1
   2
   3
   4
   5
   6
   7
   8
   9
  10
  11
  12
  13
  14
  15
  16
  17
  18
  19
  20
  21
  22
  23
  24
  25
  26
  27
  28
  29
  30
  31
  32
  33
  34
  35
  36
  37
  38
  39
  40
  41
  42
  43
  44
  45
  46
  47
  48
  49
  50
  51
  52
  53
  54
  55
  56
  57
  58
  59
  60
  61
  62
  63
  64
  65
  66
  67
  68
  69
  70
  71
  72
  73
  74
  75
  76
  77
  78
  79
  80
  81
  82
  83
  84
  85
  86
  87
  88
  89
  90
  91
  92
  93
  94
  95
  96
  97
  98
  99
 100
 101
 102
 103
 104
 105
 106
 107
 108
 109
 110
 111
 112
 113
 114
 115
 116
 117
 118
 119
 120
 121
 122
 123
 124
 125
 126
 127
 128
 129
 130
 131
 132
 133
 134
 135
 136
 137
 138
 139
 140
 141
 142
 143
 144
 145
 146
 147
 148
 149
 150
 151
 152
 153
 154
 155
 156
 157
 158
 159
 160
 161
 162
 163
 164
 165
 166
 167
 168
 169
 170
 171
 172
 173
 174
 175
 176
 177
 178
 179
 180
 181
 182
 183
 184
 185
 186
 187
 188
 189
 190
 191
 192
 193
 194
 195
 196
 197
 198
 199
 200
 201
 202
 203
 204
 205
 206
 207
 208
 209
 210
 211
 212
 213
 214
 215
 216
 217
 218
 219
 220
 221
 222
 223
 224
 225
 226
 227
 228
 229
 230
 231
 232
 233
 234
 235
 236
 237
 238
 239
 240
 241
 242
 243
 244
 245
 246
 247
 248
 249
 250
 251
 252
 253
 254
 255
 256
 257
 258
 259
 260
 261
 262
 263
 264
 265
 266
 267
 268
 269
 270
 271
 272
 273
 274
 275
 276
 277
 278
 279
 280
 281
 282
 283
 284
 285
 286
 287
 288
 289
 290
 291
 292
 293
 294
 295
 296
 297
 298
 299
 300
 301
 302
 303
 304
 305
 306
 307
 308
 309
 310
 311
 312
 313
 314
 315
 316
 317
 318
 319
 320
 321
 322
 323
 324
 325
 326
 327
 328
 329
 330
 331
 332
 333
 334
 335
 336
 337
 338
 339
 340
 341
 342
 343
 344
 345
 346
 347
 348
 349
 350
 351
 352
 353
 354
 355
 356
 357
 358
 359
 360
 361
 362
 363
 364
 365
 366
 367
 368
 369
 370
 371
 372
 373
 374
 375
 376
 377
 378
 379
 380
 381
 382
 383
 384
 385
 386
 387
 388
 389
 390
 391
 392
 393
 394
 395
 396
 397
 398
 399
 400
 401
 402
 403
 404
 405
 406
 407
 408
 409
 410
 411
 412
 413
 414
 415
 416
 417
 418
 419
 420
 421
 422
 423
 424
 425
 426
 427
 428
 429
 430
 431
 432
 433
 434
 435
 436
 437
 438
 439
 440
 441
 442
 443
 444
 445
 446
 447
 448
 449
 450
 451
 452
 453
 454
 455
 456
 457
 458
 459
 460
 461
 462
 463
 464
 465
 466
 467
 468
 469
 470
 471
 472
 473
 474
 475
 476
 477
 478
 479
 480
 481
 482
 483
 484
 485
 486
 487
 488
 489
 490
 491
 492
 493
 494
 495
 496
 497
 498
 499
 500
 501
 502
 503
 504
 505
 506
 507
 508
 509
 510
 511
 512
 513
 514
 515
 516
 517
 518
 519
 520
 521
 522
 523
 524
 525
 526
 527
 528
 529
 530
 531
 532
 533
 534
 535
 536
 537
 538
 539
 540
 541
 542
 543
 544
 545
 546
 547
 548
 549
 550
 551
 552
 553
 554
 555
 556
 557
 558
 559
 560
 561
 562
 563
 564
 565
 566
 567
 568
 569
 570
 571
 572
 573
 574
 575
 576
 577
 578
 579
 580
 581
 582
 583
 584
 585
 586
 587
 588
 589
 590
 591
 592
 593
 594
 595
 596
 597
 598
 599
 600
 601
 602
 603
 604
 605
 606
 607
 608
 609
 610
 611
 612
 613
 614
 615
 616
 617
 618
 619
 620
 621
 622
 623
 624
 625
 626
 627
 628
 629
 630
 631
 632
 633
 634
 635
 636
 637
 638
 639
 640
 641
 642
 643
 644
 645
 646
 647
 648
 649
 650
 651
 652
 653
 654
 655
 656
 657
 658
 659
 660
 661
 662
 663
 664
 665
 666
 667
 668
 669
 670
 671
 672
 673
 674
 675
 676
 677
 678
 679
 680
 681
 682
 683
 684
 685
 686
 687
 688
 689
 690
 691
 692
 693
 694
 695
 696
 697
 698
 699
 700
 701
 702
 703
 704
 705
 706
 707
 708
 709
 710
 711
 712
 713
 714
 715
 716
 717
 718
 719
 720
 721
 722
 723
 724
 725
 726
 727
 728
 729
 730
 731
 732
 733
 734
 735
 736
 737
 738
 739
 740
 741
 742
 743
 744
 745
 746
 747
 748
 749
 750
 751
 752
 753
 754
 755
 756
 757
 758
 759
 760
 761
 762
 763
 764
 765
 766
 767
 768
 769
 770
 771
 772
 773
 774
 775
 776
 777
 778
 779
 780
 781
 782
 783
 784
 785
 786
 787
 788
 789
 790
 791
 792
 793
 794
 795
 796
 797
 798
 799
 800
 801
 802
 803
 804
 805
 806
 807
 808
 809
 810
 811
 812
 813
 814
 815
 816
 817
 818
 819
 820
 821
 822
 823
 824
 825
 826
 827
 828
 829
 830
 831
 832
 833
 834
 835
 836
 837
 838
 839
 840
 841
 842
 843
 844
 845
 846
 847
 848
 849
 850
 851
 852
 853
 854
 855
 856
 857
 858
 859
 860
 861
 862
 863
 864
 865
 866
 867
 868
 869
 870
 871
 872
 873
 874
 875
 876
 877
 878
 879
 880
 881
 882
 883
 884
 885
 886
 887
 888
 889
 890
 891
 892
 893
 894
 895
 896
 897
 898
 899
 900
 901
 902
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
/* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */
/*
 * This file is part of the LibreOffice project.
 *
 * This Source Code Form is subject to the terms of the Mozilla Public
 * License, v. 2.0. If a copy of the MPL was not distributed with this
 * file, You can obtain one at http://mozilla.org/MPL/2.0/.
 *
 * This file incorporates work covered by the following license notice:
 *
 *   Licensed to the Apache Software Foundation (ASF) under one or more
 *   contributor license agreements. See the NOTICE file distributed
 *   with this work for additional information regarding copyright
 *   ownership. The ASF licenses this file to you under the Apache
 *   License, Version 2.0 (the "License"); you may not use this file
 *   except in compliance with the License. You may obtain a copy of
 *   the License at http://www.apache.org/licenses/LICENSE-2.0 .
 */

#include <msfilter.hxx>
#include "writerwordglue.hxx"
#include <doc.hxx>
#include "writerhelper.hxx"
#include <IDocumentStylePoolAccess.hxx>

#include <algorithm>

#include <rtl/tencinfo.h>
#include <sal/log.hxx>

#include <unicode/ubidi.h>
#include <tools/tenccvt.hxx>
#include <com/sun/star/i18n/ScriptType.hpp>
#include <com/sun/star/i18n/XBreakIterator.hpp>

#include <editeng/lrspitem.hxx>
#include <editeng/ulspitem.hxx>
#include <editeng/boxitem.hxx>
#include <editeng/fontitem.hxx>
#include <o3tl/sorted_vector.hxx>
#include <frmfmt.hxx>
#include <fmtclds.hxx>
#include <hfspacingitem.hxx>
#include <fmtfsize.hxx>
#include <poolfmt.hxx>
#include <swrect.hxx>
#include <fmthdft.hxx>
#include <frmatr.hxx>
#include <ndtxt.hxx>
#include <breakit.hxx>

using namespace css;

namespace myImplHelpers
{
    static SwTwips CalcHdFtDist(const SwFrameFormat& rFormat, sal_uInt16 nSpacing)
    {
        /*
        The normal case for reexporting word docs is to have dynamic spacing,
        as this is word's only setting, and the reason for the existence of the
        dynamic spacing features. If we have dynamic spacing active then we can
        add its spacing to the value height of the h/f and get the wanted total
        size for word.

        Otherwise we have to get the real layout rendered
        height, which is totally nonoptimum, but the best we can do.
        */
        long nDist=0;
        const SwFormatFrameSize& rSz = rFormat.GetFrameSize();

        const SwHeaderAndFooterEatSpacingItem &rSpacingCtrl =
            sw::util::ItemGet<SwHeaderAndFooterEatSpacingItem>
            (rFormat, RES_HEADER_FOOTER_EAT_SPACING);
        if (rSpacingCtrl.GetValue())
            nDist += rSz.GetHeight();
        else
        {
            SwRect aRect(rFormat.FindLayoutRect());
            if (aRect.Height())
                nDist += aRect.Height();
            else
            {
                const SwFormatFrameSize& rSize = rFormat.GetFrameSize();
                if (SwFrameSize::Variable != rSize.GetHeightSizeType())
                    nDist += rSize.GetHeight();
                else
                {
                    nDist += 274;       // default for 12pt text
                    nDist += nSpacing;
                }
            }
        }
        return nDist;
    }

    static SwTwips CalcHdDist(const SwFrameFormat& rFormat)
    {
        return CalcHdFtDist(rFormat, rFormat.GetULSpace().GetUpper());
    }

    static SwTwips CalcFtDist(const SwFrameFormat& rFormat)
    {
        return CalcHdFtDist(rFormat, rFormat.GetULSpace().GetLower());
    }

    /*
     SwTextFormatColl and SwCharFormat are quite distinct types and how they are
     gotten is also distinct, but the algorithm to match word's equivalents into
     them is the same, so we put the different stuff into two separate helper
     implementations and a core template that uses the helpers that uses the
     same algorithm to do the work. We'll make the helpers specializations of a
     non existing template so I can let the compiler figure out the right one
     to use from a simple argument to the algorithm class
    */
    template <class C> class MapperImpl;
    template<> class MapperImpl<SwTextFormatColl>
    {
    private:
        SwDoc &mrDoc;
    public:
        MapperImpl(SwDoc &rDoc) : mrDoc(rDoc) {}
        SwTextFormatColl* GetBuiltInStyle(ww::sti eSti);
        SwTextFormatColl* GetStyle(const OUString &rName);
        SwTextFormatColl* MakeStyle(const OUString &rName);
    };

    SwTextFormatColl* MapperImpl<SwTextFormatColl>::GetBuiltInStyle(ww::sti eSti)
    {
        const RES_POOL_COLLFMT_TYPE RES_NONE  = RES_POOLCOLL_DOC_END;
        static const RES_POOL_COLLFMT_TYPE aArr[]=
        {
            RES_POOLCOLL_STANDARD, RES_POOLCOLL_HEADLINE1,
            RES_POOLCOLL_HEADLINE2, RES_POOLCOLL_HEADLINE3,
            RES_POOLCOLL_HEADLINE4, RES_POOLCOLL_HEADLINE5,
            RES_POOLCOLL_HEADLINE6, RES_POOLCOLL_HEADLINE7,
            RES_POOLCOLL_HEADLINE8, RES_POOLCOLL_HEADLINE9,
            RES_POOLCOLL_TOX_IDX1, RES_POOLCOLL_TOX_IDX2,
            RES_POOLCOLL_TOX_IDX3, RES_NONE, RES_NONE, RES_NONE, RES_NONE,
            RES_NONE, RES_NONE, RES_POOLCOLL_TOX_CNTNT1,
            RES_POOLCOLL_TOX_CNTNT2, RES_POOLCOLL_TOX_CNTNT3,
            RES_POOLCOLL_TOX_CNTNT4, RES_POOLCOLL_TOX_CNTNT5,
            RES_POOLCOLL_TOX_CNTNT6, RES_POOLCOLL_TOX_CNTNT7,
            RES_POOLCOLL_TOX_CNTNT8, RES_POOLCOLL_TOX_CNTNT9, RES_NONE,
            RES_POOLCOLL_FOOTNOTE, RES_NONE, RES_POOLCOLL_HEADER,
            RES_POOLCOLL_FOOTER, RES_POOLCOLL_TOX_IDXH, RES_NONE, RES_NONE,
            RES_POOLCOLL_JAKETADRESS, RES_POOLCOLL_SENDADRESS, RES_NONE,
            RES_NONE, RES_NONE, RES_NONE, RES_NONE, RES_POOLCOLL_ENDNOTE,
            RES_NONE, RES_NONE, RES_NONE, RES_POOLCOLL_LISTS_BEGIN,
            RES_NONE, RES_NONE, RES_NONE, RES_NONE, RES_NONE, RES_NONE,
            RES_NONE, RES_NONE, RES_NONE, RES_NONE, RES_NONE, RES_NONE,
            RES_NONE, RES_NONE, RES_POOLCOLL_HEADLINE_BASE, RES_NONE,
            RES_POOLCOLL_SIGNATURE, RES_NONE, RES_POOLCOLL_TEXT,
            RES_POOLCOLL_TEXT_MOVE, RES_NONE, RES_NONE, RES_NONE, RES_NONE,
            RES_NONE, RES_NONE, RES_POOLCOLL_DOC_SUBTITLE
        };

        OSL_ENSURE(SAL_N_ELEMENTS(aArr) == 75, "Style Array has false size");

        SwTextFormatColl* pRet = nullptr;
        //If this is a built-in word style that has a built-in writer
        //equivalent, then map it to one of our built in styles regardless
        //of its name
        if (sal::static_int_cast< size_t >(eSti) < SAL_N_ELEMENTS(aArr) && aArr[eSti] != RES_NONE)
            pRet = mrDoc.getIDocumentStylePoolAccess().GetTextCollFromPool( static_cast< sal_uInt16 >(aArr[eSti]), false);
        return pRet;
    }

    SwTextFormatColl* MapperImpl<SwTextFormatColl>::GetStyle(const OUString &rName)
    {
        return sw::util::GetParaStyle(mrDoc, rName);
    }

    SwTextFormatColl* MapperImpl<SwTextFormatColl>::MakeStyle(const OUString &rName)
    {
        return mrDoc.MakeTextFormatColl(rName,
            mrDoc.GetDfltTextFormatColl());
    }

    template<> class MapperImpl<SwCharFormat>
    {
    private:
        SwDoc &mrDoc;
    public:
        MapperImpl(SwDoc &rDoc) : mrDoc(rDoc) {}
        SwCharFormat* GetBuiltInStyle(ww::sti eSti);
        SwCharFormat* GetStyle(const OUString &rName);
        SwCharFormat* MakeStyle(const OUString &rName);
    };

    SwCharFormat* MapperImpl<SwCharFormat>::GetBuiltInStyle(ww::sti eSti)
    {
        RES_POOL_CHRFMT_TYPE eLookup = RES_POOLCHR_NORMAL_END;
        switch (eSti)
        {
            case ww::stiFootnoteRef:
                eLookup = RES_POOLCHR_FOOTNOTE;
                break;
            case ww::stiLnn:
                eLookup = RES_POOLCHR_LINENUM;
                break;
            case ww::stiPgn:
                eLookup = RES_POOLCHR_PAGENO;
                break;
            case ww::stiEdnRef:
                eLookup = RES_POOLCHR_ENDNOTE;
                break;
            case ww::stiHyperlink:
                eLookup = RES_POOLCHR_INET_NORMAL;
                break;
            case ww::stiHyperlinkFollowed:
                eLookup = RES_POOLCHR_INET_VISIT;
                break;
            case ww::stiStrong:
                eLookup = RES_POOLCHR_HTML_STRONG;
                break;
            case ww::stiEmphasis:
                eLookup = RES_POOLCHR_HTML_EMPHASIS;
                break;
            default:
                eLookup = RES_POOLCHR_NORMAL_END;
                break;
        }
        SwCharFormat *pRet = nullptr;
        if (eLookup != RES_POOLCHR_NORMAL_END)
            pRet = mrDoc.getIDocumentStylePoolAccess().GetCharFormatFromPool( static_cast< sal_uInt16 >(eLookup) );
        return pRet;
    }

    SwCharFormat* MapperImpl<SwCharFormat>::GetStyle(const OUString &rName)
    {
        return sw::util::GetCharStyle(mrDoc, rName);
    }

    SwCharFormat* MapperImpl<SwCharFormat>::MakeStyle(const OUString &rName)
    {
        return mrDoc.MakeCharFormat(rName, mrDoc.GetDfltCharFormat());
    }

    template<class C> class StyleMapperImpl
    {
    private:
        MapperImpl<C> maHelper;
        o3tl::sorted_vector<const C*> maUsedStyles;
        C* MakeNonCollidingStyle(const OUString& rName);
    public:
        typedef std::pair<C*, bool> StyleResult;
        explicit StyleMapperImpl(SwDoc &rDoc) : maHelper(rDoc) {}
        StyleResult GetStyle(const OUString& rName, ww::sti eSti);
    };

    template<class C>
    typename StyleMapperImpl<C>::StyleResult
    StyleMapperImpl<C>::GetStyle(const OUString& rName, ww::sti eSti)
    {
        C *pRet = maHelper.GetBuiltInStyle(eSti);

        //If we've used it once, don't reuse it
        if (pRet && (maUsedStyles.end() != maUsedStyles.find(pRet)))
            pRet = nullptr;

        if (!pRet)
        {
            pRet = maHelper.GetStyle(rName);
            //If we've used it once, don't reuse it
            if (pRet && (maUsedStyles.end() != maUsedStyles.find(pRet)))
                pRet = nullptr;
        }

        bool bStyExist = pRet != nullptr;

        if (!pRet)
        {
            OUString aName(rName);
            sal_Int32 nIdx = rName.indexOf(',');
            // No commas allow in SW style names
            if (-1 != nIdx)
                aName = rName.copy( 0, nIdx );
            pRet = MakeNonCollidingStyle( aName );
        }

        if (pRet)
            maUsedStyles.insert(pRet);

        return StyleResult(pRet, bStyExist);
    }

    template<class C>
    C* StyleMapperImpl<C>::MakeNonCollidingStyle(const OUString& rName)
    {
        OUString aName(rName);
        C* pColl = 0;

        if (0 != (pColl = maHelper.GetStyle(aName)))
        {
            //If the style collides first stick WW- in front of it, unless
            //it already has it and then successively add a larger and
            //larger number after it, it's got to work at some stage!
            if (!aName.startsWith("WW-"))
                aName = "WW-" + aName;

            sal_Int32 nI = 1;
            OUString aBaseName = aName;
            while (
                    0 != (pColl = maHelper.GetStyle(aName)) &&
                    (nI < SAL_MAX_INT32)
                  )
            {
                aName = aBaseName + OUString::number(nI++);
            }
        }

        return pColl ? 0 : maHelper.MakeStyle(aName);
    }

    static OUString FindBestMSSubstituteFont(const OUString &rFont)
    {
        if (IsStarSymbol(rFont))
            return "Arial Unicode MS";
        return GetSubsFontName(rFont, SubsFontFlags::ONLYONE | SubsFontFlags::MS);
    }

    namespace {

    //Utility to remove entries before a given starting position
    class IfBeforeStart
    {
    private:
        sal_Int32 mnStart;
    public:
        explicit IfBeforeStart(sal_Int32 nStart) : mnStart(nStart) {}
        bool operator()(const sw::util::CharRunEntry &rEntry) const
        {
            return rEntry.mnEndPos < mnStart;
        }
    };

    }
}

/// Count what Word calls left/right margin from a format's LRSpace + Box.
static SvxLRSpaceItem lcl_getWordLRSpace(const SwFrameFormat& rFormat)
{
    SvxLRSpaceItem aLR(rFormat.GetLRSpace());
    const SvxBoxItem& rBox = rFormat.GetBox();

    aLR.SetLeft(aLR.GetLeft() + rBox.GetDistance(SvxBoxItemLine::LEFT));
    if (const editeng::SvxBorderLine* pLeft = rBox.GetLeft())
        aLR.SetLeft(aLR.GetLeft() + pLeft->GetWidth());

    aLR.SetRight(aLR.GetRight() + rBox.GetDistance(SvxBoxItemLine::RIGHT));
    if (const editeng::SvxBorderLine* pRight = rBox.GetRight())
        aLR.SetRight(aLR.GetRight() + pRight->GetWidth());

    return aLR;
}

namespace sw
{
    namespace util
    {

        bool IsPlausableSingleWordSection(const SwFrameFormat &rTitleFormat, const SwFrameFormat &rFollowFormat)
        {
            bool bPlausableSingleWordSection = true;

            const SwFormatCol& rFirstCols = rTitleFormat.GetCol();
            const SwFormatCol& rFollowCols = rFollowFormat.GetCol();
            const SwColumns& rFirstColumns = rFirstCols.GetColumns();
            const SwColumns& rFollowColumns = rFollowCols.GetColumns();
            SvxLRSpaceItem aOneLR = lcl_getWordLRSpace(rTitleFormat);
            SvxLRSpaceItem aTwoLR = lcl_getWordLRSpace(rFollowFormat);
            const SwFormatFrameSize& rFirstFrameSize = rTitleFormat.GetFrameSize();
            const SwFormatFrameSize& rFollowFrameSize = rFollowFormat.GetFrameSize();

            if (rFirstColumns.size() != rFollowColumns.size())
            {
                //e.g. #i4320#
                bPlausableSingleWordSection = false;
            }
            else if (aOneLR != aTwoLR)
                bPlausableSingleWordSection = false;
            else if (rFirstFrameSize != rFollowFrameSize)
                bPlausableSingleWordSection = false;
            else
            {
                HdFtDistanceGlue aOne(rTitleFormat.GetAttrSet());
                HdFtDistanceGlue aTwo(rFollowFormat.GetAttrSet());
                //e.g. #i14509#
                if (!aOne.StrictEqualTopBottom(aTwo))
                    bPlausableSingleWordSection = false;
            }
            return bPlausableSingleWordSection;
        }

        HdFtDistanceGlue::HdFtDistanceGlue(const SfxItemSet &rPage)
        {
            if (const SvxBoxItem *pBox = rPage.GetItem<SvxBoxItem>(RES_BOX))
            {
                dyaHdrTop = pBox->CalcLineSpace( SvxBoxItemLine::TOP, /*bEvenIfNoLine*/true );
                dyaHdrBottom = pBox->CalcLineSpace( SvxBoxItemLine::BOTTOM, /*bEvenIfNoLine*/true );
            }
            else
            {
                dyaHdrTop = dyaHdrBottom = 0;
            }
            const SvxULSpaceItem &rUL =
                ItemGet<SvxULSpaceItem>(rPage, RES_UL_SPACE);
            dyaHdrTop += rUL.GetUpper();
            dyaHdrBottom += rUL.GetLower();

            dyaTop = dyaHdrTop;
            dyaBottom = dyaHdrBottom;

            const SwFormatHeader *pHd = rPage.GetItem<SwFormatHeader>(RES_HEADER);
            if (pHd && pHd->IsActive() && pHd->GetHeaderFormat())
            {
                mbHasHeader = true;
                dyaTop = dyaTop + static_cast< sal_uInt16 >( (myImplHelpers::CalcHdDist(*(pHd->GetHeaderFormat()))) );
            }
            else
                mbHasHeader = false;

            const SwFormatFooter *pFt = rPage.GetItem<SwFormatFooter>(RES_FOOTER);
            if (pFt && pFt->IsActive() && pFt->GetFooterFormat())
            {
                mbHasFooter = true;
                dyaBottom = dyaBottom + static_cast< sal_uInt16 >( (myImplHelpers::CalcFtDist(*(pFt->GetFooterFormat()))) );
            }
            else
                mbHasFooter = false;
        }

        bool HdFtDistanceGlue::StrictEqualTopBottom(const HdFtDistanceGlue &rOther)
            const
        {
            // Check top only if both object have a header or if
            // both object don't have a header
            if (HasHeader() == rOther.HasHeader())
            {
                if (dyaTop != rOther.dyaTop)
                    return false;
            }

            // Check bottom only if both object have a footer or if
            // both object don't have a footer
            if (HasFooter() == rOther.HasFooter())
            {
                if (dyaBottom != rOther.dyaBottom)
                    return false;
            }

            return true;
        }

        ParaStyleMapper::ParaStyleMapper(SwDoc &rDoc)
            : mpImpl(new myImplHelpers::StyleMapperImpl<SwTextFormatColl>(rDoc))
        {
        }

        ParaStyleMapper::~ParaStyleMapper()
        {
        }

        ParaStyleMapper::StyleResult ParaStyleMapper::GetStyle(
            const OUString& rName, ww::sti eSti)
        {
            return mpImpl->GetStyle(rName, eSti);
        }

        CharStyleMapper::CharStyleMapper(SwDoc &rDoc)
            : mpImpl(new myImplHelpers::StyleMapperImpl<SwCharFormat>(rDoc))
        {
        }

        CharStyleMapper::~CharStyleMapper()
        {
        }

        CharStyleMapper::StyleResult CharStyleMapper::GetStyle(
            const OUString& rName, ww::sti eSti)
        {
            return mpImpl->GetStyle(rName, eSti);
        }

        FontMapExport::FontMapExport(const OUString &rFamilyName)
        {
            sal_Int32 nIndex = 0;
            msPrimary = GetNextFontToken(rFamilyName, nIndex);
            msSecondary = myImplHelpers::FindBestMSSubstituteFont(msPrimary);
            if (msSecondary.isEmpty() && nIndex != -1)
                msSecondary = GetNextFontToken(rFamilyName, nIndex);
        }

        bool ItemSort::operator()(sal_uInt16 nA, sal_uInt16 nB) const
        {
            /*
             #i24291#
             All we want to do is ensure for now is that if a charfmt exist
             in the character properties that it rises to the top and is
             exported first.  In the future we might find more ordering
             dependencies for export, in which case this is the place to do
             it
            */
            if (nA == nB)
                return false;
            if (nA == RES_TXTATR_CHARFMT)
                return true;
            if (nB == RES_TXTATR_CHARFMT)
                return false;
            if (nA == RES_TXTATR_INETFMT)
                return true;
            if (nB == RES_TXTATR_INETFMT)
               return false;
            return nA < nB;
        }

        CharRuns GetPseudoCharRuns(const SwTextNode& rTextNd)
        {
            const OUString &rText = rTextNd.GetText();

            bool bParaIsRTL = false;
            OSL_ENSURE(rTextNd.GetDoc(), "No document for node?, suspicious");
            if (rTextNd.GetDoc())
            {
                if (SvxFrameDirection::Horizontal_RL_TB ==
                    rTextNd.GetDoc()->GetTextDirection(SwPosition(rTextNd)))
                {
                    bParaIsRTL = true;
                }
            }

            using namespace ::com::sun::star::i18n;

            sal_uInt16 nScript = i18n::ScriptType::LATIN;
            assert(g_pBreakIt && g_pBreakIt->GetBreakIter().is());
            if (!rText.isEmpty())
                nScript = g_pBreakIt->GetBreakIter()->getScriptType(rText, 0);

            rtl_TextEncoding eChrSet = ItemGet<SvxFontItem>(rTextNd,
                GetWhichOfScript(RES_CHRATR_FONT, nScript)).GetCharSet();
            eChrSet = GetExtendedTextEncoding(eChrSet);

            CharRuns aRunChanges;

            if (rText.isEmpty())
            {
                aRunChanges.emplace_back(0, nScript, eChrSet,
                    bParaIsRTL);
                return aRunChanges;
            }

            typedef std::pair<int32_t, bool> DirEntry;
            typedef std::pair<sal_Int32, sal_uInt16> ScriptEntry;
            std::vector<DirEntry> aDirChanges;
            std::vector<ScriptEntry> aScripts;

            UBiDiDirection eDefaultDir = bParaIsRTL ? UBIDI_RTL : UBIDI_LTR;
            UErrorCode nError = U_ZERO_ERROR;
            UBiDi* pBidi = ubidi_openSized(rText.getLength(), 0, &nError);
            ubidi_setPara(pBidi, reinterpret_cast<const UChar *>(rText.getStr()), rText.getLength(),
                    static_cast< UBiDiLevel >(eDefaultDir), nullptr, &nError);

            sal_Int32 nCount = ubidi_countRuns(pBidi, &nError);
            aDirChanges.reserve(nCount);

            int32_t nStart = 0;
            int32_t nEnd;
            UBiDiLevel nCurrDir;

            for (sal_Int32 nIdx = 0; nIdx < nCount; ++nIdx)
            {
                ubidi_getLogicalRun(pBidi, nStart, &nEnd, &nCurrDir);
                /*
                UBiDiLevel is the type of the level values in this BiDi
                implementation.

                It holds an embedding level and indicates the visual direction
                by its bit 0 (even/odd value).

                The value for UBIDI_DEFAULT_LTR is even and the one for
                UBIDI_DEFAULT_RTL is odd
                */
                aDirChanges.emplace_back(nEnd, nCurrDir & 0x1);
                nStart = nEnd;
            }
            ubidi_close(pBidi);

            assert(g_pBreakIt && g_pBreakIt->GetBreakIter().is());

            sal_Int32 nLen = rText.getLength();
            sal_Int32 nPos = 0;
            while (nPos < nLen)
            {
                sal_Int32 nEnd2 = g_pBreakIt->GetBreakIter()->endOfScript(rText, nPos,
                    nScript);
                if (nEnd2 < 0)
                    break;
                nPos = nEnd2;
                aScripts.emplace_back(nPos, nScript);
                nScript = g_pBreakIt->GetBreakIter()->getScriptType(rText, nPos);
            }

            auto aBiDiEnd = aDirChanges.cend();
            auto aScriptEnd = aScripts.cend();

            auto aBiDiIter = aDirChanges.cbegin();
            auto aScriptIter = aScripts.cbegin();

            bool bCharIsRTL = bParaIsRTL;

            while (
                    aBiDiIter != aBiDiEnd ||
                    aScriptIter != aScriptEnd
                  )
            {
                sal_Int32 nMinPos = rText.getLength();

                if (aBiDiIter != aBiDiEnd)
                {
                    if (aBiDiIter->first < nMinPos)
                        nMinPos = aBiDiIter->first;
                    bCharIsRTL = aBiDiIter->second;
                }

                if (aScriptIter != aScriptEnd)
                {
                    if (aScriptIter->first < nMinPos)
                        nMinPos = aScriptIter->first;
                    nScript = aScriptIter->second;
                }

                aRunChanges.emplace_back(nMinPos, nScript, eChrSet, bCharIsRTL);

                if (aBiDiIter != aBiDiEnd)
                {
                    if (aBiDiIter->first == nMinPos)
                        ++aBiDiIter;
                }

                if (aScriptIter != aScriptEnd)
                {
                    if (aScriptIter->first == nMinPos)
                        ++aScriptIter;
                }
            }

            aRunChanges.erase(std::remove_if(aRunChanges.begin(),
                aRunChanges.end(), myImplHelpers::IfBeforeStart(0/*nTextStart*/)), aRunChanges.end());

            return aRunChanges;
        }
    }

    namespace ms
    {
        sal_uInt8 rtl_TextEncodingToWinCharset(rtl_TextEncoding eTextEncoding)
        {
            sal_uInt8 nRet =
                rtl_getBestWindowsCharsetFromTextEncoding(eTextEncoding);
            switch (eTextEncoding)
            {
                case RTL_TEXTENCODING_DONTKNOW:
                case RTL_TEXTENCODING_UCS2:
                case RTL_TEXTENCODING_UTF7:
                case RTL_TEXTENCODING_UTF8:
                case RTL_TEXTENCODING_JAVA_UTF8:
                    nRet = 0x01;
                    break;
                default:
                    break;
            }
            return nRet;
        }

        static bool
        CanEncode(OUString const& rString, rtl_TextEncoding const eEncoding)
        {
            OString tmp;
            return rString.convertToString(&tmp, eEncoding,
                    RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_ERROR |
                    RTL_TEXTTOUNICODE_FLAGS_INVALID_ERROR);
        }

        sal_uInt8 rtl_TextEncodingToWinCharsetRTF(
                OUString const& rFontName, OUString const& rAltName,
                rtl_TextEncoding eTextEncoding)
        {
            sal_uInt8 nRet =
                rtl_getBestWindowsCharsetFromTextEncoding(eTextEncoding);
            rtl_TextEncoding enc2 = rtl_getTextEncodingFromWindowsCharset(nRet);
            if (!rtl_isOctetTextEncoding(enc2) /* check to avoid asserts */ ||
                !(CanEncode(rFontName, enc2) && CanEncode(rAltName, enc2)))
            {
                static struct { rtl_TextEncoding enc; sal_uInt8 charset; }
                const s_fallbacks [] = {
                    { RTL_TEXTENCODING_MS_932, 0x80 }, // Shift-JIS
                    { RTL_TEXTENCODING_MS_936, 0x86 }, // GB-2312
                    { RTL_TEXTENCODING_MS_950, 0x88 }, // Big5
                    { RTL_TEXTENCODING_MS_949, 0x81 }, // EUC-KR
                };
                for (const auto & i : s_fallbacks)
                {
                    // fall back to a charset that can at least encode the
                    // font's name
                    if (CanEncode(rFontName, i.enc)
                        && CanEncode(rAltName, i.enc))
                    {<--- Consider using std::find_if algorithm instead of a raw loop.
                        return i.charset;
                    }
                }
                SAL_INFO("sw.rtf", "no fallback charset found for font: "
                         << rFontName << " " << rAltName);
                nRet = 0x01; // all hope lost: "default", whatever that is
            }
            return nRet;
        }

        sal_uInt32 DateTime2DTTM( const DateTime& rDT )
        {
        /*
        mint    short   :6  0000003F    minutes (0-59)
        hr      short   :5  000007C0    hours (0-23)
        dom     short   :5  0000F800    days of month (1-31)
        mon     short   :4  000F0000    months (1-12)
        yr      short   :9  1FF00000    years (1900-2411)-1900
        wdy     short   :3  E0000000    weekday(Sunday=0
                                                Monday=1
        ( wdy can be ignored )                  Tuesday=2
                                                Wednesday=3
                                                Thursday=4
                                                Friday=5
                                                Saturday=6)
        */

            if ( rDT.GetDate() == 0 )
                return 0;
            sal_uInt32 nDT = ( rDT.GetDayOfWeek() + 1 ) % 7;
            nDT <<= 9;
            nDT += ( rDT.GetYear() - 1900 ) & 0x1ff;
            nDT <<= 4;
            nDT += rDT.GetMonth() & 0xf;
            nDT <<= 5;
            nDT += rDT.GetDay() & 0x1f;
            nDT <<= 5;
            nDT += rDT.GetHour() & 0x1f;
            nDT <<= 6;
            nDT += rDT.GetMin() & 0x3f;
            return nDT;
        }


        /** Find cFind in rParams if not embedded in " double quotes.
            Will NOT find '\\' or '"'.
         */
        static sal_Int32 findUnquoted( const OUString& rParams, sal_Unicode cFind, sal_Int32 nFromPos )
        {
            const sal_Int32 nLen = rParams.getLength();
            if (nFromPos < 0 || nLen <= nFromPos)
                return -1;
            for (sal_Int32 nI = nFromPos; nI < nLen; ++nI)
            {
                const sal_Unicode c = rParams[nI];
                if (c == '\\')
                    ++nI;
                else if (c == '\"')
                {
                    ++nI;
                    // While not at the end and not at an unescaped end quote
                    while (nI < nLen)
                    {
                        if (rParams[nI] == '\"' && rParams[nI-1] != '\\')
                            break;
                        ++nI;
                    }
                }
                else //normal unquoted section
                {
                    if (c == cFind)
                        return nI;
                }
            }
            return -1;
        }

        /** Find all rFind in rParams if not embedded in " double quotes and
            replace with rReplace. Will NOT find '\\' or '"'.
         */
        static bool replaceUnquoted( OUString& rParams, const OUString& rFind, const OUString& rReplace )
        {
            bool bReplaced = false;
            if (rFind.isEmpty())
                return bReplaced;
            const sal_Unicode cFirst = rFind[0];

            sal_Int32 nLen = rParams.getLength();
            for (sal_Int32 nI = 0; nI < nLen; ++nI)
            {
                const sal_Unicode c = rParams[nI];
                if (rParams[nI] == '\\')
                    ++nI;
                else if (rParams[nI] == '\"')
                {
                    ++nI;
                    // While not at the end and not at an unescaped end quote
                    while (nI < nLen)
                    {
                        if (rParams[nI] == '\"' && rParams[nI-1] != '\\')
                            break;
                        ++nI;
                    }
                }
                else //normal unquoted section
                {
                    if (c == cFirst && rParams.match( rFind, nI))
                    {
                        const sal_Int32 nFindLen = rFind.getLength();
                        const sal_Int32 nDiff = rReplace.getLength() - nFindLen;
                        rParams = rParams.replaceAt( nI, nFindLen, rReplace);
                        nI += nFindLen + nDiff - 1;
                        nLen += nDiff;
                        bReplaced = true;
                    }
                }
            }
            return bReplaced;
        }

        sal_uLong MSDateTimeFormatToSwFormat(OUString& rParams,
            SvNumberFormatter *pFormatter, LanguageType &rLang, bool bHijri,
            LanguageType nDocLang)
        {
            // tell the Formatter about the new entry
            sal_Int32 nCheckPos = 0;
            SvNumFormatType nType = SvNumFormatType::DEFINED;
            sal_uInt32  nKey = 0;

            SwapQuotesInField(rParams);

            // Force to Japanese when finding one of 'geE'.
            // XXX This actually may not be correct, all era keywords could be
            // used in other locales as well. I just don't know about Word. But
            // this is how it was for 10 years...
            bool bForceJapanese = (-1 != findUnquoted( rParams, 'g', 0));
            // XXX Why replace? The number formatter does handle them and this
            // effectively changes from Gengou to Gregorian calendar. Legacy
            // because it wasn't supported a decade ago and now moot? Or is
            // that a Word specialty?
            bForceJapanese |= replaceUnquoted( rParams, "ee", "yyyy");
            bForceJapanese |= replaceUnquoted( rParams, "EE", "YYYY");
            if (LANGUAGE_FRENCH != nDocLang)
            {
                // Handle the 'a' case here
                sal_Int32 nLastPos = 0;
                do
                {
                    sal_Int32 nPos = findUnquoted( rParams, 'a', nLastPos + 1 );
                    bForceJapanese |= ( nPos != -1 && IsNotAM( rParams, nPos ) );
                    nLastPos = nPos;
                } while ( -1 != nLastPos );
            }

            // Force to NatNum when finding one of 'oOA'
            bool bForceNatNum  = replaceUnquoted( rParams, "o", "m")
                                 || replaceUnquoted( rParams, "O", "M");
            if (LANGUAGE_FRENCH != nDocLang)
            {
                // Handle the 'A' case here
                sal_Int32 nLastPos = 0;
                do
                {
                    sal_Int32 nPos = findUnquoted( rParams, 'A', nLastPos + 1 );
                    bool bIsCharA = ( nPos != -1 && IsNotAM( rParams, nPos ) );
                    bForceNatNum |= bIsCharA;
                    if ( bIsCharA )
                        rParams = rParams.replaceAt( nPos, 1, "D" );
                    nLastPos = nPos;
                } while ( -1 != nLastPos );
            }

            sal_Int32 nLen = rParams.getLength();
            for (sal_Int32 nI = 0; nI < nLen; ++nI)
            {
                if (rParams[nI] == '\\')
                    ++nI;
                else if (rParams[nI] == '\"')
                {
                    ++nI;
                    // While not at the end and not at an unescaped end quote
                    while (nI < nLen)
                    {
                        if (rParams[nI] == '\"' && rParams[nI-1] != '\\')
                            break;
                        ++nI;
                    }
                }
                else //normal unquoted section
                {
                    sal_Unicode nChar = rParams[nI];

                    // Change the localized word string to english
                    if ( nDocLang == LANGUAGE_FRENCH )
                    {
                        if ( ( nChar == 'a' || nChar == 'A' ) && IsNotAM(rParams, nI) )
                            rParams = rParams.replaceAt(nI, 1, "Y");
                    }
                    if (nChar == '/')
                    {
                        // MM: We have to escape '/' in case it's used as a char.
                        // But not if it's a '/' inside AM/PM
                        if (!(IsPreviousAM(rParams, nI) && IsNextPM(rParams, nI)))
                        {
                            rParams = rParams.replaceAt(nI, 1, "\\/");
                            nLen++;
                        }
                        nI++;
                    }

                    // Deal with language differences in date format expression.
                    // Should be made with i18n framework.
                    // The list of the mappings and of those "special" locales is to be found at:
                    // http://l10n.openoffice.org/i18n_framework/LocaleData.html
                    if ( !bForceJapanese && !bForceNatNum )
                    {
                        // Convert to the localized equivalent for OOo
                        if ( rLang == LANGUAGE_FINNISH )
                        {
                            if (nChar == 'y' || nChar == 'Y')
                                rParams = rParams.replaceAt(nI, 1, "V");
                            else if (nChar == 'm' || nChar == 'M')
                                rParams = rParams.replaceAt(nI, 1, "K");
                            else if (nChar == 'd' || nChar == 'D')
                                rParams = rParams.replaceAt(nI, 1, "P");
                            else if (nChar == 'h' || nChar == 'H')
                                rParams = rParams.replaceAt(nI, 1, "T");
                        }
                        else if ( rLang.anyOf(
                            LANGUAGE_DANISH,
                            LANGUAGE_NORWEGIAN,
                            LANGUAGE_NORWEGIAN_BOKMAL,
                            LANGUAGE_NORWEGIAN_NYNORSK,
                            LANGUAGE_SWEDISH,
                            LANGUAGE_SWEDISH_FINLAND))
                        {
                            if (nChar == 'h' || nChar == 'H')
                                rParams = rParams.replaceAt(nI, 1, "T");
                        }
                        else if ( rLang.anyOf(
                            LANGUAGE_PORTUGUESE,
                            LANGUAGE_PORTUGUESE_BRAZILIAN,
                            LANGUAGE_SPANISH_MODERN,
                            LANGUAGE_SPANISH_DATED,
                            LANGUAGE_SPANISH_MEXICAN,
                            LANGUAGE_SPANISH_GUATEMALA,
                            LANGUAGE_SPANISH_COSTARICA,
                            LANGUAGE_SPANISH_PANAMA,
                            LANGUAGE_SPANISH_DOMINICAN_REPUBLIC,
                            LANGUAGE_SPANISH_VENEZUELA,
                            LANGUAGE_SPANISH_COLOMBIA,
                            LANGUAGE_SPANISH_PERU,
                            LANGUAGE_SPANISH_ARGENTINA,
                            LANGUAGE_SPANISH_ECUADOR,
                            LANGUAGE_SPANISH_CHILE,
                            LANGUAGE_SPANISH_URUGUAY,
                            LANGUAGE_SPANISH_PARAGUAY,
                            LANGUAGE_SPANISH_BOLIVIA,
                            LANGUAGE_SPANISH_EL_SALVADOR,
                            LANGUAGE_SPANISH_HONDURAS,
                            LANGUAGE_SPANISH_NICARAGUA,
                            LANGUAGE_SPANISH_PUERTO_RICO))
                        {
                            if (nChar == 'a' || nChar == 'A')
                                rParams = rParams.replaceAt(nI, 1, "O");
                            else if (nChar == 'y' || nChar == 'Y')
                                rParams = rParams.replaceAt(nI, 1, "A");
                        }
                        else if ( rLang.anyOf(
                            LANGUAGE_DUTCH,
                            LANGUAGE_DUTCH_BELGIAN))
                        {
                            if (nChar == 'y' || nChar == 'Y')
                                rParams = rParams.replaceAt(nI, 1, "J");
                            else if (nChar == 'u' || nChar == 'U')
                                rParams = rParams.replaceAt(nI, 1, "H");
                        }
                        else if ( rLang.anyOf(
                                LANGUAGE_ITALIAN,
                                LANGUAGE_ITALIAN_SWISS))
                        {
                            if (nChar == 'a' || nChar == 'A')
                                rParams = rParams.replaceAt(nI, 1, "O");
                            else if (nChar == 'g' || nChar == 'G')
                                rParams = rParams.replaceAt(nI, 1, "X");
                            else if (nChar == 'y' || nChar == 'Y')
                                rParams = rParams.replaceAt(nI, 1, "A");
                            else if (nChar == 'd' || nChar == 'D')
                                rParams = rParams.replaceAt(nI, 1, "G");
                        }
                        else if ( rLang.anyOf(
                            LANGUAGE_GERMAN,
                            LANGUAGE_GERMAN_SWISS,
                            LANGUAGE_GERMAN_AUSTRIAN,
                            LANGUAGE_GERMAN_LUXEMBOURG,
                            LANGUAGE_GERMAN_LIECHTENSTEIN))
                        {
                            if (nChar == 'y' || nChar == 'Y')
                                rParams = rParams.replaceAt(nI, 1, "J");
                            else if (nChar == 'd' || nChar == 'D')
                                rParams = rParams.replaceAt(nI, 1, "T");
                        }
                        else if ( rLang.anyOf(
                            LANGUAGE_FRENCH,
                            LANGUAGE_FRENCH_BELGIAN,
                            LANGUAGE_FRENCH_CANADIAN,
                            LANGUAGE_FRENCH_SWISS,
                            LANGUAGE_FRENCH_LUXEMBOURG,
                            LANGUAGE_FRENCH_MONACO))
                        {
                            if (nChar == 'y' || nChar == 'Y' || nChar == 'a')
                                rParams = rParams.replaceAt(nI, 1, "A");
                            else if (nChar == 'd' || nChar == 'D' || nChar == 'j')
                                rParams = rParams.replaceAt(nI, 1, "J");
                        }
                    }
                }
            }

            if (bForceNatNum)
                bForceJapanese = true;

            if (bForceJapanese)
                rLang = LANGUAGE_JAPANESE;

            if (bForceNatNum)
                rParams = "[NatNum1][$-411]" + rParams;

            if (bHijri)
                rParams = "[~hijri]" + rParams;

            pFormatter->PutEntry(rParams, nCheckPos, nType, nKey, rLang);

            return nKey;
        }

        bool IsPreviousAM(OUString const & rParams, sal_Int32 nPos)
        {
            return nPos>=2 && rParams.matchIgnoreAsciiCase("am", nPos-2);
        }
        bool IsNextPM(OUString const & rParams, sal_Int32 nPos)
        {
            return nPos+2<rParams.getLength() && rParams.matchIgnoreAsciiCase("pm", nPos+1);
        }
        bool IsNotAM(OUString const & rParams, sal_Int32 nPos)
        {
            ++nPos;
            return nPos>=rParams.getLength() || (rParams[nPos]!='M' && rParams[nPos]!='m');
        }

        void SwapQuotesInField(OUString &rFormat)
        {
            //Swap unescaped " and ' with ' and "
            const sal_Int32 nLen = rFormat.getLength();
            for (sal_Int32 nI = 0; nI < nLen; ++nI)
            {
                if (!nI || rFormat[nI-1]!='\\')
                {
                    if (rFormat[nI]=='\"')
                        rFormat = rFormat.replaceAt(nI, 1, "\'");
                    else if (rFormat[nI]=='\'')
                        rFormat = rFormat.replaceAt(nI, 1, "\"");
                }
            }
        }

    }
}

/* vim:set shiftwidth=4 softtabstop=4 expandtab: */