaded by the Javascript in the browsers.
> ==
> It's quite tricky using lookahead with wildcards.
> Rather than parse the group list twice, it might be simpler to parse it once and extract all the groups; the code can the check if the group name ends with -pmc and act accordingly.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)