use of edu.uci.ics.texera.api.schema.Schema in project textdb by TextDB.
the class Join method open.
@Override
public void open() throws TexeraException {
if (cursor != CLOSED) {
return;
}
if (innerOperator == null) {
throw new DataflowException("Inner Input Operator is not set.");
}
if (outerOperator == null) {
throw new DataflowException("Outer Input Operator is not set.");
}
// generate output schema from schema of inner and outer operator
innerOperator.open();
Schema innerOperatorSchema = innerOperator.getOutputSchema();
outerOperator.open();
Schema outerOperatorSchema = outerOperator.getOutputSchema();
this.outputSchema = joinPredicate.generateOutputSchema(innerOperatorSchema, outerOperatorSchema);
cursor = OPENED;
}
use of edu.uci.ics.texera.api.schema.Schema in project textdb by TextDB.
the class JoinDistancePredicate method generateIntersectionSchema.
/**
* Create outputSchema, which is the intersection of innerOperator's schema and outerOperator's schema.
* The attributes have to be exactly the same (name and type) to be intersected.
*
* InnerOperator's attributes and outerOperator's attributes must:
* both contain the attributes to be joined.
* both contain "_ID" attribute.
* both contain "spanList" attribute.
*
* @return outputSchema
*/
private Schema generateIntersectionSchema(Schema innerOperatorSchema, Schema outerOperatorSchema) throws DataflowException {
List<Attribute> innerAttributes = innerOperatorSchema.getAttributes();
List<Attribute> outerAttributes = outerOperatorSchema.getAttributes();
List<Attribute> intersectionAttributes = innerAttributes.stream().filter(attr -> outerAttributes.contains(attr)).collect(Collectors.toList());
Schema intersectionSchema = new Schema(intersectionAttributes.stream().toArray(Attribute[]::new));
// check if output schema contain necessary attributes
if (intersectionSchema.getAttributes().isEmpty()) {
throw new DataflowException("inner operator and outer operator don't share any common attributes");
} else if (!intersectionSchema.containsAttribute(this.joinAttributeName)) {
throw new DataflowException("inner operator or outer operator doesn't contain join attribute");
} else if (!intersectionSchema.containsAttribute(SchemaConstants._ID)) {
throw new DataflowException("inner operator or outer operator doesn't contain _ID attribute");
} else if (!intersectionSchema.containsAttribute(SchemaConstants.SPAN_LIST)) {
throw new DataflowException("inner operator or outer operator doesn't contain spanList attribute");
}
// check if join attribute is TEXT or STRING
AttributeType joinAttrType = intersectionSchema.getAttribute(this.joinAttributeName).getType();
if (joinAttrType != AttributeType.TEXT && joinAttrType != AttributeType.STRING) {
throw new DataflowException(String.format("Join attribute %s must be either TEXT or STRING.", this.joinAttributeName));
}
return intersectionSchema;
}
use of edu.uci.ics.texera.api.schema.Schema in project textdb by TextDB.
the class SimilarityJoinPredicate method joinTuples.
@Override
public Tuple joinTuples(Tuple innerTuple, Tuple outerTuple, Schema outputSchema) throws DataflowException {
if (similarityThreshold == 0) {
return null;
}
// get the span list only with the joinAttributeName
ListField<Span> innerSpanListField = innerTuple.getField(SchemaConstants.SPAN_LIST);
List<Span> innerRelevantSpanList = innerSpanListField.getValue().stream().filter(span -> span.getAttributeName().equals(innerJoinAttrName)).collect(Collectors.toList());
ListField<Span> outerSpanListField = outerTuple.getField(SchemaConstants.SPAN_LIST);
List<Span> outerRelevantSpanList = outerSpanListField.getValue().stream().filter(span -> span.getAttributeName().equals(outerJoinAttrName)).collect(Collectors.toList());
// get a set of span's values (since multiple spans may have the same value)
Set<String> innerSpanValueSet = innerRelevantSpanList.stream().map(span -> span.getValue()).collect(Collectors.toSet());
Set<String> outerSpanValueSet = outerRelevantSpanList.stream().map(span -> span.getValue()).collect(Collectors.toSet());
// compute the result value set using the similarity function
Set<String> resultValueSet = new HashSet<>();
for (String innerString : innerSpanValueSet) {
for (String outerString : outerSpanValueSet) {
if (this.similarityFunc.calculateSimilarity(innerString, outerString) >= this.similarityThreshold) {
resultValueSet.add(innerString);
resultValueSet.add(outerString);
}
}
}
// return null if none of them are similar
if (resultValueSet.isEmpty()) {
return null;
}
// generate the result spans
List<Span> resultSpans = new ArrayList<>();
for (Span span : innerRelevantSpanList) {
if (resultValueSet.contains(span.getValue())) {
resultSpans.add(addFieldPrefix(span, INNER_PREFIX));
}
}
for (Span span : outerRelevantSpanList) {
if (resultValueSet.contains(span.getValue())) {
resultSpans.add(addFieldPrefix(span, OUTER_PREFIX));
}
}
return mergeTuples(innerTuple, outerTuple, outputSchema, resultSpans);
}
use of edu.uci.ics.texera.api.schema.Schema in project textdb by TextDB.
the class EmojiSentimentOperator method open.
@Override
public void open() throws TexeraException {
if (cursor != CLOSED) {
return;
}
if (inputOperator == null) {
throw new DataflowException(ErrorMessages.INPUT_OPERATOR_NOT_SPECIFIED);
}
inputOperator.open();
Schema inputSchema = inputOperator.getOutputSchema();
// check if input schema is present
if (!inputSchema.containsAttribute(predicate.getInputAttributeName())) {
throw new TexeraException(String.format("input attribute %s is not in the input schema %s", predicate.getInputAttributeName(), inputSchema.getAttributeNames()));
}
// check if attribute type is valid
AttributeType inputAttributeType = inputSchema.getAttribute(predicate.getInputAttributeName()).getType();
boolean isValidType = inputAttributeType.equals(AttributeType.STRING) || inputAttributeType.equals(AttributeType.TEXT);
if (!isValidType) {
throw new TexeraException(String.format("input attribute %s must have type String or Text, its actual type is %s", predicate.getInputAttributeName(), inputAttributeType));
}
// generate output schema by transforming the input schema
outputSchema = transformSchema(inputOperator.getOutputSchema());
cursor = OPENED;
}
use of edu.uci.ics.texera.api.schema.Schema in project textdb by TextDB.
the class NltkSentimentOperator method open.
@Override
public void open() throws TexeraException {
if (cursor != CLOSED) {
return;
}
if (inputOperator == null) {
throw new DataflowException(ErrorMessages.INPUT_OPERATOR_NOT_SPECIFIED);
}
inputOperator.open();
Schema inputSchema = inputOperator.getOutputSchema();
// check if the input schema is presented
if (!inputSchema.containsAttribute(predicate.getInputAttributeName())) {
throw new TexeraException(String.format("input attribute %s is not in the input schema %s", predicate.getInputAttributeName(), inputSchema.getAttributeNames()));
}
// check if the attribute type is valid
AttributeType inputAttributeType = inputSchema.getAttribute(predicate.getInputAttributeName()).getType();
boolean isValidType = inputAttributeType.equals(AttributeType.STRING) || inputAttributeType.equals(AttributeType.TEXT);
if (!isValidType) {
throw new TexeraException(String.format("input attribute %s must have type String or Text, its actual type is %s", predicate.getInputAttributeName(), inputAttributeType));
}
// generate output schema by transforming the input schema
outputSchema = transformSchema(inputOperator.getOutputSchema());
cursor = OPENED;
}
Aggregations