Hi, On Fri, Aug 28, 2015 at 7:03 PM, Stefan Hett <ste...@egosoft.com> wrote:
> It does and that's not quite the problem for me. However what's possible > to do with SVN are merges like these: > > merge A/B/somefile into A > > absolutely valid operation, but nothing we use here. In cases this is done > here in our company it's always an incorrect merge (because someone chose > the wrong destination folder. > So we want to prevent issues like these, because they can easily result in > polluting the mergeinfo records. > > Don't now how many out there appreciate the 'subtree merge' behavior, but in our and our clients, a potential scenario of tenths or hundreds of developers branching&merging at any level of source tree will become a unmanageable version control system, in addition to a confusing user experience for the most of developers. Take a look to this snippet of a unix/bash pre-commit hook, just in case it inspires you: Disclaimer: We don´t have any measure of performance of this script. We simply haven't need to measure it yet. <code> # Check forbidden mergeinfo paths TXN_PATHS=$($SVNLOOK changed -t "$TXN" "$REPOS") # Grep for paths *not* in merge-allowed levels: /trunk, /branches/release/*,... TXN_PATHS=$(echo "$TXN_PATHS" | grep -Ev "^....trunk/?$" | grep -Ev "^....branches/release/\w+/?$" | grep -Ev "^....branches/feature/\w+/?$") for TXN_PATH in "$TXN_PATHS" do ELEM_PATH=$(echo "$TXN_PATH" | cut -c 5-) MERGEINFO=$($SVNLOOK propget "$REPOS" svn:mergeinfo -t "$TXN" "$ELEM_PATH" 2>/dev/null) if [ ! "$MERGEINFO" = "" ]; then fail_forbidden_mergeinfo_path;fi done </code> Regards. _ Joseba
