fdo#85812: RTF import: fix run type in new groups

Apparently the run type resets to LTR in a new group.

(regression from fc49c052dbdbb5ab3b0a02a13143705f769b9662)

(cherry picked from commit 88d3f9e4cf64e4ef037063b26ddf347fd42d8d84)

fdo#85812: RTF import: better fix for run type in new groups

Apparently Word treats \ltrch \rtlch differently from
\loch \hich \dbch when groups are opened.

Change-Id: I257712521e8e77fa66e76857489797ecc675506e
(cherry picked from commit a9a9718bb1f64318429562ecdaa3d3763d9e2f4a)
Reviewed-on: https://gerrit.libreoffice.org/12269
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
Tested-by: Miklos Vajna <vmiklos@collabora.co.uk>

3 files changed, 114 insertions, 0 deletions

diff --git a/sw/qa/extras/rtfimport/data/fdo85812.rtf b/sw/qa/extras/rtfimport/data/fdo85812.rtf
new file mode 100644
index 000000000000..152971f4cf89
--- /dev/null
+++ b/sw/qa/extras/rtfimport/data/fdo85812.rtf
@@ -0,0 +1,39 @@
+{\rtf1\ansi\deflang1049\deflangfe2052
+{\*\generator Microsoft Word 11.0.5604;}
+
+\pard\plain
+\lang1049\langfe2052\loch\hich\dbch\langnp1049\langfenp2052
+{\lang1033\langfe2052\dbch\langnp1033 \hich\dbch\loch This }
+
+\par\pard\plain
+\lang1049\langfe2052\loch\hich\dbch
+{CharGroup}
+\lang1033 AfterChar
+
+\par\pard\plain
+\lang1049\langfe2052\loch\hich\dbch
+{\*\bkmkstart foomark}{\*\bkmkend foomark}
+\lang1033 AfterBookmark
+
+\par\pard\plain
+\lang1049\langfe2052\loch\hich\lang5121\dbch\lang1049
+{CharGroup}
+\lang1033 AfterChar
+
+\par\pard\plain
+\lang1049\langfe2052\loch\hich\lang5121\dbch\lang1049
+{\*\bkmkstart foomark}{\*\bkmkend foomark}
+\lang1033 AfterBookmark
+
+\par\pard\plain
+\lang1049\langfe2052\ltrch\lang1033\rtlch\lang5121
+{CharGroup}
+\lang3073 AfterChar
+
+\par\pard\plain
+\lang1049\langfe2052\ltrch\lang1033\rtlch\lang5121
+{\*\bkmkstart foomark}{\*\bkmkend foomark}
+\lang3073 AfterBookmark
+
+{\lang1036\langfe2052\langnp1036 \par }
+}
diff --git a/sw/qa/extras/rtfimport/rtfimport.cxx b/sw/qa/extras/rtfimport/rtfimport.cxx
index 1c3ed00a34ba..93e1904c29a2 100644
--- a/sw/qa/extras/rtfimport/rtfimport.cxx
+++ b/sw/qa/extras/rtfimport/rtfimport.cxx
@@ -475,6 +475,78 @@ DECLARE_RTFIMPORT_TEST(testFdo48037, "fdo48037.rtf")
     CPPUNIT_ASSERT_EQUAL(nExpected, nActual);
 }
 
+DECLARE_RTFIMPORT_TEST(testFdo85812, "fdo85812.rtf")
+{
+    lang::Locale locale(getProperty<lang::Locale>(
+                getRun(getParagraph(1), 1, "This "), "CharLocale"));
+    // the \lang inside the group was applied to CJK not Western
+    CPPUNIT_ASSERT_EQUAL(OUString("en"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("US"), locale.Country);
+    // further testing indicates that Word is doing really weird stuff
+    // \loch \hich \dbch is reset by opening a group
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(2), 1, "CharGroup"), "CharLocale");
+    CPPUNIT_ASSERT_EQUAL(OUString("ru"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("RU"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(2), 2, "AfterChar"), "CharLocale");
+    CPPUNIT_ASSERT_EQUAL(OUString("en"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("US"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(3), 2, "AfterBookmark"), "CharLocale");
+    CPPUNIT_ASSERT_EQUAL(OUString("en"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("US"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(4), 1, "CharGroup"), "CharLocale");
+    CPPUNIT_ASSERT_EQUAL(OUString("ru"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("RU"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(4), 1, "CharGroup"), "CharLocaleComplex");
+    CPPUNIT_ASSERT_EQUAL(OUString("ar"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("DZ"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(4), 2, "AfterChar"), "CharLocale");
+    CPPUNIT_ASSERT_EQUAL(OUString("en"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("US"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(4), 2, "AfterChar"), "CharLocaleComplex");
+    CPPUNIT_ASSERT_EQUAL(OUString("ar"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("DZ"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(5), 2, "AfterBookmark"), "CharLocale");
+    CPPUNIT_ASSERT_EQUAL(OUString("en"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("US"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(5), 2, "AfterBookmark"), "CharLocaleComplex");
+    CPPUNIT_ASSERT_EQUAL(OUString("ar"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("DZ"), locale.Country);
+    // \ltrch \rtlch works differently - it is sticky across groups
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(6), 1, "CharGroup"), "CharLocale");
+    CPPUNIT_ASSERT_EQUAL(OUString("en"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("US"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(6), 1, "CharGroup"), "CharLocaleComplex");
+    CPPUNIT_ASSERT_EQUAL(OUString("ar"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("DZ"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(6), 2, "AfterChar"), "CharLocale");
+    CPPUNIT_ASSERT_EQUAL(OUString("en"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("US"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(6), 2, "AfterChar"), "CharLocaleComplex");
+    CPPUNIT_ASSERT_EQUAL(OUString("ar"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("EG"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(7), 2, "AfterBookmark"), "CharLocale");
+    CPPUNIT_ASSERT_EQUAL(OUString("en"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("US"), locale.Country);
+    locale = getProperty<lang::Locale>(
+                getRun(getParagraph(7), 2, "AfterBookmark"), "CharLocaleComplex");
+    CPPUNIT_ASSERT_EQUAL(OUString("ar"), locale.Language);
+    CPPUNIT_ASSERT_EQUAL(OUString("EG"), locale.Country);
+}
+
 DECLARE_RTFIMPORT_TEST(testFdo47764, "fdo47764.rtf")
 {
     // \cbpat with zero argument should mean the auto (-1) color, not a default color (black)
diff --git a/writerfilter/source/rtftok/rtfdocumentimpl.cxx b/writerfilter/source/rtftok/rtfdocumentimpl.cxx
index 145fd4a3d67f..6c6e282f9845 100644
--- a/writerfilter/source/rtftok/rtfdocumentimpl.cxx
+++ b/writerfilter/source/rtftok/rtfdocumentimpl.cxx
@@ -4770,6 +4770,9 @@ int RTFDocumentImpl::pushState()
         m_aStates.push(m_aDefaultState);
     else
     {
+        // fdo#85812 group resets run type of _current_ and new state (but not RTL)
+        m_aStates.top().eRunType = RTFParserState::LOCH;
+
         if (m_aStates.top().nDestinationState == DESTINATION_MR)
             lcl_DestinationToMath(*m_aStates.top().pDestinationText, m_aMathBuffer, m_bMathNor);
         m_aStates.push(m_aStates.top());