Sergi Siso

Another idea is for the fparser2 frontend to split this statements to the declaration and the initialisation, so it becomes: ``` INTEGER :: y y=MAX(1,2,3) ```

Yes, well spotted. This is the issue. It does a pretty simple substitution without looking if `ref` has a subtree which currently gets lost. If it has a subtree (due...

As discussed this morning, the files we exclude are due to errors in some of the processing steps. To create a more robust/generic script we could recover from this errors...

I will give my take on this: > Or perhaps DataNode would not be needed We thought `DataNode` as a higher abstraction virtual PSyIR node (not a concrete node) to...

To be clear by > ... these properties should be in the specific object DataSymbol and Literal ... I mean DataSymbol and Literal should have a self._type/datatype pointing to a...

So the proper check could be something like: ``` if (.NOT.ALLOCATED(X) .or. LBOUND(X,1) != L1D .or. UBOUND(X,1) != U1D .or. LBOUND(X,2) != L2D .or. UBOUND(X,2) != U2D ) then if...

To close this issue by moduleinlining the majority of LFric kernels we need: - [ ] (Issue #1824) Failed to retrieve with get_kernel_schedule - can be the metadata validation that...

Remaining issues after merging #1968 [gravitywave_not_inlined.txt](https://github.com/stfc/PSyclone/files/10176399/gravitywave_not_inlined.txt) [gungho_not_inlined.txt](https://github.com/stfc/PSyclone/files/10176400/gungho_not_inlined.txt)

If this uses the reference_access method, this could be the same issue as #2271

>if I add a reference_access method to the symbol table I was expecting it to be all implemented in ScopedNode, since Node has the reference_access method that needs to be...