Related tdf#104597, tdf#151546: Introduce comphelper::string::reverseCodePoints

69e9925ded584113e52f84ef0ed7c224079fa061 "sdext.pdfimport: resolves tdf#104597:
RTL script text runs are reversed" and f6004e1c457ddab5e0c91e6159875d25130b108a
"tdf#151546: RTL text is reversed (Writer pdfimport)" had introduced two calls
to comphelper::string::reverseString into sdext.  That function reverts on the
basis of individual UTF-16 code units, not on the basis of Unicode code points.
And while at least some pre-existing callers of that function want the former
semantics (see below), these two new callers in sdext apparently want the latter
semantics.  Therefore, introduce an additional function
comphelper::string::reverseCodePoints with the latter semantics.

I identified three other places that call comphelper::string::reverseString:
* SbRtl_StrReverse in basic/source/runtime/methods1.cxx apparently implements
some StrReverse Basic function, where a (presumably non-existing) Basic spec
would need to decide which of the two semantics is called for.  So leave it
alone for now.
* SvtFileDialog::IsolateFilterFromPath_Impl in fpicker/source/office/iodlg.cxx
reverts a string, operates on it, then reverts (parts of) it back.  Whether or
not that is the most elegant code, using the latter semantics here would
apparently be wrong, as double invocation of
comphelper::string::reverseCodePoints is not idempotent when the input is a
malformed sequence of UTF-16 code units containing a low surrogate followed by
a high surrogate.
* AccessibleCell::getCellName in svx/source/table/accessiblecell.cxx apparently
always operates on a string consisting only of Latin uppercase letters A--Z,
for which both semantics are equivalent.  (So we can just as well stick with
the simpler comphelper::string::reverseString here.)

(Extending the tests in comphelper/qa/string/test_string.cxx ran into an issue
where loplugin:stringliteralvar warns about deliberate uses of sal_Unicode
arrays rather than UTF-16 string literals wrapped in OUStringLiteral, as those
arrays deliberately contain malformed UTF-16 code unit sequences and thus
converting them into UTF-16 string literals might be considered inappropriate,
see the newly added comment at
StringLiteralVar::isPotentiallyInitializedWithMalformedUtf16 in
compilerplugins/clang/stringliteralvar.cxx for details.  So that loplugin had to
be improved here, too.)

Change-Id: I641cc32c76b0c5f6339ae44d8aa85df0022ffb05
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/142949
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>

7 files changed, 128 insertions, 5 deletions

diff --git a/comphelper/qa/string/test_string.cxx b/comphelper/qa/string/test_string.cxx
index 0b2f3ee05479..58f9c3f63c16 100644
--- a/comphelper/qa/string/test_string.cxx
+++ b/comphelper/qa/string/test_string.cxx
@@ -17,6 +17,10 @@
  *   the License at http://www.apache.org/licenses/LICENSE-2.0 .
  */
 
+#include <sal/config.h>
+
+#include <iterator>
+
 #include <comphelper/string.hxx>
 #include <cppuhelper/implbase.hxx>
 #include <com/sun/star/i18n/CharType.hpp>
@@ -43,6 +47,7 @@ public:
     void testDecimalStringToNumber();
     void testIsdigitAsciiString();
     void testReverseString();
+    void testReverseCodePoints();
     void testSplit();
     void testRemoveAny();
 
@@ -55,6 +60,7 @@ public:
     CPPUNIT_TEST(testDecimalStringToNumber);
     CPPUNIT_TEST(testIsdigitAsciiString);
     CPPUNIT_TEST(testReverseString);
+    CPPUNIT_TEST(testReverseCodePoints);
     CPPUNIT_TEST(testSplit);
     CPPUNIT_TEST(testRemoveAny);
     CPPUNIT_TEST_SUITE_END();
@@ -178,9 +184,28 @@ void TestString::testTokenCount()
 
 void TestString::testReverseString()
 {
-    OUString aOut = ::comphelper::string::reverseString(u"ABC");
+    CPPUNIT_ASSERT_EQUAL(OUString(), comphelper::string::reverseString(u""));
+    CPPUNIT_ASSERT_EQUAL(OUString("cba"), comphelper::string::reverseString(u"abc"));
+    static sal_Unicode const rev[] = {'w', 0xDFFF, 0xDBFF, 'v', 0xDC00, 0xD800, 'u'};
+    CPPUNIT_ASSERT_EQUAL(
+        OUString(rev, std::size(rev)),
+        comphelper::string::reverseString(u"u\U00010000v\U0010FFFFw"));
+    static sal_Unicode const malformed[] = {0xDC00, 0xD800};
+    CPPUNIT_ASSERT_EQUAL(
+        OUString(u"\U00010000"),
+        comphelper::string::reverseString(std::u16string_view(malformed, std::size(malformed))));
+}
 
-    CPPUNIT_ASSERT_EQUAL(OUString("CBA"), aOut);
+void TestString::testReverseCodePoints() {
+    CPPUNIT_ASSERT_EQUAL(OUString(), comphelper::string::reverseCodePoints(""));
+    CPPUNIT_ASSERT_EQUAL(OUString("cba"), comphelper::string::reverseCodePoints("abc"));
+    CPPUNIT_ASSERT_EQUAL(
+        OUString(u"w\U0010FFFFv\U00010000u"),
+        comphelper::string::reverseCodePoints(u"u\U00010000v\U0010FFFFw"));
+    static sal_Unicode const malformed[] = {0xDC00, 0xD800};
+    CPPUNIT_ASSERT_EQUAL(
+        OUString(u"\U00010000"),
+        comphelper::string::reverseCodePoints(OUString(malformed, std::size(malformed))));
 }
 
 void TestString::testSplit()
diff --git a/comphelper/source/misc/string.cxx b/comphelper/source/misc/string.cxx
index 95e936e60142..d0f7cb9ca423 100644
--- a/comphelper/source/misc/string.cxx
+++ b/comphelper/source/misc/string.cxx
@@ -530,6 +530,15 @@ OUString reverseString(std::u16string_view rStr)
     return sBuf.makeStringAndClear();
 }
 
+OUString reverseCodePoints(OUString const & str) {
+    auto const len = str.getLength();
+    OUStringBuffer buf(len);
+    for (auto i = len; i != 0;) {
+        buf.appendUtf32(str.iterateCodePoints(&i, -1));
+    }
+    return buf.makeStringAndClear();
+}
+
 sal_Int32 indexOfAny(std::u16string_view rIn,
         sal_Unicode const*const pChars, sal_Int32 const nPos)
 {
diff --git a/compilerplugins/clang/stringliteralvar.cxx b/compilerplugins/clang/stringliteralvar.cxx
index 5ace384f1e16..fcd3690669e7 100644
--- a/compilerplugins/clang/stringliteralvar.cxx
+++ b/compilerplugins/clang/stringliteralvar.cxx
@@ -28,6 +28,7 @@
 #include <cassert>
 
 #include "check.hxx"
+#include "compat.hxx"
 #include "plugin.hxx"
 
 namespace
@@ -137,6 +138,10 @@ public:
                     return true;
                 }
                 auto const d = e1->getDecl();
+                if (isPotentiallyInitializedWithMalformedUtf16(d))
+                {
+                    return true;
+                }
                 if (!reportedArray_.insert(d).second)
                 {
                     return true;
@@ -188,6 +193,10 @@ public:
             return true;
         }
         auto const d = e->getDecl();
+        if (isPotentiallyInitializedWithMalformedUtf16(d))
+        {
+            return true;
+        }
         if (!reportedArray_.insert(d).second)
         {
             return true;
@@ -246,6 +255,61 @@ private:
         }
     }
 
+    // There is some confusion on the semantics of numeric-escape-sequences in string literals, see
+    // <https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2020/p2029r4.html> "Proposed resolution
+    // for core issues 411, 1656, and 2333; numeric and universal character escapes in character and
+    // string literals", so suppress warnings about arrays that are deliberately not written as
+    // UTF-16 string literals because they contain lone surrogates:
+    bool isPotentiallyInitializedWithMalformedUtf16(ValueDecl const* decl) const
+    {
+        if (!decl->getType()->getArrayElementTypeNoTypeQual()->isChar16Type())
+        {
+            return false;
+        }
+        auto const init = cast<VarDecl>(decl)->getAnyInitializer();
+        if (init == nullptr)
+        {
+            return true;
+        }
+        auto const list = dyn_cast<InitListExpr>(init);
+        if (list == nullptr)
+        {
+            // Assuming that the initializer already is a string literal, assume that that string
+            // literal has no issues with malformed UTF-16:
+            if (isDebugMode())
+            {
+                assert(isa<clang::StringLiteral>(init));
+            }
+            return false;
+        }
+        auto highSurrogate = false;
+        for (auto const e : list->inits())
+        {
+            llvm::APSInt v;
+            if (!compat::EvaluateAsInt(e, v, compiler.getASTContext()))
+            {
+                return true;
+            }
+            if (highSurrogate)
+            {
+                if (v < 0xDC00 || v > 0xDFFF)
+                {
+                    return true;
+                }
+                highSurrogate = false;
+            }
+            else if (v >= 0xD800 && v <= 0xDBFF)
+            {
+                highSurrogate = true;
+            }
+            else if (v >= 0xDC00 && v <= 0xDFFF)
+            {
+                return true;
+            }
+        }
+        return highSurrogate;
+    }
+
     std::set<Decl const*> reportedAutomatic_;
     std::set<Decl const*> reportedArray_;
 };
diff --git a/compilerplugins/clang/test/stringliteralvar.cxx b/compilerplugins/clang/test/stringliteralvar.cxx
index 3c0eaaccae04..d0fdcedb0668 100644
--- a/compilerplugins/clang/test/stringliteralvar.cxx
+++ b/compilerplugins/clang/test/stringliteralvar.cxx
@@ -110,4 +110,25 @@ void f11(int nStreamType)
     (void)sStreamType;
 }
 
+extern sal_Unicode const extarr[1];
+
+sal_Unicode init();
+
+void f12()
+{
+    // Suppress warnings if the array contains a malformed sequence of UTF-16 code units...:
+    static sal_Unicode const arr1[] = { 0xD800 };
+    f(OUString(arr1, 1));
+    // ...Or potentially contains a malformed sequence of UTF-16 code units...:
+    f(OUString(extarr, 1));
+    sal_Unicode const arr2[] = { init() };
+    f(OUString(arr2, 1));
+    // ...But generate a warning if the array contains a well-formed sequence of UTF-16 code units
+    // containing surrogates:
+    // expected-error-re@+1 {{change type of variable 'arr3' from constant character array ('const sal_Unicode{{ ?}}[2]'{{( \(aka 'const char16_t\[2\]'\))?}}) to OUStringLiteral [loplugin:stringliteralvar]}}
+    static sal_Unicode const arr3[] = { 0xD800, 0xDC00 };
+    // expected-note-re@+1 {{first passed into a '{{(rtl::)?}}OUString' constructor here [loplugin:stringliteralvar]}}
+    f(OUString(arr3, 2));
+}
+
 /* vim:set shiftwidth=4 softtabstop=4 expandtab cinoptions=b1,g0,N-s cinkeys+=0=break: */
diff --git a/include/comphelper/string.hxx b/include/comphelper/string.hxx
index cee741bc865e..b43086a77d30 100644
--- a/include/comphelper/string.hxx
+++ b/include/comphelper/string.hxx
@@ -150,13 +150,17 @@ COMPHELPER_DLLPUBLIC sal_Int32 getTokenCount(std::string_view rIn, char cTok);
 */
 COMPHELPER_DLLPUBLIC sal_Int32 getTokenCount(std::u16string_view rIn, sal_Unicode cTok);
 
-/** Reverse an OUString
+/** Reverse an OUString's UTF-16 code units.
 
   @param    rIn     the input OUString
   @return   the reversed input
 */
 COMPHELPER_DLLPUBLIC OUString reverseString(std::u16string_view rStr);
 
+/** Reverse an OUString's Unicode code points.
+*/
+COMPHELPER_DLLPUBLIC OUString reverseCodePoints(OUString const & str);
+
 
 namespace detail
 {
diff --git a/sdext/source/pdfimport/tree/drawtreevisiting.cxx b/sdext/source/pdfimport/tree/drawtreevisiting.cxx
index 9f760150343c..ead2fd432452 100644
--- a/sdext/source/pdfimport/tree/drawtreevisiting.cxx
+++ b/sdext/source/pdfimport/tree/drawtreevisiting.cxx
@@ -123,7 +123,7 @@ void DrawXmlEmitter::visit( TextElement& elem, const std::list< std::unique_ptr<
     }
 
     if (isRTL)  // If so, reverse string
-        str = ::comphelper::string::reverseString(str);
+        str = ::comphelper::string::reverseCodePoints(str);
 
     m_rEmitContext.rEmitter.beginTag( "text:span", aProps );
 
diff --git a/sdext/source/pdfimport/tree/writertreevisiting.cxx b/sdext/source/pdfimport/tree/writertreevisiting.cxx
index 510689be1588..78c1cf00c865 100644
--- a/sdext/source/pdfimport/tree/writertreevisiting.cxx
+++ b/sdext/source/pdfimport/tree/writertreevisiting.cxx
@@ -112,7 +112,7 @@ void WriterXmlEmitter::visit( TextElement& elem, const std::list< std::unique_pt
     }
 
     if (isRTL)  // If so, reverse string
-        str = ::comphelper::string::reverseString(str);
+        str = ::comphelper::string::reverseCodePoints(str);
 
     m_rEmitContext.rEmitter.beginTag( "text:span", aProps );