I could run with copy_tree using this workaround
=======================================
diff --git a/rtemstoolkit/path.py b/rtemstoolkit/path.py
index 760f4bd..df553b3 100644
--- a/rtemstoolkit/path.py
+++ b/rtemstoolkit/path.py
@@ -191,10 +191,10 @@ def copy_tree(src, dst):
hsrc = host(src)
hdst = host(dst)
- if os.path.exists(src):
+ if os.path.exists(src) and os.path.isdir(src):
names = os.listdir(src)
else:
- name = []
+ names = []
if not os.path.isdir(dst):
os.makedirs(dst)
=====================================
But there's still an issue with the file permission as it uses
copystat which
copies the permission bits which is not required in this case.
It requires a manual `rm` with root permission before rerunning the
coverage.
On Sun, 28 Oct 2018 at 16:31, Vijay Kumar Banerjee <vijaykumar9...@gmail.com>
wrote:
>
>
>
> On Sun, 28 Oct 2018 at 15:46, Chris Johns <chr...@rtems.org> wrote:
>
>> On 28/10/18 8:33 pm, Vijay Kumar Banerjee wrote:
>> > On Sun, 28 Oct 2018 at 09:00, Chris Johns <chr...@rtems.org
>> > <mailto:chr...@rtems.org>> wrote:
>> >
>> > On 27/10/18 5:13 am, Vijay Kumar Banerjee wrote:
>> > > ---
>> > > tester/rt/coverage.py | 17 ++++-------------
>> > > 1 file changed, 4 insertions(+), 13 deletions(-)
>> > >
>> > > diff --git a/tester/rt/coverage.py b/tester/rt/coverage.py
>> > > index 978a8c4..4200dad 100644
>> > > --- a/tester/rt/coverage.py
>> > > +++ b/tester/rt/coverage.py
>> > > @@ -201,23 +201,14 @@ class report_gen_html:
>> > > index_content =
>> > self._prepare_index_content(partial_reports)
>> > > self._create_index_file(head_section,index_content)
>> > >
>> > > - def add_covoar_src_path(self):
>> > > + def add_covoar_css(self):
>> > > table_js_path = path.join(self.covoar_src_path,
>> 'table.js')
>> > > covoar_css_path = path.join(self.covoar_src_path,
>> > 'covoar.css')
>> > > for symbol_set in self.symbol_sets:
>> > > symbol_set_dir = path.join(self.build_dir,
>> > > self.bsp + '-coverage',
>> > symbol_set)
>> > > - html_files = os.listdir(symbol_set_dir)
>> > > - for html_file in html_files:
>> > > - html_file = path.join(symbol_set_dir, html_file)
>> > > - if path.exists(html_file) and 'html' in
>> html_file:
>> > > - with open(html_file, 'r') as f:
>> > > - file_data = f.read()
>> > > - file_data = file_data.replace('table.js',
>> > table_js_path)
>> > > - file_data = file_data.replace('covoar.css',
>> > > -
>> covoar_css_path)
>> > > - with open(html_file, 'w') as f:
>> > > - f.write(file_data)
>> > > + shutil.copy(covoar_css_path, symbol_set_dir)
>> > > + shutil.copy(table_js_path, symbol_set_dir)
>> >
>> > How portable are these calls?
>> >
>> > Please have a look at 'path.copy_tree' in the rtemstoolkit's
>> path.py. I
>> > think you can use that call instead.
>> >
>> > shutil is portable as it comes with python.
>>
>> It does not handle some issues on Windows.
>>
>> > i looked into the path.copy_tree
>> > it uses shutil as well,
>>
>> It does but it handles some issues expose in the RSB over many years.
>>
> Understood.
>
>>
>> > but it requires the source-name to be a directory
>> > wherefrom it will copy all the files to destination directory.
>>
>> It can be extended.
>>
> I'm getting a 'Not a directory' error as it uses listdir on the srcpath
> and then iterates over the
> files. Shall I write a path.copy for it ? with the WindowsError handling
> similar to
> copy_tree ?
>
>>
>> > Is there a reason that path.copy_tree has been used instead of
>> > shutil.copytree ?
>> > https://docs.python.org/2/library/shutil.html#shutil.copytree
>>
>> It can break on Windows.
>>
>> Please use the path module for all file operations to we can be
>> portable. I think one call was pushed in the other patch which Joel
>> merged that needs to be fixed.
>
> Sorry about that. I didn't know that it breaks on Windows.
> Can you please comment on the patch so that I can fix it?
> I'll ping you there. :)
>
>>
>>
> Chris
>>
>
_______________________________________________
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel
