Codebase list ciftilib / fresh-snapshots/main src / CiftiFile.cxx
fresh-snapshots/main

Tree @fresh-snapshots/main (Download .tar.gz)

CiftiFile.cxx @fresh-snapshots/mainraw · history · blame

  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
/*LICENSE_START*/ 
/*
 *  Copyright (c) 2014, Washington University School of Medicine
 *  All rights reserved.
 *
 *  Redistribution and use in source and binary forms, with or without modification,
 *  are permitted provided that the following conditions are met:
 *
 *  1. Redistributions of source code must retain the above copyright notice,
 *  this list of conditions and the following disclaimer.
 *
 *  2. Redistributions in binary form must reproduce the above copyright notice,
 *  this list of conditions and the following disclaimer in the documentation
 *  and/or other materials provided with the distribution.
 *
 *  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
 *  AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,
 *  THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
 *  ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
 *  FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 *  (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 *  LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
 *  ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 *  (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE,
 *  EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */

#include "CiftiFile.h"

#include "Common/CiftiAssert.h"
#include "Common/MultiDimArray.h"
#include "NiftiIO.h"

#ifdef CIFTILIB_USE_QT
    #include <QFileInfo>
#else
    //use boost filesystem, because cross-platform filesystem support with POSIX is absurd
    #define BOOST_FILESYSTEM_VERSION 3
    #include "boost/filesystem.hpp"
#endif

#include <iostream>

using namespace std;
using namespace boost;
using namespace cifti;

//private implementation classes, helpers
namespace
{
    class CiftiOnDiskImpl : public CiftiFile::WriteImplInterface
    {
        mutable NiftiIO m_nifti;//because file objects aren't stateless (current position), so reading "changes" them
        CiftiXML m_xml;//because we need to parse it to set up the dimensions anyway
    public:
        CiftiOnDiskImpl(const AString& filename);//read-only
        CiftiOnDiskImpl(const AString& filename, const CiftiXML& xml, const CiftiVersion& version, const bool& swapEndian,
                        const int16_t& datatype, const bool& rescale, const double& minval, const double& maxval);//make new empty file with read/write
        void getRow(float* dataOut, const std::vector<int64_t>& indexSelect, const bool& tolerateShortRead) const;
        void getColumn(float* dataOut, const int64_t& index) const;
        const CiftiXML& getCiftiXML() const { return m_xml; }
        AString getFilename() const { return m_nifti.getFilename(); }
        bool isSwapped() const { return m_nifti.getHeader().isSwapped(); }
        void setRow(const float* dataIn, const std::vector<int64_t>& indexSelect);
        void setColumn(const float* dataIn, const int64_t& index);
        void close();
    };
    
    class CiftiMemoryImpl : public CiftiFile::WriteImplInterface
    {
        MultiDimArray<float> m_array;
    public:
        CiftiMemoryImpl(const CiftiXML& xml);
        void getRow(float* dataOut, const std::vector<int64_t>& indexSelect, const bool& tolerateShortRead) const;
        void getColumn(float* dataOut, const int64_t& index) const;
        bool isInMemory() const { return true; }
        void setRow(const float* dataIn, const std::vector<int64_t>& indexSelect);
        void setColumn(const float* dataIn, const int64_t& index);
    };
    
    bool shouldSwap(const CiftiFile::ENDIAN& endian)
    {
        if (ByteSwapping::isBigEndian())
        {
            if (endian == CiftiFile::LITTLE) return true;
        } else {
            if (endian == CiftiFile::BIG) return true;
        }
        return false;//default for all other enum values is to write native endian
    }
    
    bool dontRewrite(const CiftiFile::ENDIAN& endian)
    {
        return (endian == CiftiFile::ANY);
    }
    
    AString pathToAbsolute(const AString& mypath)
    {
#ifdef CIFTILIB_USE_QT
        return QFileInfo(mypath).absoluteFilePath();
#else
#ifdef CIFTILIB_BOOST_NO_FSV3
	return filesystem::complete(AString_to_std_string(mypath)).file_string();
#else
        return filesystem::absolute(AString_to_std_string(mypath)).native();
#endif
#endif
    }
    
    AString pathToCanonical(const AString& mypath)
    {
#ifdef CIFTILIB_USE_QT
        return QFileInfo(mypath).canonicalFilePath();
#else
#ifdef CIFTILIB_BOOST_NO_FSV3
        return filesystem::complete(AString_to_std_string(mypath)).file_string();
#else
#ifdef CIFTILIB_BOOST_NO_CANONICAL
        filesystem::path temp = AString_to_std_string(mypath);
        if (!filesystem::exists(temp)) return "";
        return absolute(temp).normalize().native();
#else
        string temp = AString_to_std_string(mypath);
        if (!filesystem::exists(temp)) return "";
        return filesystem::canonical(temp).native();
#endif
#endif
#endif
    }
}

CiftiFile::ReadImplInterface::~ReadImplInterface()
{
}

CiftiFile::WriteImplInterface::~WriteImplInterface()
{
}

CiftiFile::CiftiFile()
{
    m_endianPref = NATIVE;
    setWritingDataTypeNoScaling();//default argument is float32
}

CiftiFile::CiftiFile(const AString& fileName)
{
    m_endianPref = NATIVE;
    setWritingDataTypeNoScaling();//default argument is float32
    openFile(fileName);
}

void CiftiFile::openFile(const AString& fileName)
{
    close();//to make sure it closes everything first, even if the open throws
    boost::shared_ptr<CiftiOnDiskImpl> newRead(new CiftiOnDiskImpl(pathToAbsolute(fileName)));//this constructor opens existing file read-only
    m_readingImpl = newRead;//it should be noted that if the constructor throws (if the file isn't readable), new guarantees the memory allocated for the object will be freed
    m_xml = newRead->getCiftiXML();
    m_dims = m_xml.getDimensions();
    m_onDiskVersion = m_xml.getParsedVersion();
}

void CiftiFile::setWritingFile(const AString& fileName, const CiftiVersion& writingVersion, const ENDIAN& endian)
{
    m_writingFile = pathToAbsolute(fileName);//always resolve paths as soon as they enter CiftiFile, in case some clown changes directory before writing data
    m_writingImpl.reset();//prevent writing to previous writing implementation, let the next set...() set up for writing
    m_onDiskVersion = writingVersion;
    m_endianPref = endian;
}

void CiftiFile::setWritingDataTypeNoScaling(const int16_t& type)
{
    m_writingDataType = type;//could do some validation here
    m_doWriteScaling = false;
    m_minScalingVal = -1.0;//these scaling values should never be used, but don't leave them uninitialized
    m_maxScalingVal = 1.0;
    m_writingImpl.reset();//prevent writing to previous writing implementation, let the next set...() set up for writing
}

void CiftiFile::setWritingDataTypeAndScaling(const int16_t& type, const double& minval, const double& maxval)
{
    m_writingDataType = type;//could do some validation here
    m_doWriteScaling = true;
    m_minScalingVal = minval;
    m_maxScalingVal = maxval;
    m_writingImpl.reset();//prevent writing to previous writing implementation, let the next set...() set up for writing
}

void CiftiFile::writeFile(const AString& fileName, const CiftiVersion& writingVersion, const ENDIAN& endian)
{
    if (m_readingImpl == NULL || m_dims.empty()) throw CiftiException("writeFile called on uninitialized CiftiFile");
    bool writeSwapped = shouldSwap(endian);
    AString canonicalFilename = pathToCanonical(fileName);//NOTE: returns EMPTY STRING for nonexistent file
    const CiftiOnDiskImpl* testImpl = dynamic_cast<CiftiOnDiskImpl*>(m_readingImpl.get());
    bool collision = false, hadWriter = (m_writingImpl != NULL);
    if (testImpl != NULL && canonicalFilename != "" && pathToCanonical(testImpl->getFilename()) == canonicalFilename)
    {//empty string test is so that we don't say collision if both are nonexistent - could happen if file is removed/unlinked while reading on some filesystems
        if (m_onDiskVersion == writingVersion && (dontRewrite(endian) || writeSwapped == testImpl->isSwapped())) return;//don't need to copy to itself
        collision = true;//we need to copy to memory temporarily
        boost::shared_ptr<WriteImplInterface> tempMemory(new CiftiMemoryImpl(m_xml));//because tempRead is a ReadImpl, can't be used to copy to
        copyImplData(m_readingImpl.get(), tempMemory.get(), m_dims);
        m_readingImpl = tempMemory;//we are about to make the old reading impl very unhappy, replace it so that if we get an error while writing, we hang onto the memory version
        m_writingImpl.reset();//and make it re-magic the writing implementation again if it tries to write again
    }
    boost::shared_ptr<WriteImplInterface> tempWrite(new CiftiOnDiskImpl(pathToAbsolute(fileName), m_xml, writingVersion, writeSwapped,
                                                                        m_writingDataType, m_doWriteScaling, m_minScalingVal, m_maxScalingVal));
    copyImplData(m_readingImpl.get(), tempWrite.get(), m_dims);
    if (collision)//if we rewrote the file, we need the handle to the new file, and to dump the temporary in-memory version
    {
        m_onDiskVersion = writingVersion;//also record the current version number
        m_readingImpl = tempWrite;//replace the temporary memory version
        if (hadWriter)//if it was in read-write mode
        {
            m_writingImpl = tempWrite;//set the writer too
        }
    }
}

void CiftiFile::close()
{
    if (m_writingImpl != NULL)
    {
        m_writingImpl->close();//only writing implementations should ever throw errors on close, and specifically only on-disk
    }
    m_writingImpl.reset();
    m_readingImpl.reset();
    m_dims.clear();
    m_xml = CiftiXML();
    m_writingFile = "";
    m_onDiskVersion = CiftiVersion();//for completeness, it gets reset on open anyway
    m_endianPref = NATIVE;//reset things to defaults
    setWritingDataTypeNoScaling();//default argument is float32
}

void CiftiFile::convertToInMemory()
{
    if (isInMemory()) return;
    if (m_readingImpl == NULL || m_dims.empty())//not set up yet
    {
        m_writingFile = "";//make sure it doesn't do on-disk when set...() is called
        return;
    }
    boost::shared_ptr<WriteImplInterface> tempWrite(new CiftiMemoryImpl(m_xml));//if we get an error while reading, free the memory immediately, and don't leave m_readingImpl and m_writingImpl pointing to different things
    copyImplData(m_readingImpl.get(), tempWrite.get(), m_dims);
    m_writingImpl = tempWrite;
    m_readingImpl = tempWrite;
}

bool CiftiFile::isInMemory() const
{
    if (m_readingImpl == NULL)
    {
        return (m_writingFile == "");//return what it would be if verifyWriteImpl() was called
    } else {
        return m_readingImpl->isInMemory();
    }
}

void CiftiFile::getRow(float* dataOut, const vector<int64_t>& indexSelect, const bool& tolerateShortRead) const
{
    if (m_dims.empty()) throw CiftiException("getRow called on uninitialized CiftiFile");
    if (m_readingImpl == NULL) return;//NOT an error because we are pretending to have a matrix already, while we are waiting for setRow to actually start writing the file
    m_readingImpl->getRow(dataOut, indexSelect, tolerateShortRead);
}

void CiftiFile::getColumn(float* dataOut, const int64_t& index) const
{
    if (m_dims.empty()) throw CiftiException("getColumn called on uninitialized CiftiFile");
    if (m_dims.size() != 2) throw CiftiException("getColumn called on non-2D CiftiFile");
    if (m_readingImpl == NULL) return;//NOT an error because we are pretending to have a matrix already, while we are waiting for setRow to actually start writing the file
    m_readingImpl->getColumn(dataOut, index);
}

void CiftiFile::setCiftiXML(const CiftiXML& xml, const bool useOldMetadata)
{
    if (xml.getNumberOfDimensions() == 0) throw CiftiException("setCiftiXML called with 0-dimensional CiftiXML");
    vector<int64_t> xmlDims = xml.getDimensions();
    for (size_t i = 0; i < xmlDims.size(); ++i)
    {
        if (xmlDims[i] < 1) throw CiftiException("cifti xml dimensions must be greater than zero");
    }
    m_readingImpl.reset();//drop old implementation, as it is now invalid due to XML (and therefore matrix size) change
    m_writingImpl.reset();
    if (useOldMetadata)
    {
        MetaData newmd = m_xml.getFileMetaData();//make a copy
        m_xml = xml;//because this will overwrite the metadata
        m_xml.setFileMetaData(newmd);
    } else {
        m_xml = xml;
    }
    m_dims = xmlDims;
}

void CiftiFile::setRow(const float* dataIn, const vector<int64_t>& indexSelect)
{
    verifyWriteImpl();
    m_writingImpl->setRow(dataIn, indexSelect);
}

void CiftiFile::setColumn(const float* dataIn, const int64_t& index)
{
    verifyWriteImpl();
    if (m_dims.size() != 2) throw CiftiException("setColumn called on non-2D CiftiFile");
    m_writingImpl->setColumn(dataIn, index);
}

//single-index functions
void CiftiFile::getRow(float* dataOut, const int64_t& index, const bool& tolerateShortRead) const
{
    if (m_dims.empty()) throw CiftiException("getRow called on uninitialized CiftiFile");
    if (m_dims.size() != 2) throw CiftiException("getRow with single index called on non-2D CiftiFile");
    if (m_readingImpl == NULL) return;//NOT an error because we are pretending to have a matrix already, while we are waiting for setRow to actually start writing the file
    vector<int64_t> tempvec(1, index);//could use a member if we need more speed
    m_readingImpl->getRow(dataOut, tempvec, tolerateShortRead);
}

void CiftiFile::setRow(const float* dataIn, const int64_t& index)
{
    verifyWriteImpl();
    if (m_dims.size() != 2) throw CiftiException("setRow with single index called on non-2D CiftiFile");
    vector<int64_t> tempvec(1, index);//could use a member if we need more speed
    m_writingImpl->setRow(dataIn, tempvec);
}
//*///end single-index functions

void CiftiFile::verifyWriteImpl()
{//this is where the magic happens - we want to emulate being a simple in-memory file, but actually be reading/writing on-disk when possible
    if (m_writingImpl != NULL) return;
    CiftiAssert(!m_dims.empty());//if the xml hasn't been set, then we can't do anything meaningful
    if (m_dims.empty()) throw CiftiException("setRow or setColumn attempted on uninitialized CiftiFile");
    if (m_writingFile == "")
    {
        if (m_readingImpl != NULL)
        {
            convertToInMemory();
        } else {
            m_writingImpl = boost::shared_ptr<CiftiMemoryImpl>(new CiftiMemoryImpl(m_xml));
        }
    } else {//NOTE: m_onDiskVersion gets set in setWritingFile
        if (m_readingImpl != NULL)
        {
            CiftiOnDiskImpl* testImpl = dynamic_cast<CiftiOnDiskImpl*>(m_readingImpl.get());
            if (testImpl != NULL)
            {
                AString canonicalCurrent = pathToCanonical(testImpl->getFilename());//returns "" if nonexistent, if unlinked while open
                if (canonicalCurrent != "" && canonicalCurrent == pathToCanonical(m_writingFile))//these were already absolute
                {
                    convertToInMemory();//save existing data in memory before we clobber file
                }
            }
        }
        m_writingImpl = boost::shared_ptr<CiftiOnDiskImpl>(new CiftiOnDiskImpl(m_writingFile, m_xml, m_onDiskVersion, shouldSwap(m_endianPref),
                                                                               m_writingDataType, m_doWriteScaling, m_minScalingVal, m_maxScalingVal));//this constructor makes new file for writing
        if (m_readingImpl != NULL)
        {
            copyImplData(m_readingImpl.get(), m_writingImpl.get(), m_dims);
        }
    }
    m_readingImpl = m_writingImpl;//read-only implementations are set up in specialized functions
}

void CiftiFile::copyImplData(const ReadImplInterface* from, WriteImplInterface* to, const vector<int64_t>& dims)
{
    vector<int64_t> iterateDims(dims.begin() + 1, dims.end());
    vector<float> scratchRow(dims[0]);
    for (MultiDimIterator<int64_t> iter(iterateDims); !iter.atEnd(); ++iter)
    {
        from->getRow(scratchRow.data(), *iter, false);
        to->setRow(scratchRow.data(), *iter);
    }
}

CiftiMemoryImpl::CiftiMemoryImpl(const CiftiXML& xml)
{
    CiftiAssert(xml.getNumberOfDimensions() != 0);
    m_array.resize(xml.getDimensions());
}

void CiftiMemoryImpl::getRow(float* dataOut, const vector<int64_t>& indexSelect, const bool&) const
{
    const float* ref = m_array.get(1, indexSelect);
    int64_t rowSize = m_array.getDimensions()[0];//we don't accept 0-D CiftiXML, so this will always work
    for (int64_t i = 0; i < rowSize; ++i)
    {
        dataOut[i] = ref[i];
    }
}

void CiftiMemoryImpl::getColumn(float* dataOut, const int64_t& index) const
{
    CiftiAssert(m_array.getDimensions().size() == 2);//otherwise, CiftiFile shouldn't have called this
    const float* ref = m_array.get(2, vector<int64_t>());//empty vector is intentional, only 2 dimensions exist, so no more to select from
    int64_t rowSize = m_array.getDimensions()[0];
    int64_t colSize = m_array.getDimensions()[1];
    CiftiAssert(index >= 0 && index < rowSize);//because we are doing the indexing math manually for speed
    for (int64_t i = 0; i < colSize; ++i)
    {
        dataOut[i] = ref[index + rowSize * i];
    }
}

void CiftiMemoryImpl::setRow(const float* dataIn, const vector<int64_t>& indexSelect)
{
    float* ref = m_array.get(1, indexSelect);
    int64_t rowSize = m_array.getDimensions()[0];//we don't accept 0-D CiftiXML, so this will always work
    for (int64_t i = 0; i < rowSize; ++i)
    {
        ref[i] = dataIn[i];
    }
}

void CiftiMemoryImpl::setColumn(const float* dataIn, const int64_t& index)
{
    CiftiAssert(m_array.getDimensions().size() == 2);//otherwise, CiftiFile shouldn't have called this
    float* ref = m_array.get(2, vector<int64_t>());//empty vector is intentional, only 2 dimensions exist, so no more to select from
    int64_t rowSize = m_array.getDimensions()[0];
    int64_t colSize = m_array.getDimensions()[1];
    CiftiAssert(index >= 0 && index < rowSize);//because we are doing the indexing math manually for speed
    for (int64_t i = 0; i < colSize; ++i)
    {
        ref[index + rowSize * i] = dataIn[i];
    }
}

CiftiOnDiskImpl::CiftiOnDiskImpl(const AString& filename)
{//opens existing file for reading
    m_nifti.openRead(filename);//read-only, so we don't need write permission to read a cifti file
    if (m_nifti.getNumComponents() != 1) throw CiftiException("complex or rgb datatype found in file '" + filename + "', these are not supported in cifti");
    const NiftiHeader& myHeader = m_nifti.getHeader();
    int numExts = (int)myHeader.m_extensions.size(), whichExt = -1;
    for (int i = 0; i < numExts; ++i)
    {
        if (myHeader.m_extensions[i]->m_ecode == NIFTI_ECODE_CIFTI)
        {
            whichExt = i;
            break;
        }
    }
    if (whichExt == -1) throw CiftiException("no cifti extension found in file '" + filename + "'");
    m_xml.readXML(myHeader.m_extensions[whichExt]->m_bytes);
    vector<int64_t> dimCheck = m_nifti.getDimensions();
    if (dimCheck.size() < 5) throw CiftiException("invalid dimensions in cifti file '" + filename + "'");
    for (int i = 0; i < 4; ++i)
    {
        if (dimCheck[i] != 1) throw CiftiException("non-singular dimension #" + AString_number(i + 1) + " in cifti file '" + filename + "'");
    }
    if (m_xml.getParsedVersion().hasReversedFirstDims())
    {
        while (dimCheck.size() < 6) dimCheck.push_back(1);//just in case
        int64_t temp = dimCheck[4];//note: nifti dim[5] is the 5th dimension, index 4 in this vector
        dimCheck[4] = dimCheck[5];
        dimCheck[5] = temp;
        m_nifti.overrideDimensions(dimCheck);
    }
    if (m_xml.getNumberOfDimensions() + 4 != (int)dimCheck.size()) throw CiftiException("XML does not match number of nifti dimensions in file " + filename + "'");
    for (int i = 4; i < (int)dimCheck.size(); ++i)
    {
        if (m_xml.getDimensionLength(i - 4) < 0)//CiftiXML will only let this happen with cifti-1
        {
            m_xml.getSeriesMap(i - 4).setLength(dimCheck[i]);//and only in a series map
        } else {
            if (m_xml.getDimensionLength(i - 4) != dimCheck[i])
            {
                throw CiftiException("xml and nifti header disagree on matrix dimensions");
            }
        }
    }
}

namespace
{
    void warnForBadExtension(const AString& filename, const CiftiXML& myXML)
    {
        char junk[16];
        int32_t intent_code = myXML.getIntentInfo(CiftiVersion(), junk);//use default writing version to check file extension, older version is missing some intent codes
        switch (intent_code)
        {
            default:
                cerr << "warning: unhandled cifti type in extension warning check, tell the developers what you just tried to do" << endl;
                CiftiAssert(0);//yes, let it fall through to "unknown" in release so that it at least looks for .nii
                //-fallthrough
            case 3000://unknown
                if (!AString_endsWith(filename, ".nii"))
                {
                    cerr << "warning: cifti file of nonstandard mapping combination '" << AString_to_std_string(filename) << "' should be saved ending in .<something>.nii" << endl;
                }
                if (AString_endsWith(filename, ".dconn.nii") ||
                    AString_endsWith(filename, ".dtseries.nii") ||
                    AString_endsWith(filename, ".pconn.nii") ||
                    AString_endsWith(filename, ".ptseries.nii") ||
                    AString_endsWith(filename, ".dscalar.nii") ||
                    AString_endsWith(filename, ".dfan.nii") ||
                    AString_endsWith(filename, ".fiberTemp.nii") ||
                    AString_endsWith(filename, ".dlabel.nii") ||
                    AString_endsWith(filename, ".pscalar.nii") ||
                    AString_endsWith(filename, ".pdconn.nii") ||
                    AString_endsWith(filename, ".dpconn.nii") ||
                    AString_endsWith(filename, ".pconnseries.nii") ||
                    AString_endsWith(filename, ".pconnscalar.nii"))
                {
                    cerr << "warning: cifti file of nonstandard mapping combination '" << AString_to_std_string(filename) << "' should NOT be saved using an already-used cifti extension, "
                            << "please choose a different, reasonable cifti extension of the form .<something>.nii" << endl;
                }
                break;
            case 3001:
                if (!AString_endsWith(filename, ".dconn.nii"))
                {
                    cerr << "warning: dense by dense cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .dconn.nii" << endl;
                }
                break;
            case 3002:
                if (!AString_endsWith(filename, ".dtseries.nii"))
                {
                    cerr << "warning: series by dense cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .dtseries.nii" << endl;
                }
                break;
            case 3003:
                if (!AString_endsWith(filename, ".pconn.nii"))
                {
                    cerr << "warning: parcels by parcels cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .pconn.nii" << endl;
                }
                break;
            case 3004:
                if (!AString_endsWith(filename, ".ptseries.nii"))
                {
                    cerr << "warning: series by parcels cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .ptseries.nii" << endl;
                }
                break;
            case 3006://3005 unused in practice
                if (!(AString_endsWith(filename, ".dscalar.nii") || AString_endsWith(filename, ".dfan.nii") || AString_endsWith(filename, ".fiberTEMP.nii")))
                {//there are additional special extensions in the standard for this mapping combination (specializations of scalar maps)
                    //also include workbench's fiberTEMP special extension
                    cerr << "warning: scalars by dense cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .dscalar.nii" << endl;
                }
                break;
            case 3007:
                if (!AString_endsWith(filename, ".dlabel.nii"))
                {
                    cerr << "warning: labels by dense cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .dlabel.nii" << endl;
                }
                break;
            case 3008:
                if (!AString_endsWith(filename, ".pscalar.nii"))
                {
                    cerr << "warning: scalars by parcels cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .pscalar.nii" << endl;
                }
                break;
            case 3009:
                if (!AString_endsWith(filename, ".pdconn.nii"))
                {
                    cerr << "warning: dense by parcels cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .pdconn.nii" << endl;
                }
                break;
            case 3010:
                if (!AString_endsWith(filename, ".dpconn.nii"))
                {
                    cerr << "warning: parcels by dense cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .dpconn.nii" << endl;
                }
                break;
            case 3011:
                if (!AString_endsWith(filename, ".pconnseries.nii"))
                {
                    cerr << "warning: parcels by parcels by series cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .pconnseries.nii" << endl;
                }
                break;
            case 3012:
                if (!AString_endsWith(filename, ".pconnscalar.nii"))
                {
                    cerr << "warning: parcels by parcels by scalar cifti file '" << AString_to_std_string(filename) << "' should be saved ending in .pconnscalar.nii" << endl;
                }
                break;
        }
    }
}

CiftiOnDiskImpl::CiftiOnDiskImpl(const AString& filename, const CiftiXML& xml, const CiftiVersion& version, const bool& swapEndian,
                                 const int16_t& datatype, const bool& rescale, const double& minval, const double& maxval)
{//starts writing new file
    warnForBadExtension(filename, xml);
    NiftiHeader outHeader;
    if (rescale)
    {
        outHeader.setDataTypeAndScaleRange(datatype, minval, maxval);
    } else {
        outHeader.setDataType(datatype);
    }
    if (outHeader.getNumComponents() != 1)
    {
        throw CiftiException("cifti cannot be written with multi-component nifti datatypes (i.e., complex, RGB)");
    }
    char intentName[16];
    int32_t intentCode = xml.getIntentInfo(version, intentName);
    outHeader.setIntent(intentCode, intentName);
    boost::shared_ptr<NiftiExtension> outExtension(new NiftiExtension());
    outExtension->m_ecode = NIFTI_ECODE_CIFTI;
    outExtension->m_bytes = xml.writeXMLToVector(version);
    outHeader.m_extensions.push_back(outExtension);
    vector<int64_t> matrixDims = xml.getDimensions();
    vector<int64_t> niftiDims(4, 1);//the reserved space and time dims
    niftiDims.insert(niftiDims.end(), matrixDims.begin(), matrixDims.end());
    if (version.hasReversedFirstDims())
    {
        vector<int64_t> headerDims = niftiDims;
        while (headerDims.size() < 6) headerDims.push_back(1);//just in case
        int64_t temp = headerDims[4];
        headerDims[4] = headerDims[5];
        headerDims[5] = temp;
        outHeader.setDimensions(headerDims);//give the header the reversed dimensions
        m_nifti.writeNew(filename, outHeader, 2, true, swapEndian);
        m_nifti.overrideDimensions(niftiDims);//and then tell the nifti reader to use the correct dimensions
    } else {
        outHeader.setDimensions(niftiDims);
        m_nifti.writeNew(filename, outHeader, 2, true, swapEndian);
    }
    m_xml = xml;
}

void CiftiOnDiskImpl::close()
{
    m_nifti.close();//lets this throw when there is a writing problem
}//don't bother resetting m_xml, this instance is about to be destroyed

void CiftiOnDiskImpl::getRow(float* dataOut, const vector<int64_t>& indexSelect, const bool& tolerateShortRead) const
{
    m_nifti.readData(dataOut, 5, indexSelect, tolerateShortRead);//5 means 4 reserved (space and time) plus the first cifti dimension
}

void CiftiOnDiskImpl::getColumn(float* dataOut, const int64_t& index) const
{
    CiftiAssert(m_xml.getNumberOfDimensions() == 2);//otherwise this shouldn't be called
    CiftiAssert(index >= 0 && index < m_xml.getDimensionLength(CiftiXML::ALONG_ROW));
    vector<int64_t> indexSelect(2);
    indexSelect[0] = index;
    int64_t colLength = m_xml.getDimensionLength(CiftiXML::ALONG_COLUMN);
    for (int64_t i = 0; i < colLength; ++i)//assume if they really want getColumn on disk, they don't want their pagecache obliterated, so read it 1 element at a time
    {
        indexSelect[1] = i;
        m_nifti.readData(dataOut + i, 4, indexSelect);//4 means just the 4 reserved dimensions, so 1 element of the matrix
    }
}

void CiftiOnDiskImpl::setRow(const float* dataIn, const vector<int64_t>& indexSelect)
{
    m_nifti.writeData(dataIn, 5, indexSelect);
}

void CiftiOnDiskImpl::setColumn(const float* dataIn, const int64_t& index)
{
    CiftiAssert(m_xml.getNumberOfDimensions() == 2);//otherwise this shouldn't be called
    CiftiAssert(index >= 0 && index < m_xml.getDimensionLength(CiftiXML::ALONG_ROW));
    vector<int64_t> indexSelect(2);
    indexSelect[0] = index;
    int64_t colLength = m_xml.getDimensionLength(CiftiXML::ALONG_COLUMN);
    for (int64_t i = 0; i < colLength; ++i)//don't do RMW, so write it 1 element at a time
    {
        indexSelect[1] = i;
        m_nifti.writeData(dataIn + i, 4, indexSelect);//4 means just the 4 reserved dimensions, so 1 element of the matrix
    }
}