Re: [PATCH] Fix issue #877 - work around in test suite bug

[kf...@collab.net]

"Marcos Chaves" <mc...@hotmail.com> writes:
> Fix issue #877 by adding regular expression for the 'Restored' output
> from 'svn update'.
> 
> * update_tests.py
>    (update_missing): Fixed commented out test cases.
> 
> * tree.py
>    (build_tree_from_checkout): Regular expression for 'Restored'
>    during 'svn update' added.

Thanks for the patch!  I committed it in r10488.

A couple of minor nits: the paths in the log message should be
relative to the top of your source tree.  (In this case especially so,
as the two files you changed are not actually in the same directory.)

Also, your patch got munged somehow, see below:

> Index: update_tests.py
> ===================================================================
> --- update_tests.py	(revision 10462)
> +++ update_tests.py	(working copy)
> @@ -310,10 +310,9 @@
>    E_path = os.path.join(wc_dir, 'A', 'B', 'E')
>    H_path = os.path.join(wc_dir, 'A', 'D', 'H')
> 
> -  ### FIXME run_and_verify_update doesn't appear to understand 'Restored'
> -  ### feedback
> -  #os.remove(mu_path)
> -  #os.remove(rho_path)
> +  # remove two files to verify if they will be restored
> +  os.remove(mu_path)
> +  os.remove(rho_path)
> 
>    ### FIXME I think directories work because they generate 'A'
>    ### feedback, is this the correct feedback?
> @@ -322,8 +321,8 @@
> 
>    # Create expected output tree for an update of the missing items by name
>    expected_output = svntest.wc.State(wc_dir, {
> -    #'A/mu'        : Item(status='A '),
> -    #'A/D/G/rho'   : Item(status='A '),
> +    'A/mu'        : Item(verb='Restored'),
> +    'A/D/G/rho'   : Item(verb='Restored'),
>      'A/B/E' : Item(status='A '),
>      'A/B/E/alpha' : Item(status='A '),
>      'A/B/E/beta' : Item(status='A '),
> Index: svntest/tree.py
> ===================================================================
> --- svntest/tree.py	(revision 10462)
> +++ svntest/tree.py	(working copy)
> @@ -104,8 +104,10 @@
> # depending on the parsing context:
> 
> #   - in the 'svn co/up' use-case, each line of output starts with two
> -#     characters from the set of (A, D, G, U, C, _).  This status code
> -#     is stored in a attribute named 'status'.
> +#     characters from the set of (A, D, G, U, C, _) or 'Restored'.  The
> +#     status code is stored in a attribute named 'status'. In the case
> +#     of a restored file, the word 'Restored' is stored in a attribute
> +#     named 'verb'.
> 
> #   - in the 'svn ci/im' use-case, each line of output starts with one
> #      of the words (Adding, Deleting, Sending).  This verb is stored in

Indentation problem: those "#"s in the first column should be in the
second column.  I fixed it by hand before applying.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org