From d08940016b0511b04e7f090cd2e856c3a85df774 Mon Sep 17 00:00:00 2001 From: Peter Edberg Date: Thu, 25 Sep 2008 05:48:27 +0000 Subject: [PATCH] ICU-5858 Add constructors for RuleBasedBreakIterator and RBBIDataWrapper that take RBBIDataHeader* but do not adopt it. X-SVN-Rev: 24641 --- icu4c/source/common/rbbi.cpp | 14 ++++++ icu4c/source/common/rbbidata.cpp | 10 ++++- icu4c/source/common/rbbidata.h | 7 ++- icu4c/source/common/unicode/rbbi.h | 24 +++++++++- icu4c/source/test/intltest/rbbiapts.cpp | 60 +++++++++++++++++++++++++ icu4c/source/test/intltest/rbbiapts.h | 20 ++++++++- 6 files changed, 130 insertions(+), 5 deletions(-) diff --git a/icu4c/source/common/rbbi.cpp b/icu4c/source/common/rbbi.cpp index 8d5fdfc0f7b..b06a01c0936 100644 --- a/icu4c/source/common/rbbi.cpp +++ b/icu4c/source/common/rbbi.cpp @@ -70,6 +70,20 @@ RuleBasedBreakIterator::RuleBasedBreakIterator(RBBIDataHeader* data, UErrorCode } } +/** + * Same as above but does not adopt memory + */ +RuleBasedBreakIterator::RuleBasedBreakIterator(const RBBIDataHeader* data, enum EDontAdopt, UErrorCode &status) +{ + init(); + fData = new RBBIDataWrapper(data, RBBIDataWrapper::kDontAdopt, status); // status checked in constructor + if (U_FAILURE(status)) {return;} + if(fData == 0) { + status = U_MEMORY_ALLOCATION_ERROR; + return; + } +} + //------------------------------------------------------------------------------- // // Constructor from a UDataMemory handle to precompiled break rules diff --git a/icu4c/source/common/rbbidata.cpp b/icu4c/source/common/rbbidata.cpp index 7897b44e7b9..442fb3d4b1c 100644 --- a/icu4c/source/common/rbbidata.cpp +++ b/icu4c/source/common/rbbidata.cpp @@ -1,6 +1,6 @@ /* *************************************************************************** -* Copyright (C) 1999-2007 International Business Machines Corporation * +* Copyright (C) 1999-2008 International Business Machines Corporation * * and others. All rights reserved. * *************************************************************************** */ @@ -49,6 +49,11 @@ RBBIDataWrapper::RBBIDataWrapper(const RBBIDataHeader *data, UErrorCode &status) init(data, status); } +RBBIDataWrapper::RBBIDataWrapper(const RBBIDataHeader *data, enum EDontAdopt, UErrorCode &status) { + init(data, status); + fDontFreeData = TRUE; +} + RBBIDataWrapper::RBBIDataWrapper(UDataMemory* udm, UErrorCode &status) { const RBBIDataHeader *d = (const RBBIDataHeader *) // ((char *)&(udm->pHeader->info) + udm->pHeader->info.size); @@ -77,6 +82,7 @@ void RBBIDataWrapper::init(const RBBIDataHeader *data, UErrorCode &status) { return; } + fDontFreeData = FALSE; fUDataMem = NULL; fReverseTable = NULL; fSafeFwdTable = NULL; @@ -130,7 +136,7 @@ RBBIDataWrapper::~RBBIDataWrapper() { U_ASSERT(fRefCount == 0); if (fUDataMem) { udata_close(fUDataMem); - } else { + } else if (!fDontFreeData) { uprv_free((void *)fHeader); } } diff --git a/icu4c/source/common/rbbidata.h b/icu4c/source/common/rbbidata.h index 7ba1db05d65..ee6aa486160 100644 --- a/icu4c/source/common/rbbidata.h +++ b/icu4c/source/common/rbbidata.h @@ -1,7 +1,7 @@ /* ******************************************************************************* * -* Copyright (C) 1999-2005, International Business Machines +* Copyright (C) 1999-2005,2008 International Business Machines * Corporation and others. All Rights Reserved. * ******************************************************************************* @@ -141,7 +141,11 @@ typedef enum { /* */ class RBBIDataWrapper : public UMemory { public: + enum EDontAdopt { + kDontAdopt + }; RBBIDataWrapper(const RBBIDataHeader *data, UErrorCode &status); + RBBIDataWrapper(const RBBIDataHeader *data, enum EDontAdopt dontAdopt, UErrorCode &status); RBBIDataWrapper(UDataMemory* udm, UErrorCode &status); ~RBBIDataWrapper(); @@ -179,6 +183,7 @@ private: int32_t fRefCount; UDataMemory *fUDataMem; UnicodeString fRuleString; + UBool fDontFreeData; RBBIDataWrapper(const RBBIDataWrapper &other); /* forbid copying of this class */ RBBIDataWrapper &operator=(const RBBIDataWrapper &other); /* forbid copying of this class */ diff --git a/icu4c/source/common/unicode/rbbi.h b/icu4c/source/common/unicode/rbbi.h index 747ea44ce1c..90ec6e6bd6a 100644 --- a/icu4c/source/common/unicode/rbbi.h +++ b/icu4c/source/common/unicode/rbbi.h @@ -1,6 +1,6 @@ /* *************************************************************************** -* Copyright (C) 1999-2007 International Business Machines Corporation * +* Copyright (C) 1999-2008 International Business Machines Corporation * * and others. All rights reserved. * *************************************************************************** @@ -170,6 +170,18 @@ protected: // constructors //======================================================================= + /** + * Constant to be used in the constructor + * RuleBasedBreakIterator(RBBIDataHeader*, EDontAdopt, UErrorCode &); + * which does not adopt the memory indicated by the RBBIDataHeader* + * parameter. + * + * @internal + */ + enum EDontAdopt { + kDontAdopt + }; + /** * Constructor from a flattened set of RBBI data in malloced memory. * RulesBasedBreakIterators built from a custom set of rules @@ -182,6 +194,16 @@ protected: */ RuleBasedBreakIterator(RBBIDataHeader* data, UErrorCode &status); + /** + * Constructor from a flattened set of RBBI data in memory which need not + * be malloced (e.g. it may be a memory-mapped file, etc.). + * + * This version does not adopt the memory, and does not + * free it when done. + * @internal + */ + RuleBasedBreakIterator(const RBBIDataHeader* data, enum EDontAdopt dontAdopt, UErrorCode &status); + friend class RBBIRuleBuilder; /** @internal */ diff --git a/icu4c/source/test/intltest/rbbiapts.cpp b/icu4c/source/test/intltest/rbbiapts.cpp index f9ffcf84657..db6266beee4 100644 --- a/icu4c/source/test/intltest/rbbiapts.cpp +++ b/icu4c/source/test/intltest/rbbiapts.cpp @@ -21,6 +21,7 @@ #include "ubrkimpl.h" #include "unicode/ustring.h" #include "unicode/utext.h" +#include "cmemory.h" /** * API Test the RuleBasedBreakIterator class @@ -1047,6 +1048,50 @@ void RBBIAPITest::TestRoundtripRules() { } } +// Try out the RuleBasedBreakIterator constructors that take RBBIDataHeader* +// (these are protected so we access them via a local class RBBIWithProtectedFunctions). +// This is just a sanity check, not a thorough test (e.g. we don't check that the +// first delete actually frees rulesCopy). +void RBBIAPITest::TestCreateFromRBBIData() { + // Get some handy RBBIData + const char *brkName = "word"; // or "sent", "line", "char", etc. + UErrorCode status = U_ZERO_ERROR; + UDataMemory * data = udata_open(U_ICUDATA_BRKITR, "brk", brkName, &status); + if ( U_SUCCESS(status) ) { + const RBBIDataHeader * builtRules = (const RBBIDataHeader *)udata_getMemory(data); + uint32_t length = builtRules->fLength; + RBBIWithProtectedFunctions * brkItr; + + // Try the memory-adopting constructor, need to copy the data first + RBBIDataHeader * rulesCopy = (RBBIDataHeader *) uprv_malloc(length); + if ( rulesCopy ) { + uprv_memcpy( rulesCopy, builtRules, length ); + + brkItr = new RBBIWithProtectedFunctions(rulesCopy, status); + if ( U_SUCCESS(status) ) { + delete brkItr; // this should free rulesCopy + } else { + errln("create RuleBasedBreakIterator from RBBIData (adopted): ICU Error \"%s\"\n", u_errorName(status) ); + status = U_ZERO_ERROR;// reset for the next test + uprv_free( rulesCopy ); + } + } + + // Now try the non-adopting constructor + brkItr = new RBBIWithProtectedFunctions(builtRules, RBBIWithProtectedFunctions::kDontAdopt, status); + if ( U_SUCCESS(status) ) { + delete brkItr; // this should NOT attempt to free builtRules + if (builtRules->fLength != length) { // sanity check + errln("create RuleBasedBreakIterator from RBBIData (non-adopted): delete affects data\n" ); + } + } else { + errln("create RuleBasedBreakIterator from RBBIData (non-adopted): ICU Error \"%s\"\n", u_errorName(status) ); + } + + udata_close(data); + } +} + //--------------------------------------------- // runIndexedTest //--------------------------------------------- @@ -1069,6 +1114,7 @@ void RBBIAPITest::runIndexedTest( int32_t index, UBool exec, const char* &name, case 10: name = "TestRegistration"; if (exec) TestRegistration(); break; case 11: name = "TestBoilerPlate"; if (exec) TestBoilerPlate(); break; case 12: name = "TestRoundtripRules"; if (exec) TestRoundtripRules(); break; + case 13: name = "TestCreateFromRBBIData"; if (exec) TestCreateFromRBBIData(); break; default: name = ""; break; // needed to end loop } @@ -1115,4 +1161,18 @@ void RBBIAPITest::doTest(UnicodeString& testString, int32_t start, int32_t gotof logln(prettify("****selected \"" + selected + "\"")); } +//--------------------------------------------- +//RBBIWithProtectedFunctions class functions +//--------------------------------------------- + +RBBIWithProtectedFunctions::RBBIWithProtectedFunctions(RBBIDataHeader* data, UErrorCode &status) + : RuleBasedBreakIterator(data, status) +{ +} + +RBBIWithProtectedFunctions::RBBIWithProtectedFunctions(const RBBIDataHeader* data, enum EDontAdopt, UErrorCode &status) + : RuleBasedBreakIterator(data, RuleBasedBreakIterator::kDontAdopt, status) +{ +} + #endif /* #if !UCONFIG_NO_BREAK_ITERATION */ diff --git a/icu4c/source/test/intltest/rbbiapts.h b/icu4c/source/test/intltest/rbbiapts.h index ae759248cea..0ce64ac3bad 100644 --- a/icu4c/source/test/intltest/rbbiapts.h +++ b/icu4c/source/test/intltest/rbbiapts.h @@ -1,6 +1,6 @@ /******************************************************************** * COPYRIGHT: - * Copyright (c) 1999-2004, International Business Machines Corporation and + * Copyright (c) 1999-2004,2008 International Business Machines Corporation and * others. All Rights Reserved. ********************************************************************/ /************************************************************************ @@ -63,6 +63,12 @@ public: void RoundtripRule(const char *dataFile); + /** + * Test creating RuleBasedBreakIterator from RBBIData. + **/ + + void TestCreateFromRBBIData(void); + /** * Tests grouping effect of 'single quotes' in rules. **/ @@ -92,6 +98,18 @@ public: }; +/** + * Special class to enable testing of protected functions in RuleBasedBreakIterator + */ +class RBBIWithProtectedFunctions: public RuleBasedBreakIterator { +public: + enum EDontAdopt { + kDontAdopt + }; + RBBIWithProtectedFunctions(RBBIDataHeader* data, UErrorCode &status); + RBBIWithProtectedFunctions(const RBBIDataHeader* data, enum EDontAdopt dontAdopt, UErrorCode &status); +}; + #endif /* #if !UCONFIG_NO_BREAK_ITERATION */ #endif