tdf#64038 ODF import: fix handling of space following <text:s>

The whitespace collapse algorithm given in ODF 1.2 part 1,
"6.1.2 White Space Characters", was not implemented as well as it could be.

Although the problematic case is arguably invalid, because "6.1.3
<text:s>" says:

This element shall be used to represent the second and all following
“ “ (U+0020, SPACE) characters in a sequence of “ “ (U+0020, SPACE)
characters.

Hence it is probably irrelevant in practice that a space following <text:s>
was ignored, because a conforming document cannot have such content.

Change-Id: Ic30347fff27176c511ea317d46a1011b410e57d5

2 files changed, 79 insertions, 0 deletions

diff --git a/sw/qa/extras/odfimport/data/space.odt b/sw/qa/extras/odfimport/data/space.odt
new file mode 100644
index 000000000000..66b5109efca0
--- /dev/null
+++ b/sw/qa/extras/odfimport/data/space.odt
diff --git a/sw/qa/extras/odfimport/odfimport.cxx b/sw/qa/extras/odfimport/odfimport.cxx
index e8bc1cf41f51..90f1da1f96f2 100644
--- a/sw/qa/extras/odfimport/odfimport.cxx
+++ b/sw/qa/extras/odfimport/odfimport.cxx
@@ -249,6 +249,85 @@ DECLARE_ODFIMPORT_TEST(testPageStyleLayoutDefault, "hello.odt")
     CPPUNIT_ASSERT_EQUAL(style::PageStyleLayout_ALL, getProperty<style::PageStyleLayout>(xPropertySet, "PageStyleLayout"));
 }
 
+DECLARE_ODFIMPORT_TEST(testTdf64038, "space.odt")
+{
+    // no space
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(4), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(4), 2)->getString());
+    // one space
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(6), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(6), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(7), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" "), getRun(getParagraph(7), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(7), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(8), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" b"), getRun(getParagraph(8), 2)->getString());
+    // two spaces
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(10), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" "), getRun(getParagraph(10), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(10), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(11), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(11), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(12), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(12), 2)->getString());
+    // three spaces
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(14), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(14), 2)->getString());
+    // no space
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(17), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" "), getRun(getParagraph(17), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(17), 3)->getString());
+    // one space
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(19), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" "), getRun(getParagraph(19), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(19), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(20), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("  "), getRun(getParagraph(20), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(20), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(21), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("  "), getRun(getParagraph(21), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(21), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(22), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" "), getRun(getParagraph(22), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" b"), getRun(getParagraph(22), 3)->getString());
+    // two spaces
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(24), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" "), getRun(getParagraph(24), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(24), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(25), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("  "), getRun(getParagraph(25), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(25), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(26), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" "), getRun(getParagraph(26), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" b"), getRun(getParagraph(26), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(27), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("   "), getRun(getParagraph(27), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(27), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(28), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("  "), getRun(getParagraph(28), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" b"), getRun(getParagraph(28), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(29), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("  "), getRun(getParagraph(29), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(29), 3)->getString());
+    // three spaces
+    CPPUNIT_ASSERT_EQUAL(OUString("a"), getRun(getParagraph(31), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("   "), getRun(getParagraph(31), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(31), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(32), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("  "), getRun(getParagraph(32), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(32), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(33), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" "), getRun(getParagraph(33), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString(" b"), getRun(getParagraph(33), 3)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(34), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("  "), getRun(getParagraph(34), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(34), 3)->getString());
+    // four spaces
+    CPPUNIT_ASSERT_EQUAL(OUString("a "), getRun(getParagraph(36), 1)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("  "), getRun(getParagraph(36), 2)->getString());
+    CPPUNIT_ASSERT_EQUAL(OUString("b"), getRun(getParagraph(36), 3)->getString());
+}
+
 DECLARE_ODFIMPORT_TEST(testTdf74524, "tdf74524.odt")
 {
     uno::Reference<text::XTextFieldsSupplier> xTextFieldsSupplier(mxComponent, uno::UNO_QUERY);