This is an interesting topic. I had started implementing this, but never managed to finish, because of higher priorities (migration to ASF, Codehaus shutdown). It means the implementation of effectively final variables inference is not ready, and buggy. I doubt I will have time to fix, which probably means disabling in 2.5, unless someone else takes it over.
2017-03-05 21:45 GMT+01:00 Paul King <pa...@asert.com.au>: > I noticed a few similar problems recently (GROOVY-8093, GROOVY-8094) > but yours looks slightly different again. Probably worth creating an > issue. > > On Sun, Mar 5, 2017 at 11:16 PM, Alain Stalder <astal...@span.ch> wrote: > > Not sure if this is the right/best place to report this, but I got this > with > > the current head of the master branch of > > https://github.com/apache/groovy.git (ec3179a455): > > > > -- > > final File file = new File('whatever') > > final List<String> lines > > try { > > lines = file.readLines() > > } catch (IOException e) { > > return null > > } > > -- > > > > => > > > > 1 compilation error: > > > > The variable [lines] may be uninitialized > > . At [2:20] at line: 2, column: 20 > > > > -- > > final File file = new File('whatever') > > final List<String> lines > > try { > > lines = file.readLines() > > } catch (IOException e) { > > lines = null > > } > > -- > > > > => > > > > 1 compilation error: > > > > The variable [lines] is declared final but is reassigned > > . At [6:3] at line: 6, column: 3 > > > > >
