"Robert Shearman" R.J.Shearman@warwick.ac.uk wrote:
While you are at it, perhaps it would be a better idea to use CompareStringA/W with appropriate flags for all those StrCmpNIA/W, StrCmpNA/W, etc.
Like this?
Not really. Why do you need ugly helpers/wrappers? Why to not use CompareStringA/W directly?
StrCmpNIA/W need to be changed as well.
On Tue, 2004-02-17 at 15:08, Dmitry Timoshkov wrote:
"Robert Shearman" R.J.Shearman@warwick.ac.uk wrote:
While you are at it, perhaps it would be a better idea to use CompareStringA/W with appropriate flags for all those StrCmpNIA/W, StrCmpNA/W, etc.
Like this?
Not really. Why do you need ugly helpers/wrappers? Why to not use CompareStringA/W directly?
Because this implementation is copied from the SHLWAPI implementation. I did try to submit a test case to test for DBCS compatibility there, but as was noted by several people, DBCS depends on the current locale so the test was wrong. Therefore, I couldn't change any of the functions without risking breaking them.
StrCmpNIA/W need to be changed as well.
Rob
"Robert Shearman" R.J.Shearman@warwick.ac.uk wrote:
Because this implementation is copied from the SHLWAPI implementation. I did try to submit a test case to test for DBCS compatibility there, but as was noted by several people, DBCS depends on the current locale so the test was wrong. Therefore, I couldn't change any of the functions without risking breaking them.
That's true for Chr* APIs, but APIs that take strings as parameters can be easily changed to use CompareString.
Actually you *can* submit DBCS tests, just call GetCPInfoEx before doing an actual testing, and if MaxCharSize != 2 bail out with a message explaining why the test is skipped.
-
Dmitry Timoshkov -
Robert Shearman