private List<Tuple<Text, Integer>> getExpectedFacetEntries(
Set<String> fieldValues,
Map<String, Integer> controlFacetsField,
int size,
TermsFacet.ComparatorType sort,
List<String> excludes,
String regex,
boolean allTerms) {
Pattern pattern = null;
if (regex != null) {
pattern = Regex.compile(regex, null);
}
List<Tuple<Text, Integer>> entries = new ArrayList<Tuple<Text, Integer>>();
for (Map.Entry<String, Integer> e : controlFacetsField.entrySet()) {
if (excludes.contains(e.getKey())) {
continue;
}
if (pattern != null && !pattern.matcher(e.getKey()).matches()) {
continue;
}
entries.add(new Tuple<Text, Integer>(new StringText(e.getKey()), e.getValue()));
}
if (allTerms) {
for (String fieldValue : fieldValues) {
if (!controlFacetsField.containsKey(fieldValue)) {
if (excludes.contains(fieldValue)) {
continue;
}
if (pattern != null && !pattern.matcher(fieldValue).matches()) {
continue;
}
entries.add(new Tuple<Text, Integer>(new StringText(fieldValue), 0));
}
}
}
switch (sort) {
case COUNT:
Collections.sort(entries, count);
break;
case REVERSE_COUNT:
Collections.sort(entries, count_reverse);
break;
case TERM:
Collections.sort(entries, term);
break;
case REVERSE_TERM:
Collections.sort(entries, term_reverse);
break;
}
return size >= entries.size() ? entries : entries.subList(0, size);
}
@Inject
public PatternTokenizerFactory(
IndexSettings indexSettings, @Assisted String name, @Assisted Settings settings) {
super(indexSettings, name, settings);
String sPattern = settings.get("pattern", "\\W+" /*PatternAnalyzer.NON_WORD_PATTERN*/);
if (sPattern == null) {
throw new IllegalArgumentException(
"pattern is missing for [" + name + "] tokenizer of type 'pattern'");
}
this.pattern = Regex.compile(sPattern, settings.get("flags"));
this.group = settings.getAsInt("group", -1);
}
@Override
public FacetExecutor parse(String facetName, XContentParser parser, SearchContext context)
throws IOException {
String field = null;
int size = 10;
int shardSize = -1;
String[] fieldsNames = null;
ImmutableSet<BytesRef> excluded = ImmutableSet.of();
String regex = null;
String regexFlags = null;
TermsFacet.ComparatorType comparatorType = TermsFacet.ComparatorType.COUNT;
String scriptLang = null;
String script = null;
ScriptService.ScriptType scriptType = null;
Map<String, Object> params = null;
boolean allTerms = false;
String executionHint = null;
String currentFieldName = null;
XContentParser.Token token;
while ((token = parser.nextToken()) != XContentParser.Token.END_OBJECT) {
if (token == XContentParser.Token.FIELD_NAME) {
currentFieldName = parser.currentName();
} else if (token == XContentParser.Token.START_OBJECT) {
if ("params".equals(currentFieldName)) {
params = parser.map();
} else {
throw new ElasticsearchParseException(
"unknown parameter ["
+ currentFieldName
+ "] while parsing terms facet ["
+ facetName
+ "]");
}
} else if (token == XContentParser.Token.START_ARRAY) {
if ("exclude".equals(currentFieldName)) {
ImmutableSet.Builder<BytesRef> builder = ImmutableSet.builder();
while ((token = parser.nextToken()) != XContentParser.Token.END_ARRAY) {
builder.add(parser.bytes());
}
excluded = builder.build();
} else if ("fields".equals(currentFieldName)) {
List<String> fields = Lists.newArrayListWithCapacity(4);
while ((token = parser.nextToken()) != XContentParser.Token.END_ARRAY) {
fields.add(parser.text());
}
fieldsNames = fields.toArray(new String[fields.size()]);
} else {
throw new ElasticsearchParseException(
"unknown parameter ["
+ currentFieldName
+ "] while parsing terms facet ["
+ facetName
+ "]");
}
} else if (token.isValue()) {
if ("field".equals(currentFieldName)) {
field = parser.text();
} else if (ScriptService.SCRIPT_INLINE.match(currentFieldName)) {
script = parser.text();
scriptType = ScriptService.ScriptType.INLINE;
} else if (ScriptService.SCRIPT_ID.match(currentFieldName)) {
script = parser.text();
scriptType = ScriptService.ScriptType.INDEXED;
} else if (ScriptService.SCRIPT_FILE.match(currentFieldName)) {
script = parser.text();
scriptType = ScriptService.ScriptType.FILE;
} else if (ScriptService.SCRIPT_LANG.match(currentFieldName)) {
scriptLang = parser.text();
} else if ("size".equals(currentFieldName)) {
size = parser.intValue();
} else if ("shard_size".equals(currentFieldName) || "shardSize".equals(currentFieldName)) {
shardSize = parser.intValue();
} else if ("all_terms".equals(currentFieldName) || "allTerms".equals(currentFieldName)) {
allTerms = parser.booleanValue();
} else if ("regex".equals(currentFieldName)) {
regex = parser.text();
} else if ("regex_flags".equals(currentFieldName)
|| "regexFlags".equals(currentFieldName)) {
regexFlags = parser.text();
} else if ("order".equals(currentFieldName) || "comparator".equals(currentFieldName)) {
comparatorType = TermsFacet.ComparatorType.fromString(parser.text());
} else if ("execution_hint".equals(currentFieldName)
|| "executionHint".equals(currentFieldName)) {
executionHint = parser.textOrNull();
} else {
throw new ElasticsearchParseException(
"unknown parameter ["
+ currentFieldName
+ "] while parsing terms facet ["
+ facetName
+ "]");
}
}
}
if (fieldsNames != null && fieldsNames.length == 1) {
field = fieldsNames[0];
fieldsNames = null;
}
Pattern pattern = null;
if (regex != null) {
pattern = Regex.compile(regex, regexFlags);
}
SearchScript searchScript = null;
if (script != null) {
searchScript =
context.scriptService().search(context.lookup(), scriptLang, script, scriptType, params);
}
// shard_size cannot be smaller than size as we need to at least fetch <size> entries from every
// shards in order to return <size>
if (shardSize < size) {
shardSize = size;
}
if (fieldsNames != null) {
// in case of multi files, we only collect the fields that are mapped and facet on them.
ArrayList<FieldMapper> mappers = new ArrayList<>(fieldsNames.length);
for (int i = 0; i < fieldsNames.length; i++) {
FieldMapper mapper = context.smartNameFieldMapper(fieldsNames[i]);
if (mapper != null) {
mappers.add(mapper);
}
}
if (mappers.isEmpty()) {
// non of the fields is mapped
return new UnmappedFieldExecutor(size, comparatorType);
}
return new FieldsTermsStringFacetExecutor(
mappers.toArray(new FieldMapper[mappers.size()]),
size,
shardSize,
comparatorType,
allTerms,
context,
excluded,
pattern,
searchScript);
}
if (field == null && script != null) {
return new ScriptTermsStringFieldFacetExecutor(
size,
shardSize,
comparatorType,
context,
excluded,
pattern,
scriptLang,
script,
scriptType,
params,
context.cacheRecycler());
}
if (field == null) {
throw new ElasticsearchParseException(
"terms facet [" + facetName + "] must have a field, fields or script parameter");
}
FieldMapper fieldMapper = context.smartNameFieldMapper(field);
if (fieldMapper == null) {
return new UnmappedFieldExecutor(size, comparatorType);
}
IndexFieldData indexFieldData = context.fieldData().getForField(fieldMapper);
if (indexFieldData instanceof IndexNumericFieldData) {
IndexNumericFieldData indexNumericFieldData = (IndexNumericFieldData) indexFieldData;
if (indexNumericFieldData.getNumericType().isFloatingPoint()) {
return new TermsDoubleFacetExecutor(
indexNumericFieldData,
size,
shardSize,
comparatorType,
allTerms,
context,
excluded,
searchScript,
context.cacheRecycler());
} else {
return new TermsLongFacetExecutor(
indexNumericFieldData,
size,
shardSize,
comparatorType,
allTerms,
context,
excluded,
searchScript,
context.cacheRecycler());
}
} else {
if (script != null || "map".equals(executionHint)) {
return new TermsStringFacetExecutor(
indexFieldData,
size,
shardSize,
comparatorType,
allTerms,
context,
excluded,
pattern,
searchScript);
} else if (indexFieldData instanceof IndexOrdinalsFieldData) {
return new TermsStringOrdinalsFacetExecutor(
(IndexOrdinalsFieldData) indexFieldData,
size,
shardSize,
comparatorType,
allTerms,
context,
excluded,
pattern,
ordinalsCacheAbove);
} else {
return new TermsStringFacetExecutor(
indexFieldData,
size,
shardSize,
comparatorType,
allTerms,
context,
excluded,
pattern,
searchScript);
}
}
}