RE: [platform-core-dev] Latest on path character issues

[John Arthorne <John_Arthorne@xxxxxxxxxx>]

Thanks for the comments, Ed.

1) isValidSegment/isValidPath will not
be deprecated, and they will enforce platform-specific segment and device
separators. I have clarified this in the proposal.
2) This was not mentioned in the proposal.
The two argument constructor never specified that it looked for the device
in the second argument if not found in the first.  This "feature"
will be removed, so it will take the device from the first argument only.
If the second argument contains literal colons, they will be tolerated
on all platforms.
3) I'm not sure how else I can clarify
it. I have attached an implementation to bug 24152 if you want to look
at the code or try it.  Here are a couple of important examples:

The absolute path with device "C:"
and single segment "foo" would look like:

toString -> C:/foo
toPortableString -> C:/foo
toOSString (windows) -> C:\foo
toOSString (Linux) -> C:/foo


The relative path with null device,
and two segments "C:" and "foo" would look like:

toString -> C:/foo
toPortableString -> C::/foo
toOSString (windows) -> C:\foo
toOSString (Linux) -> C:/foo

"Ed Burnette" <Ed.Burnette@xxxxxxx>
Sent by: platform-core-dev-admin@xxxxxxxxxxx

10/18/2004 04:18 PM

Please respond to
platform-core-dev

To	<platform-core-dev@xxxxxxxxxxx>
cc
Subject	RE: [platform-core-dev] Latest on path character issues

This looks better. Notes:
 
1. You have isValidSegment() and
isValidPath() deprecated in one section (Other migration issues) and not
deprecated in an earlier section (API Details). I think you are intending
for nothing to be deprecated in this latest version.
2. It's not clear what change,
if any, is proposed for the two-argument Path() contstructor. On Windows
if the device is null does it look for device in the second argument? How
about on Unix? If device is not null does it still look for and strip off
(but otherwise ignore) a device in the second argument on Windows? Unix?
3. It's still not clear how the
output of toString(), toOSString(), and toPortableString() will vary for
various paths on various hosts.

---

From: platform-core-dev-admin@xxxxxxxxxxx
[mailto:platform-core-dev-admin@xxxxxxxxxxx] On Behalf Of John Arthorne
Sent: Monday, October 18, 2004 2:13 PM
To: platform-core-dev@xxxxxxxxxxx
Subject: [platform-core-dev] Latest on path character issues


Based on feedback from the previous proposal, and from the experience of
implementing and testing various solutions, I have made some changes to
the proposal to allow colon and backslash characters in paths.  ...