xmloff: ODF import: improve meta:generator checks

Instead of a hard-coded check for (effectively) one project name
"LibreOffice" (which is build-time configurable), check for the string
"LibreOffice_project", which has been produced hard-coded ever since
LO 3.3.0.

This now recognises additional downstreams "LibreOffice_Vanilla"
and "Collabora_Office", and also historic "BrOffice".

An important point here is that the build-time configurable version
numbers of any downstream that retains the hard-coded
"LibreOffice_project" *MUST* be the same as the upstream's at least
in their major and minor versions (micro and further digits are
currently not used), and that such downstreams don't backport
changes with ODF export compatibility impact further than upstream.

Add a unit test too, with a representative sample of 4372 distinct
generators in bugzilla attachments.

This revealed that StarOffice 6 and AOO 4.0.1 were falling through
the cracks and not recognised, so fix that too.

Change-Id: I8105222d3428e7b20cc4a6b8e76732c697812594
Reviewed-on: https://gerrit.libreoffice.org/51171
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>

3 files changed, 148 insertions, 9 deletions

diff --git a/xmloff/qa/unit/uxmloff.cxx b/xmloff/qa/unit/uxmloff.cxx
index bb4a15cc8852..71a1b371f006 100644
--- a/xmloff/qa/unit/uxmloff.cxx
+++ b/xmloff/qa/unit/uxmloff.cxx
@@ -10,6 +10,12 @@
 #include <sal/config.h>
 #include <test/bootstrapfixture.hxx>
 
+#include <com/sun/star/beans/PropertyAttribute.hpp>
+
+#include <comphelper/genericpropertyset.hxx>
+
+#include <xmloff/xmlimp.hxx>
+#include <xmloff/xmlmetai.hxx>
 #include <xmloff/xmlexp.hxx>
 #include <xmloff/xmltoken.hxx>
 #include <xmloff/xmlaustp.hxx>
@@ -30,9 +36,11 @@ public:
     virtual void tearDown() override;
 
     void testAutoStylePool();
+    void testMetaGenerator();
 
     CPPUNIT_TEST_SUITE(Test);
     CPPUNIT_TEST(testAutoStylePool);
+    CPPUNIT_TEST(testMetaGenerator);
     CPPUNIT_TEST_SUITE_END();
 private:
     SvXMLExport *pExport;
@@ -87,6 +95,134 @@ void Test::testAutoStylePool()
     CPPUNIT_ASSERT_EQUAL_MESSAGE( "same style not found", aName, aSameName );
 }
 
+void Test::testMetaGenerator()
+{
+    comphelper::PropertyMapEntry const aInfoMap[] = {
+        { OUString("BuildId"), 0, ::cppu::UnoType<OUString>::get(), beans::PropertyAttribute::MAYBEVOID, 0 },
+        { OUString(), 0, css::uno::Type(), 0, 0 }
+    };
+    uno::Reference<beans::XPropertySet> const xInfoSet(
+        comphelper::GenericPropertySet_CreateInstance(
+            new comphelper::PropertySetInfo(aInfoMap)));
+
+    struct {
+        char const*const generator;
+        char const*const buildId;
+        sal_uInt16 const result;
+    } const tests [] = {
+        // foreign
+        { "AbiWord/2.8.6 (unix, gtk)", "", SvXMLImport::ProductVersionUnknown },
+        { "Aspose.Words for Java 13.10.0.0", "", SvXMLImport::ProductVersionUnknown },
+        { "CIB jsmerge 1.0.0", "", SvXMLImport::ProductVersionUnknown },
+        { "Calligra/2.4.3", "", SvXMLImport::ProductVersionUnknown },
+        { "CocoaODFWriter/1339", "", SvXMLImport::ProductVersionUnknown },
+        { "KOffice/1.4.1", "", SvXMLImport::ProductVersionUnknown },
+        { "KPresenter 1.3", "", SvXMLImport::ProductVersionUnknown },
+        { "KSpread 1.3.2", "", SvXMLImport::ProductVersionUnknown },
+        { "Lotus Symphony/1.2.0_20081023.1730/Win32", "", SvXMLImport::ProductVersionUnknown },
+        { "Microsoft Excel Online", "", SvXMLImport::ProductVersionUnknown },
+        { "MicrosoftOffice/12.0 MicrosoftExcel/CalculationVersion-4518", "", SvXMLImport::ProductVersionUnknown },
+        { "MicrosoftOffice/15.0 MicrosoftWord", "", SvXMLImport::ProductVersionUnknown },
+        { "ODF Converter v1.0.0", "", SvXMLImport::ProductVersionUnknown },
+        { "ODF::lpOD 1.121", "", SvXMLImport::ProductVersionUnknown },
+        { "ODFDOM/0.6.1$Build-1", "", SvXMLImport::ProductVersionUnknown },
+        { "ODFPY/0.9.6", "", SvXMLImport::ProductVersionUnknown },
+        { "OpenXML/ODF Translator Command Line Tool 3.0 2.0.0", "", SvXMLImport::ProductVersionUnknown },
+        { "Org-7.8.03/Emacs-24.0.93.1", "", SvXMLImport::ProductVersionUnknown },
+        { "TeX4ht from eqns_long.tex, options: xhtml,ooffice,refcaption", "", SvXMLImport::ProductVersionUnknown },
+        { "TextMaker", "", SvXMLImport::ProductVersionUnknown },
+        { "docbook2odf generator (http://open.comsultia.com/docbook2odf/)", "", SvXMLImport::ProductVersionUnknown },
+        { "fig2sxd", "", SvXMLImport::ProductVersionUnknown },
+        { "gnumeric/1.10.9", "", SvXMLImport::ProductVersionUnknown },
+        { "libodfgen/0.1.6", "", SvXMLImport::ProductVersionUnknown },
+
+        // OOo 1.x
+        { "StarSuite 6.0 (Linux)", "645$8687", SvXMLImport::OOo_1x },
+        { "StarOffice 6.1 (Win32)", "645$8687", SvXMLImport::OOo_1x },
+        { "OpenOffice.org 1.1.2RC3.DE (Win32)", "645$8687", SvXMLImport::OOo_1x },
+        { "OpenOffice.org 1.1.5 (Win32)", "645$8687", SvXMLImport::OOo_1x },
+        { "StarOffice 7 (Win32)", "645$8687", SvXMLImport::OOo_1x },
+
+        // OOo 2.x
+        { "Sun_ODF_Plugin_for_Microsoft_Office/1.1$Win32 OpenOffice.org_project/680m5$Build-9221", "680$9221", SvXMLImport::OOo_2x },
+        { "StarSuite/8$Win32 OpenOffice.org_project/680m6$Build-9095", "680$9095", SvXMLImport::OOo_2x },
+        { "StarOffice/8$Win32 OpenOffice.org_project/680m93$Build-8897", "680$8897", SvXMLImport::OOo_2x },
+        { "OpenOffice.org/2.0$Linux OpenOffice.org_project/680m3$Build-8968", "680$8968", SvXMLImport::OOo_2x },
+        { "OpenOffice.org/2.1$Win32 OpenOffice.org_project/680m6$Build-9095", "680$9095", SvXMLImport::OOo_2x },
+        { "OpenOffice.org/2.4$Win32 OpenOffice.org_project/680m248$Build-9274", "680$9274", SvXMLImport::OOo_2x },
+
+        // OOo 3.x
+        { "OpenOffice.org/3.0$Solaris_Sparc OpenOffice.org_project/300m9$Build-9358", "300$9358", SvXMLImport::OOo_30x },
+        { "StarSuite/9$Unix OpenOffice.org_project/300m9$Build-9358", "300$9358", SvXMLImport::OOo_30x },
+        { "StarOffice/9$Win32 OpenOffice.org_project/300m14$Build-9376", "300$9376", SvXMLImport::OOo_30x },
+        { "OpenOffice.org/3.1$Solaris_x86 OpenOffice.org_project/310m11$Build-9399", "310$9399", SvXMLImport::OOo_31x },
+        { "IBM_Lotus_Symphony/2.0$Win32 OpenOffice.org_project/310m11$Build-9399", "310$9399", SvXMLImport::OOo_31x },
+        { "BrOffice.org/3.1$Linux OpenOffice.org_project/310m11$Build-9399", "310$9399", SvXMLImport::OOo_31x },
+        { "StarOffice/9$Solaris_Sparc OpenOffice.org_project/310m19$Build-9420", "310$9420", SvXMLImport::OOo_31x },
+        { "OpenOffice.org/3.2$Linux OpenOffice.org_project/320m12$Build-9483", "320$9483", SvXMLImport::OOo_32x },
+        { "StarOffice/9$Win32 OpenOffice.org_project/320m12$Build-9483", "320$9483", SvXMLImport::OOo_32x },
+        { "OpenOffice.org/3.3$Linux OpenOffice.org_project/330m20$Build-9567", "330$9567", SvXMLImport::OOo_33x },
+        { "Oracle_Open_Office/3.3$Win32 OpenOffice.org_project/330m7$Build-9552", "330$9552", SvXMLImport::OOo_33x },
+        { "OpenOffice.org/3.4$Unix OpenOffice.org_project/340m1$Build-9590", "340$9590", SvXMLImport::OOo_34x },
+
+        // AOO versions
+        { "OpenOffice/4.0.0$Win32 OpenOffice.org_project/400m3$Build-9702", "400$9702", SvXMLImport::AOO_40x },
+        { "OpenOffice/4.0.1$Linux OpenOffice.org_project/401m4$Build-9713", "401$9713", SvXMLImport::AOO_40x },
+        { "OpenOffice/4.1.1$FreeBSD/amd64 OpenOffice.org_project/411m6$Build-9775", "411$9775", SvXMLImport::AOO_4x },
+        { "OpenOffice/4.1.2$OS/2 OpenOffice.org_project/412m3$Build-9782-bww", "412$9782-bww", SvXMLImport::AOO_4x },
+        { "OpenOffice/4.1.4$Unix OpenOffice.org_project/414m2$Build-9785", "414$9785", SvXMLImport::AOO_4x },
+
+        // LO versions
+        { "LibreOffice/3.3$Linux LibreOffice_project/330m17$Build-3", "330$3;33", SvXMLImport::LO_3x },
+        { "BrOffice/3.3$Win32 LibreOffice_project/330m19$Build-8", "330$8;33", SvXMLImport::LO_3x },
+        { "LibreOffice/3.4$Linux LibreOffice_project/340m1$Build-1206", "340$1206;34", SvXMLImport::LO_3x },
+        { "LibreOffice/3.5$Linux_X86_64 LibreOffice_project/3fa2330-e49ffd2-90d118b-705e248-051e21c", ";35", SvXMLImport::LO_3x },
+        { "LibreOffice/3.6$Windows_x86 LibreOffice_project/a9a0717-273e462-768e6e3-978247f-65e65f", ";36", SvXMLImport::LO_3x },
+        { "LibreOffice/4.0.2.2$Windows_x86 LibreOffice_project/4c82dcdd6efcd48b1d8bba66bfe1989deee49c3", ";4022", SvXMLImport::LO_41x },
+        { "LibreOffice/4.1.2.3$MacOSX_x86 LibreOffice_project/40b2d7fde7e8d2d7bc5a449dc65df4d08a7dd38", ";4123", SvXMLImport::LO_41x },
+        { "LibreOffice/4.2.8.2$Windows_x86 LibreOffice_project/48d50dbfc06349262c9d50868e5c1f630a573ebd", ";4282", SvXMLImport::LO_42x },
+        { "LibreOffice_from_Collabora_4.2-8/4.2.10.8$Linux_x86 LibreOffice_project/84584cc237b2eb93f7684d8fcd063bb37e87b5fb", ";42108", SvXMLImport::LO_42x },
+        { "LibreOffice/4.3.3.2$Linux_x86 LibreOffice_project/9bb7eadab57b6755b1265afa86e04bf45fbfc644", ";4332", SvXMLImport::LO_43x },
+        { "LibreOffice_from_Collabora_4.4-10/4.4.10.9$Linux_x86 LibreOffice_project/5600b19b88a01bbb669b0900100760758dff8c26", ";44109", SvXMLImport::LO_44x },
+        { "LibreOffice/4.3.3.2$Linux_X86_64 LibreOffice_project/430m0$Build-2", "430$2;4332", SvXMLImport::LO_43x },
+        { "LibreOffice/4.4.3.2$Linux_x86 LibreOffice_project/88805f81e9fe61362df02b9941de8e38a9b5fd16", ";4432", SvXMLImport::LO_44x },
+        { "LibreOffice/5.0.1.1$Linux_x86 LibreOffice_project/00m0$Build-1", "00$1;5011", SvXMLImport::LO_5x },
+        { "LibreOffice/5.0.3.2$Windows_X86_64 LibreOffice_project/e5f16313668ac592c1bfb310f4390624e3dbfb75", ";5032", SvXMLImport::LO_5x },
+        { "Collabora_Office/5.0.10.19$Linux_X86_64 LibreOffice_project/95060d44300d8866fa81c16fc8fe2afe22d63777", ";501019", SvXMLImport::LO_5x },
+        { "LibreOffice/5.1.6.2.0$Linux_X86_64 LibreOffice_project/10$Build-2", ";51620", SvXMLImport::LO_5x },
+        { "Collabora_Office/5.1.10.17$Linux_X86_64 LibreOffice_project/a104cbe76eefca3cf23973da68893d2225fd718b", ";511017", SvXMLImport::LO_5x },
+        { "LibreOffice/5.2.1.2$Windows_X86_64 LibreOffice_project/31dd62db80d4e60af04904455ec9c9219178d620", ";5212", SvXMLImport::LO_5x },
+        { "LibreOffice_Vanilla/5.2.3.5$MacOSX_X86_64 LibreOffice_project/83adc9c35c74e0badc710d981405858b1179a327", ";5235", SvXMLImport::LO_5x },
+        { "LibreOffice/5.3.4.2$Windows_X86_64 LibreOffice_project/f82d347ccc0be322489bf7da61d7e4ad13fe2ff3", ";5342", SvXMLImport::LO_5x },
+        { "Collabora_Office/5.3.10.27$Linux_X86_64 LibreOffice_project/7a5a5378661e338a44666c08773cc796b8d1c84a", ";531027", SvXMLImport::LO_5x },
+        { "LibreOfficeDev/5.4.7.0.0$Linux_X86_64 LibreOffice_project/ba7461fc88c08e75e315f786020a2946e56166c9", ";54700", SvXMLImport::LO_5x },
+        { "LibreOfficeDev/6.0.3.0.0$Linux_X86_64 LibreOffice_project/34442b85bfb0c451738b4db023345a7484463321", ";60300", SvXMLImport::LO_6x },
+    };
+
+    for (size_t i = 0; i < SAL_N_ELEMENTS(tests); ++i)
+    {
+        // the DocumentInfo instance is cached so need fresh SvXMLImport
+        rtl::Reference<SvXMLImport> const pImport(new SvXMLImport(
+            comphelper::getProcessComponentContext(), "testdummy",
+            SvXMLImportFlags::ALL));
+
+        pImport->initialize(uno::Sequence<uno::Any>{ uno::Any(xInfoSet) });
+
+        SvXMLMetaDocumentContext::setBuildId(
+                OUString::createFromAscii(tests[i].generator), xInfoSet);
+        if (std::strlen(tests[i].buildId) != 0)
+        {
+            CPPUNIT_ASSERT_EQUAL(OUString::createFromAscii(tests[i].buildId),
+                    xInfoSet->getPropertyValue("BuildId").get<OUString>());
+        }
+        else
+        {
+            CPPUNIT_ASSERT(!xInfoSet->getPropertyValue("BuildId").hasValue());
+        }
+        CPPUNIT_ASSERT_EQUAL(tests[i].result, pImport->getGeneratorVersion());
+    }
+}
+
 CPPUNIT_TEST_SUITE_REGISTRATION(Test);
 
 CPPUNIT_PLUGIN_IMPLEMENT();
diff --git a/xmloff/source/core/xmlimp.cxx b/xmloff/source/core/xmlimp.cxx
index 3c1e8bb0c6ea..b92d5edc64c1 100644
--- a/xmloff/source/core/xmlimp.cxx
+++ b/xmloff/source/core/xmlimp.cxx
@@ -229,7 +229,7 @@ public:
             {
                 mnGeneratorVersion = SvXMLImport::OOo_34x;
             }
-            else if (nUPD == 400)
+            else if (nUPD == 400 || nUPD == 401)
             {
                 mnGeneratorVersion = SvXMLImport::AOO_40x;
             }
diff --git a/xmloff/source/meta/xmlmetai.cxx b/xmloff/source/meta/xmlmetai.cxx
index ba733bd73a20..1f9a1b9604e2 100644
--- a/xmloff/source/meta/xmlmetai.cxx
+++ b/xmloff/source/meta/xmlmetai.cxx
@@ -256,6 +256,8 @@ void SvXMLMetaDocumentContext::setBuildId(OUString const& i_rBuildId, const uno:
     {
         if (    i_rBuildId.startsWith("StarOffice 7")
             ||  i_rBuildId.startsWith("StarSuite 7")
+            ||  i_rBuildId.startsWith("StarOffice 6")
+            ||  i_rBuildId.startsWith("StarSuite 6")
             ||  i_rBuildId.startsWith("OpenOffice.org 1"))
         {
             sBuildId = "645$8687";
@@ -266,19 +268,20 @@ void SvXMLMetaDocumentContext::setBuildId(OUString const& i_rBuildId, const uno:
         }
     }
 
-    OUString rest;
-    if (i_rBuildId.startsWith("LibreOffice/", &rest) ||
-        i_rBuildId.startsWith("LibreOfficeDev/", &rest) ||
-        i_rBuildId.startsWith("LOdev/", &rest))
+    // "LibreOffice_project" was hard-coded since LO 3.3.0
+    // see utl::DocInfoHelper::GetGeneratorString()
+    if (i_rBuildId.indexOf("LibreOffice_project/") != -1)
     {
         OUStringBuffer sNumber;
-        for (sal_Int32 i = 0; i < rest.getLength(); ++i)
+        auto const firstSlash = i_rBuildId.indexOf("/");
+        assert(firstSlash != -1);
+        for (sal_Int32 i = firstSlash + 1; i < i_rBuildId.getLength(); ++i)
         {
-            if (rtl::isAsciiDigit(rest[i]))
+            if (rtl::isAsciiDigit(i_rBuildId[i]))
             {
-                sNumber.append(rest[i]);
+                sNumber.append(i_rBuildId[i]);
             }
-            else if ('.' != rest[i])
+            else if ('.' != i_rBuildId[i])
             {
                 break;
             }