private void serializeObject(final ParseContext context, String currentFieldName)
throws IOException {
if (currentFieldName == null) {
throw new MapperParsingException(
"object mapping ["
+ name
+ "] trying to serialize an object with no field associated with it, current value ["
+ context.parser().textOrNull()
+ "]");
}
context.path().add(currentFieldName);
Mapper objectMapper = mappers.get(currentFieldName);
if (objectMapper != null) {
objectMapper.parse(context);
} else {
Dynamic dynamic = this.dynamic;
if (dynamic == null) {
dynamic = context.root().dynamic();
}
if (dynamic == Dynamic.STRICT) {
throw new StrictDynamicMappingException(fullPath, currentFieldName);
} else if (dynamic == Dynamic.TRUE) {
// we sync here just so we won't add it twice. Its not the end of the world
// to sync here since next operations will get it before
synchronized (mutex) {
objectMapper = mappers.get(currentFieldName);
if (objectMapper == null) {
// remove the current field name from path, since template search and the object builder
// add it as well...
context.path().remove();
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "object");
if (builder == null) {
builder = MapperBuilders.object(currentFieldName).enabled(true).pathType(pathType);
// if this is a non root object, then explicitly set the dynamic behavior if set
if (!(this instanceof RootObjectMapper) && this.dynamic != Defaults.DYNAMIC) {
((Builder) builder).dynamic(this.dynamic);
}
}
BuilderContext builderContext =
new BuilderContext(context.indexSettings(), context.path());
objectMapper = builder.build(builderContext);
putDynamicMapper(context, currentFieldName, objectMapper);
} else {
objectMapper.parse(context);
}
}
} else {
// not dynamic, read everything up to end object
context.parser().skipChildren();
}
}
context.path().remove();
}
private void serializeArray(ParseContext context, String lastFieldName) throws IOException {
String arrayFieldName = lastFieldName;
Mapper mapper = mappers.get(lastFieldName);
if (mapper != null && mapper instanceof ArrayValueMapperParser) {
mapper.parse(context);
} else {
XContentParser parser = context.parser();
XContentParser.Token token;
while ((token = parser.nextToken()) != XContentParser.Token.END_ARRAY) {
if (token == XContentParser.Token.START_OBJECT) {
serializeObject(context, lastFieldName);
} else if (token == XContentParser.Token.START_ARRAY) {
serializeArray(context, lastFieldName);
} else if (token == XContentParser.Token.FIELD_NAME) {
lastFieldName = parser.currentName();
} else if (token == XContentParser.Token.VALUE_NULL) {
serializeNullValue(context, lastFieldName);
} else if (token == null) {
throw new MapperParsingException(
"object mapping for ["
+ name
+ "] with array for ["
+ arrayFieldName
+ "] tried to parse as array, but got EOF, is there a mismatch in types for the same field?");
} else {
serializeValue(context, lastFieldName, token);
}
}
}
}
private void serializeValue(
final ParseContext context, String currentFieldName, XContentParser.Token token)
throws IOException {
if (currentFieldName == null) {
throw new MapperParsingException(
"object mapping ["
+ name
+ "] trying to serialize a value with no field associated with it, current value ["
+ context.parser().textOrNull()
+ "]");
}
Mapper mapper = mappers.get(currentFieldName);
if (mapper != null) {
mapper.parse(context);
} else {
parseDynamicValue(context, currentFieldName, token);
}
}
private void serializeValue(
final ParseContext context, String currentFieldName, XContentParser.Token token)
throws IOException {
if (currentFieldName == null) {
throw new MapperParsingException(
"object mapping ["
+ name
+ "] trying to serialize a value with no field associated with it, current value ["
+ context.parser().textOrNull()
+ "]");
}
Mapper mapper = mappers.get(currentFieldName);
if (mapper != null) {
mapper.parse(context);
return;
}
Dynamic dynamic = this.dynamic;
if (dynamic == null) {
dynamic = context.root().dynamic();
}
if (dynamic == Dynamic.STRICT) {
throw new StrictDynamicMappingException(fullPath, currentFieldName);
}
if (dynamic == Dynamic.FALSE) {
return;
}
// we sync here since we don't want to add this field twice to the document mapper
// its not the end of the world, since we add it to the mappers once we create it
// so next time we won't even get here for this field
boolean newMapper = false;
synchronized (mutex) {
mapper = mappers.get(currentFieldName);
if (mapper == null) {
newMapper = true;
BuilderContext builderContext = new BuilderContext(context.indexSettings(), context.path());
if (token == XContentParser.Token.VALUE_STRING) {
boolean resolved = false;
// do a quick test to see if its fits a dynamic template, if so, use it.
// we need to do it here so we can handle things like attachment templates, where calling
// text (to see if its a date) causes the binary value to be cleared
if (!resolved) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "string", null);
if (builder != null) {
mapper = builder.build(builderContext);
resolved = true;
}
}
if (!resolved && context.parser().textLength() == 0) {
// empty string with no mapping, treat it like null value
return;
}
if (!resolved && context.root().dateDetection()) {
String text = context.parser().text();
// a safe check since "1" gets parsed as well
if (Strings.countOccurrencesOf(text, ":") > 1
|| Strings.countOccurrencesOf(text, "-") > 1
|| Strings.countOccurrencesOf(text, "/") > 1) {
for (FormatDateTimeFormatter dateTimeFormatter :
context.root().dynamicDateTimeFormatters()) {
try {
dateTimeFormatter.parser().parseMillis(text);
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "date");
if (builder == null) {
builder = dateField(currentFieldName).dateTimeFormatter(dateTimeFormatter);
}
mapper = builder.build(builderContext);
resolved = true;
break;
} catch (Exception e) {
// failure to parse this, continue
}
}
}
}
if (!resolved && context.root().numericDetection()) {
String text = context.parser().text();
try {
Long.parseLong(text);
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "long");
if (builder == null) {
builder = longField(currentFieldName);
}
mapper = builder.build(builderContext);
resolved = true;
} catch (Exception e) {
// not a long number
}
if (!resolved) {
try {
Double.parseDouble(text);
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "double");
if (builder == null) {
builder = doubleField(currentFieldName);
}
mapper = builder.build(builderContext);
resolved = true;
} catch (Exception e) {
// not a long number
}
}
}
// DON'T do automatic ip detection logic, since it messes up with docs that have hosts and
// ips
// check if its an ip
// if (!resolved && text.indexOf('.') != -1) {
// try {
// IpFieldMapper.ipToLong(text);
// XContentMapper.Builder builder =
// context.root().findTemplateBuilder(context, currentFieldName, "ip");
// if (builder == null) {
// builder = ipField(currentFieldName);
// }
// mapper = builder.build(builderContext);
// resolved = true;
// } catch (Exception e) {
// // failure to parse, not ip...
// }
// }
if (!resolved) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "string");
if (builder == null) {
builder = stringField(currentFieldName);
}
mapper = builder.build(builderContext);
}
} else if (token == XContentParser.Token.VALUE_NUMBER) {
XContentParser.NumberType numberType = context.parser().numberType();
if (numberType == XContentParser.NumberType.INT) {
if (context.parser().estimatedNumberType()) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "long");
if (builder == null) {
builder = longField(currentFieldName);
}
mapper = builder.build(builderContext);
} else {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "integer");
if (builder == null) {
builder = integerField(currentFieldName);
}
mapper = builder.build(builderContext);
}
} else if (numberType == XContentParser.NumberType.LONG) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "long");
if (builder == null) {
builder = longField(currentFieldName);
}
mapper = builder.build(builderContext);
} else if (numberType == XContentParser.NumberType.FLOAT) {
if (context.parser().estimatedNumberType()) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "double");
if (builder == null) {
builder = doubleField(currentFieldName);
}
mapper = builder.build(builderContext);
} else {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "float");
if (builder == null) {
builder = floatField(currentFieldName);
}
mapper = builder.build(builderContext);
}
} else if (numberType == XContentParser.NumberType.DOUBLE) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "double");
if (builder == null) {
builder = doubleField(currentFieldName);
}
mapper = builder.build(builderContext);
}
} else if (token == XContentParser.Token.VALUE_BOOLEAN) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "boolean");
if (builder == null) {
builder = booleanField(currentFieldName);
}
mapper = builder.build(builderContext);
} else if (token == XContentParser.Token.VALUE_EMBEDDED_OBJECT) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "binary");
if (builder == null) {
builder = binaryField(currentFieldName);
}
mapper = builder.build(builderContext);
} else {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, null);
if (builder != null) {
mapper = builder.build(builderContext);
} else {
// TODO how do we identify dynamically that its a binary value?
throw new ElasticSearchIllegalStateException(
"Can't handle serializing a dynamic type with content token ["
+ token
+ "] and field name ["
+ currentFieldName
+ "]");
}
}
putMapper(mapper);
context.setMappingsModified();
}
}
if (newMapper) {
mapper.traverse(context.newFieldMappers());
}
mapper.parse(context);
}
public void parse(ParseContext context) throws IOException {
if (!enabled) {
context.parser().skipChildren();
return;
}
XContentParser parser = context.parser();
String currentFieldName = parser.currentName();
XContentParser.Token token = parser.currentToken();
if (token == XContentParser.Token.VALUE_NULL) {
// the object is null ("obj1" : null), simply bail
return;
}
if (token.isValue() && !allowValue()) {
// if we are parsing an object but it is just a value, its only allowed on root level parsers
// with there
// is a field name with the same name as the type
throw new MapperParsingException(
"object mapping for [" + name + "] tried to parse as object, but found a concrete value");
}
Document restoreDoc = null;
if (nested.isNested()) {
Document nestedDoc = new Document();
// pre add the uid field if possible (id was already provided)
IndexableField uidField = context.doc().getField(UidFieldMapper.NAME);
if (uidField != null) {
// we don't need to add it as a full uid field in nested docs, since we don't need
// versioning
// we also rely on this for UidField#loadVersion
// this is a deeply nested field
if (uidField.stringValue() != null) {
nestedDoc.add(
new Field(
UidFieldMapper.NAME,
uidField.stringValue(),
UidFieldMapper.Defaults.NESTED_FIELD_TYPE));
} else {
nestedDoc.add(
new Field(
UidFieldMapper.NAME,
((UidField) uidField).uid(),
UidFieldMapper.Defaults.NESTED_FIELD_TYPE));
}
}
// the type of the nested doc starts with __, so we can identify that its a nested one in
// filters
// note, we don't prefix it with the type of the doc since it allows us to execute a nested
// query
// across types (for example, with similar nested objects)
nestedDoc.add(
new Field(
TypeFieldMapper.NAME, nestedTypePathAsString, TypeFieldMapper.Defaults.FIELD_TYPE));
restoreDoc = context.switchDoc(nestedDoc);
context.addDoc(nestedDoc);
}
ContentPath.Type origPathType = context.path().pathType();
context.path().pathType(pathType);
// if we are at the end of the previous object, advance
if (token == XContentParser.Token.END_OBJECT) {
token = parser.nextToken();
}
if (token == XContentParser.Token.START_OBJECT) {
// if we are just starting an OBJECT, advance, this is the object we are parsing, we need the
// name first
token = parser.nextToken();
}
while (token != XContentParser.Token.END_OBJECT) {
if (token == XContentParser.Token.START_OBJECT) {
serializeObject(context, currentFieldName);
} else if (token == XContentParser.Token.START_ARRAY) {
serializeArray(context, currentFieldName);
} else if (token == XContentParser.Token.FIELD_NAME) {
currentFieldName = parser.currentName();
} else if (token == XContentParser.Token.VALUE_NULL) {
serializeNullValue(context, currentFieldName);
} else if (token == null) {
throw new MapperParsingException(
"object mapping for ["
+ name
+ "] tried to parse as object, but got EOF, has a concrete value been provided to it?");
} else if (token.isValue()) {
serializeValue(context, currentFieldName, token);
}
token = parser.nextToken();
}
// restore the enable path flag
context.path().pathType(origPathType);
if (nested.isNested()) {
Document nestedDoc = context.switchDoc(restoreDoc);
if (nested.isIncludeInParent()) {
for (IndexableField field : nestedDoc.getFields()) {
if (field.name().equals(UidFieldMapper.NAME)
|| field.name().equals(TypeFieldMapper.NAME)) {
continue;
} else {
context.doc().add(field);
}
}
}
if (nested.isIncludeInRoot()) {
// don't add it twice, if its included in parent, and we are handling the master doc...
if (!(nested.isIncludeInParent() && context.doc() == context.rootDoc())) {
for (IndexableField field : nestedDoc.getFields()) {
if (field.name().equals(UidFieldMapper.NAME)
|| field.name().equals(TypeFieldMapper.NAME)) {
continue;
} else {
context.rootDoc().add(field);
}
}
}
}
}
}
@Override
public void parse(ParseContext context) throws IOException {
String iri;
XContentParser parser = context.parser();
XContentParser.Token token = parser.currentToken();
if (token == XContentParser.Token.VALUE_STRING) {
iri = parser.text();
} else {
throw new MapperParsingException(name() + " does not contain String value");
}
ContentPath.Type origPathType = context.path().pathType();
context.path().pathType(ContentPath.Type.FULL);
context.path().add(names.name());
try {
OntologyHelper helper = getHelper(ontologySettings, threadPool);
OntologyData data = findOntologyData(helper, iri);
if (data == null) {
logger.debug("Cannot find OWL class for IRI {}", iri);
} else {
addFieldData(
context,
getPredefinedMapper(FieldMappings.URI, context),
Collections.singletonList(iri));
// Look up the label(s)
addFieldData(context, getPredefinedMapper(FieldMappings.LABEL, context), data.getLabels());
// Look up the synonyms
addFieldData(
context, getPredefinedMapper(FieldMappings.SYNONYMS, context), data.getLabels());
// Add the child details
addRelatedNodesWithLabels(
context,
data.getChildIris(),
getPredefinedMapper(FieldMappings.CHILD_URI, context),
data.getChildLabels(),
getPredefinedMapper(FieldMappings.CHILD_LABEL, context));
// Add the parent details
addRelatedNodesWithLabels(
context,
data.getParentIris(),
getPredefinedMapper(FieldMappings.PARENT_URI, context),
data.getParentLabels(),
getPredefinedMapper(FieldMappings.PARENT_LABEL, context));
if (ontologySettings.isIncludeIndirect()) {
// Add the descendant details
addRelatedNodesWithLabels(
context,
data.getDescendantIris(),
getPredefinedMapper(FieldMappings.DESCENDANT_URI, context),
data.getDescendantLabels(),
getPredefinedMapper(FieldMappings.DESCENDANT_LABEL, context));
// Add the ancestor details
addRelatedNodesWithLabels(
context,
data.getAncestorIris(),
getPredefinedMapper(FieldMappings.ANCESTOR_URI, context),
data.getAncestorLabels(),
getPredefinedMapper(FieldMappings.ANCESTOR_LABEL, context));
}
if (ontologySettings.isIncludeRelations()) {
// Add the related nodes
Map<String, Collection<String>> relations = data.getRelationIris();
for (String relation : relations.keySet()) {
// Sanitise the relation name
String sanRelation = relation.replaceAll("\\W+", "_");
String uriMapperName = sanRelation + DYNAMIC_URI_FIELD_SUFFIX;
String labelMapperName = sanRelation + DYNAMIC_LABEL_FIELD_SUFFIX;
// Get the mapper for the relation
FieldMapper<String> uriMapper =
mappers.get(context.path().fullPathAsText(uriMapperName));
FieldMapper<String> labelMapper =
mappers.get(context.path().fullPathAsText(labelMapperName));
if (uriMapper == null) {
// No mappers created yet - build new ones for URI and label
BuilderContext builderContext =
new BuilderContext(context.indexSettings(), context.path());
uriMapper =
MapperBuilders.stringField(uriMapperName)
.store(true)
.index(true)
.tokenized(false)
.build(builderContext);
labelMapper =
MapperBuilders.stringField(labelMapperName)
.store(true)
.index(true)
.tokenized(true)
.build(builderContext);
}
addRelatedNodesWithLabels(
context,
relations.get(relation),
uriMapper,
helper.findLabelsForIRIs(relations.get(relation)),
labelMapper);
}
}
}
helper.updateLastCallTime();
} catch (OntologyHelperException e) {
throw new ElasticsearchException("Could not initialise ontology helper", e);
} finally {
context.path().remove();
context.path().pathType(origPathType);
}
}
public void parseDynamicValue(
final ParseContext context, String currentFieldName, XContentParser.Token token)
throws IOException {
Dynamic dynamic = this.dynamic;
if (dynamic == null) {
dynamic = context.root().dynamic();
}
if (dynamic == Dynamic.STRICT) {
throw new StrictDynamicMappingException(fullPath, currentFieldName);
}
if (dynamic == Dynamic.FALSE) {
return;
}
// we sync here since we don't want to add this field twice to the document mapper
// its not the end of the world, since we add it to the mappers once we create it
// so next time we won't even get here for this field
synchronized (mutex) {
Mapper mapper = mappers.get(currentFieldName);
if (mapper == null) {
BuilderContext builderContext = new BuilderContext(context.indexSettings(), context.path());
if (token == XContentParser.Token.VALUE_STRING) {
boolean resolved = false;
// do a quick test to see if its fits a dynamic template, if so, use it.
// we need to do it here so we can handle things like attachment templates, where calling
// text (to see if its a date) causes the binary value to be cleared
if (!resolved) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "string", null);
if (builder != null) {
mapper = builder.build(builderContext);
resolved = true;
}
}
if (!resolved && context.parser().textLength() == 0) {
// empty string with no mapping, treat it like null value
return;
}
if (!resolved && context.root().dateDetection()) {
String text = context.parser().text();
// a safe check since "1" gets parsed as well
if (Strings.countOccurrencesOf(text, ":") > 1
|| Strings.countOccurrencesOf(text, "-") > 1
|| Strings.countOccurrencesOf(text, "/") > 1) {
for (FormatDateTimeFormatter dateTimeFormatter :
context.root().dynamicDateTimeFormatters()) {
try {
dateTimeFormatter.parser().parseMillis(text);
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "date");
if (builder == null) {
builder = dateField(currentFieldName).dateTimeFormatter(dateTimeFormatter);
}
mapper = builder.build(builderContext);
resolved = true;
break;
} catch (Exception e) {
// failure to parse this, continue
}
}
}
}
if (!resolved && context.root().numericDetection()) {
String text = context.parser().text();
try {
Long.parseLong(text);
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "long");
if (builder == null) {
builder = longField(currentFieldName);
}
mapper = builder.build(builderContext);
resolved = true;
} catch (Exception e) {
// not a long number
}
if (!resolved) {
try {
Double.parseDouble(text);
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "double");
if (builder == null) {
builder = doubleField(currentFieldName);
}
mapper = builder.build(builderContext);
resolved = true;
} catch (Exception e) {
// not a long number
}
}
}
if (!resolved) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "string");
if (builder == null) {
builder = stringField(currentFieldName);
}
mapper = builder.build(builderContext);
}
} else if (token == XContentParser.Token.VALUE_NUMBER) {
XContentParser.NumberType numberType = context.parser().numberType();
if (numberType == XContentParser.NumberType.INT) {
if (context.parser().estimatedNumberType()) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "long");
if (builder == null) {
builder = longField(currentFieldName);
}
mapper = builder.build(builderContext);
} else {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "integer");
if (builder == null) {
builder = integerField(currentFieldName);
}
mapper = builder.build(builderContext);
}
} else if (numberType == XContentParser.NumberType.LONG) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "long");
if (builder == null) {
builder = longField(currentFieldName);
}
mapper = builder.build(builderContext);
} else if (numberType == XContentParser.NumberType.FLOAT) {
if (context.parser().estimatedNumberType()) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "double");
if (builder == null) {
builder = doubleField(currentFieldName);
}
mapper = builder.build(builderContext);
} else {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "float");
if (builder == null) {
builder = floatField(currentFieldName);
}
mapper = builder.build(builderContext);
}
} else if (numberType == XContentParser.NumberType.DOUBLE) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "double");
if (builder == null) {
builder = doubleField(currentFieldName);
}
mapper = builder.build(builderContext);
}
} else if (token == XContentParser.Token.VALUE_BOOLEAN) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "boolean");
if (builder == null) {
builder = booleanField(currentFieldName);
}
mapper = builder.build(builderContext);
} else if (token == XContentParser.Token.VALUE_EMBEDDED_OBJECT) {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, "binary");
if (builder == null) {
builder = binaryField(currentFieldName);
}
mapper = builder.build(builderContext);
} else {
Mapper.Builder builder =
context.root().findTemplateBuilder(context, currentFieldName, null);
if (builder != null) {
mapper = builder.build(builderContext);
} else {
// TODO how do we identify dynamically that its a binary value?
throw new ElasticsearchIllegalStateException(
"Can't handle serializing a dynamic type with content token ["
+ token
+ "] and field name ["
+ currentFieldName
+ "]");
}
}
if (context.isWithinNewMapper()) {
mapper.parse(context);
} else {
context.setWithinNewMapper();
try {
mapper.parse(context);
FieldMapperListener.Aggregator newFields = new FieldMapperListener.Aggregator();
mapper.traverse(newFields);
context.docMapper().addFieldMappers(newFields.mappers);
} finally {
context.clearWithinNewMapper();
}
}
// only put after we traversed and did the callbacks, so other parsing won't see it only
// after we
// properly traversed it and adding the mappers
putMapper(mapper);
context.setMappingsModified();
} else {
mapper.parse(context);
}
}
}
@Override
public Mapper parse(ParseContext context) throws IOException {
byte[] content = null;
String contentType = null;
int indexedChars = defaultIndexedChars;
boolean langDetect = defaultLangDetect;
String name = null;
String language = null;
XContentParser parser = context.parser();
XContentParser.Token token = parser.currentToken();
if (token == XContentParser.Token.VALUE_STRING) {
content = parser.binaryValue();
} else {
String currentFieldName = null;
while ((token = parser.nextToken()) != XContentParser.Token.END_OBJECT) {
if (token == XContentParser.Token.FIELD_NAME) {
currentFieldName = parser.currentName();
} else if (token == XContentParser.Token.VALUE_STRING) {
if ("_content".equals(currentFieldName)) {
content = parser.binaryValue();
} else if ("_content_type".equals(currentFieldName)) {
contentType = parser.text();
} else if ("_name".equals(currentFieldName)) {
name = parser.text();
} else if ("_language".equals(currentFieldName)) {
language = parser.text();
}
} else if (token == XContentParser.Token.VALUE_NUMBER) {
if ("_indexed_chars".equals(currentFieldName)
|| "_indexedChars".equals(currentFieldName)) {
indexedChars = parser.intValue();
}
} else if (token == XContentParser.Token.VALUE_BOOLEAN) {
if ("_detect_language".equals(currentFieldName)
|| "_detectLanguage".equals(currentFieldName)) {
langDetect = parser.booleanValue();
}
}
}
}
// Throw clean exception when no content is provided Fix #23
if (content == null) {
throw new MapperParsingException("No content is provided.");
}
Metadata metadata = new Metadata();
if (contentType != null) {
metadata.add(Metadata.CONTENT_TYPE, contentType);
}
if (name != null) {
metadata.add(Metadata.RESOURCE_NAME_KEY, name);
}
String parsedContent;
try {
parsedContent = TikaImpl.parse(content, metadata, indexedChars);
} catch (Throwable e) {
// #18: we could ignore errors when Tika does not parse data
if (!ignoreErrors) {
logger.trace("exception caught", e);
throw new MapperParsingException(
"Failed to extract ["
+ indexedChars
+ "] characters of text for ["
+ name
+ "] : "
+ e.getMessage(),
e);
} else {
logger.debug(
"Failed to extract [{}] characters of text for [{}]: [{}]",
indexedChars,
name,
e.getMessage());
logger.trace("exception caught", e);
}
return null;
}
context = context.createExternalValueContext(parsedContent);
contentMapper.parse(context);
if (langDetect) {
try {
if (language != null) {
metadata.add(Metadata.CONTENT_LANGUAGE, language);
} else {
LanguageIdentifier identifier = new LanguageIdentifier(parsedContent);
language = identifier.getLanguage();
}
context = context.createExternalValueContext(language);
languageMapper.parse(context);
} catch (Throwable t) {
logger.debug("Cannot detect language: [{}]", t.getMessage());
}
}
if (name != null) {
try {
context = context.createExternalValueContext(name);
nameMapper.parse(context);
} catch (MapperParsingException e) {
if (!ignoreErrors) throw e;
if (logger.isDebugEnabled())
logger.debug(
"Ignoring MapperParsingException catch while parsing name: [{}]", e.getMessage());
}
}
if (metadata.get(Metadata.DATE) != null) {
try {
context = context.createExternalValueContext(metadata.get(Metadata.DATE));
dateMapper.parse(context);
} catch (MapperParsingException e) {
if (!ignoreErrors) throw e;
if (logger.isDebugEnabled())
logger.debug(
"Ignoring MapperParsingException catch while parsing date: [{}]: [{}]",
e.getMessage(),
context.externalValue());
}
}
if (metadata.get(Metadata.TITLE) != null) {
try {
context = context.createExternalValueContext(metadata.get(Metadata.TITLE));
titleMapper.parse(context);
} catch (MapperParsingException e) {
if (!ignoreErrors) throw e;
if (logger.isDebugEnabled())
logger.debug(
"Ignoring MapperParsingException catch while parsing title: [{}]: [{}]",
e.getMessage(),
context.externalValue());
}
}
if (metadata.get(Metadata.AUTHOR) != null) {
try {
context = context.createExternalValueContext(metadata.get(Metadata.AUTHOR));
authorMapper.parse(context);
} catch (MapperParsingException e) {
if (!ignoreErrors) throw e;
if (logger.isDebugEnabled())
logger.debug(
"Ignoring MapperParsingException catch while parsing author: [{}]: [{}]",
e.getMessage(),
context.externalValue());
}
}
if (metadata.get(Metadata.KEYWORDS) != null) {
try {
context = context.createExternalValueContext(metadata.get(Metadata.KEYWORDS));
keywordsMapper.parse(context);
} catch (MapperParsingException e) {
if (!ignoreErrors) throw e;
if (logger.isDebugEnabled())
logger.debug(
"Ignoring MapperParsingException catch while parsing keywords: [{}]: [{}]",
e.getMessage(),
context.externalValue());
}
}
if (contentType == null) {
contentType = metadata.get(Metadata.CONTENT_TYPE);
}
if (contentType != null) {
try {
context = context.createExternalValueContext(contentType);
contentTypeMapper.parse(context);
} catch (MapperParsingException e) {
if (!ignoreErrors) throw e;
if (logger.isDebugEnabled())
logger.debug(
"Ignoring MapperParsingException catch while parsing content_type: [{}]: [{}]",
e.getMessage(),
context.externalValue());
}
}
int length = content.length;
// If we have CONTENT_LENGTH from Tika we use it
if (metadata.get(Metadata.CONTENT_LENGTH) != null) {
length = Integer.parseInt(metadata.get(Metadata.CONTENT_LENGTH));
}
try {
context = context.createExternalValueContext(length);
contentLengthMapper.parse(context);
} catch (MapperParsingException e) {
if (!ignoreErrors) throw e;
if (logger.isDebugEnabled())
logger.debug(
"Ignoring MapperParsingException catch while parsing content_length: [{}]: [{}]",
e.getMessage(),
context.externalValue());
}
// multiFields.parse(this, context);
return null;
}
