tdf#136306 offapi linguistic: add options to disable rule-based compounding

Add two new spell checking options to disable rule-based closed
and hyphenated compound word recognition with Hunspell dictionaries:

com::sun::star::linguistic2::XLinguProperties::IsSpellClosedCompound
com::sun::star::linguistic2::XLinguProperties::IsSpellHyphenatedCompound

For professional proofreaders, it can be more important to avoid
of the mistakes of the rule-based compound word recognition, than
to speed up proofreading. Disabling the following two new options
will report all rule-based closed compound words (default in
Dutch, German, Hungarian etc. dictionaries) and rule-based
hyphenated compound words (all languages with BREAK usage in
their Hunspell dictionaries):

- "Accept possible closed compound words"

- "Accept possible hyphenated compound words"

For example, disabling the second one, dictionary word "scot-free"
will be still correct word in English spell checking, but not
the previously accepted compound "arbitrary-word-with-hyphen".

Note: the second option works with the update to Hunspell 1.7.2.

Change-Id: Id879610927d5e8269fda5ad207c1c2fe1f57a0b6
Reviewed-on: https://gerrit.libreoffice.org/c/core/+/144875
Tested-by: Jenkins
Reviewed-by: László Németh <nemeth@numbertext.org>

2 files changed, 20 insertions, 7 deletions

diff --git a/lingucomponent/source/spellcheck/spell/sspellimp.cxx b/lingucomponent/source/spellcheck/spell/sspellimp.cxx
index 95b264157533..506d2c9dd014 100644
--- a/lingucomponent/source/spellcheck/spell/sspellimp.cxx
+++ b/lingucomponent/source/spellcheck/spell/sspellimp.cxx
@@ -243,7 +243,7 @@ sal_Bool SAL_CALL SpellChecker::hasLocale(const Locale& rLocale)
     return bRes;
 }
 
-sal_Int16 SpellChecker::GetSpellFailure(const OUString &rWord, const Locale &rLocale)
+sal_Int16 SpellChecker::GetSpellFailure(const OUString &rWord, const Locale &rLocale, int& rInfo)
 {
     if (rWord.getLength() > MAXWORDLEN)
         return -1;
@@ -332,9 +332,9 @@ sal_Int16 SpellChecker::GetSpellFailure(const OUString &rWord, const Locale &rLo
 
                 OString aWrd(OU2ENC(nWord,eEnc));
 #if defined(H_DEPRECATED)
-                bool bVal = pMS->spell(std::string(aWrd.getStr()));
+                bool bVal = pMS->spell(std::string(aWrd.getStr()), &rInfo);
 #else
-                bool bVal = pMS->spell(aWrd.getStr()) != 0;
+                bool bVal = pMS->spell(aWrd.getStr(), &rInfo) != 0;
 #endif
                 if (!bVal) {
                     if (extrachar && (eEnc != RTL_TEXTENCODING_UTF8)) {
@@ -355,9 +355,9 @@ sal_Int16 SpellChecker::GetSpellFailure(const OUString &rWord, const Locale &rLo
                         OUString aWord(aBuf.makeStringAndClear());
                         OString bWrd(OU2ENC(aWord, eEnc));
 #if defined(H_DEPRECATED)
-                        bVal = pMS->spell(std::string(bWrd.getStr()));
+                        bVal = pMS->spell(std::string(bWrd.getStr()), &rInfo);
 #else
-                        bVal = pMS->spell(bWrd.getStr()) != 0;
+                        bVal = pMS->spell(bWrd.getStr(), &rInfo) != 0;
 #endif
                         if (bVal) return -1;
                     }
@@ -396,7 +396,8 @@ sal_Bool SAL_CALL SpellChecker::isValid( const OUString& rWord, const Locale& rL
     PropertyHelper_Spelling& rHelper = GetPropHelper();
     rHelper.SetTmpPropVals( rProperties );
 
-    sal_Int16 nFailure = GetSpellFailure( rWord, rLocale );
+    int nInfo = 0;
+    sal_Int16 nFailure = GetSpellFailure( rWord, rLocale, nInfo );
     if (nFailure != -1 && !rWord.match(SPELL_XML, 0))
     {
         LanguageType nLang = LinguLocaleToLanguage( rLocale );
@@ -408,6 +409,18 @@ sal_Bool SAL_CALL SpellChecker::isValid( const OUString& rWord, const Locale& rL
         if (bIgnoreError)
             nFailure = -1;
     }
+//#define SPELL_COMPOUND 1 << 0
+
+    // valid word, but it's a rule-based compound word
+    if ( nFailure == -1 && (nInfo & SPELL_COMPOUND) )
+    {
+        bool bHasHyphen = rWord.indexOf('-') > -1;
+        if ( (bHasHyphen && !rHelper.IsSpellHyphenatedCompound()) ||
+             (!bHasHyphen && !rHelper.IsSpellClosedCompound()) )
+        {
+            return false;
+        }
+    }
 
     return (nFailure == -1);
 }
diff --git a/lingucomponent/source/spellcheck/spell/sspellimp.hxx b/lingucomponent/source/spellcheck/spell/sspellimp.hxx
index 000f1756f5bc..68ddc69b3c56 100644
--- a/lingucomponent/source/spellcheck/spell/sspellimp.hxx
+++ b/lingucomponent/source/spellcheck/spell/sspellimp.hxx
@@ -79,7 +79,7 @@ class SpellChecker :
         return m_pPropHelper ? *m_pPropHelper : GetPropHelper_Impl();
     }
 
-    sal_Int16   GetSpellFailure( const OUString &rWord, const Locale &rLocale );
+    sal_Int16   GetSpellFailure( const OUString &rWord, const Locale &rLocale, int& rInfo );
     Reference< XSpellAlternatives > GetProposals( const OUString &rWord, const Locale &rLocale );
 
 public: