The difference between the 2 charactersets:
* 27 codepoints are NOT existing in P1 that are filled in / used in WE8MSWIN1252
(note that the WE8MSWIN1252 codepoints
91 = U+2018 : LEFT SINGLE QUOTATION MARK, 92 = U+2019 : RIGHT SINGLE QUOTATION MARK,
93 = U+201C : LEFT DOUBLE QUOTATION MARK, 94 = U+201D : RIGHT DOUBLE QUOTATION MARK
are the default quotation marks of the Microsoft Word product, so if you have data that comes from
Microsoft Office products you *need* a WE8MSWIN1252 database characterset.
Make sure you clients NLS_LANG is also correct:
Note 179133.1 The correct NLS_LANG in a Windows Environment
* all characters included in the P1 characterset are included in WE8MSWIN1252 so 1252 is
a *logical* superset of P1
* no codepoints have a different symbol in WE8MSWIN1252 than in P1 so
1252 is a *binary* superset of P1.
* up to 8i the *default* NLS_LANG of a windows (!) client installation was
WE8IOS8895P1, wich is in fact incorrect, the correct value is WE8MSWIN1252.