Closed hiker closed 3 months ago
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 99.86%. Comparing base (
93083d8
) to head (42eec07
).
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
I've also updated the dev guide (thanks for pointing this out in #2658)
Thanks for this Joerg, I think it's a good first step. I am wondering whether we should take note of: ... i.e. that any symbols initialised in a DATA statement effectively have the SAVE attribute. This is true for both Fortran2003 and 2008. I'm not sure about earlier standards. I'll check integration tests next time. Appropriate doc has been updated.
OK, I have added this, including support for nested do loops, and derived statements, and added tests based on the F90 standard. You know fparser better than me, I hope I got the possible node type for variables correct (there seems to be a bit of a difference between fparser and the standard:
R526: data-stmt-object is variable
or data-implied-do
I assume the variable
is skipped? So I hope I capture all possible ways a variable might be specified. Similar for:
R528 data-i-do-object is array-element
or scalar-structure-component
or data-implied-do
I don't see array-element or scalar-structure component, so I search for PartRef.
I've triggered CI.
All done, CI triggered and it finished successfully.
Adds some support for Data statements - they are turned into
UnsupportedFortranType
s, but this means that the rest of files containing data statements are properly handled now.