Add note to XCharacterClassification::getStringType() about meaninglessness

Given how it works that's almost deprecated, almost..

Change-Id: Ife2fe012db5568e4646b98b0ba55df8418069062
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/139534
Reviewed-by: Eike Rathke <erack@redhat.com>
Tested-by: Jenkins

1 files changed, 10 insertions, 0 deletions

diff --git a/offapi/com/sun/star/i18n/XCharacterClassification.idl b/offapi/com/sun/star/i18n/XCharacterClassification.idl
index 72f69f850f49..2d8da610f4f7 100644
--- a/offapi/com/sun/star/i18n/XCharacterClassification.idl
+++ b/offapi/com/sun/star/i18n/XCharacterClassification.idl
@@ -106,6 +106,16 @@ published interface XCharacterClassification : com::sun::star::uno::XInterface
             A number with appropriate flags set to indicate what type of
             characters the string contains, each flag value being one of
             KCharacterType values.
+
+        @note The accumulated bits of several characters are meaningless
+            as soon as characters of different classifications are
+            involved, which even may have a common subset like
+            KCharacterType::LETTER or KCharacterType::PRINTABLE, unless
+            it is to be determined what overall character properties are
+            present in the string. Use getCharacterType() of single
+            characters instead and handle bits as needed if sets of
+            character properties are to be obtained.
+
     */
     long getStringType( [in] string aText, [in] long nPos, [in] long nCount,
                         [in] com::sun::star::lang::Locale aLocale );