Alright then, I will prepare a branch with the changes so they can be merged before an appropriate release.
Cheers! On Sat, Sep 30, 2023 at 8:57 PM Thiago H. de Paula Figueiredo < thiag...@gmail.com> wrote: > Hello, Ben! > > Great catch! > > I suggest we make the change to throw an exception in Tapestry 5.9.0 > with a big warning in the release notes. > > Cheers! > > On Sat, Jul 29, 2023 at 7:55 AM Ben Weidig <b...@netzgut.net> wrote: > > > > Hi there! > > > > I was trying to add "static empty()" methods to JSONArray and JSONObject > to > > reduce the creation of new instances if we know it's supposed to be > empty, > > like Collections does with emptyList()/emptyMap() > > > > Adding the methods isn't the issue (creating private constructor > accepting > > the underlying storage, using the appropriate Collections.empty... > method), > > but what I discovered while writing the unit tests. > > > > The JSONArray method "Object remove(int index)" doesn't explode when used > > with an invalid index. > > The documentation states this behavior, but I think that's wrong or at > > least inconsistent with the rest of the type. > > There's even a JSONArrayIndexOutOfBoundsException that's used by "Object > > get(int index)" and, therefore, by all other index-based methods. > > > > This discrepancy in expected behavior seems weird to me, and if we decide > > to elevate JSONArray to implement List<Object> in the future, as it's > > almost there anyway, it goes against the documented contract of List<T>. > > > > However, it would be a possible breaking change for people relying on the > > non-explody behavior. > > > > Thoughts? > > > > > > Cheers > > Ben > > > > -- > Thiago > > --------------------------------------------------------------------- > To unsubscribe, e-mail: dev-unsubscr...@tapestry.apache.org > For additional commands, e-mail: dev-h...@tapestry.apache.org > >
