root / tmp / org.txm.groovy.core / src / groovy / org / txm / scripts / importer / xtz / XTZImporter.groovy @ 2126
History | View | Annotate | Download (14.9 kB)
| 1 |
package org.txm.scripts.importer.xtz
|
|---|---|
| 2 |
|
| 3 |
import org.txm.scripts.filters.Tokeniser.SimpleTokenizerXml |
| 4 |
|
| 5 |
import java.io.File; |
| 6 |
|
| 7 |
import net.sf.saxon.style.XSLParam; |
| 8 |
|
| 9 |
import org.txm.core.preferences.TBXPreferences; |
| 10 |
import org.txm.importer.ApplyXsl2 |
| 11 |
import org.txm.scripts.importer.CleanFile |
| 12 |
import org.txm.scripts.importer.MileStoneProjection |
| 13 |
import org.txm.importer.ValidateXml |
| 14 |
import org.txm.* |
| 15 |
|
| 16 |
import javax.xml.parsers.SAXParser |
| 17 |
import javax.xml.parsers.SAXParserFactory |
| 18 |
import javax.xml.stream.* |
| 19 |
|
| 20 |
import org.w3c.dom.Element |
| 21 |
import org.xml.sax.XMLReader |
| 22 |
import org.txm.utils.xml.DomUtils |
| 23 |
import org.txm.metadatas.Metadatas |
| 24 |
import org.txm.objects.* |
| 25 |
import org.txm.importer.scripts.xmltxm.Xml2Ana |
| 26 |
import org.txm.utils.ConsoleProgressBar |
| 27 |
import org.txm.utils.io.* |
| 28 |
import org.txm.importer.xtz.* |
| 29 |
|
| 30 |
class XTZImporter extends Importer { |
| 31 |
|
| 32 |
public boolean cancelNow = false; |
| 33 |
public def filesToProcess = [] |
| 34 |
|
| 35 |
public File frontXSLResultDirectory; |
| 36 |
public File tokenizedDirectory; |
| 37 |
public File sentencedDirectory; |
| 38 |
|
| 39 |
public String xslPath; |
| 40 |
public def xslParams; |
| 41 |
public String lang; |
| 42 |
public String wordTag; |
| 43 |
public boolean doTokenizeStep = false; |
| 44 |
|
| 45 |
public Metadatas metadata = null; // text metadata |
| 46 |
|
| 47 |
public XTZImporter(ImportModule module) {
|
| 48 |
super(module);
|
| 49 |
} |
| 50 |
|
| 51 |
public Metadatas getMetadata() {
|
| 52 |
return metadata;
|
| 53 |
} |
| 54 |
|
| 55 |
public void process() { |
| 56 |
|
| 57 |
Project project = module.getProject(); |
| 58 |
File binDir = module.getBinaryDirectory();
|
| 59 |
|
| 60 |
String corpusname = project.getName();
|
| 61 |
|
| 62 |
def srcFiles = []; |
| 63 |
def files = inputDirectory.listFiles();
|
| 64 |
if (files == null) { |
| 65 |
reason = "No file to process in "+inputDirectory
|
| 66 |
return; // no file to process in the directory |
| 67 |
} |
| 68 |
|
| 69 |
for (File f : files) { |
| 70 |
if (f.isHidden() || f.isDirectory())
|
| 71 |
continue;
|
| 72 |
else if (f.getName().endsWith(".properties")) |
| 73 |
continue;
|
| 74 |
else if (f.getName().equals("import.xml")) |
| 75 |
continue;
|
| 76 |
else if (f.getName().endsWith(".csv")) |
| 77 |
continue;
|
| 78 |
else if (f.getName().endsWith(".dtd")) |
| 79 |
continue;
|
| 80 |
else if (f.getName().endsWith(".xsl")) |
| 81 |
continue;
|
| 82 |
else if (f.getName().endsWith("~")) |
| 83 |
continue;
|
| 84 |
else if (f.getName().startsWith(".")) |
| 85 |
continue;
|
| 86 |
|
| 87 |
srcFiles << f |
| 88 |
} |
| 89 |
|
| 90 |
if (srcFiles.size() == 0) { |
| 91 |
reason = "No suitable file to process in "+inputDirectory
|
| 92 |
return; // no file to process in the directory |
| 93 |
} |
| 94 |
|
| 95 |
// prepare front XSL if any
|
| 96 |
xslPath = project.getFrontXSL() |
| 97 |
xslParams = project.getXsltParameters() |
| 98 |
|
| 99 |
lang = project.getLang(); |
| 100 |
|
| 101 |
wordTag = project.getTokenizerWordElement() |
| 102 |
this.doTokenizeStep = project.getDoTokenizerStep()
|
| 103 |
|
| 104 |
//prepare metadata if any
|
| 105 |
File allMetadataFile = Metadatas.findMetadataFile(inputDirectory);
|
| 106 |
if (allMetadataFile.exists()) {
|
| 107 |
File copy = new File(binDir, allMetadataFile.getName()) |
| 108 |
if (!FileCopy.copy(allMetadataFile, copy)) {
|
| 109 |
println "Error: could not create a copy of the metadata file "+allMetadataFile.getAbsoluteFile();
|
| 110 |
return;
|
| 111 |
} |
| 112 |
metadata = new Metadatas(copy,
|
| 113 |
Toolbox.getPreference(TBXPreferences.METADATA_ENCODING), |
| 114 |
Toolbox.getPreference(TBXPreferences.METADATA_COLSEPARATOR), |
| 115 |
Toolbox.getPreference(TBXPreferences.METADATA_TXTSEPARATOR), 1)
|
| 116 |
} |
| 117 |
|
| 118 |
String cleanDirectories = project.getCleanAfterBuild();
|
| 119 |
|
| 120 |
if (!doFixSurrogates()) return; |
| 121 |
if (!doSplitMergeXSLStep()) return; |
| 122 |
if (!doFrontXSLStep()) return; |
| 123 |
if (!doCheckXMLFileStep()) return; |
| 124 |
if (!doTokenizeStep()) return; |
| 125 |
if (!doPostTokenizeXSLStep()) return; |
| 126 |
if (!doEncodeMileStonesStep()) return; |
| 127 |
if (!doToXMLTXMStep()) return; |
| 128 |
if (!doInjectMetadataStep()) return; |
| 129 |
|
| 130 |
isSuccessFul = filesToProcess.size() > 0
|
| 131 |
|
| 132 |
if ("true".equals(cleanDirectories)) { |
| 133 |
new File(module.getBinaryDirectory(), "tokenized").deleteDir() |
| 134 |
new File(module.getBinaryDirectory(), "src").deleteDir() |
| 135 |
new File(module.getBinaryDirectory(), "split").deleteDir() |
| 136 |
} |
| 137 |
} |
| 138 |
|
| 139 |
/**
|
| 140 |
* read from source directory and remove the surrogate chars not well managed
|
| 141 |
* save the result in $bindir/src
|
| 142 |
*
|
| 143 |
*
|
| 144 |
*
|
| 145 |
*/
|
| 146 |
public boolean doFixSurrogates() { |
| 147 |
File srcDirectory = new File(module.getBinaryDirectory(), "src") |
| 148 |
|
| 149 |
def files = inputDirectory.listFiles()
|
| 150 |
if (files == null) { |
| 151 |
reason = "No file to process in $inputDirectory"
|
| 152 |
return false; |
| 153 |
} |
| 154 |
srcDirectory.deleteDir() // clean before copying
|
| 155 |
srcDirectory.mkdir() |
| 156 |
|
| 157 |
for (File file : files) { |
| 158 |
def outputFile = new File(srcDirectory, file.getName()); |
| 159 |
if (file.getName().toLowerCase().endsWith(".xml")) { |
| 160 |
CleanFile.removeSurrogateFromXmlFile(file, outputFile) |
| 161 |
} else {
|
| 162 |
FileCopy.copyFiles(file, outputFile) |
| 163 |
} |
| 164 |
} |
| 165 |
|
| 166 |
inputDirectory = srcDirectory; // the files to process are now in the "src" directory
|
| 167 |
return true; |
| 168 |
} |
| 169 |
|
| 170 |
/**
|
| 171 |
* read from source directory and write the result in $bindir/txm
|
| 172 |
*
|
| 173 |
* only one XSL is applied
|
| 174 |
*
|
| 175 |
*/
|
| 176 |
public boolean doSplitMergeXSLStep() { |
| 177 |
|
| 178 |
File splitedDirectory = new File(module.getBinaryDirectory(), "split") |
| 179 |
splitedDirectory.deleteDir() |
| 180 |
|
| 181 |
File xslDirectory = new File(module.getSourceDirectory(), "xsl/1-split-merge") |
| 182 |
println "-- Split-Merge XSL Step with $xslDirectory"
|
| 183 |
def xslFiles = xslDirectory.listFiles()
|
| 184 |
if (xslDirectory.exists() && xslFiles != null && xslFiles.size() > 0) { |
| 185 |
|
| 186 |
splitedDirectory.mkdir() |
| 187 |
xslParams["output-directory"] = splitedDirectory.getAbsoluteFile().toURI().toString();
|
| 188 |
|
| 189 |
xslFiles.sort() |
| 190 |
for (File xslFile : xslFiles) { |
| 191 |
if (xslFile.isDirectory() || xslFile.isHidden() || !xslFile.getName().endsWith(".xsl")) continue; |
| 192 |
//if (!xslFile.getName().matches("[1-9]-.+")) continue;
|
| 193 |
|
| 194 |
if (ApplyXsl2.processImportSources(xslFile, inputDirectory.listFiles(), xslParams)) {
|
| 195 |
println ""
|
| 196 |
} else {
|
| 197 |
reason = "Fail to apply split-merge XSL: $xslPath"
|
| 198 |
} |
| 199 |
break; // THERE IS ONLY ONE XSL FILE TO APPLY |
| 200 |
} |
| 201 |
inputDirectory = splitedDirectory; // the files to process are now in the "src" directory
|
| 202 |
} else {
|
| 203 |
//println "Nothing to do."
|
| 204 |
} |
| 205 |
return true; |
| 206 |
} |
| 207 |
|
| 208 |
public boolean doEncodeMileStonesStep() { |
| 209 |
filesToProcess = new File(module.getBinaryDirectory(), "tokenized").listFiles() |
| 210 |
def milestonesString = module.getProject().getTextualPlan("MileStones") |
| 211 |
if (milestonesString.length() == 0) return true; |
| 212 |
|
| 213 |
def milestones = milestonesString.split(",") |
| 214 |
//def milestones = ["lb", "pb", "cb"]
|
| 215 |
//println "milestones=$milestones type="+milestones.getClass()+" size="+milestones.size()
|
| 216 |
if (milestones.size() == 0) return true; |
| 217 |
|
| 218 |
println "-- Encoding milestone $milestones into XML files..."
|
| 219 |
ConsoleProgressBar cpb = new ConsoleProgressBar(milestones.size())
|
| 220 |
for (def milestone : milestones) { |
| 221 |
milestone = milestone.trim()// just in case
|
| 222 |
cpb.tick() |
| 223 |
|
| 224 |
for (File inputFile : filesToProcess) { |
| 225 |
File outputFile = File.createTempFile("tmp", inputFile.getName(),inputFile.getParentFile()); |
| 226 |
println "$inputFile, 'text', $wordTag, $milestone, $outputFile"
|
| 227 |
MileStoneProjection msp = new MileStoneProjection(inputFile, "text", wordTag, milestone) |
| 228 |
if (!msp.process(outputFile)) {
|
| 229 |
println "Fail to encode $milestone in $inputFile"
|
| 230 |
return false |
| 231 |
} else {
|
| 232 |
if (inputFile.delete()) {
|
| 233 |
FileCopy.copy(outputFile, new File(outputFile.getParent(), "copy.xml")) |
| 234 |
outputFile.renameTo(inputFile) |
| 235 |
} else {
|
| 236 |
println "Fail to encode $milestone in ${inputFile}: could not replace the file."
|
| 237 |
return false |
| 238 |
} |
| 239 |
} |
| 240 |
} |
| 241 |
} |
| 242 |
println ""
|
| 243 |
return true; |
| 244 |
} |
| 245 |
|
| 246 |
/**
|
| 247 |
* read from $inputDirectory and write the result in $bindir/txm
|
| 248 |
*
|
| 249 |
*/
|
| 250 |
public boolean doFrontXSLStep() { |
| 251 |
|
| 252 |
//filesToProcess = inputDirectory.listFiles();
|
| 253 |
|
| 254 |
File frontXSLdirectory = new File(module.getSourceDirectory(), "xsl/2-front") |
| 255 |
println "-- Front XSL Step with the $frontXSLdirectory directory."
|
| 256 |
def xslFiles = frontXSLdirectory.listFiles()
|
| 257 |
|
| 258 |
xslParams["output-directory"] = outputDirectory.getAbsoluteFile().toURI().toString();
|
| 259 |
|
| 260 |
if (frontXSLdirectory.exists() && xslFiles != null && xslFiles.size() > 0) { |
| 261 |
|
| 262 |
xslFiles.sort() |
| 263 |
for (File xslFile : xslFiles) { |
| 264 |
if (xslFile.isDirectory() || xslFile.isHidden() || !xslFile.getName().endsWith(".xsl")) continue; |
| 265 |
//if (!xslFile.getName().matches("[1-9]-.+")) continue;
|
| 266 |
|
| 267 |
if (ApplyXsl2.processImportSources(xslFile, inputDirectory, outputDirectory, xslParams)) {
|
| 268 |
inputDirectory = outputDirectory; // the files to process are now in the "txm" directory
|
| 269 |
println ""
|
| 270 |
} else {
|
| 271 |
reason = "Fail to apply front XSL: $xslPath"
|
| 272 |
return false; |
| 273 |
} |
| 274 |
} |
| 275 |
} else {
|
| 276 |
//println "Nothing to do."
|
| 277 |
} |
| 278 |
return true; |
| 279 |
} |
| 280 |
|
| 281 |
public boolean doCheckXMLFileStep() { |
| 282 |
filesToProcess = []
|
| 283 |
println "-- Checking XML files for well-formedness."
|
| 284 |
def files = inputDirectory.listFiles()
|
| 285 |
|
| 286 |
if (files == null || files.size() == 0) { |
| 287 |
println "No XML file (*.xml) to process. Aborting"
|
| 288 |
return false; |
| 289 |
} |
| 290 |
files = files.sort() |
| 291 |
|
| 292 |
ConsoleProgressBar cpb = new ConsoleProgressBar(files.size())
|
| 293 |
for (File f : files) { |
| 294 |
cpb.tick() |
| 295 |
if (f.getName().toLowerCase().endsWith(".xml")) |
| 296 |
if (!f.isHidden() && f.canRead() && !f.getName().endsWith(".properties") |
| 297 |
&& !f.getName().startsWith("metadata") && !f.isDirectory())
|
| 298 |
{
|
| 299 |
|
| 300 |
if (!ValidateXml.test(f)) {
|
| 301 |
continue;
|
| 302 |
} else {
|
| 303 |
if (!f.getName().equals("import.xml")) |
| 304 |
filesToProcess << f |
| 305 |
} |
| 306 |
} |
| 307 |
} |
| 308 |
println ""
|
| 309 |
if (filesToProcess.size() == 0) { |
| 310 |
reason = "No file suitable to process in "+inputDirectory
|
| 311 |
return false; |
| 312 |
} else {
|
| 313 |
return true; |
| 314 |
} |
| 315 |
} |
| 316 |
|
| 317 |
|
| 318 |
public boolean doTokenizeStep() { |
| 319 |
|
| 320 |
new File(module.getBinaryDirectory(),"tokenized").mkdir() |
| 321 |
|
| 322 |
String outSideTextTagsRegex = ""; |
| 323 |
String outSideTextTagsAndKeepContentRegex = ""; |
| 324 |
String noteRegex = ""; |
| 325 |
// get the element names to ignore
|
| 326 |
|
| 327 |
String e1 = module.getProject().getTextualPlan("OutSideTextTags") |
| 328 |
def split = e1.split(",") |
| 329 |
for (String s : split) { |
| 330 |
outSideTextTagsRegex += "|"+s.trim()
|
| 331 |
} |
| 332 |
if (outSideTextTagsRegex.trim().length() > 0) { |
| 333 |
outSideTextTagsRegex = outSideTextTagsRegex.substring(1) // remove the first "|" |
| 334 |
} |
| 335 |
|
| 336 |
String e2 = module.getProject().getTextualPlan("OutSideTextTagsAndKeepContent") |
| 337 |
def split2 = e2.split(",") |
| 338 |
for (String s : split2) { |
| 339 |
outSideTextTagsAndKeepContentRegex += "|"+s.trim()
|
| 340 |
} |
| 341 |
if (outSideTextTagsAndKeepContentRegex.trim().length() > 0) { |
| 342 |
outSideTextTagsAndKeepContentRegex = outSideTextTagsAndKeepContentRegex.substring(1) // remove the first "|" |
| 343 |
} |
| 344 |
|
| 345 |
String e3 = module.getProject().getTextualPlan("Note") |
| 346 |
def split3 = e3.split(",") |
| 347 |
for (String s : split3) { |
| 348 |
noteRegex += "|"+s.trim()
|
| 349 |
} |
| 350 |
if (noteRegex.trim().length() > 0) { |
| 351 |
noteRegex = noteRegex.substring(1) // remove the first "|" |
| 352 |
} |
| 353 |
|
| 354 |
//if (wordTag != "w") {
|
| 355 |
if (!doTokenizeStep) {
|
| 356 |
println "No tokenization do to."
|
| 357 |
// ConsoleProgressBar cpb = new ConsoleProgressBar(filesToProcess.size())
|
| 358 |
for (File f : filesToProcess) { |
| 359 |
File outfile = new File(module.getBinaryDirectory(),"tokenized/"+f.getName()); |
| 360 |
FileCopy.copy(f, outfile); |
| 361 |
} |
| 362 |
return true; |
| 363 |
} else {
|
| 364 |
println "-- Tokenizing "+filesToProcess.size()+" files" |
| 365 |
ConsoleProgressBar cpb = new ConsoleProgressBar(filesToProcess.size())
|
| 366 |
for (File f : filesToProcess) { |
| 367 |
cpb.tick() |
| 368 |
File infile = f;
|
| 369 |
File outfile = new File(module.getBinaryDirectory(),"tokenized/"+f.getName()); |
| 370 |
SimpleTokenizerXml tokenizer = new SimpleTokenizerXml(infile, outfile, lang)
|
| 371 |
|
| 372 |
if (outSideTextTagsRegex != null && outSideTextTagsRegex.trim().length() > 0) { |
| 373 |
tokenizer.setOutSideTextTags(outSideTextTagsRegex) |
| 374 |
} |
| 375 |
|
| 376 |
if (outSideTextTagsAndKeepContentRegex != null && outSideTextTagsAndKeepContentRegex.trim().length() > 0) { |
| 377 |
tokenizer.setOutSideTextTagsAndKeepContent(outSideTextTagsAndKeepContentRegex) |
| 378 |
} |
| 379 |
|
| 380 |
if (noteRegex != null && noteRegex.trim().length() > 0) { |
| 381 |
tokenizer.setNote(noteRegex) |
| 382 |
} |
| 383 |
|
| 384 |
// tokenize !
|
| 385 |
if (!tokenizer.process()) {
|
| 386 |
println("Failed to process "+f)
|
| 387 |
outfile.delete() |
| 388 |
} |
| 389 |
} |
| 390 |
} |
| 391 |
|
| 392 |
filesToProcess = new File(module.getBinaryDirectory(),"tokenized").listFiles() |
| 393 |
println ""
|
| 394 |
|
| 395 |
return true; |
| 396 |
} |
| 397 |
|
| 398 |
/**
|
| 399 |
* read from $bindir/tokenized and write the result in $bindir/tokenized
|
| 400 |
*
|
| 401 |
*/
|
| 402 |
public boolean doPostTokenizeXSLStep() { |
| 403 |
|
| 404 |
filesToProcess = new File(module.getBinaryDirectory(),"tokenized").listFiles() |
| 405 |
File xslDirectory = new File(module.getSourceDirectory(), "xsl/3-posttok") |
| 406 |
def xslFiles = xslDirectory.listFiles()
|
| 407 |
if (!xslDirectory.exists() || xslFiles == null || xslFiles.size() == 0) return true; |
| 408 |
|
| 409 |
println "-- Posttokenize XSL Step with $xslDirectory"
|
| 410 |
return ApplyXsl2.processWithMultipleXSL(filesToProcess, xslDirectory, xslParams)
|
| 411 |
} |
| 412 |
|
| 413 |
public boolean doInjectMetadataStep() { |
| 414 |
if (metadata != null) { |
| 415 |
println("-- INJECTING METADATA - "+metadata.getHeadersList()+" in texts of directory "+outputDirectory) |
| 416 |
ConsoleProgressBar cpb = new ConsoleProgressBar(filesToProcess.size())
|
| 417 |
for (File infile : filesToProcess) { |
| 418 |
cpb.tick() |
| 419 |
File outfile = File.createTempFile("temp", ".xml", infile.getParentFile()); |
| 420 |
if (metadata.injectMetadatasInXml(infile, outfile, "text", null)) { |
| 421 |
if (!(infile.delete() && outfile.renameTo(infile))) println "Warning can't rename file "+outfile+" to "+infile |
| 422 |
if (!infile.exists()) {
|
| 423 |
println "Error: could not replace $infile by $outfile"
|
| 424 |
return false; |
| 425 |
} |
| 426 |
} else {
|
| 427 |
outfile.delete(); // fail
|
| 428 |
} |
| 429 |
} |
| 430 |
cpb.done() |
| 431 |
} |
| 432 |
return true; |
| 433 |
} |
| 434 |
|
| 435 |
public boolean doToXMLTXMStep() { |
| 436 |
filesToProcess = new File(module.getBinaryDirectory(),"tokenized").listFiles() |
| 437 |
ConsoleProgressBar cpb = new ConsoleProgressBar(filesToProcess.size())
|
| 438 |
println("-- Building XML-TXM ("+filesToProcess.size()+" files)") |
| 439 |
for (File f : filesToProcess) { |
| 440 |
//ArrayList<String> milestones = new ArrayList<String>();
|
| 441 |
cpb.tick() |
| 442 |
File file = f;
|
| 443 |
String txmfile = f.getName();
|
| 444 |
|
| 445 |
def correspType = new HashMap<String,String>() |
| 446 |
def correspRef = new HashMap<String,String>() |
| 447 |
//il faut lister les id de tous les respStmt
|
| 448 |
def respId = []; |
| 449 |
//fait la correspondance entre le respId et le rapport d'execution de l'outil
|
| 450 |
def applications = new HashMap<String,HashMap<String,String>>(); |
| 451 |
//fait la correspondance entre le respId et les attributs type de la propriété ana du w txm
|
| 452 |
//pour construire les ref vers les taxonomies
|
| 453 |
def taxonomiesUtilisees = new HashMap<String,String[]>(); |
| 454 |
//associe un id d'item avec sa description et son URI
|
| 455 |
def itemsURI = new HashMap<String,HashMap<String,String>>(); |
| 456 |
//informations de respStmt
|
| 457 |
//resps (respId <voir ci-dessus>, [description, person, date])
|
| 458 |
def resps = new HashMap<String,String[]>(); |
| 459 |
//lance le traitement
|
| 460 |
def builder = new Xml2Ana(file); |
| 461 |
builder.setWordTag(module.getProject().getTokenizerWordElement()) |
| 462 |
builder.setConvertAllAtrtibutes true;
|
| 463 |
builder.setCorrespondances(correspRef, correspType); |
| 464 |
builder.setHeaderInfos(respId,resps, applications, taxonomiesUtilisees, itemsURI) |
| 465 |
//builder.setAddTEIHeader();
|
| 466 |
if (!builder.process(new File(outputDirectory, txmfile))) { |
| 467 |
println("Failed to process "+f);
|
| 468 |
new File(outputDirectory,txmfile).delete(); |
| 469 |
} |
| 470 |
} |
| 471 |
filesToProcess = outputDirectory.listFiles(); |
| 472 |
cpb.done(); |
| 473 |
return true; |
| 474 |
} |
| 475 |
|
| 476 |
@Override
|
| 477 |
public void checkFiles() { |
| 478 |
//isSuccessFul = isSuccessFul && outputDirectory.listFiles() != null;
|
| 479 |
} |
| 480 |
|
| 481 |
@Override
|
| 482 |
public void cancel() { |
| 483 |
cancelNow = true;
|
| 484 |
} |
| 485 |
} |