use of com.ibm.icu.text.RuleBasedCollator in project lucene-solr by apache.
the class TestICUCollationField method setupSolrHome.
/**
* Ugly: but what to do? We want to test custom sort, which reads rules in as a resource.
* These are largish files, and jvm-specific (as our documentation says, you should always
* look out for jvm differences with collation).
* So it's preferable to create this file on-the-fly.
*/
public static String setupSolrHome() throws Exception {
String tmpFile = createTempDir().toFile().getAbsolutePath();
// make data and conf dirs
new File(tmpFile + "/collection1", "data").mkdirs();
File confDir = new File(tmpFile + "/collection1", "conf");
confDir.mkdirs();
// copy over configuration files
FileUtils.copyFile(getFile("analysis-extras/solr/collection1/conf/solrconfig-icucollate.xml"), new File(confDir, "solrconfig.xml"));
FileUtils.copyFile(getFile("analysis-extras/solr/collection1/conf/schema-icucollate.xml"), new File(confDir, "schema.xml"));
// generate custom collation rules (DIN 5007-2), saving to customrules.dat
RuleBasedCollator baseCollator = (RuleBasedCollator) Collator.getInstance(new ULocale("de", "DE"));
String DIN5007_2_tailorings = "& ae , ä & AE , Ä" + "& oe , ö & OE , Ö" + "& ue , ü & UE , ü";
RuleBasedCollator tailoredCollator = new RuleBasedCollator(baseCollator.getRules() + DIN5007_2_tailorings);
String tailoredRules = tailoredCollator.getRules();
final String osFileName = "customrules.dat";
final FileOutputStream os = new FileOutputStream(new File(confDir, osFileName));
IOUtils.write(tailoredRules, os, "UTF-8");
os.close();
final ResourceLoader loader;
if (random().nextBoolean()) {
loader = new StringMockResourceLoader(tailoredRules);
} else {
loader = new FilesystemResourceLoader(confDir.toPath());
}
final Collator readCollator = ICUCollationField.createFromRules(osFileName, loader);
assertEquals(tailoredCollator, readCollator);
return tmpFile;
}
Aggregations