| 1 | // ================================================================ // |
|---|
| 2 | // // |
|---|
| 3 | // File : Importer.cxx // |
|---|
| 4 | // Purpose : Genome importer core // |
|---|
| 5 | // // |
|---|
| 6 | // Coded by Ralf Westram (coder@reallysoft.de) in November 2006 // |
|---|
| 7 | // Institute of Microbiology (Technical University Munich) // |
|---|
| 8 | // http://www.arb-home.de/ // |
|---|
| 9 | // // |
|---|
| 10 | // ================================================================ // |
|---|
| 11 | |
|---|
| 12 | #include "tools.h" |
|---|
| 13 | #include "DBwriter.h" |
|---|
| 14 | #include <arbdb.h> |
|---|
| 15 | #include <arb_stdstr.h> |
|---|
| 16 | |
|---|
| 17 | using namespace std; |
|---|
| 18 | |
|---|
| 19 | // -------------------------------------------------------------------------------- |
|---|
| 20 | |
|---|
| 21 | static bool is_escaped(const string& str, size_t pos) { |
|---|
| 22 | // returns true, if position 'pos' in string 'str' is escaped by '\\' |
|---|
| 23 | |
|---|
| 24 | bool escaped = false; |
|---|
| 25 | if (pos != 0) { // pos 0 can't be escaped |
|---|
| 26 | if (str[pos-1] == '\\') { // is an escape before pos ? |
|---|
| 27 | escaped = !is_escaped(str, pos-1); // pos is escaped, if the escape isn't! |
|---|
| 28 | } |
|---|
| 29 | } |
|---|
| 30 | return escaped; |
|---|
| 31 | } |
|---|
| 32 | |
|---|
| 33 | FeatureLine::FeatureLine(const string& line) { |
|---|
| 34 | // start parsing at position 5 |
|---|
| 35 | string::size_type first_char = line.find_first_not_of(' ', 5); |
|---|
| 36 | |
|---|
| 37 | orgLine = line; |
|---|
| 38 | |
|---|
| 39 | if (first_char == 5) { // feature start |
|---|
| 40 | string::size_type behind_name = line.find_first_of(' ', first_char); |
|---|
| 41 | string::size_type rest_start = line.find_first_not_of(' ', behind_name); |
|---|
| 42 | |
|---|
| 43 | if (rest_start == string::npos) { |
|---|
| 44 | if (behind_name == string::npos) throw "Expected space behind feature name"; |
|---|
| 45 | throw "Expected some content behind feature name"; |
|---|
| 46 | } |
|---|
| 47 | |
|---|
| 48 | name = line.substr(first_char, behind_name-first_char); |
|---|
| 49 | rest = line.substr(rest_start); |
|---|
| 50 | type = FL_START; |
|---|
| 51 | } |
|---|
| 52 | else if (first_char >= 21) { // not feature start |
|---|
| 53 | if (first_char == 21 && line[first_char] == '/') { // qualifier start |
|---|
| 54 | string::size_type equal_pos = line.find_first_of('=', first_char); |
|---|
| 55 | if (equal_pos == string::npos) { |
|---|
| 56 | // qualifier w/o data (i.e. "/pseudo") |
|---|
| 57 | name = line.substr(first_char+1); |
|---|
| 58 | rest = "true"; |
|---|
| 59 | type = FL_QUALIFIER_NODATA; |
|---|
| 60 | } |
|---|
| 61 | else { |
|---|
| 62 | name = line.substr(first_char+1, equal_pos-first_char-1); |
|---|
| 63 | rest = line.substr(equal_pos+1); |
|---|
| 64 | |
|---|
| 65 | if (rest[0] == '"') { |
|---|
| 66 | size_t rlen = rest.length(); |
|---|
| 67 | |
|---|
| 68 | if (rlen == 1) { // special case: only one open quote behind qualifier |
|---|
| 69 | type = FL_QUALIFIER_QUOTE_OPENED; |
|---|
| 70 | } |
|---|
| 71 | else if (rest[rlen-1] == '"' && !is_escaped(rest, rlen-1)) { // closing non-escaped quote at eol |
|---|
| 72 | type = FL_QUALIFIER_QUOTED; |
|---|
| 73 | } |
|---|
| 74 | else { |
|---|
| 75 | type = FL_QUALIFIER_QUOTE_OPENED; |
|---|
| 76 | } |
|---|
| 77 | } |
|---|
| 78 | else { |
|---|
| 79 | type = FL_QUALIFIER; |
|---|
| 80 | } |
|---|
| 81 | } |
|---|
| 82 | } |
|---|
| 83 | else { // continued line |
|---|
| 84 | interpret_as_continued_line(); |
|---|
| 85 | } |
|---|
| 86 | } |
|---|
| 87 | else { |
|---|
| 88 | if (first_char == string::npos) { |
|---|
| 89 | throw "Expected feature line, found empty line"; |
|---|
| 90 | } |
|---|
| 91 | throw GBS_global_string("Expected feature line (first char at pos=%zu unexpected)", first_char); |
|---|
| 92 | } |
|---|
| 93 | } |
|---|
| 94 | |
|---|
| 95 | void FeatureLine::interpret_as_continued_line() { |
|---|
| 96 | rest = orgLine.substr(21); |
|---|
| 97 | if (rest[rest.length()-1] == '"') { |
|---|
| 98 | type = FL_CONTINUED_QUOTE_CLOSED; |
|---|
| 99 | } |
|---|
| 100 | else { |
|---|
| 101 | type = FL_CONTINUED; |
|---|
| 102 | } |
|---|
| 103 | } |
|---|
| 104 | |
|---|
| 105 | bool FeatureLine::reinterpret_as_continued_line() { |
|---|
| 106 | bool ok = false; |
|---|
| 107 | |
|---|
| 108 | if (type == FL_QUALIFIER || type == FL_QUALIFIER_NODATA) { |
|---|
| 109 | string::size_type first_char = orgLine.find_first_not_of(' ', 5); |
|---|
| 110 | if (first_char >= 21) { |
|---|
| 111 | interpret_as_continued_line(); |
|---|
| 112 | ok = true; |
|---|
| 113 | } |
|---|
| 114 | } |
|---|
| 115 | |
|---|
| 116 | return ok; |
|---|
| 117 | } |
|---|
| 118 | |
|---|
| 119 | // -------------------------------------------------------------------------------- |
|---|
| 120 | |
|---|
| 121 | Importer::Importer(LineReader& Flatfile, DBwriter& DB_writer, const MetaTag *meta_description) |
|---|
| 122 | : db_writer(DB_writer), |
|---|
| 123 | flatfile(Flatfile), |
|---|
| 124 | tagTranslator(meta_description), |
|---|
| 125 | expectedSeqLength(-1) |
|---|
| 126 | {} |
|---|
| 127 | |
|---|
| 128 | void Importer::warning(const char *msg) { |
|---|
| 129 | warnings.push_back(msg); |
|---|
| 130 | } |
|---|
| 131 | |
|---|
| 132 | FeatureLinePtr Importer::getFeatureTableLine() { |
|---|
| 133 | FeatureLinePtr fline; |
|---|
| 134 | |
|---|
| 135 | if (pushedFeatureLines.empty()) { // nothing on stack -> read new |
|---|
| 136 | string line; |
|---|
| 137 | if (readFeatureTableLine(line)) fline = new FeatureLine(line); |
|---|
| 138 | } |
|---|
| 139 | else { |
|---|
| 140 | fline = pushedFeatureLines.back(); |
|---|
| 141 | pushedFeatureLines.pop_back(); |
|---|
| 142 | } |
|---|
| 143 | return fline; |
|---|
| 144 | } |
|---|
| 145 | |
|---|
| 146 | FeatureLinePtr Importer::getUnwrappedFeatureTableLine() { |
|---|
| 147 | FeatureLinePtr fline = getFeatureTableLine(); |
|---|
| 148 | if (!fline.isNull()) { |
|---|
| 149 | if (fline->type & FL_META_CONTINUED) throw "Expected start of feature or qualifier"; |
|---|
| 150 | |
|---|
| 151 | if (0 == (fline->type & (FL_QUALIFIER_NODATA|FL_QUALIFIER_QUOTED))) { |
|---|
| 152 | // qualifier/featurestart may be wrapped |
|---|
| 153 | FeatureLinePtr next_fline = getFeatureTableLine(); |
|---|
| 154 | |
|---|
| 155 | while (!next_fline.isNull() && |
|---|
| 156 | fline->type != FL_QUALIFIER_QUOTED) // already seen closing quote |
|---|
| 157 | { |
|---|
| 158 | if ((next_fline->type&FL_META_CONTINUED) == 0) { |
|---|
| 159 | // special case: a wrapped line of a quoted qualifier may start with /xxx |
|---|
| 160 | // (in that case it is misinterpreted as qualifier start) |
|---|
| 161 | if (fline->type == FL_QUALIFIER_QUOTE_OPENED) { |
|---|
| 162 | if (!next_fline->reinterpret_as_continued_line()) { |
|---|
| 163 | throw "did not see end of quoted qualifier (instead found next qualifiert)"; |
|---|
| 164 | } |
|---|
| 165 | gi_assert(next_fline->type & FL_META_CONTINUED); |
|---|
| 166 | } |
|---|
| 167 | else { |
|---|
| 168 | break; |
|---|
| 169 | } |
|---|
| 170 | } |
|---|
| 171 | |
|---|
| 172 | if (next_fline->type == FL_CONTINUED_QUOTE_CLOSED) { |
|---|
| 173 | if (fline->type != FL_QUALIFIER_QUOTE_OPENED) throw "Unexpected closing quote"; |
|---|
| 174 | fline->type = FL_QUALIFIER_QUOTED; |
|---|
| 175 | } |
|---|
| 176 | else { |
|---|
| 177 | gi_assert(next_fline->type == FL_CONTINUED); |
|---|
| 178 | gi_assert(fline->type == FL_START || fline->type == FL_QUALIFIER || fline->type == FL_QUALIFIER_QUOTE_OPENED); |
|---|
| 179 | } |
|---|
| 180 | |
|---|
| 181 | fline->rest.append(next_fline->rest); |
|---|
| 182 | next_fline = getFeatureTableLine(); |
|---|
| 183 | } |
|---|
| 184 | |
|---|
| 185 | if (!next_fline.isNull()) backFeatureTableLine(next_fline); |
|---|
| 186 | } |
|---|
| 187 | } |
|---|
| 188 | return fline; |
|---|
| 189 | } |
|---|
| 190 | |
|---|
| 191 | FeaturePtr Importer::parseFeature() { |
|---|
| 192 | FeaturePtr feature; |
|---|
| 193 | FeatureLinePtr fline = getUnwrappedFeatureTableLine(); |
|---|
| 194 | |
|---|
| 195 | if (!fline.isNull()) { // found a feature table line |
|---|
| 196 | if (fline->type != FL_START) throw "Expected feature start"; |
|---|
| 197 | |
|---|
| 198 | feature = new Feature(fline->name, fline->rest); |
|---|
| 199 | |
|---|
| 200 | fline = getUnwrappedFeatureTableLine(); |
|---|
| 201 | while (!fline.isNull() && (fline->type & FL_META_QUALIFIER)) { |
|---|
| 202 | feature->addQualifiedEntry(fline->name, fline->rest); |
|---|
| 203 | fline = getUnwrappedFeatureTableLine(); |
|---|
| 204 | } |
|---|
| 205 | if (!fline.isNull()) backFeatureTableLine(fline); |
|---|
| 206 | } |
|---|
| 207 | |
|---|
| 208 | return feature; |
|---|
| 209 | } |
|---|
| 210 | |
|---|
| 211 | void Importer::parseFeatureTable() { |
|---|
| 212 | FeaturePtr feature = parseFeature(); |
|---|
| 213 | |
|---|
| 214 | while (!feature.isNull()) { |
|---|
| 215 | feature->expectLocationInSequence(expectedSeqLength); |
|---|
| 216 | feature->fixEmptyQualifiers(); |
|---|
| 217 | db_writer.writeFeature(*feature, expectedSeqLength); |
|---|
| 218 | feature = parseFeature(); |
|---|
| 219 | } |
|---|
| 220 | } |
|---|
| 221 | |
|---|
| 222 | void Importer::show_warnings(const string& import_of_what) { |
|---|
| 223 | if (!warnings.empty()) { |
|---|
| 224 | const char *what = import_of_what.c_str(); |
|---|
| 225 | stringVectorCRIter e = warnings.rend(); |
|---|
| 226 | for (stringVectorCRIter i = warnings.rbegin(); i != e; ++i) { |
|---|
| 227 | GB_warningf("Warning: %s: %s", what, i->c_str()); |
|---|
| 228 | } |
|---|
| 229 | warnings.clear(); |
|---|
| 230 | } |
|---|
| 231 | } |
|---|
| 232 | |
|---|
| 233 | |
|---|
| 234 | void Importer::import() { |
|---|
| 235 | try { |
|---|
| 236 | string line; |
|---|
| 237 | while (flatfile.getLine(line)) { |
|---|
| 238 | if (!line.empty()) { // silently skip empty lines before or after section |
|---|
| 239 | flatfile.backLine(line); |
|---|
| 240 | |
|---|
| 241 | // cleanup from import of previous section |
|---|
| 242 | gi_assert(pushedFeatureLines.empty()); // oops - somehow forgot a feature |
|---|
| 243 | pushedFeatureLines.clear(); |
|---|
| 244 | warnings.clear(); |
|---|
| 245 | |
|---|
| 246 | expectedSeqLength = 0; // reset expected seq. length |
|---|
| 247 | import_section(); |
|---|
| 248 | |
|---|
| 249 | gi_assert(warnings.empty()); |
|---|
| 250 | gi_assert(pushedFeatureLines.empty()); // oops - somehow forgot a feature |
|---|
| 251 | } |
|---|
| 252 | } |
|---|
| 253 | } |
|---|
| 254 | catch (const DBerror& err) { throw err.getMessage(); } |
|---|
| 255 | catch (const string& err) { throw flatfile.lineError(err); } |
|---|
| 256 | catch (const char *err) { throw flatfile.lineError(err); } |
|---|
| 257 | } |
|---|
| 258 | |
|---|
| 259 | void Importer::check_base_counters(const SequenceBuffer& seqData, const BaseCounter *headerCount) { |
|---|
| 260 | const BaseCounter& baseCounter = seqData.getBaseCounter(); |
|---|
| 261 | if (baseCounter.getCount(BC_ALL)<1) { |
|---|
| 262 | warning("Sequence data is empty (only metadata found)."); |
|---|
| 263 | } |
|---|
| 264 | if (!headerCount) { |
|---|
| 265 | gi_assert(dynamic_cast<GenebankImporter*>(this) != NULp); // this case shall only happen with genebank files. |
|---|
| 266 | warning("No 'BASE COUNT' found. Base counts have not been validated."); |
|---|
| 267 | } |
|---|
| 268 | else { |
|---|
| 269 | headerCount->expectEqual(baseCounter); |
|---|
| 270 | } |
|---|
| 271 | } |
|---|
| 272 | |
|---|
| 273 | // -------------------------------------------------------------------------------- |
|---|
| 274 | // Meta information definitions |
|---|
| 275 | // |
|---|
| 276 | // |
|---|
| 277 | // [ please keep the list of common entries in |
|---|
| 278 | // ../HELP_SOURCE/oldhelp/sp_info.hlp |
|---|
| 279 | // up to date! ] |
|---|
| 280 | |
|---|
| 281 | static MetaTag genebank_meta_description[] = { |
|---|
| 282 | { "LOCUS", "org_locus", MT_HEADER }, |
|---|
| 283 | |
|---|
| 284 | { "REFERENCE", "", MT_REF_START }, |
|---|
| 285 | { " AUTHORS", "author", MT_REF }, |
|---|
| 286 | { " TITLE", "title", MT_REF }, |
|---|
| 287 | { " CONSRTM", "refgrp", MT_REF }, |
|---|
| 288 | { " JOURNAL", "journal", MT_REF }, |
|---|
| 289 | { " PUBMED", "pubmed_id", MT_REF }, |
|---|
| 290 | { " MEDLINE", "medline_id", MT_REF }, |
|---|
| 291 | { " REMARK", "refremark", MT_REF }, |
|---|
| 292 | |
|---|
| 293 | { "DEFINITION", "definition", MT_BASIC }, |
|---|
| 294 | { "ACCESSION", "acc", MT_BASIC }, |
|---|
| 295 | { "VERSION", "version", MT_BASIC }, |
|---|
| 296 | { "DBLINK", "db_xref", MT_BASIC }, |
|---|
| 297 | { "KEYWORDS", "keywd", MT_BASIC }, |
|---|
| 298 | { "SOURCE", "full_name", MT_BASIC }, |
|---|
| 299 | { " ORGANISM", "tax", MT_BASIC }, |
|---|
| 300 | { "COMMENT", "comment", MT_BASIC }, |
|---|
| 301 | { "PROJECT", "projref", MT_BASIC }, |
|---|
| 302 | |
|---|
| 303 | { "FEATURES", "", MT_FEATURE_START }, |
|---|
| 304 | { "CONTIG", "", MT_CONTIG }, |
|---|
| 305 | { "BASE", "", MT_SEQUENCE_START }, // BASE COUNT (sometimes missing) |
|---|
| 306 | { "ORIGIN", "", MT_SEQUENCE_START }, // only used if BASE COUNT is missing |
|---|
| 307 | { "//", "", MT_END }, |
|---|
| 308 | |
|---|
| 309 | { "", "", MT_IGNORE }, // End of array |
|---|
| 310 | }; |
|---|
| 311 | |
|---|
| 312 | static MetaTag embl_meta_description[] = { |
|---|
| 313 | { "ID", "org_id", MT_HEADER }, |
|---|
| 314 | |
|---|
| 315 | { "RN", "", MT_REF_START }, |
|---|
| 316 | { "RA", "author", MT_REF }, |
|---|
| 317 | { "RC", "auth_comm", MT_REF }, |
|---|
| 318 | { "RG", "refgrp", MT_REF }, |
|---|
| 319 | { "RL", "journal", MT_REF }, |
|---|
| 320 | { "RP", "nuc_rp", MT_REF }, |
|---|
| 321 | { "RT", "title", MT_REF }, |
|---|
| 322 | { "RX", "", MT_REF_DBID }, // @@@ extract field 'pubmed_id' ? |
|---|
| 323 | |
|---|
| 324 | { "AC", "acc", MT_BASIC }, |
|---|
| 325 | { "AH", "assembly_header", MT_BASIC }, |
|---|
| 326 | { "AS", "assembly_info", MT_BASIC }, |
|---|
| 327 | { "CC", "comment", MT_BASIC }, |
|---|
| 328 | { "CO", "contig", MT_BASIC }, |
|---|
| 329 | { "DE", "description", MT_BASIC }, |
|---|
| 330 | { "DR", "db_xref", MT_BASIC }, |
|---|
| 331 | { "DT", "date", MT_BASIC }, |
|---|
| 332 | { "SV", "version", MT_BASIC }, |
|---|
| 333 | { "KW", "keywd", MT_BASIC }, |
|---|
| 334 | { "OS", "full_name", MT_BASIC }, |
|---|
| 335 | { "OC", "tax", MT_BASIC }, |
|---|
| 336 | { "OG", "organelle", MT_BASIC }, |
|---|
| 337 | { "PR", "projref", MT_BASIC }, |
|---|
| 338 | |
|---|
| 339 | { "FH", "", MT_FEATURE_START }, |
|---|
| 340 | { "FT", "", MT_FEATURE }, |
|---|
| 341 | { "SQ", "", MT_SEQUENCE_START }, |
|---|
| 342 | { "//", "", MT_END }, |
|---|
| 343 | |
|---|
| 344 | { "XX", "", MT_IGNORE }, // spacer |
|---|
| 345 | |
|---|
| 346 | { "", "", MT_IGNORE }, // End of array |
|---|
| 347 | }; |
|---|
| 348 | |
|---|
| 349 | // -------------------------------------------------------------------------------- |
|---|
| 350 | |
|---|
| 351 | |
|---|
| 352 | GenebankImporter::GenebankImporter(LineReader& Flatfile, DBwriter& DB_writer) |
|---|
| 353 | : Importer(Flatfile, DB_writer, genebank_meta_description) |
|---|
| 354 | {} |
|---|
| 355 | |
|---|
| 356 | bool GenebankImporter::readFeatureTableLine(string& line) { |
|---|
| 357 | if (flatfile.getLine(line)) { |
|---|
| 358 | if (beginsWith(line, " ")) { |
|---|
| 359 | return true; |
|---|
| 360 | } |
|---|
| 361 | flatfile.backLine(line); |
|---|
| 362 | } |
|---|
| 363 | return false; |
|---|
| 364 | } |
|---|
| 365 | |
|---|
| 366 | static bool splitGenebankTag(const string& line, string& tag, string& content) { |
|---|
| 367 | // split a line into tag (incl. preceding spaces) and content |
|---|
| 368 | // returns true, if line suffices the format requirements (currently never returns false!) |
|---|
| 369 | // Note: returns tag="" at wrapped lines |
|---|
| 370 | |
|---|
| 371 | string::size_type first_non_space = line.find_first_not_of(' '); |
|---|
| 372 | |
|---|
| 373 | if (first_non_space >= 12 || // no tag, only content |
|---|
| 374 | (first_non_space == string::npos && line.length() == 12)) { // same with empty content |
|---|
| 375 | tag = ""; |
|---|
| 376 | content = line.substr(12); |
|---|
| 377 | return true; |
|---|
| 378 | } |
|---|
| 379 | |
|---|
| 380 | gi_assert(first_non_space<12); |
|---|
| 381 | |
|---|
| 382 | string::size_type behind_tag = line.find_first_of(' ', first_non_space); |
|---|
| 383 | if (behind_tag == string::npos) { // only tag w/o spaces behind |
|---|
| 384 | tag = line; |
|---|
| 385 | content = ""; |
|---|
| 386 | return true; |
|---|
| 387 | } |
|---|
| 388 | |
|---|
| 389 | string::size_type content_start = line.find_first_not_of(' ', behind_tag); |
|---|
| 390 | if (content_start == string::npos) { // line w/o content |
|---|
| 391 | content = ""; |
|---|
| 392 | } |
|---|
| 393 | else { |
|---|
| 394 | content = line.substr(content_start); |
|---|
| 395 | } |
|---|
| 396 | |
|---|
| 397 | tag = line.substr(0, behind_tag); |
|---|
| 398 | return true; |
|---|
| 399 | } |
|---|
| 400 | |
|---|
| 401 | static long scanSeqlenFromLOCUS(const string& locusContent) { |
|---|
| 402 | StringParser parser(locusContent); |
|---|
| 403 | parser.extractWord(); // id |
|---|
| 404 | parser.eatSpaces(); |
|---|
| 405 | |
|---|
| 406 | long bp = parser.extractNumber(); |
|---|
| 407 | parser.eatSpaces(); |
|---|
| 408 | parser.expectContent("bp"); |
|---|
| 409 | |
|---|
| 410 | return bp; |
|---|
| 411 | } |
|---|
| 412 | |
|---|
| 413 | void GenebankImporter::import_section() { |
|---|
| 414 | MetaInfo meta; |
|---|
| 415 | References refs; |
|---|
| 416 | |
|---|
| 417 | const MetaTag *prevTag = NULp; // previously handled tag |
|---|
| 418 | string prevContent; // previously found content |
|---|
| 419 | |
|---|
| 420 | bool seenHeaderLine = false; |
|---|
| 421 | bool EOS = false; // end of section ? |
|---|
| 422 | |
|---|
| 423 | // read header of file |
|---|
| 424 | while (!EOS) { |
|---|
| 425 | string line, tag, content; |
|---|
| 426 | expectLine(line); |
|---|
| 427 | if (!splitGenebankTag(line, tag, content)) { |
|---|
| 428 | gi_assert(0); |
|---|
| 429 | } |
|---|
| 430 | |
|---|
| 431 | if (tag.empty()) { // no tag - happens at wrapped lines |
|---|
| 432 | if (!content.empty()) { // do not append empty 'lines' from wrapped tag-entries. |
|---|
| 433 | prevContent.append(1, ' '); |
|---|
| 434 | prevContent.append(content); |
|---|
| 435 | } |
|---|
| 436 | } |
|---|
| 437 | else { // start of new tag |
|---|
| 438 | const MetaTag *knownTag = findTag(tag); |
|---|
| 439 | if (!knownTag) throw GBS_global_string("Invalid tag '%s'", tag.c_str()); |
|---|
| 440 | |
|---|
| 441 | if (prevTag) { // save previous tag |
|---|
| 442 | switch (prevTag->type) { |
|---|
| 443 | case MT_REF: refs.add(prevTag->field, prevContent); break; |
|---|
| 444 | case MT_BASIC: meta.add(prevTag, prevContent, true); break; |
|---|
| 445 | case MT_HEADER: |
|---|
| 446 | meta.add(prevTag, prevContent, true); // save header line |
|---|
| 447 | expectedSeqLength = scanSeqlenFromLOCUS(prevContent); |
|---|
| 448 | break; |
|---|
| 449 | case MT_REF_DBID: // embl only |
|---|
| 450 | default: gi_assert(0); break; |
|---|
| 451 | } |
|---|
| 452 | prevTag = NULp; |
|---|
| 453 | } |
|---|
| 454 | |
|---|
| 455 | switch (knownTag->type) { |
|---|
| 456 | case MT_HEADER: |
|---|
| 457 | if (seenHeaderLine) throw GBS_global_string("Multiple occurrences of tag '%s'", tag.c_str()); |
|---|
| 458 | seenHeaderLine = true; |
|---|
| 459 | // fall-through |
|---|
| 460 | case MT_BASIC: |
|---|
| 461 | case MT_REF: |
|---|
| 462 | prevTag = knownTag; |
|---|
| 463 | prevContent = content; |
|---|
| 464 | break; |
|---|
| 465 | |
|---|
| 466 | case MT_REF_START: |
|---|
| 467 | refs.start(); // start a new reference |
|---|
| 468 | break; |
|---|
| 469 | |
|---|
| 470 | case MT_FEATURE_START: |
|---|
| 471 | db_writer.createOrganism(flatfile.getFilename(), "NCBI"); |
|---|
| 472 | parseFeatureTable(); |
|---|
| 473 | break; |
|---|
| 474 | |
|---|
| 475 | case MT_SEQUENCE_START: |
|---|
| 476 | parseSequence(knownTag->tag, content); |
|---|
| 477 | EOS = true; // end of section |
|---|
| 478 | break; |
|---|
| 479 | |
|---|
| 480 | case MT_IGNORE: |
|---|
| 481 | break; |
|---|
| 482 | |
|---|
| 483 | case MT_END: |
|---|
| 484 | EOS = true; |
|---|
| 485 | break; |
|---|
| 486 | |
|---|
| 487 | case MT_CONTIG: |
|---|
| 488 | throw GBS_global_string("Cannot import files containing CONTIG"); |
|---|
| 489 | |
|---|
| 490 | case MT_REF_DBID: // embl only |
|---|
| 491 | default: |
|---|
| 492 | gi_assert(0); |
|---|
| 493 | throw GBS_global_string("Tag '%s' not expected here", knownTag->tag.c_str()); |
|---|
| 494 | } |
|---|
| 495 | } |
|---|
| 496 | } |
|---|
| 497 | |
|---|
| 498 | db_writer.finalizeOrganism(meta, refs, *this); |
|---|
| 499 | show_warnings(meta.getAccessionNumber()); |
|---|
| 500 | } |
|---|
| 501 | |
|---|
| 502 | // -------------------------------------------------------------------------------- |
|---|
| 503 | |
|---|
| 504 | |
|---|
| 505 | EmblImporter::EmblImporter(LineReader& Flatfile, DBwriter& DB_writer) |
|---|
| 506 | : Importer(Flatfile, DB_writer, embl_meta_description) |
|---|
| 507 | {} |
|---|
| 508 | |
|---|
| 509 | static bool splitEmblTag(const string& line, string& tag, string& content) { |
|---|
| 510 | // split a line into 2-character tag and content |
|---|
| 511 | // return true on success (i.e. if line suffices the required format) |
|---|
| 512 | |
|---|
| 513 | if (line.length() == 2) { |
|---|
| 514 | tag = line; |
|---|
| 515 | content = ""; |
|---|
| 516 | } |
|---|
| 517 | else { |
|---|
| 518 | string::size_type spacer = line.find(" "); // separator between tag and content |
|---|
| 519 | if (spacer != 2) return false; // expect spacer at pos 2-4 |
|---|
| 520 | |
|---|
| 521 | tag = line.substr(0, 2); |
|---|
| 522 | content = line.substr(5); |
|---|
| 523 | } |
|---|
| 524 | |
|---|
| 525 | return true; |
|---|
| 526 | } |
|---|
| 527 | |
|---|
| 528 | bool EmblImporter::readFeatureTableLine(string& line) { |
|---|
| 529 | if (flatfile.getLine(line)) { |
|---|
| 530 | if (beginsWith(line, "FT ")) { |
|---|
| 531 | return true; |
|---|
| 532 | } |
|---|
| 533 | flatfile.backLine(line); |
|---|
| 534 | } |
|---|
| 535 | return false; |
|---|
| 536 | } |
|---|
| 537 | |
|---|
| 538 | static long scanSeqlenFromID(const string& idContent) { |
|---|
| 539 | StringParser parser(idContent); |
|---|
| 540 | string lastWord = parser.extractWord(); // eat id |
|---|
| 541 | bool bpseen = false; |
|---|
| 542 | long bp = -1; |
|---|
| 543 | |
|---|
| 544 | while (!bpseen) { |
|---|
| 545 | parser.eatSpaces(); |
|---|
| 546 | string word = parser.extractWord(); |
|---|
| 547 | if (word == "BP.") { |
|---|
| 548 | // basecount is in word before "BP." |
|---|
| 549 | bp = atol(lastWord.c_str()); |
|---|
| 550 | bpseen = true; |
|---|
| 551 | } |
|---|
| 552 | else { |
|---|
| 553 | lastWord = word; |
|---|
| 554 | } |
|---|
| 555 | } |
|---|
| 556 | |
|---|
| 557 | if (bp == -1) throw "Could not parse bp from header"; |
|---|
| 558 | |
|---|
| 559 | return bp; |
|---|
| 560 | } |
|---|
| 561 | |
|---|
| 562 | void EmblImporter::import_section() { |
|---|
| 563 | MetaInfo meta; |
|---|
| 564 | References refs; |
|---|
| 565 | |
|---|
| 566 | const MetaTag *prevTag = NULp; // previously handled tag |
|---|
| 567 | string prevContent; // previously found content |
|---|
| 568 | bool prevAppendNL = false; // append '\n' into multiline tags |
|---|
| 569 | |
|---|
| 570 | bool seenHeaderLine = false; |
|---|
| 571 | bool EOS = false; // end of section ? |
|---|
| 572 | |
|---|
| 573 | // read header of file |
|---|
| 574 | while (!EOS) { |
|---|
| 575 | string line, tag, content; |
|---|
| 576 | expectLine(line); |
|---|
| 577 | if (!splitEmblTag(line, tag, content)) { |
|---|
| 578 | throw "Expected two-character tag at start of line"; |
|---|
| 579 | } |
|---|
| 580 | |
|---|
| 581 | const MetaTag *knownTag = findTag(tag); |
|---|
| 582 | if (!knownTag) throw GBS_global_string("Invalid tag '%s'", tag.c_str()); |
|---|
| 583 | |
|---|
| 584 | if (knownTag == prevTag) { // multiline tag |
|---|
| 585 | if (prevAppendNL) prevContent.append("\n"); // append a newline to make parsing in add_dbid() more easy |
|---|
| 586 | prevContent.append(content); // append w/o space - EMBL flatfiles have spaces at EOL when needed |
|---|
| 587 | } |
|---|
| 588 | else { // start of new tag |
|---|
| 589 | if (prevTag) { // save previous tag |
|---|
| 590 | switch (prevTag->type) { |
|---|
| 591 | case MT_REF: refs.add(prevTag->field, prevContent); break; |
|---|
| 592 | case MT_REF_DBID: refs.add_dbid(prevContent); prevAppendNL = false; break; |
|---|
| 593 | case MT_BASIC: meta.add(prevTag, prevContent, true); break; |
|---|
| 594 | case MT_HEADER: |
|---|
| 595 | meta.add(prevTag, prevContent, true); |
|---|
| 596 | expectedSeqLength = scanSeqlenFromID(prevContent); |
|---|
| 597 | break; |
|---|
| 598 | default: gi_assert(0); break; |
|---|
| 599 | } |
|---|
| 600 | prevTag = NULp; |
|---|
| 601 | } |
|---|
| 602 | |
|---|
| 603 | switch (knownTag->type) { |
|---|
| 604 | case MT_HEADER: |
|---|
| 605 | if (seenHeaderLine) throw GBS_global_string("Multiple occurrences of tag '%s'", tag.c_str()); |
|---|
| 606 | seenHeaderLine = true; |
|---|
| 607 | // fall-through |
|---|
| 608 | case MT_BASIC: |
|---|
| 609 | case MT_REF: |
|---|
| 610 | prevTag = knownTag; |
|---|
| 611 | prevContent = content; |
|---|
| 612 | break; |
|---|
| 613 | |
|---|
| 614 | case MT_REF_DBID: |
|---|
| 615 | prevTag = knownTag; |
|---|
| 616 | prevContent = content; |
|---|
| 617 | prevAppendNL = true; |
|---|
| 618 | break; |
|---|
| 619 | |
|---|
| 620 | case MT_REF_START: |
|---|
| 621 | refs.start(); // start a new reference |
|---|
| 622 | break; |
|---|
| 623 | |
|---|
| 624 | case MT_FEATURE: |
|---|
| 625 | flatfile.backLine(line); |
|---|
| 626 | db_writer.createOrganism(flatfile.getFilename(), "EMBL"); |
|---|
| 627 | parseFeatureTable(); |
|---|
| 628 | break; |
|---|
| 629 | |
|---|
| 630 | case MT_SEQUENCE_START: |
|---|
| 631 | parseSequence(content); |
|---|
| 632 | EOS = true; // end of section |
|---|
| 633 | break; |
|---|
| 634 | |
|---|
| 635 | case MT_FEATURE_START: |
|---|
| 636 | case MT_IGNORE: |
|---|
| 637 | break; |
|---|
| 638 | |
|---|
| 639 | default: |
|---|
| 640 | gi_assert(0); |
|---|
| 641 | throw GBS_global_string("Tag '%s' not expected here", knownTag->tag.c_str()); |
|---|
| 642 | } |
|---|
| 643 | } |
|---|
| 644 | } |
|---|
| 645 | db_writer.finalizeOrganism(meta, refs, *this); |
|---|
| 646 | show_warnings(meta.getAccessionNumber()); |
|---|
| 647 | } |
|---|
| 648 | |
|---|
| 649 | // -------------------------------------------------------------------------------- |
|---|
| 650 | // sequence readers: |
|---|
| 651 | |
|---|
| 652 | inline bool parseCounter(bool expect, BaseCounter& headerCount, StringParser& parser, Base base, const char *word) { |
|---|
| 653 | // parses part of string (e.g. " 6021225 BP;" or " 878196 A;") |
|---|
| 654 | // if 'expect' == true -> throw exception if missing |
|---|
| 655 | // if 'expect' == false -> return false if missing |
|---|
| 656 | |
|---|
| 657 | bool found = false; |
|---|
| 658 | stringCIter start = parser.getPosition(); |
|---|
| 659 | |
|---|
| 660 | parser.expectSpaces(0); |
|---|
| 661 | |
|---|
| 662 | bool seen_number; |
|---|
| 663 | long count = parser.eatNumber(seen_number); |
|---|
| 664 | |
|---|
| 665 | if (seen_number) { |
|---|
| 666 | headerCount.addCount(base, count); |
|---|
| 667 | size_t spaces = parser.eatSpaces(); |
|---|
| 668 | if (spaces>0) { |
|---|
| 669 | size_t len = parser.lookingAt(word); |
|---|
| 670 | if (len>0) { // seen |
|---|
| 671 | parser.advance(len); |
|---|
| 672 | found = true; |
|---|
| 673 | } |
|---|
| 674 | } |
|---|
| 675 | } |
|---|
| 676 | |
|---|
| 677 | if (!found) { |
|---|
| 678 | parser.setPosition(start); // reset position |
|---|
| 679 | if (expect) throw GBS_global_string("Expected counter '### %s', found '%s'", word, parser.rest().c_str()); |
|---|
| 680 | } |
|---|
| 681 | return found; |
|---|
| 682 | } |
|---|
| 683 | |
|---|
| 684 | void GenebankImporter::parseSequence(const string& tag, const string& headerline) { |
|---|
| 685 | SmartPtr<BaseCounter> headerCount; |
|---|
| 686 | |
|---|
| 687 | if (tag == "BASE") { // base count not always present |
|---|
| 688 | // parse headerline : |
|---|
| 689 | headerCount = new BaseCounter("sequence header"); |
|---|
| 690 | { |
|---|
| 691 | StringParser parser(headerline); |
|---|
| 692 | |
|---|
| 693 | parser.expectContent("COUNT"); |
|---|
| 694 | |
|---|
| 695 | parseCounter(true, *headerCount, parser, BC_A, "a"); |
|---|
| 696 | parseCounter(true, *headerCount, parser, BC_C, "c"); |
|---|
| 697 | parseCounter(true, *headerCount, parser, BC_G, "g"); |
|---|
| 698 | parseCounter(true, *headerCount, parser, BC_T, "t"); |
|---|
| 699 | parseCounter(false, *headerCount, parser, BC_OTHER, "others"); // not always present |
|---|
| 700 | |
|---|
| 701 | headerCount->calcOverallCounter(); |
|---|
| 702 | } |
|---|
| 703 | } |
|---|
| 704 | |
|---|
| 705 | // parse sequence data |
|---|
| 706 | size_t est_seq_size = headerCount.isNull() ? 500000 : headerCount->getCount(BC_ALL); |
|---|
| 707 | SequenceBuffer seqData(est_seq_size); |
|---|
| 708 | { |
|---|
| 709 | string line; |
|---|
| 710 | |
|---|
| 711 | if (!headerCount.isNull()) { |
|---|
| 712 | // if BASE COUNT was present, check ORIGIN line |
|---|
| 713 | // otherwise ORIGIN line has already been read |
|---|
| 714 | expectLine(line); |
|---|
| 715 | if (!beginsWith(line, "ORIGIN")) throw "Expected 'ORIGIN'"; |
|---|
| 716 | } |
|---|
| 717 | |
|---|
| 718 | bool eos_seen = false; |
|---|
| 719 | while (!eos_seen) { |
|---|
| 720 | expectLine(line); |
|---|
| 721 | if (beginsWith(line, "//")) { |
|---|
| 722 | eos_seen = true; |
|---|
| 723 | } |
|---|
| 724 | else { |
|---|
| 725 | string data; |
|---|
| 726 | data.reserve(60); |
|---|
| 727 | StringParser parser(line); |
|---|
| 728 | |
|---|
| 729 | parser.eatSpaces(); // not sure whether there really have to be spaces if number has 9 digits or more |
|---|
| 730 | size_t cur_pos = (size_t)parser.extractNumber(); |
|---|
| 731 | size_t datasize = seqData.getBaseCounter().getCount(BC_ALL); |
|---|
| 732 | |
|---|
| 733 | if (cur_pos != (datasize+1)) { |
|---|
| 734 | throw GBS_global_string("Got wrong base position (found=%zu, expected=%zu)", cur_pos, size_t(datasize+1)); |
|---|
| 735 | } |
|---|
| 736 | |
|---|
| 737 | int blocks = 0; |
|---|
| 738 | while (!parser.atEnd() && parser.at() == ' ') { |
|---|
| 739 | parser.expectSpaces(1); |
|---|
| 740 | |
|---|
| 741 | stringCIter start = parser.pos; |
|---|
| 742 | stringCIter end = parser.find(' '); |
|---|
| 743 | |
|---|
| 744 | data.append(start, end); |
|---|
| 745 | blocks++; |
|---|
| 746 | } |
|---|
| 747 | |
|---|
| 748 | if (blocks>6) throw "Found more than 6 parts of sequence data"; |
|---|
| 749 | seqData.addLine(data); |
|---|
| 750 | } |
|---|
| 751 | } |
|---|
| 752 | } |
|---|
| 753 | |
|---|
| 754 | check_base_counters(seqData, headerCount.content()); |
|---|
| 755 | db_writer.writeSequence(seqData); |
|---|
| 756 | } |
|---|
| 757 | |
|---|
| 758 | void EmblImporter::parseSequence(const string& headerline) { |
|---|
| 759 | // parse headerline: |
|---|
| 760 | BaseCounter headerCount("sequence header"); |
|---|
| 761 | { |
|---|
| 762 | StringParser parser(headerline); |
|---|
| 763 | |
|---|
| 764 | parser.expectContent("Sequence"); |
|---|
| 765 | |
|---|
| 766 | parseCounter(true, headerCount, parser, BC_ALL, "BP;"); |
|---|
| 767 | parseCounter(true, headerCount, parser, BC_A, "A;"); |
|---|
| 768 | parseCounter(true, headerCount, parser, BC_C, "C;"); |
|---|
| 769 | parseCounter(true, headerCount, parser, BC_G, "G;"); |
|---|
| 770 | parseCounter(true, headerCount, parser, BC_T, "T;"); |
|---|
| 771 | parseCounter(true, headerCount, parser, BC_OTHER, "other;"); |
|---|
| 772 | |
|---|
| 773 | headerCount.checkOverallCounter(); |
|---|
| 774 | } |
|---|
| 775 | |
|---|
| 776 | // parse sequence data |
|---|
| 777 | SequenceBuffer seqData(headerCount.getCount(BC_ALL)); |
|---|
| 778 | { |
|---|
| 779 | bool eos_seen = false; |
|---|
| 780 | string line; |
|---|
| 781 | |
|---|
| 782 | while (!eos_seen) { |
|---|
| 783 | expectLine(line); |
|---|
| 784 | if (beginsWith(line, "//")) { |
|---|
| 785 | eos_seen = true; |
|---|
| 786 | } |
|---|
| 787 | else { |
|---|
| 788 | string data; |
|---|
| 789 | data.reserve(60); |
|---|
| 790 | StringParser parser(line); |
|---|
| 791 | |
|---|
| 792 | parser.expectSpaces(5, false); |
|---|
| 793 | int blocks = 0; |
|---|
| 794 | while (!parser.atEnd() && isalpha(parser.at())) { |
|---|
| 795 | stringCIter start = parser.pos; |
|---|
| 796 | stringCIter end = parser.find(' '); |
|---|
| 797 | |
|---|
| 798 | data.append(start, end); |
|---|
| 799 | blocks++; |
|---|
| 800 | parser.expectSpaces(1); |
|---|
| 801 | } |
|---|
| 802 | |
|---|
| 803 | if (blocks>6) throw "Found more than 6 parts of sequence data"; |
|---|
| 804 | |
|---|
| 805 | size_t basecount = (size_t)parser.extractNumber(); |
|---|
| 806 | |
|---|
| 807 | seqData.addLine(data); |
|---|
| 808 | size_t datasize = seqData.getBaseCounter().getCount(BC_ALL); |
|---|
| 809 | |
|---|
| 810 | if (basecount != datasize) { |
|---|
| 811 | throw GBS_global_string("Got wrong base counter(found=%zu, expected=%zu)", basecount, datasize); |
|---|
| 812 | } |
|---|
| 813 | } |
|---|
| 814 | } |
|---|
| 815 | } |
|---|
| 816 | |
|---|
| 817 | check_base_counters(seqData, &headerCount); |
|---|
| 818 | db_writer.writeSequence(seqData); |
|---|
| 819 | } |
|---|
| 820 | |
|---|