[ https://issues.apache.org/jira/browse/HADOOP-11897?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14737411#comment-14737411 ]
Allen Wittenauer commented on HADOOP-11897: ------------------------------------------- So basically, this is about the output in the actual linked to log files. For the vast majority of patches, they stay within one module, so printing the entire patch from base is sort of waste. It might be better to just list the paths up to the point that they conflict. For example, let's say you have a patch that touches two files: {code} hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-resourcemanager/pom.xml hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server/hadoop-yarn-server-sharedcachemanager/pom.xml {code} and it introduces whitespace. Rather than list hadoop-yarn-project/hadoop-yarn/hadoop-yarn-server every time, maybe just list hadoop-yarn-server-sharedcachemanager/pom.xml and hadoop-yarn-server-resourcemanager/pom.xml . This is just a food-for-thought type of idea. It would almost certainly have to be done as a post processing step in each plugin. > test-patch.sh plugins should abbreviate the path > ------------------------------------------------ > > Key: HADOOP-11897 > URL: https://issues.apache.org/jira/browse/HADOOP-11897 > Project: Hadoop Common > Issue Type: Sub-task > Components: test, yetus > Affects Versions: HADOOP-12111 > Reporter: Allen Wittenauer > Priority: Trivial > Labels: newbie > > The current output of checkstyle, shellcheck, and whitespace results in very > long file name paths. It might be useful to abbreviate them in some way, > maybe removing the entire path and leaving just the filename or last > directory+filename. -- This message was sent by Atlassian JIRA (v6.3.4#6332)