Package org.apache.tika.detect

Class EncodingResult

java.lang.Object

org.apache.tika.detect.EncodingResult

public class EncodingResult
extends Object

A charset detection result pairing a Charset with a confidence score and a EncodingResult.ResultType indicating the nature of the evidence.

Result types

• EncodingResult.ResultType.DECLARATIVE — the document explicitly stated its encoding (BOM, HTML <meta charset>). These are authoritative claims about author intent and get preference over inferred results when consistent with the actual bytes.
• EncodingResult.ResultType.STRUCTURAL — byte-grammar proof (ISO-2022 escape sequences, UTF-8 multibyte validation). The encoding is proven by the byte structure itself, independent of any declaration.
• EncodingResult.ResultType.STATISTICAL — probabilistic inference from a statistical model. The confidence float is meaningful here for ranking among candidates; for DECLARATIVE and STRUCTURAL results it is conventionally 1.0 but carries no additional information.

Since:

Apache Tika 4.0

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static enum	EncodingResult.ResultType	The nature of the evidence that produced this result.

Constructor Summary

Constructors

Constructor	Description
EncodingResult(Charset charset, float confidence)	Constructs a STATISTICAL result.
EncodingResult(Charset charset, float confidence, String label)	Constructs a STATISTICAL result with a detector-specific label.
EncodingResult(Charset charset, float confidence, String label, EncodingResult.ResultType resultType)	Constructs a result with an explicit EncodingResult.ResultType.

Method Summary

Modifier and Type	Method	Description
Charset	getCharset()
float	getConfidence()	Detection confidence in [0.0, 1.0].
String	getLabel()	The detector's original label for this result.
EncodingResult.ResultType	getResultType()	The nature of the evidence that produced this result.
String	toString()

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Details

EncodingResult

public EncodingResult(Charset charset,
 float confidence)

Constructs a STATISTICAL result. Existing detectors that do not yet classify their evidence type default to statistical (probabilistic) treatment, which is the safe, arbitratable assumption.

Parameters:

charset - the detected charset; must not be null

confidence - detection confidence in [0.0, 1.0]

EncodingResult

public EncodingResult(Charset charset,
 float confidence,
 String label)

Constructs a STATISTICAL result with a detector-specific label.

Parameters:

charset - the detected charset; must not be null

confidence - detection confidence in [0.0, 1.0]

label - the detector's original label (e.g. "IBM420-ltr"); if null, defaults to charset.name()

EncodingResult

public EncodingResult(Charset charset,
 float confidence,
 String label,
 EncodingResult.ResultType resultType)

Constructs a result with an explicit EncodingResult.ResultType.

Parameters:

charset - the detected charset; must not be null

confidence - detection confidence in [0.0, 1.0]

label - the detector's original label; if null, defaults to charset.name()

resultType - the nature of the evidence; must not be null

Method Details

getCharset

public Charset getCharset()

getConfidence

public float getConfidence()

Detection confidence in [0.0, 1.0]. Meaningful for ranking among EncodingResult.ResultType.STATISTICAL candidates. For EncodingResult.ResultType.DECLARATIVE and EncodingResult.ResultType.STRUCTURAL results the value is conventionally 1.0 but carries no additional information beyond the type itself.

getResultType

public EncodingResult.ResultType getResultType()

The nature of the evidence that produced this result.

See Also:

• EncodingResult.ResultType

getLabel

public String getLabel()

The detector's original label for this result. Usually identical to getCharset().name(), but preserved when the detector uses finer-grained labels than the Java charset registry supports (e.g. "IBM420-ltr", "IBM420-rtl", "windows-874").

toString

public String toString()

Overrides:

toString in class Object