[trunk] Docstrings Byte Literal contains non-ascii characters

[Guewen Baconnier @ Camptocamp]

Hello,

In the docstring methods, the following line is often present :
  @param uid: the current user’s ID for security checks

The char ’ is not an ascii char and the docstrings are not declared as unicode (u"""docstring content""") [1]

With the command below, 254 lines match.
grep -hcr "the current user’s ID for security checks" * | awk '{ count += $0} END {print count}'

I think we should do some sed on all addons to correct that (and server as there is the same issue) like :
grep -hlr "the current user’s ID for security checks" * | xargs sed -i "s/the current user’s ID for security checks/the current user's ID for security checks/g"

Some others with the same problem:
grep -hcr "’" * | awk '{ count += $0} END {print count}' (result : 688)
@param ids: List of open wiki page’s IDs
@param ids: List of calendar event import’s IDs
@param data_id: Get Data’s ID or False
@param ids: List of auction lots enable’s IDs.

Thanks
Guewen

[1] http://www.python.org/dev/peps/pep-0257/

[xrg (xrg)]

On Thursday 22 September 2011, you wrote:
> Public bug reported:
>
> Hello,
>
> In the docstring methods, the following line is often present :
> @param uid: the current user’s ID for security checks
>
++

indeed, have noted this and complained a few weeks ago.

Even more, the (self, cr, uid, ids , context) parameters should not be
documented at each model.method. They are standard API convention and thus
need no more explanation.

(yes, there is several methods documenting "self", but not all those cryptic
parameters after 'ids' ..)

[Guewen Baconnier @ Camptocamp (gbaconnier-c2c)]

I do agree with you.

[Vishal Parmar(Open ERP) (vpa-openerp)]

Hello,

This issue is generalize issue so I am assigning to framework.

Thanks.

[Olivier Dony (Odoo) (odo-openerp)]

Note that all our Python source files are UTF-8 encoded, as shown in the coding hint at the top of each file:
  # -*- coding: utf-8 -*-
So all unicode characters are just fine in any string, including docstrings: they're gonna be utf8-encoded.
And that's fine because no-one should ever assume that strings are ascii, even docstrings.

Now I do agree that there's no need for the funky apostrophe here, or for documenting self,cr, uid or ids at all (I think it's in our guidelines, and if not I will add it). So we should simply get rid of that useless stuff :-)

PS: definitely a wishlist