Changeset 33515 for gs3-extensions
- Timestamp:
- 2019-09-24T19:50:40+12:00 (5 years ago)
- File:
-
- 1 edited
Legend:
- Unmodified
- Added
- Removed
-
gs3-extensions/maori-lang-detection/src/org/greenstone/atea/CCWETProcessor.java
r33503 r33515 334 334 335 335 } 336 /*337 public boolean isInDiscardFilter(String url) {338 String discardFilterFile = "url-discard-filter.txt"; // in conf folder339 340 try (341 BufferedReader reader = new BufferedReader(new InputStreamReader(org.greenstone.atea.CCWETProcessor.class.getClassLoader().getResourceAsStream(discardFilterFile), "UTF-8"));342 ) {343 String filter = null;344 while((filter = reader.readLine()) != null) {345 if(filter.trim().equals("")) {346 continue;347 }348 //System.err.println("Got filter: " + filter);349 if(filter.startsWith("^") && filter.endsWith("$") && url.equals(filter.substring(1, filter.length()-1))) {350 System.err.println("*** Discarding url " + url + "\n\tas it MATCHES filter " + filter);351 }352 else if(filter.startsWith("^") && url.startsWith(filter.substring(1))) {353 System.err.println("*** Discarding url " + url + "\n\tas it STARTS WITH filter " + filter);354 return true;355 }356 else if(filter.endsWith("$") && url.endsWith(filter.substring(0, filter.length()-1))) {357 System.err.println("*** Discarding url " + url + "\n\tas it ENDS WITH filter " + filter);358 return true;359 }360 else if(url.contains(filter)) {361 System.err.println("*** Discarding url " + url + "\n\tas it CONTAINS filter " + filter);362 return true;363 }364 365 }366 367 } catch (IOException ioe) {368 ioe.printStackTrace();369 System.err.println("\n@@@@@@@@@ Error reading from " + discardFilterFile);370 }371 372 return false;373 }*/374 336 375 337 /** Maintain a count of all WET files processed. */
Note:
See TracChangeset
for help on using the changeset viewer.