Hello,
I am walking an AST, looking for specific comments in the tree. I encountered some unexpected behavior regarding the hierarchy of the comments. This impacts the printer too.
My sample program looks like this, with 3 comments outside the do loop, and one comment inside the loop.
program test
integer :: arg1
integer :: iterator
!comment_out 1
arg1 = 10
!comment_out 2
do iterator = 0,arg1
!comment_in
print *, iterator
end do
!comment_out 3
end program testI applied the following script to expose the unexpected behavior:
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
ARG_FORTRAN_FILE_INPUT_PATH = "test_fail.f90"
ARG_FORTRAN_STANDARD = 'f2008'
from fparser.common.readfortran import FortranFileReader
from fparser.two.parser import ParserFactory
from fparser.common.readfortran import FortranFileReader
from fparser.two.utils import walk
from fparser.two.Fortran2003 import Comment
print("## Parsing input")
reader = FortranFileReader("{0}".format(ARG_FORTRAN_FILE_INPUT_PATH), ignore_comments=False)
f_parser = ParserFactory().create(std=ARG_FORTRAN_STANDARD)
parse_tree = f_parser(reader)
print("## Visiting AST")
cmtlist = walk(parse_tree, (Comment))
for cmt in cmtlist:
if cmt.items[0]:
print("## ---" + cmt.items[0])
print(type(cmt.parent))
print(cmt.parent)
print("## Done")and got the following output:
## Parsing input
## Visiting AST
## ---!comment_out 1
<class 'fparser.two.Fortran2003.Implicit_Part'>
!comment_out 1
## ---!comment_out 2
<class 'fparser.two.Fortran2003.Block_Nonlabel_Do_Construct'>
!comment_out 2
DO iterator = 0, arg1
!comment_in
PRINT *, iterator
END DO
## ---!comment_in
<class 'fparser.two.Fortran2003.Block_Nonlabel_Do_Construct'>
!comment_out 2
DO iterator = 0, arg1
!comment_in
PRINT *, iterator
END DO
## ---!comment_out 3
<class 'fparser.two.Fortran2003.Execution_Part'>
arg1 = 10
!comment_out 2
DO iterator = 0, arg1
!comment_in
PRINT *, iterator
END DO
!comment_out 3
## Done
We can observe that the comment_out 2 is considered as a children of the do loop, whereas I expect it to be at the same level as comment_out 3, that is as part of the execution part.
I've run this with fparser2 version 0.0.16.
We can also observe that the fparser2 script fparser2 --std=f2008 test_fail.f90 prints a wrong indentation when parsing this sample program:
PROGRAM test
INTEGER :: arg1
INTEGER :: iterator
!comment_out 1
arg1 = 10
!comment_out 2
DO iterator = 0, arg1
!comment_in
PRINT *, iterator
END DO
!comment_out 3
END PROGRAM testBest.
Hello,
I am walking an AST, looking for specific comments in the tree. I encountered some unexpected behavior regarding the hierarchy of the comments. This impacts the printer too.
My sample program looks like this, with 3 comments outside the do loop, and one comment inside the loop.
I applied the following script to expose the unexpected behavior:
and got the following output:
We can observe that the
comment_out 2is considered as a children of the do loop, whereas I expect it to be at the same level ascomment_out 3, that is as part of the execution part.I've run this with fparser2 version 0.0.16.
We can also observe that the fparser2 script
fparser2 --std=f2008 test_fail.f90prints a wrong indentation when parsing this sample program:Best.