Add MasgnNode class for masgn nodes
#203
Conversation
|
Sorry I missed this in my previous PR @marcandre 😅 🙌 |
There was a problem hiding this comment.
masgn is complicated, isn't it? 😅
PR looks like a good start, but you seem to have overlooked 2 important cases: foo.bar, foo[:baz], ... = ....
Please describe results of your proposed methods and add tests.
Also:
-
each_descendantshould only be used on the first child. Please add a test forx, y = 1, z+=2for example. Even then, some corner cases could be problematic:foo[x+=1], bar = ... -
I doubt
each_assignmentis necessary. Performance of enumerators is typically quite poor, especially for short lists
|
I knew there were things I wasn't thinking of, thanks @marcandre I'll fix up for those cases! |
|
@marcandre updated now. I have now introduced a
I also added For symmetry, Please let me know your thoughts! |
dvandersluis
force-pushed
the
masgn
branch
2 times, most recently
from
0476aa2
to
8b5f9b3
Compare
|
The "modern" spec was failing on |
dvandersluis
force-pushed
the
masgn
branch
2 times, most recently
from
96e0836
to
ff20438
Compare
| context 'with nested assignment on LHS' do | ||
| let(:source) { 'a, b[c+=1] = z' } | ||
|
|
||
| it { is_expected.to eq([:a, 'b[c+=1]']) } |
There was a problem hiding this comment.
Do you have a use for 'b[c+=1]'? It seems like a bad result for a name, and string vs symbol is a bit ugly. What about :"b[]" instead?
| context 'with a method chain on LHS' do | ||
| let(:source) { 'a, b.c = z' } | ||
|
|
||
| it { is_expected.to eq([:a, 'b.c']) } |
There was a problem hiding this comment.
Same with 'b.c'. How about :c= or :"#c"?
| it { is_expected.to eq(%i[a b c]) } | ||
| end | ||
|
|
||
| context 'with nested assignment on LHS' do |
| context 'with variables' do | ||
| it { is_expected.to eq(s(:send, nil, :z)) } | ||
| end | ||
|
|
||
| context 'with a LHS splat' do | ||
| let(:source) { 'x, *y = z' } | ||
|
|
||
| it { is_expected.to eq(s(:send, nil, :z)) } | ||
| end | ||
|
|
||
| context 'with multiple RHS values' do | ||
| let(:source) { 'x, y = 1, 2' } | ||
|
|
||
| it { is_expected.to eq(s(:array, s(:int, 1), s(:int, 2))) } | ||
| end | ||
|
|
||
| context 'with an RHS splat' do | ||
| let(:source) { 'x, y = *z' } | ||
|
|
||
| it { is_expected.to eq(s(:array, s(:splat, s(:send, nil, :z)))) } | ||
| end | ||
|
|
||
| context 'with assignment on RHS' do | ||
| let(:source) { 'x, y = 1, z+=2' } | ||
|
|
||
| it { is_expected.to eq(s(:array, s(:int, 1), s(:op_asgn, s(:lvasgn, :z), :+, s(:int, 2)))) } | ||
| end |
There was a problem hiding this comment.
This is overkill for such a simple method. How about a single test here?
| # @return [Array<Node>] values being assigned on the RHS of the multiple assignment | ||
| def values | ||
| array? ? expression.children : [expression] | ||
| end |
There was a problem hiding this comment.
Could you justify when / how this is useful? I find it a bad idea that x, y = z and x, y = [z] produce the same values even though they are very different
| context 'when the RHS has a single value' do | ||
| let(:source) { 'x, y = z' } | ||
|
|
||
| it { is_expected.to eq([s(:send, nil, :z)]) } | ||
| end | ||
|
|
There was a problem hiding this comment.
If this method is to remain, then needs a test for 'x, y = [z]'
| end | ||
|
|
||
| # @return [Boolean] whether the expression has multiple values | ||
| def array? |
There was a problem hiding this comment.
Do you have a need for this public method? I find the definition dubious, as x, y = z and x, y = *z are basically the same yet have a different array? result.
Follows #201. I actually ended up needing
masgnto have its own class withexpressiondefined in order to do the refactoring I wanted to, so this adds it.Because
masgnnodes are more involved than the other assignment types, this class has a different interface than the other assignments, except forexpressionwhich works the same. Rather thannamethat the other assignment classes have,MasgnNodehasnames(since there will inherently be more than one). It also is able to handle nestedmlhses.