SEGYImport.cpp 92.9 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
/****************************************************************************
** Copyright 2019 The Open Group
** Copyright 2019 Bluware, Inc.
**
** Licensed 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
**
** Unless required by applicable law or agreed to in writing, software
** distributed under the License is distributed on an "AS IS" BASIS,
** WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
** See the License for the specific language governing permissions and
** limitations under the License.
****************************************************************************/

Jørgen Lind's avatar
Jørgen Lind committed
18
#define _CRT_SECURE_NO_WARNINGS 1
19
#include <SEGYUtils/SEGYFileInfo.h>
20
#include "IO/File.h"
21
#include "VDS/Hash.h"
22
#include <SEGYUtils/DataProvider.h>
23
#include <SEGYUtils/TraceDataManager.h>
24
25

#include <OpenVDS/OpenVDS.h>
26
#include <OpenVDS/MetadataContainer.h>
27
28
29
30
31
32
#include <OpenVDS/VolumeDataLayoutDescriptor.h>
#include <OpenVDS/VolumeDataAxisDescriptor.h>
#include <OpenVDS/VolumeDataChannelDescriptor.h>
#include <OpenVDS/VolumeDataAccess.h>
#include <OpenVDS/Range.h>
#include <OpenVDS/VolumeDataLayout.h>
33
#include <OpenVDS/KnownMetadata.h>
Morten Ofstad's avatar
Morten Ofstad committed
34
#include <OpenVDS/GlobalMetadataCommon.h>
35

36
37
#include "IO/IOManager.h"

38
39
40
41
42
43
44
45
46
47
#include <mutex>
#include <cstdlib>
#include <climits>
#include <cassert>
#include <algorithm>

#include "cxxopts.hpp"
#include <json/json.h>
#include <fmt/format.h>

48
#include "SplitUrl.h"
49
#include <PrintHelpers.h>
50

51
#include <chrono>
52
#include <numeric>
53

Jørgen Lind's avatar
Jørgen Lind committed
54
55
56
57
58
59
#if defined(WIN32)
#undef WIN32_LEAN_AND_MEAN // avoid warnings if defined on command line
#define WIN32_LEAN_AND_MEAN 1
#define NOMINMAX 1
#include <io.h>
#include <windows.h>
60
61
62
63
64
65
66
67
68

int64_t GetTotalSystemMemory()
{
    MEMORYSTATUSEX status;
    status.dwLength = sizeof(status);
    GlobalMemoryStatusEx(&status);
    return int64_t(status.ullTotalPhys);
}

Jørgen Lind's avatar
Jørgen Lind committed
69
70
#else
#include <unistd.h>
71
72
73
74
75
76
77

int64_t GetTotalSystemMemory()
{
    long pages = sysconf(_SC_PHYS_PAGES);
    long page_size = sysconf(_SC_PAGE_SIZE);
    return int64_t(pages) * int64_t(page_size);
}
Jørgen Lind's avatar
Jørgen Lind committed
78
79
#endif

80
81
82
83
84
85
inline char asciitolower(char in) {
  if (in <= 'Z' && in >= 'A')
    return in - ('Z' - 'z');
  return in;
}

86
DataProvider CreateDataProviderFromFile(const std::string &filename, OpenVDS::Error &error)
Jørgen Lind's avatar
Jørgen Lind committed
87
88
89
90
91
92
93
{
  std::unique_ptr<OpenVDS::File> file(new OpenVDS::File());
  if (!file->Open(filename, false, false, false, error))
    return DataProvider((OpenVDS::File *)nullptr);
  return DataProvider(file.release());
}

94
DataProvider CreateDataProviderFromOpenOptions(const std::string &url, const std::string &connectionString, const std::string &objectId, OpenVDS::Error &error)
Jørgen Lind's avatar
Jørgen Lind committed
95
{
Jørgen Lind's avatar
Jørgen Lind committed
96
  std::unique_ptr<OpenVDS::IOManager> ioManager(OpenVDS::IOManager::CreateIOManager(url, connectionString, OpenVDS::IOManager::AccessPattern::ReadOnly, error));
Jørgen Lind's avatar
Jørgen Lind committed
97
  if (error.code)
98
99
    return DataProvider((OpenVDS::IOManager *)nullptr, "", error);
  return DataProvider(ioManager.release(), objectId, error);
Jørgen Lind's avatar
Jørgen Lind committed
100
101
}

102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
DataProvider CreateDataProvider(const std::string& name, const std::string& connection, OpenVDS::Error& error)
{
  if (OpenVDS::IsSupportedProtocol(name))
  {
    std::string dirname;
    std::string basename;
    std::string parameters;
    splitUrl(name, dirname, basename, parameters, error);
    if (error.code)
      return DataProvider(nullptr);

    std::string url = dirname + parameters;
    return CreateDataProviderFromOpenOptions(url, connection, basename, error);
  }
  else
  {
    return CreateDataProviderFromFile(name, error);
  }
  return DataProvider(nullptr);
}

Jørgen Lind's avatar
Jørgen Lind committed
123
static std::vector<DataProvider> CreateDataProviders(const std::vector<std::string> &fileNames, const std::string &connection, OpenVDS::Error &error, std::string &errorFileName)
124
125
126
127
128
129
130
131
132
133
{
  std::vector<DataProvider>
    dataProviders;

  for (const auto& fileName : fileNames)
  {
    dataProviders.push_back(CreateDataProvider(fileName, connection, error));

    if (error.code != 0)
    {
Jørgen Lind's avatar
Jørgen Lind committed
134
      errorFileName = fileName;
135
136
137
138
139
140
141
      dataProviders.clear();
      break;
    }
  }
  return dataProviders;
}

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
Json::Value
SerializeSEGYBinInfo(SEGYBinInfo const& binInfo)
{
  Json::Value
    jsonBinInfo;

  jsonBinInfo["inlineNumber"] = binInfo.m_inlineNumber;
  jsonBinInfo["crosslineNumber"] = binInfo.m_crosslineNumber;

  jsonBinInfo["ensembleXCoordinate"] = binInfo.m_ensembleXCoordinate;
  jsonBinInfo["ensembleYCoordinate"] = binInfo.m_ensembleYCoordinate;

  return jsonBinInfo;
}

Json::Value
SerializeSEGYSegmentInfo(SEGYSegmentInfo const& segmentInfo)
{
  Json::Value
    jsonSegmentInfo;

  jsonSegmentInfo["primaryKey"] = segmentInfo.m_primaryKey;
  jsonSegmentInfo["traceStart"] = segmentInfo.m_traceStart;
  jsonSegmentInfo["traceStop"] = segmentInfo.m_traceStop;

  jsonSegmentInfo["binInfoStart"] = SerializeSEGYBinInfo(segmentInfo.m_binInfoStart);
  jsonSegmentInfo["binInfoStop"] = SerializeSEGYBinInfo(segmentInfo.m_binInfoStop);

  return jsonSegmentInfo;
}

std::string
ToString(SEGY::Endianness endiannness)
{
  switch (endiannness)
  {
  case SEGY::Endianness::BigEndian:  return "BigEndian";
  case SEGY::Endianness::LittleEndian: return "LittleEndian";
  default:
    assert(0); return "";
  }
}

std::string
ToString(SEGY::FieldWidth fieldWidth)
{
  switch (fieldWidth)
  {
  case SEGY::FieldWidth::TwoByte:  return "TwoByte";
  case SEGY::FieldWidth::FourByte: return "FourByte";
  default:
    assert(0); return "";
  }
}

Json::Value
SerializeSEGYHeaderField(SEGY::HeaderField const& headerField)
{
  Json::Value
    jsonHeaderField(Json::ValueType::arrayValue);

  jsonHeaderField.append(headerField.byteLocation);
  jsonHeaderField.append(ToString(headerField.fieldWidth));

  return jsonHeaderField;
}

Json::Value
Jim King's avatar
Jim King committed
210
SerializeSEGYFileInfo(SEGYFileInfo const& fileInfo, const int fileIndex)
211
212
213
214
215
216
217
218
{
  Json::Value
    jsonFileInfo;

  jsonFileInfo["persistentID"] = fmt::format("{:X}", fileInfo.m_persistentID);
  jsonFileInfo["headerEndianness"] = ToString(fileInfo.m_headerEndianness);
  jsonFileInfo["dataSampleFormatCode"] = (int)fileInfo.m_dataSampleFormatCode;
  jsonFileInfo["sampleCount"] = fileInfo.m_sampleCount;
219
  jsonFileInfo["startTime"] = fileInfo.m_startTimeMilliseconds;
220
  jsonFileInfo["sampleInterval"] = fileInfo.m_sampleIntervalMilliseconds;
Jim King's avatar
Jim King committed
221
  jsonFileInfo["traceCount"] = fileInfo.m_traceCounts[fileIndex];
222
223
224
225
226
227
  jsonFileInfo["primaryKey"] = SerializeSEGYHeaderField(fileInfo.m_primaryKey);
  jsonFileInfo["secondaryKey"] = SerializeSEGYHeaderField(fileInfo.m_secondaryKey);

  Json::Value
    jsonSegmentInfoArray(Json::ValueType::arrayValue);

Jim King's avatar
Jim King committed
228
  for (auto const& segmentInfo : fileInfo.m_segmentInfoLists[fileIndex])
229
230
231
232
233
234
235
236
237
238
239
240
  {
    jsonSegmentInfoArray.append(SerializeSEGYSegmentInfo(segmentInfo));
  }

  jsonFileInfo["segmentInfo"] = jsonSegmentInfoArray;

  return jsonFileInfo;
}

std::map<std::string, SEGY::HeaderField>
g_traceHeaderFields =
{
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
 { "tracesequencenumber",           SEGY::TraceHeader::TraceSequenceNumberHeaderField },
 { "tracesequencenumberwithinfile", SEGY::TraceHeader::TraceSequenceNumberWithinFileHeaderField },
 { "energysourcepointnumber",       SEGY::TraceHeader::EnergySourcePointNumberHeaderField },
 { "ensemblenumber",                SEGY::TraceHeader::EnsembleNumberHeaderField },
 { "tracenumberwithinensemble",     SEGY::TraceHeader::TraceNumberWithinEnsembleHeaderField },
 { "traceidentificationcode",       SEGY::TraceHeader::TraceIdentificationCodeHeaderField },
 { "coordinatescale",               SEGY::TraceHeader::CoordinateScaleHeaderField },
 { "sourcexcoordinate",             SEGY::TraceHeader::SourceXCoordinateHeaderField },
 { "sourceycoordinate",             SEGY::TraceHeader::SourceYCoordinateHeaderField },
 { "groupxcoordinate",              SEGY::TraceHeader::GroupXCoordinateHeaderField },
 { "groupycoordinate",              SEGY::TraceHeader::GroupYCoordinateHeaderField },
 { "coordinateunits",               SEGY::TraceHeader::CoordinateUnitsHeaderField },
 { "starttime",                     SEGY::TraceHeader::StartTimeHeaderField },
 { "numsamples",                    SEGY::TraceHeader::NumSamplesHeaderField },
 { "sampleinterval",                SEGY::TraceHeader::SampleIntervalHeaderField },
 { "ensemblexcoordinate",           SEGY::TraceHeader::EnsembleXCoordinateHeaderField },
 { "ensembleycoordinate",           SEGY::TraceHeader::EnsembleYCoordinateHeaderField },
 { "inlinenumber",                  SEGY::TraceHeader::InlineNumberHeaderField },
 { "crosslinenumber",               SEGY::TraceHeader::CrosslineNumberHeaderField },
 { "receiver",                      SEGY::TraceHeader::ReceiverHeaderField },
 { "offset",                        SEGY::TraceHeader::OffsetHeaderField }
262
263
264
265
266
};

std::map<std::string, std::string>
g_aliases =
{
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
 { "inline",              "inlinenumber" },
 { "crossline",           "crosslinenumber" },
 { "shot",                "energysourcepointnumber" },
 { "sp",                  "energysourcepointnumber" },
 { "cdp",                 "ensemblenumber" },
 { "cmp",                 "ensemblenumber" },
 { "easting",             "ensemblexcoordinate" },
 { "northing",            "ensembleycoordinate" },
 { "cdpxcoordinate",      "ensemblexcoordinate" },
 { "cdpycoordinate",      "ensembleycoordinate" },
 { "cdp-x",               "ensemblexcoordinate" },
 { "cdp-y",               "ensembleycoordinate" },
 { "source-x",            "sourcexcoordinate" },
 { "source-y",            "sourceycoordinate" },
 { "group-x",             "groupxcoordinate" },
 { "group-y",             "groupycoordinate" },
 { "receiverxcoordinate", "groupxcoordinate" },
 { "receiverycoordinate", "groupycoordinate" },
 { "receiver-x",          "groupxcoordinate" },
 { "receiver-y",          "groupycoordinate" },
 { "scalar",              "coordinatescale" }
288
289
290
291
292
};

void
ResolveAlias(std::string& fieldName)
{
293
294

  std::transform(fieldName.begin(), fieldName.end(), fieldName.begin(), asciitolower);
295
296
297
298
299
300
301
302
303
304
305
306
  if (g_aliases.find(fieldName) != g_aliases.end())
  {
    fieldName = g_aliases[fieldName];
  }
}

SEGY::Endianness
EndiannessFromJson(Json::Value const& jsonEndianness)
{
  std::string
    endiannessString = jsonEndianness.asString();

307
308
  std::transform(endiannessString.begin(), endiannessString.end(), endiannessString.begin(), asciitolower);
  if (endiannessString == "bigendian")
309
310
311
  {
    return SEGY::Endianness::BigEndian;
  }
312
  else if (endiannessString == "littleendian")
313
314
315
316
317
318
319
320
321
322
323
324
  {
    return SEGY::Endianness::LittleEndian;
  }

  throw Json::Exception("Illegal endianness");
}

SEGY::FieldWidth
FieldWidthFromJson(Json::Value const& jsonFieldWidth)
{
  std::string
    fieldWidthString = jsonFieldWidth.asString();
325
  std::transform(fieldWidthString.begin(), fieldWidthString.end(), fieldWidthString.begin(), asciitolower);
326

327
  if (fieldWidthString == "twobyte")
328
329
330
  {
    return SEGY::FieldWidth::TwoByte;
  }
331
  else if (fieldWidthString == "fourbyte")
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
  {
    return SEGY::FieldWidth::FourByte;
  }

  throw Json::Exception("Illegal field width");
}

SEGY::HeaderField
HeaderFieldFromJson(Json::Value const& jsonHeaderField)
{
  int
    bytePosition = jsonHeaderField[0].asInt();

  SEGY::FieldWidth
    fieldWidth = FieldWidthFromJson(jsonHeaderField[1]);

  if (bytePosition < 1 || bytePosition > SEGY::TraceHeaderSize - ((fieldWidth == SEGY::FieldWidth::TwoByte) ? 2 : 4))
  {
    throw Json::Exception(std::string("Illegal field definition: ") + jsonHeaderField.toStyledString());
  }

  return SEGY::HeaderField(bytePosition, fieldWidth);
}

bool
357
ParseHeaderFormatFile(DataProvider &dataProvider, std::map<std::string, SEGY::HeaderField>& traceHeaderFields, SEGY::Endianness& headerEndianness, OpenVDS::Error &error)
358
{
359
  int64_t dataSize = dataProvider.Size(error);
360
361
362
363
364
365

  if (error.code != 0)
  {
    return false;
  }

366
  if (dataSize > INT_MAX)
367
368
369
370
371
  {
    return false;
  }

  std::unique_ptr<char[]>
372
    buffer(new char[dataSize]);
373

374
  dataProvider.Read(buffer.get(), 0, (int32_t)dataSize, error);
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397

  if (error.code != 0)
  {
    return false;
  }

  try
  {
    Json::CharReaderBuilder
      rbuilder;

    rbuilder["collectComments"] = false;

    std::string
      errs;

    std::unique_ptr<Json::CharReader>
      reader(rbuilder.newCharReader());

    Json::Value
      root;

    bool
398
      success = reader->parse(buffer.get(), buffer.get() + dataSize, &root, &errs);
399
400
401
402
403
404
405
406
407
408
409

    if (!success)
    {
      throw Json::Exception(errs);
    }

    for (std::string const& fieldName : root.getMemberNames())
    {
      std::string canonicalFieldName = fieldName;
      ResolveAlias(canonicalFieldName);

410
      if (canonicalFieldName == "endianness")
411
412
413
414
415
416
417
418
419
      {
        headerEndianness = EndiannessFromJson(root[fieldName]);
      }
      else
      {
        traceHeaderFields[canonicalFieldName] = HeaderFieldFromJson(root[fieldName]);
      }
    }
  }
Jørgen Lind's avatar
Jørgen Lind committed
420
  catch (Json::Exception &e)
421
  {
422
423
    error.code = -1;
    error.string = e.what();
424
425
426
427
428
429
    return false;
  }

  return true;
}

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
bool OnlyDigits(const std::string& str)
{
  for (auto a : str)
  {
    if (a < '0' || a > '9')
      return false;
  }
  return true;
}

bool
ParseHeaderFieldArgs(const std::vector<std::string> &header_fields_args, std::map<std::string, SEGY::HeaderField>& traceHeaderFields, SEGY::Endianness& headerEndianness, OpenVDS::Error& error)
{
  for (auto& header_field : header_fields_args)
  {
    if (header_field.empty())
    {
      error.code = -1;
      error.string = "Cannot parse empty header-field";
      return false;
    }
    auto it = std::find(header_field.begin(), header_field.end(), '=');
    if (it == header_field.end())
    {
      error.code = -1;
      error.string = fmt::format("Failed to parse header-field {}.", header_field);
      return false;
    }
    std::string header_name(header_field.begin(), it);
    if (it + 1 == header_field.end())
    {
      error.code = -1;
      error.string = fmt::format("Can not find value for header-field {}.", header_name);
      return false;
    }
    std::string header_value(it + 1, header_field.end());
    auto min_delimiter = std::find(header_value.begin(), header_value.end(), '-');
    int field_width = -1;
    int offset = -1;
    if (min_delimiter != header_value.end())
    {
      if (min_delimiter + 1 == header_value.end())
      {
        error.code = -1;
        error.string = fmt::format("unable to parse value for header-field {} with value {}.", header_name, header_value);
        return false;
      }
      std::string value_start(header_value.begin(), min_delimiter);
      std::string value_end(min_delimiter + 1, header_value.end());
      if (!OnlyDigits(value_start) || !OnlyDigits(value_end))
      {
        error.code = -1;
        error.string = fmt::format("unable to parse header-field {} value range {}.", header_name, header_value);
        return false;
      }
      int value_start_value = atoi(value_start.c_str());
      int value_end_value = atoi(value_end.c_str());
      offset = value_start_value;
      field_width = value_end_value - value_start_value;
    }
    else
    {
      auto colon_delimiter = std::find(header_value.begin(), header_value.end(), ':');
      std::string offset_str(header_value.begin(), colon_delimiter);
      if (!OnlyDigits(offset_str))
      {
        error.code = -1;
        error.string = fmt::format("unable to parse offset for header-field {}: {}.", header_name, header_value);
        return false;
      }
      offset = atoi(offset_str.c_str());
      if (colon_delimiter < header_value.end() && colon_delimiter + 1 < header_value.end())
      {
        std::string width_str(colon_delimiter + 1, header_value.end());
        if (!OnlyDigits(width_str))
        {
          error.code = -1;
          error.string = fmt::format("unable to parse width specifier for header-field {}: {}.", header_name, width_str);
          return false;
        }
        field_width = atoi(width_str.c_str());
      }
    }
    if (offset < 0)
    {
      error.code = -1;
      error.string = fmt::format("unable to find offset for header-field {}: {}.", header_name, header_value);
      return false;
    }
    ResolveAlias(header_name);
    auto& traceHeaderField = traceHeaderFields[header_name];
    traceHeaderField.byteLocation = offset;
    if (field_width != -1)
    {
      if (field_width == 2)
      {
        traceHeaderField.fieldWidth = SEGY::FieldWidth::TwoByte;
      }
      else if (field_width == 4)
      {
        traceHeaderField.fieldWidth = SEGY::FieldWidth::FourByte;
      }
      else
      {
        error.code = -1;
        error.string = fmt::format("header-field {} has illegal field width of {}. Only widths of 2 or 4 are accepted.", header_name, field_width);
        return false;
      }
    }
  }
  return true;
}

543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
SEGYBinInfo
binInfoFromJson(Json::Value const& jsonBinInfo)
{
  int inlineNumber = jsonBinInfo["inlineNumber"].asInt();
  int crosslineNumber = jsonBinInfo["crosslineNumber"].asInt();
  double ensembleXCoordinate = jsonBinInfo["ensembleXCoordinate"].asDouble();
  double ensembleYCoordinate = jsonBinInfo["ensembleYCoordinate"].asDouble();

  return SEGYBinInfo(inlineNumber, crosslineNumber, ensembleXCoordinate, ensembleYCoordinate);
}

SEGYSegmentInfo
segmentInfoFromJson(Json::Value const& jsonSegmentInfo)
{
  int primaryKey = jsonSegmentInfo["primaryKey"].asInt();
  int traceStart = jsonSegmentInfo["traceStart"].asInt();
  int traceStop = jsonSegmentInfo["traceStop"].asInt();
  SEGYBinInfo binInfoStart = binInfoFromJson(jsonSegmentInfo["binInfoStart"]);
  SEGYBinInfo binInfoStop = binInfoFromJson(jsonSegmentInfo["binInfoStop"]);

  return SEGYSegmentInfo(primaryKey, traceStart, traceStop, binInfoStart, binInfoStop);
}

566
567
568
569
570
571
572
573
std::vector<int>
getOrderedSegmentListIndices(SEGYFileInfo const& fileInfo, size_t& globalTotalSegments)
{
  // Generate a list of indices that will traverse m_segyFileInfo.m_segmentInfoLists in primary key order, which
  // may be different from the order that the files were given.

  std::vector<int>
    orderedListIndices;
Jim King's avatar
Jim King committed
574
  size_t
575
576
    longestList = 0;
  globalTotalSegments = 0;
Jim King's avatar
Jim King committed
577
  for (size_t i = 0; i < fileInfo.m_segmentInfoLists.size(); ++i)
578
  {
Jim King's avatar
Jim King committed
579
    orderedListIndices.push_back(static_cast<int>(i));
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
    globalTotalSegments += fileInfo.m_segmentInfoLists[i].size();
    if (fileInfo.m_segmentInfoLists[i].size() > fileInfo.m_segmentInfoLists[longestList].size())
    {
      longestList = i;
    }
  }
  const bool
    isAscending = fileInfo.m_segmentInfoLists[longestList].front().m_binInfoStart.m_inlineNumber <= fileInfo.m_segmentInfoLists[longestList].back().m_binInfoStart.m_inlineNumber;
  auto
    comparator = [&](int i1, int i2)
  {
    const auto
      & v1 = fileInfo.m_segmentInfoLists[i1],
      & v2 = fileInfo.m_segmentInfoLists[i2];
    return isAscending ? v1.front().m_binInfoStart.m_inlineNumber < v2.front().m_binInfoStart.m_inlineNumber : v2.front().m_binInfoStart.m_inlineNumber < v1.front().m_binInfoStart.m_inlineNumber;
  };
  std::sort(orderedListIndices.begin(), orderedListIndices.end(), comparator);

  return orderedListIndices;
}

601
SEGYSegmentInfo const&
602
findRepresentativeSegment(SEGYFileInfo const& fileInfo, int& primaryStep, int& bestListIndex)
603
{
604
605
606
607
608
609
610
  // Since we give more weight to segments near the center of the data we need a sorted index of segment lists so that we can
  // traverse the lists in data order, instead of the arbitrary order given by the filename ordering.
  size_t
    globalTotalSegments = 0;
  auto
    orderedListIndices = getOrderedSegmentListIndices(fileInfo, globalTotalSegments);

611
  primaryStep = 0;
612
  bestListIndex = 0;
613

614
  float bestScore = 0.0f;
Jim King's avatar
Jim King committed
615
  size_t bestIndex = 0;
616
617

  int segmentPrimaryStep = 0;
618

619
620
621
622
  size_t
    globalOffset = 0;

  for (const auto listIndex : orderedListIndices)
623
  {
624
625
    const auto&
      segmentInfoList = fileInfo.m_segmentInfoLists[listIndex];
626

Jim King's avatar
Jim King committed
627
    for (size_t i = 0; i < segmentInfoList.size(); i++)
628
629
630
    {
      int64_t
        numTraces = (segmentInfoList[i].m_traceStop - segmentInfoList[i].m_traceStart + 1);
631

632
633
634
      // index of this segment within the entirety of segments from all input files
      const auto
        globalIndex = globalOffset + i;
635

636
637
      float
        multiplier = 1.5f - abs(globalIndex - (float)globalTotalSegments / 2) / (float)globalTotalSegments; // give 50% more importance to a segment in the middle of the dataset
638

639
640
      float
        score = float(numTraces) * multiplier;
641

642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
      if (score > bestScore)
      {
        bestScore = score;
        bestListIndex = listIndex;
        bestIndex = i;
      }

      // Updating the primary step with the step for the previous segment intentionally ignores the step of the last segment since it can be anomalous
      if (segmentPrimaryStep && (!primaryStep || std::abs(segmentPrimaryStep) < std::abs(primaryStep)))
      {
        primaryStep = segmentPrimaryStep;
      }

      if (i > 0)
      {
        segmentPrimaryStep = segmentInfoList[i].m_primaryKey - segmentInfoList[i - 1].m_primaryKey;
      }
659
    }
660
661

    globalOffset += segmentInfoList.size();
662
663
  }

664
665
666
  // If the primary step couldn't be determined, set it to the last step or 1
  primaryStep = primaryStep ? primaryStep : std::max(segmentPrimaryStep, 1);

667
  return fileInfo.m_segmentInfoLists[bestListIndex][bestIndex];
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
void
copySamples(const void* data, SEGY::BinaryHeader::DataSampleFormatCode dataSampleFormatCode, SEGY::Endianness endianness, float* target, int sampleStart, int sampleCount)
{
  if (dataSampleFormatCode == SEGY::BinaryHeader::DataSampleFormatCode::IBMFloat)
  {
    if(endianness == SEGY::Endianness::LittleEndian)
    {
      // Reverse endianness since Ibm2ieee expects big endian data
      const char * source = reinterpret_cast<const char*>((intptr_t)data + (size_t)sampleStart * 4);
      std::unique_ptr<char[]> temp(new char[sampleCount * 4]);
      for(int sample = 0; sample < sampleCount; sample++)
      {
        temp[sample * 4 + 0] = source[sample * 4 + 3];
        temp[sample * 4 + 1] = source[sample * 4 + 2];
        temp[sample * 4 + 2] = source[sample * 4 + 1];
        temp[sample * 4 + 3] = source[sample * 4 + 0];
      }
      SEGY::Ibm2ieee(target, temp.get(), sampleCount);
    }
    else
    {
      assert(endianness == SEGY::Endianness::BigEndian);
      SEGY::Ibm2ieee(target, reinterpret_cast<const uint32_t*>((intptr_t)data + (size_t)sampleStart * 4), sampleCount);
    }
  }
  else
  {
    assert(dataSampleFormatCode == SEGY::BinaryHeader::DataSampleFormatCode::IEEEFloat);
    if(endianness == SEGY::Endianness::LittleEndian)
    {
      SEGY::ConvertFromEndianness<SEGY::Endianness::LittleEndian>(target, reinterpret_cast<const char*>((intptr_t)data + (size_t)sampleStart * 4), sampleCount);
    }
    else
    {
      assert(endianness == SEGY::Endianness::BigEndian);
      SEGY::ConvertFromEndianness<SEGY::Endianness::BigEndian>(target, reinterpret_cast<const char*>((intptr_t)data + (size_t)sampleStart * 4), sampleCount);
    }
  }
}

710
bool
Jørgen Lind's avatar
Jørgen Lind committed
711
analyzeSegment(DataProvider &dataProvider, SEGYFileInfo const& fileInfo, SEGYSegmentInfo const& segmentInfo, float valueRangePercentile, OpenVDS::FloatRange& valueRange, int& fold, int& secondaryStep, const SEGY::SEGYType segyType, int& offsetStart, int& offsetEnd, int& offsetStep, bool jsonOutput, OpenVDS::Error& error)
712
{
713
  assert(segmentInfo.m_traceStop >= segmentInfo.m_traceStart && "A valid segment info should always have a stop trace greater or equal to the start trace");
714

715
  bool success = true;
716

717
718
719
  valueRange = OpenVDS::FloatRange(0.0f, 1.0f);
  secondaryStep = 0;
  fold = 1;
720
721
722
  offsetStart = 0;
  offsetEnd = 0;
  offsetStep = 0;
723

724
  const int traceByteSize = fileInfo.TraceByteSize();
725

726
727
728
  int64_t traceBufferStart = 0;
  int traceBufferSize = 0;
  std::unique_ptr<char[]> buffer;
729

730
  // Create min/max heaps for determining value range
731
  int heapSizeMax = int(((100.0f - valueRangePercentile) / 100.0f) * (segmentInfo.m_traceStop - segmentInfo.m_traceStart + 1) * fileInfo.m_sampleCount / 2) + 1;
732

733
734
735
736
737
738
739
740
741
742
743
744
  std::vector<float> minHeap, maxHeap;

  minHeap.reserve(heapSizeMax);
  maxHeap.reserve(heapSizeMax);

  // Allocate sample buffer for converting samples to float
  std::unique_ptr<float[]> sampleBuffer(new float[fileInfo.m_sampleCount]);
  float* samples = sampleBuffer.get();

  // Determine fold and secondary step
  int gatherSecondaryKey = 0, gatherFold = 0, gatherSecondaryStep = 0;

745
746
747
748
749
  bool
    hasPreviousGatherOffset = false;
  int
    previousGatherOffset = 0;

750
  for (int64_t trace = segmentInfo.m_traceStart; trace <= segmentInfo.m_traceStop; trace++)
751
  {
752
753
754
    if(trace - traceBufferStart >= traceBufferSize)
    {
      traceBufferStart = trace;
755
      traceBufferSize = (segmentInfo.m_traceStop - trace + 1) < 1000 ? int(segmentInfo.m_traceStop - trace + 1) : 1000;
756
757
758
759

      buffer.reset(new char[traceByteSize * traceBufferSize]);
      int64_t offset = SEGY::TextualFileHeaderSize + SEGY::BinaryFileHeaderSize + traceByteSize * traceBufferStart;
      success = dataProvider.Read(buffer.get(), offset, traceByteSize * traceBufferSize, error);
760

761
762
763
764
765
      if (!success)
      {
        break;
      }
    }
766

767
768
    const void *header = buffer.get() + traceByteSize * (trace - traceBufferStart);
    const void *data   = buffer.get() + traceByteSize * (trace - traceBufferStart) + SEGY::TraceHeaderSize;
769

770
    int tracePrimaryKey = SEGY::ReadFieldFromHeader(header, fileInfo.m_primaryKey, fileInfo.m_headerEndianness);
771
    int traceSecondaryKey = fileInfo.IsUnbinned() ? static_cast<int>(trace - segmentInfo.m_traceStart + 1) : SEGY::ReadFieldFromHeader(header, fileInfo.m_secondaryKey, fileInfo.m_headerEndianness);
772

773
    if(tracePrimaryKey != segmentInfo.m_primaryKey)
774
    {
775
      OpenVDS::printWarning(jsonOutput, "SEGY", fmt::format("trace {} has a primary key that doesn't match with the segment. This trace will be ignored.", segmentInfo.m_traceStart + trace));
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
      continue;
    }

    if(gatherFold > 0 && traceSecondaryKey == gatherSecondaryKey)
    {
      gatherFold++;
      fold = std::max(fold, gatherFold);
    }
    else
    {
      // Updating the secondary step with the step for the previous gather intentionally ignores the step of the last gather since it can be anomalous
      if(gatherSecondaryStep && (!secondaryStep || std::abs(gatherSecondaryStep) < std::abs(secondaryStep)))
      {
        secondaryStep = gatherSecondaryStep;
      }
791

792
      if(gatherFold > 0)
793
      {
794
        gatherSecondaryStep = traceSecondaryKey - gatherSecondaryKey;
795
      }
796
797
798
799
      gatherSecondaryKey = traceSecondaryKey;
      gatherFold = 1;
    }

800
    if (fileInfo.HasGatherOffset())
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
    {
      auto
        thisOffset = SEGY::ReadFieldFromHeader(header, g_traceHeaderFields["Offset"], fileInfo.m_headerEndianness);
      if (hasPreviousGatherOffset)
      {
        offsetStart = std::min(offsetStart, thisOffset);
        offsetEnd = std::max(offsetEnd, thisOffset);
        if (thisOffset != previousGatherOffset)
        {
          offsetStep = std::min(offsetStep, std::abs(thisOffset - previousGatherOffset));
        }
      }
      else
      {
        offsetStart = thisOffset;
        offsetEnd = thisOffset;
        offsetStep = INT32_MAX;
        hasPreviousGatherOffset = true;
      }
      previousGatherOffset = thisOffset;
    }

823
824
825
826
    // Update value range
    if (fileInfo.m_dataSampleFormatCode == SEGY::BinaryHeader::DataSampleFormatCode::IBMFloat || fileInfo.m_dataSampleFormatCode == SEGY::BinaryHeader::DataSampleFormatCode::IEEEFloat)
    {
      copySamples(data, fileInfo.m_dataSampleFormatCode, fileInfo.m_headerEndianness, samples, 0, fileInfo.m_sampleCount);
827
828
829

      for (int sample = 0; sample < fileInfo.m_sampleCount; sample++)
      {
Jørgen Lind's avatar
Jørgen Lind committed
830
        if (int(minHeap.size()) < heapSizeMax)
831
        {
832
833
834
835
836
837
838
          minHeap.push_back(samples[sample]);
          std::push_heap(minHeap.begin(), minHeap.end(), std::less<float>());
        }
        else if (samples[sample] < minHeap[0])
        {
          std::pop_heap(minHeap.begin(), minHeap.end(), std::less<float>());
          minHeap.back() = samples[sample];
839
          std::push_heap(minHeap.begin(), minHeap.end(), std::less<float>());
840
841
        }

Jørgen Lind's avatar
Jørgen Lind committed
842
        if (int(maxHeap.size()) < heapSizeMax)
843
        {
844
845
846
847
848
849
850
          maxHeap.push_back(samples[sample]);
          std::push_heap(maxHeap.begin(), maxHeap.end(), std::greater<float>());
        }
        else if (samples[sample] > maxHeap[0])
        {
          std::pop_heap(maxHeap.begin(), maxHeap.end(), std::greater<float>());
          maxHeap.back() = samples[sample];
851
          std::push_heap(maxHeap.begin(), maxHeap.end(), std::greater<float>());
852
853
854
        }
      }
    }
855
856
  }

857
  if (fileInfo.HasGatherOffset() && !fileInfo.IsUnbinned())
858
  {
859
860
861
862
863
    // check that offset start/end/step is consistent
    if (offsetStart + (fold - 1) * offsetStep != offsetEnd)
    {
      const auto
        msgFormat = "The detected gather offset start/end/step of '{0}/{1}/{2}' is not consistent with the detected fold of '{3}'. This usually indicates using the wrong header format for the input dataset.\n.";
Jørgen Lind's avatar
Jørgen Lind committed
864
865
      error.string = fmt::format(msgFormat, offsetStart, offsetEnd, offsetStep, fold);
      error.code = -1;
866
      return false;
867

868
    }
869
870
  }

871
872
873
874
875
876
877
  // If the secondary step couldn't be determined, set it to the last step or 1
  secondaryStep = secondaryStep ? secondaryStep : std::max(gatherSecondaryStep, 1);

  // Set value range
  if (!minHeap.empty())
  {
    assert(!maxHeap.empty());
878
879
880
881
882
883
884
885
886
887
888

    if (minHeap[0] != maxHeap[0])
    {
      valueRange = OpenVDS::FloatRange(minHeap[0], maxHeap[0]);
    }
    else
    {
      valueRange = OpenVDS::FloatRange(minHeap[0], minHeap[0] + 1.0f);
    }
  }

889
  return success;
890
891
892
}

bool
893
createSEGYMetadata(DataProvider &dataProvider, SEGYFileInfo const &fileInfo, OpenVDS::MetadataContainer& metadataContainer, SEGY::BinaryHeader::MeasurementSystem &measurementSystem, OpenVDS::Error& error)
894
895
896
897
898
{
  std::vector<uint8_t> textHeader(SEGY::TextualFileHeaderSize);
  std::vector<uint8_t> binaryHeader(SEGY::BinaryFileHeaderSize);

  // Read headers
Jørgen Lind's avatar
Jørgen Lind committed
899
900
  bool success = dataProvider.Read(textHeader.data(), 0, SEGY::TextualFileHeaderSize, error) &&
    dataProvider.Read(binaryHeader.data(), SEGY::TextualFileHeaderSize, SEGY::BinaryFileHeaderSize, error);
901
902
903
904
905
906
907
908

  if (!success) return false;

  // Create metadata
  metadataContainer.SetMetadataBLOB("SEGY", "TextHeader", textHeader);

  metadataContainer.SetMetadataBLOB("SEGY", "BinaryHeader", binaryHeader);

909
910
  metadataContainer.SetMetadataInt("SEGY", "Endianness", int(fileInfo.m_headerEndianness));
  metadataContainer.SetMetadataInt("SEGY", "DataSampleFormatCode", int(fileInfo.m_dataSampleFormatCode));
911
912

  measurementSystem = SEGY::BinaryHeader::MeasurementSystem(SEGY::ReadFieldFromHeader(binaryHeader.data(), SEGY::BinaryHeader::MeasurementSystemHeaderField, fileInfo.m_headerEndianness));
913
914
915
916
  return success;
}

void
917
createSurveyCoordinateSystemMetadata(SEGYFileInfo const& fileInfo, SEGY::BinaryHeader::MeasurementSystem measurementSystem, std::string const &crsWkt, OpenVDS::MetadataContainer& metadataContainer)
918
{
919
  if (fileInfo.m_segmentInfoLists.empty()) return;
920
921
922
923

  double inlineSpacing[2] = { 0, 0 };
  double crosslineSpacing[2] = { 0, 0 };

924
925
926
927
928
  size_t
    globalTotalSegments;
  auto
    orderedListIndices = getOrderedSegmentListIndices(fileInfo, globalTotalSegments);

929
930
931
  // Determine crossline spacing
  int countedCrosslineSpacings = 0;

932
  for (size_t listIndex : orderedListIndices)
933
  {
934
935
    const auto
      & segmentInfoList = fileInfo.m_segmentInfoLists[listIndex];
936

937
938
939
940
941
    for (auto const& segmentInfo : segmentInfoList)
    {
      int crosslineCount = segmentInfo.m_binInfoStop.m_crosslineNumber - segmentInfo.m_binInfoStart.m_crosslineNumber;

      if (crosslineCount == 0 || segmentInfo.m_binInfoStart.m_inlineNumber != segmentInfo.m_binInfoStop.m_inlineNumber) continue;
942

943
      double segmentCrosslineSpacing[3];
944

945
946
      segmentCrosslineSpacing[0] = (segmentInfo.m_binInfoStop.m_ensembleXCoordinate - segmentInfo.m_binInfoStart.m_ensembleXCoordinate) / crosslineCount;
      segmentCrosslineSpacing[1] = (segmentInfo.m_binInfoStop.m_ensembleYCoordinate - segmentInfo.m_binInfoStart.m_ensembleYCoordinate) / crosslineCount;
947

948
949
      crosslineSpacing[0] += segmentCrosslineSpacing[0];
      crosslineSpacing[1] += segmentCrosslineSpacing[1];
950

951
952
      countedCrosslineSpacings++;
    }
953
954
955
956
957
958
959
960
961
962
963
964
965
966
  }

  if (countedCrosslineSpacings > 0)
  {
    crosslineSpacing[0] /= countedCrosslineSpacings;
    crosslineSpacing[1] /= countedCrosslineSpacings;
  }
  else
  {
    crosslineSpacing[0] = 0;
    crosslineSpacing[1] = 1;
  }

  // Determine inline spacing
967
968
  SEGYSegmentInfo const& firstSegmentInfo = fileInfo.m_segmentInfoLists[orderedListIndices.front()].front();
  SEGYSegmentInfo const& lastSegmentInfo = fileInfo.m_segmentInfoLists[orderedListIndices.back()].back();
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

  if (firstSegmentInfo.m_binInfoStart.m_inlineNumber != lastSegmentInfo.m_binInfoStart.m_inlineNumber)
  {
    int inlineNunberDelta = lastSegmentInfo.m_binInfoStart.m_inlineNumber - firstSegmentInfo.m_binInfoStart.m_inlineNumber;
    int crosslineNunberDelta = lastSegmentInfo.m_binInfoStart.m_crosslineNumber - firstSegmentInfo.m_binInfoStart.m_crosslineNumber;

    double offset[2] = { crosslineSpacing[0] * crosslineNunberDelta,
               crosslineSpacing[1] * crosslineNunberDelta };

    inlineSpacing[0] = (lastSegmentInfo.m_binInfoStart.m_ensembleXCoordinate - firstSegmentInfo.m_binInfoStart.m_ensembleXCoordinate - offset[0]) / inlineNunberDelta;
    inlineSpacing[1] = (lastSegmentInfo.m_binInfoStart.m_ensembleYCoordinate - firstSegmentInfo.m_binInfoStart.m_ensembleYCoordinate - offset[1]) / inlineNunberDelta;
  }
  else
  {
    // make square voxels
    inlineSpacing[0] = crosslineSpacing[1];
    inlineSpacing[1] = -crosslineSpacing[0];
  }

  // Determine origin
  double origin[2];

  origin[0] = firstSegmentInfo.m_binInfoStart.m_ensembleXCoordinate;
  origin[1] = firstSegmentInfo.m_binInfoStart.m_ensembleYCoordinate;

  origin[0] -= inlineSpacing[0] * firstSegmentInfo.m_binInfoStart.m_inlineNumber;
  origin[1] -= inlineSpacing[1] * firstSegmentInfo.m_binInfoStart.m_inlineNumber;

  origin[0] -= crosslineSpacing[0] * firstSegmentInfo.m_binInfoStart.m_crosslineNumber;
  origin[1] -= crosslineSpacing[1] * firstSegmentInfo.m_binInfoStart.m_crosslineNumber;

  // Set coordinate system
For faster browsing, not all history is shown. View entire blame