Home | History | Annotate | Download | only in unittest 
      1 package org.unicode.cldr.unittest;
      2 
      3 import java.io.BufferedReader;
      4 import java.io.IOException;
      5 import java.util.Collection;
      6 import java.util.List;
      7 import java.util.Map;
      8 import java.util.Map.Entry;
      9 import java.util.Set;
     10 
     11 import org.unicode.cldr.draft.FileUtilities;
     12 import org.unicode.cldr.util.CLDRConfig;
     13 import org.unicode.cldr.util.CLDRFile;
     14 import org.unicode.cldr.util.CldrUtility;
     15 import org.unicode.cldr.util.SupplementalDataInfo;
     16 
     17 import com.google.common.base.Splitter;
     18 import com.google.common.collect.ImmutableSet;
     19 import com.google.common.collect.ImmutableSetMultimap;
     20 import com.google.common.collect.Multimap;
     21 import com.google.common.collect.TreeMultimap;
     22 import com.ibm.icu.impl.Row.R2;
     23 import com.ibm.icu.util.ICUUncheckedIOException;
     24 
     25 public class TestUnContainment extends TestFmwkPlus {
     26     static CLDRConfig testInfo = CLDRConfig.getInstance();
     27     private static final SupplementalDataInfo SUPPLEMENTAL_DATA_INFO = testInfo.getSupplementalDataInfo();
     28     Map<String, R2<List<String>, String>> regionToInfo = SUPPLEMENTAL_DATA_INFO
     29         .getLocaleAliasInfo()
     30         .get("territory");
     31 
     32     private static final Set<String> NOT_CLDR_TERRITORY_CODES = ImmutableSet.of("830"); // Channel Islands
     33     private static final Set<String> KNOWN_CONTAINMENT_EXCEPTIONS = ImmutableSet.of("AQ","680"); // Antarctica, Sark
     34 
     35     final Multimap<String, String> UnChildToParent;
     36     {
     37         Multimap<String, String> _UnChildToParent = TreeMultimap.create();
     38         Splitter tab = Splitter.on('\t').trimResults();
     39         try (BufferedReader unCodes = CldrUtility.getUTF8Data("external/UnCodes.txt");) {
     40             for (String line : FileUtilities.in(unCodes)) {
     41                 List<String> items = tab.splitToList(line);
     42                 if (line.isEmpty() || line.startsWith("Global Code")) {
     43                     continue;
     44                 }
     45                 String parent = null;
     46                 for (int i = 0; i < 10; i += 2) {
     47                     String region = items.get(i);
     48                     if (!region.isEmpty()) {
     49                         region = unToCldrCode(region);
     50                         if (parent != null && region != null){
     51                             _UnChildToParent.put(region, parent);
     52                         }
     53                         if (region != null) {
     54                             parent = region;
     55                         }
     56                     }
     57                     if (i == 6) {
     58                         ++i; // hack because last two are out of order
     59                     }
     60                 }
     61             }
     62             UnChildToParent = ImmutableSetMultimap.copyOf(_UnChildToParent);
     63         } catch (IOException e) {
     64             throw new ICUUncheckedIOException(e);
     65         }
     66     }
     67 
     68     public static void main(String[] args) {
     69         new TestUnContainment().run(args);
     70     }
     71 
     72     private String name(Collection<String> codes) {
     73         StringBuilder result = new StringBuilder();
     74         for (String code : codes) {
     75             if (result.length() != 0) {
     76                 result.append(", ");
     77             }
     78             result.append(name(code));
     79         }
     80         return result.toString();
     81     }
     82 
     83     private String name(String code) {
     84         String name = testInfo.getEnglish().getName(CLDRFile.TERRITORY_NAME, code);
     85         return name + " (" + code + ")";
     86     }
     87 
     88     private String unToCldrCode(String code) {
     89 
     90         if (NOT_CLDR_TERRITORY_CODES.contains(code)) {
     91             return null;
     92         }
     93 
     94         R2<List<String>, String> codeInfo = regionToInfo.get(code);
     95         if (codeInfo != null) {
     96             if (codeInfo.get0() != null && !codeInfo.get0().isEmpty()) {
     97                 code = codeInfo.get0().get(0);
     98             }
     99         }
    100         return code;
    101     }
    102 
    103     public void TestContainment() {
    104 
    105         /*
    106         CLDR
    107         <group type="001" contains="019 002 150 142 009"/> <!--World -->
    108         <group type="001" contains="EU EZ UN" status="grouping"/> <!--European Union, Eurozone, United Nations -->
    109         <group type="001" contains="QU" status="deprecated"/> <!--European Union -->
    110         <group type="011" contains="BF BJ CI CV GH GM GN GW LR ML MR NE NG SH SL SN TG"/> <!--Western Africa -->
    111          */
    112         for (Entry<String, Collection<String>> entry : UnChildToParent.asMap().entrySet()) {
    113             Collection<String> unParents = entry.getValue();
    114             String unChild = entry.getKey();
    115             //System.out.println(name(unParents) + "\t" + name(unChild));
    116             for (String unParent : unParents) {
    117                 Set<String> children = SUPPLEMENTAL_DATA_INFO.getContained(unParent);
    118                 if (children != null && children.contains(unChild)) {
    119                     continue;
    120                 }
    121                 // See CLDR ticket 10187 for rationalization on the known containment exceptions.
    122                 if (KNOWN_CONTAINMENT_EXCEPTIONS.contains(unChild)) {
    123                     continue;
    124                 }
    125                 msg("UN containment doesn't match CLDR for " + name(unParent)
    126                     + ": cldr children " + children
    127                     + " don't contain UN " + name(unChild), ERR, true, true);
    128             }
    129         }
    130     }
    131 }
    132