sw/source/core/tox/ToxTextGenerator.cxx

   1 /* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
   2 /*
   3  * This file is part of the LibreOffice project.
   4  *
   5  * This Source Code Form is subject to the terms of the Mozilla Public
   6  * License, v. 2.0. If a copy of the MPL was not distributed with this
   7  * file, You can obtain one at http://mozilla.org/MPL/2.0/.
   8  *
   9  * This file incorporates work covered by the following license notice:
  10  *
  11  *   Licensed to the Apache Software Foundation (ASF) under one or more
  12  *   contributor license agreements. See the NOTICE file distributed
  13  *   with this work for additional information regarding copyright
  14  *   ownership. The ASF licenses this file to you under the Apache
  15  *   License, Version 2.0 (the "License"); you may not use this file
  16  *   except in compliance with the License. You may obtain a copy of
  17  *   the License at http://www.apache.org/licenses/LICENSE-2.0 .
  18  */
  19
  20 #include <ToxTextGenerator.hxx>
  21
  22 #include <chpfld.hxx>
  23 #include <cntfrm.hxx>
  24 #include <txtfrm.hxx>
  25 #include <rootfrm.hxx>
  26 #include <ndindex.hxx>
  27 #include <fchrfmt.hxx>
  28 #include <doc.hxx>
  29 #include <IDocumentLayoutAccess.hxx>
  30 #include <IDocumentStylePoolAccess.hxx>
  31 #include <ndtxt.hxx>
  32 #include <tox.hxx>
  33 #include <txmsrt.hxx>
  34 #include <fmtautofmt.hxx>
  35 #include <swatrset.hxx>
  36 #include <ToxWhitespaceStripper.hxx>
  37 #include <ToxLinkProcessor.hxx>
  38 #include <ToxTabStopTokenHandler.hxx>
  39 #include <txatbase.hxx>
  40 #include <modeltoviewhelper.hxx>
  41 #include <strings.hrc>
  42
  43 #include <osl/diagnose.h>
  44 #include <rtl/ustrbuf.hxx>
  45 #include <svl/itemiter.hxx>
  46
  47 #include <cassert>
  48 #include <memory>
  49 #include <utility>
  50
  51 namespace {
  52
  53 bool sortTabHasNoToxSourcesOrFirstToxSourceHasNoNode(const SwTOXSortTabBase& sortTab)
  54 {
  55     if (sortTab.aTOXSources.empty()) {
  56         return true;
  57     }
  58     if (sortTab.aTOXSources.at(0).pNd == nullptr) {
  59         return true;
  60     }
  61     return false;
  62 }
  63
  64 // Similar to rtl::isAsciiWhiteSpace, but applicable to ToC entry number
  65 bool isWhiteSpace(sal_Unicode ch) { return ch == ' ' || ch == '\t'; }
  66
  67 } // end anonymous namespace
  68
  69 namespace sw {
  70
  71 OUString
  72 ToxTextGenerator::GetNumStringOfFirstNode(const SwTOXSortTabBase& rBase,
  73         bool bUsePrefix, sal_uInt8 nLevel,
  74         SwRootFrame const*const pLayout, bool bAddSpace)
  75 {
  76     if (sortTabHasNoToxSourcesOrFirstToxSourceHasNoNode(rBase)) {
  77         return OUString();
  78     }
  79
  80     OUString sRet;
  81     if (rBase.pTextMark) { // only if it's not a Mark
  82         return sRet;
  83     }
  84
  85     const SwTextNode* pNd = rBase.aTOXSources[0].pNd->GetTextNode();
  86     if (!pNd) {
  87         return sRet;
  88     }
  89     if (pLayout && pLayout->HasMergedParas())
  90     {   // note: pNd could be any node, since it could be Sequence etc.
  91         pNd = sw::GetParaPropsNode(*pLayout, *pNd);
  92     }
  93
  94     const SwNumRule* pRule = pNd->GetNumRule();
  95     if (!pRule) {
  96         return sRet;
  97     }
  98
  99     if (pNd->GetActualListLevel() < MAXLEVEL) {
 100         sRet = pNd->GetNumString(bUsePrefix, nLevel, pLayout);
 101     }
 102
 103     if (bAddSpace && !sRet.isEmpty() && !isWhiteSpace(sRet[sRet.getLength() - 1])) {
 104         sRet += " ";// Makes sure spacing is done only when there is outline numbering
 105     }
 106
 107     return sRet;
 108 }
 109
 110
 111 ToxTextGenerator::ToxTextGenerator(const SwForm& toxForm,
 112         std::shared_ptr<ToxTabStopTokenHandler> tabStopHandler)
 113 : mToxForm(toxForm),
 114   mLinkProcessor(std::make_shared<ToxLinkProcessor>()),
 115   mTabStopTokenHandler(std::move(tabStopHandler))
 116 {}
 117
 118 ToxTextGenerator::~ToxTextGenerator()
 119 {}
 120
 121 OUString
 122 ToxTextGenerator::HandleChapterToken(const SwTOXSortTabBase& rBase,
 123         const SwFormToken& aToken, SwRootFrame const*const pLayout) const
 124 {
 125     if (sortTabHasNoToxSourcesOrFirstToxSourceHasNoNode(rBase)) {
 126         return OUString();
 127     }
 128
 129     // A bit tricky: Find a random Frame
 130     const SwContentNode* contentNode = rBase.aTOXSources.at(0).pNd->GetContentNode();
 131     if (!contentNode) {
 132         return OUString();
 133     }
 134
 135     // #i53420#
 136     const SwContentFrame* contentFrame = contentNode->getLayoutFrame(pLayout);
 137     if (!contentFrame) {
 138         return OUString();
 139     }
 140
 141     return GenerateTextForChapterToken(aToken, contentFrame, contentNode, pLayout);
 142 }
 143
 144 OUString
 145 ToxTextGenerator::GenerateTextForChapterToken(const SwFormToken& chapterToken, const SwContentFrame* contentFrame,
 146         const SwContentNode *contentNode,
 147         SwRootFrame const*const pLayout) const
 148 {
 149     OUString retval;
 150
 151     SwChapterFieldType chapterFieldType;
 152     SwChapterField aField = ObtainChapterField(&chapterFieldType, &chapterToken, contentFrame, contentNode);
 153
 154     //---> #i89791#
 155     // continue to support CF_NUMBER and CF_NUM_TITLE in order to handle ODF 1.0/1.1 written by OOo 3.x
 156     // in the same way as OOo 2.x would handle them.
 157     if (CF_NUM_NOPREPST_TITLE == chapterToken.nChapterFormat || CF_NUMBER == chapterToken.nChapterFormat) {
 158         retval += aField.GetNumber(pLayout); // get the string number without pre/postfix
 159     }
 160     else if (CF_NUMBER_NOPREPST == chapterToken.nChapterFormat || CF_NUM_TITLE == chapterToken.nChapterFormat) {
 161         retval += aField.GetNumber(pLayout) + " " + aField.GetTitle(pLayout);
 162     } else if (CF_TITLE == chapterToken.nChapterFormat) {
 163         retval += aField.GetTitle(pLayout);
 164     }
 165     return retval;
 166 }
 167
 168 std::optional<std::pair<SwTextNode *, SvxTabStopItem>>
 169 ToxTextGenerator::GenerateText(SwDoc* pDoc,
 170         std::unordered_map<OUString, int> & rMarkURLs,
 171         const std::vector<std::unique_ptr<SwTOXSortTabBase>> &entries,
 172         sal_uInt16 indexOfEntryToProcess, sal_uInt16 numberOfEntriesToProcess,
 173         SwRootFrame const*const pLayout)
 174 {
 175     std::optional<std::pair<SwTextNode *, SvxTabStopItem>> oRet;
 176     // pTOXNd is only set at the first mark
 177     SwTextNode* pTOXNd = const_cast<SwTextNode*>(entries.at(indexOfEntryToProcess)->pTOXNd);
 178     // FIXME this operates directly on the node text
 179     OUString & rText = const_cast<OUString&>(pTOXNd->GetText());
 180     rText.clear();
 181     for(sal_uInt16 nIndex = indexOfEntryToProcess; nIndex < indexOfEntryToProcess + numberOfEntriesToProcess; nIndex++)
 182     {
 183         if(nIndex > indexOfEntryToProcess)
 184             rText += ", "; // comma separation
 185         // Initialize String with the Pattern from the form
 186         const SwTOXSortTabBase& rBase = *entries.at(nIndex);
 187         sal_uInt16 nLvl = rBase.GetLevel();
 188         OSL_ENSURE( nLvl < mToxForm.GetFormMax(), "invalid FORM_LEVEL");
 189
 190         oRet.emplace(pTOXNd, SvxTabStopItem(0, 0, SvxTabAdjust::Default, RES_PARATR_TABSTOP));
 191         // create an enumerator
 192         // #i21237#
 193         SwFormTokens aPattern = mToxForm.GetPattern(nLvl);
 194         // remove text from node
 195         for (size_t i = 0; i < aPattern.size(); ++i) // #i21237#
 196         {
 197             const auto& aToken = aPattern[i];
 198             sal_Int32 nStartCharStyle = rText.getLength();
 199             OUString aCharStyleName = aToken.sCharStyleName;
 200             switch( aToken.eTokenType )
 201             {
 202             case TOKEN_ENTRY_NO:
 203                 // for TOC numbering
 204                 // Only add space when there is outline numbering, and also when the next token
 205                 // is the entry text: it can also be e.g. a tab, or the entry number can be used
 206                 // in page number area like "2-15" for chapter 2, page 15.
 207                 rText += GetNumStringOfFirstNode(rBase,
 208                     aToken.nChapterFormat == CF_NUMBER,
 209                     static_cast<sal_uInt8>(aToken.nOutlineLevel - 1), pLayout,
 210                     i < aPattern.size() - 1 && aPattern[i + 1].eTokenType == TOKEN_ENTRY_TEXT);
 211                 break;
 212
 213             case TOKEN_ENTRY_TEXT: {
 214                 HandledTextToken htt = HandleTextToken(rBase, pDoc->GetAttrPool(), pLayout);
 215                 ApplyHandledTextToken(htt, *pTOXNd);
 216             }
 217                 break;
 218
 219             case TOKEN_ENTRY:
 220                 {
 221                     // for TOC numbering
 222                     rText += GetNumStringOfFirstNode(rBase, true, MAXLEVEL, pLayout);
 223                     HandledTextToken htt = HandleTextToken(rBase, pDoc->GetAttrPool(), pLayout);
 224                     ApplyHandledTextToken(htt, *pTOXNd);
 225                 }
 226                 break;
 227
 228             case TOKEN_TAB_STOP: {
 229                 ToxTabStopTokenHandler::HandledTabStopToken htst =
 230                     mTabStopTokenHandler->HandleTabStopToken(aToken, *pTOXNd);
 231                 rText += htst.text;
 232                 oRet->second.Insert(htst.tabStop);
 233                 break;
 234             }
 235
 236             case TOKEN_TEXT:
 237                 rText += aToken.sText;
 238                 break;
 239
 240             case TOKEN_PAGE_NUMS:
 241                 rText += ConstructPageNumberPlaceholder(rBase.aTOXSources.size());
 242                 break;
 243
 244             case TOKEN_CHAPTER_INFO:
 245                 rText += HandleChapterToken(rBase, aToken, pLayout);
 246                 break;
 247
 248             case TOKEN_LINK_START:
 249                 mLinkProcessor->StartNewLink(rText.getLength(), aToken.sCharStyleName);
 250                 break;
 251
 252             case TOKEN_LINK_END:
 253                 {
 254                     auto [url, isMark] = rBase.GetURL(pLayout);
 255                     if (isMark)
 256                     {
 257                         auto [iter, _] = rMarkURLs.emplace(url, 0);
 258                         (void) _; // sigh... ignore it more explicitly
 259                         ++iter->second;
 260                         url = "#" + OUString::number(iter->second) + url;
 261                     }
 262                     mLinkProcessor->CloseLink(rText.getLength(), url, /*bRelative=*/true);
 263                 }
 264                 break;
 265
 266             case TOKEN_AUTHORITY:
 267                 {
 268                     ToxAuthorityField eField = static_cast<ToxAuthorityField>(aToken.nAuthorityField);
 269                     SwContentIndex aIdx( pTOXNd, rText.getLength() );
 270                     if (eField == ToxAuthorityField::AUTH_FIELD_URL)
 271                     {
 272                         aCharStyleName = SwResId(STR_POOLCHR_INET_NORMAL);
 273                         mLinkProcessor->StartNewLink(rText.getLength(), aCharStyleName);
 274                     }
 275                     rBase.FillText( *pTOXNd, aIdx, o3tl::narrowing<sal_uInt16>(eField), pLayout );
 276                     if (eField == ToxAuthorityField::AUTH_FIELD_URL)
 277                     {
 278                         // Get the absolute URL, the text may be a relative one.
 279                         const auto& rAuthority = static_cast<const SwTOXAuthority&>(rBase);
 280                         OUString aURL = SwTOXAuthority::GetSourceURL(
 281                             rAuthority.GetText(AUTH_FIELD_URL, pLayout));
 282
 283                         mLinkProcessor->CloseLink(rText.getLength(), aURL, /*bRelative=*/false);
 284                     }
 285                 }
 286                 break;
 287             case TOKEN_END: break;
 288             }
 289
 290             if (!aCharStyleName.isEmpty())
 291             {
 292                 SwCharFormat* pCharFormat;
 293                 if( USHRT_MAX != aToken.nPoolId )
 294                     pCharFormat = pDoc->getIDocumentStylePoolAccess().GetCharFormatFromPool( aToken.nPoolId );
 295                 else
 296                     pCharFormat = pDoc->FindCharFormatByName(aCharStyleName);
 297
 298                 if (pCharFormat)
 299                 {
 300                     SwFormatCharFormat aFormat( pCharFormat );
 301                     pTOXNd->InsertItem( aFormat, nStartCharStyle,
 302                         rText.getLength(), SetAttrMode::DONTEXPAND );
 303                 }
 304             }
 305         }
 306     }
 307     mLinkProcessor->InsertLinkAttributes(*pTOXNd);
 308     return oRet;
 309 }
 310
 311 /*static*/ std::shared_ptr<SfxItemSet>
 312 ToxTextGenerator::CollectAttributesForTox(const SwTextAttr& hint, SwAttrPool& pool)
 313 {
 314     auto retval = std::make_shared<SfxItemSet>(pool);
 315     if (hint.Which() != RES_TXTATR_AUTOFMT) {
 316         return retval;
 317     }
 318     const SwFormatAutoFormat& afmt = hint.GetAutoFormat();
 319     SfxItemIter aIter( *afmt.GetStyleHandle());
 320     const SfxPoolItem* pItem = aIter.GetCurItem();
 321     do
 322     {
 323         if (pItem->Which() == RES_CHRATR_ESCAPEMENT ||
 324             pItem->Which() == RES_CHRATR_POSTURE ||
 325             pItem->Which() == RES_CHRATR_CJK_POSTURE ||
 326             pItem->Which() == RES_CHRATR_CTL_POSTURE)
 327         {
 328             retval->Put(std::unique_ptr<SfxPoolItem>(pItem->Clone()));
 329         }
 330         pItem = aIter.NextItem();
 331     } while (pItem);
 332     return retval;
 333 }
 334
 335 void ToxTextGenerator::GetAttributesForNode(
 336     ToxTextGenerator::HandledTextToken & rResult,
 337     sal_Int32 & rOffset,
 338     SwTextNode const& rNode,
 339     ToxWhitespaceStripper const& rStripper,
 340     SwAttrPool & rPool,
 341     SwRootFrame const*const pLayout)
 342 {
 343     // note: this *must* use the same flags as SwTextNode::GetExpandText()
 344     // or indexes will be off!
 345     ExpandMode eMode = ExpandMode::ExpandFields | ExpandMode::HideFieldmarkCommands;
 346     if (pLayout && pLayout->IsHideRedlines())
 347     {
 348         eMode |= ExpandMode::HideDeletions;
 349     }
 350     ModelToViewHelper aConversionMap(rNode, pLayout, eMode);
 351     if (SwpHints const*const pHints = rNode.GetpSwpHints())
 352     {
 353         for (size_t i = 0; i < pHints->Count(); ++i)
 354         {
 355             const SwTextAttr* pHint = pHints->Get(i);
 356             std::shared_ptr<SfxItemSet> attributesToClone =
 357                 CollectAttributesForTox(*pHint, rPool);
 358             if (attributesToClone->Count() <= 0) {
 359                 continue;
 360             }
 361
 362             // sw_redlinehide: due to the ... interesting ... multi-level index
 363             // mapping going on here, can't use the usual merged attr iterators :(
 364
 365             sal_Int32 const nStart(aConversionMap.ConvertToViewPosition(pHint->GetStart()));
 366             sal_Int32 const nEnd(aConversionMap.ConvertToViewPosition(pHint->GetAnyEnd()));
 367             if (nStart != nEnd) // might be in delete redline, and useless anyway
 368             {
 369                 std::unique_ptr<SwFormatAutoFormat> pClone(pHint->GetAutoFormat().Clone());
 370                 pClone->SetStyleHandle(attributesToClone);
 371                 rResult.autoFormats.push_back(std::move(pClone));
 372                 // note the rStripper is on the whole merged text, so need rOffset
 373                 rResult.startPositions.push_back(
 374                     rStripper.GetPositionInStrippedString(rOffset + nStart));
 375                 rResult.endPositions.push_back(
 376                     rStripper.GetPositionInStrippedString(rOffset + nEnd));
 377             }
 378         }
 379     }
 380     rOffset += aConversionMap.getViewText().getLength();
 381 }
 382
 383 ToxTextGenerator::HandledTextToken
 384 ToxTextGenerator::HandleTextToken(const SwTOXSortTabBase& source,
 385         SwAttrPool& pool, SwRootFrame const*const pLayout)
 386 {
 387     HandledTextToken result;
 388     ToxWhitespaceStripper stripper(source.GetText().sText);
 389     result.text = stripper.GetStrippedString();
 390
 391     // FIXME: there is a pre-existing problem that the index mapping of the
 392     // attributes only works if the paragraph is fully selected
 393     if (!source.IsFullPara() || source.aTOXSources.empty())
 394         return result;
 395
 396     const SwTextNode* pSrc = source.aTOXSources.front().pNd->GetTextNode();
 397     if (!pSrc)
 398     {
 399         return result;
 400     }
 401
 402     sal_Int32 nOffset(0);
 403     GetAttributesForNode(result, nOffset, *pSrc, stripper, pool, pLayout);
 404     if (pLayout && pLayout->HasMergedParas())
 405     {
 406         if (SwTextFrame const*const pFrame = static_cast<SwTextFrame*>(pSrc->getLayoutFrame(pLayout)))
 407         {
 408             if (sw::MergedPara const*const pMerged = pFrame->GetMergedPara())
 409             {
 410                 // pSrc already copied above
 411                 assert(pSrc == pMerged->pParaPropsNode);
 412                 for (SwNodeOffset i = pSrc->GetIndex() + 1;
 413                      i <= pMerged->pLastNode->GetIndex(); ++i)
 414                 {
 415                     SwNode *const pTmp(pSrc->GetNodes()[i]);
 416                     if (pTmp->GetRedlineMergeFlag() == SwNode::Merge::NonFirst)
 417                     {
 418                         GetAttributesForNode(result, nOffset,
 419                                 *pTmp->GetTextNode(), stripper, pool, pLayout);
 420                     }
 421                 }
 422             }
 423         }
 424     }
 425
 426     return result;
 427 }
 428
 429 /*static*/ void
 430 ToxTextGenerator::ApplyHandledTextToken(const HandledTextToken& htt, SwTextNode& targetNode)
 431 {
 432     sal_Int32 offset = targetNode.GetText().getLength();
 433     SwContentIndex aIdx(&targetNode, offset);
 434     targetNode.InsertText(htt.text, aIdx);
 435     for (size_t i=0; i < htt.autoFormats.size(); ++i) {
 436         targetNode.InsertItem(*htt.autoFormats.at(i),
 437                 htt.startPositions.at(i) + offset,
 438                 htt.endPositions.at(i) + offset);
 439     }
 440 }
 441
 442 /*static*/ OUString
 443 ToxTextGenerator::ConstructPageNumberPlaceholder(size_t numberOfToxSources)
 444 {
 445     if (numberOfToxSources == 0) {
 446         return OUString();
 447     }
 448     OUStringBuffer retval;
 449     // Place holder for the PageNumber; we only respect the first one
 450     retval.append(C_NUM_REPL);
 451     for (size_t i = 1; i < numberOfToxSources; ++i) {
 452         retval.append(SwTOXMark::S_PAGE_DELI + OUStringChar(C_NUM_REPL));
 453     }
 454     retval.append(C_END_PAGE_NUM);
 455     return retval.makeStringAndClear();
 456 }
 457
 458 /*virtual*/ SwChapterField
 459 ToxTextGenerator::ObtainChapterField(SwChapterFieldType* chapterFieldType,
 460         const SwFormToken* chapterToken, const SwContentFrame* contentFrame,
 461         const SwContentNode* contentNode) const
 462 {
 463     assert(chapterToken);
 464     assert(chapterToken->nOutlineLevel >= 1);
 465
 466     SwChapterField retval(chapterFieldType, chapterToken->nChapterFormat);
 467     retval.SetLevel(static_cast<sal_uInt8>(chapterToken->nOutlineLevel - 1));
 468     // #i53420#
 469     retval.ChangeExpansion(*contentFrame, contentNode, true);
 470     return retval;
 471 }
 472 } // end namespace sw
 473
 474 /* vim:set shiftwidth=4 softtabstop=4 expandtab: */