org.egothor.html
Class HTMLPrinter

java.lang.Object
  org.cyberneko.html.filters.DefaultFilter
      org.egothor.html.HTMLPrinter

All Implemented Interfaces:

org.apache.xerces.xni.parser.XMLComponent, org.apache.xerces.xni.parser.XMLDocumentFilter, org.apache.xerces.xni.parser.XMLDocumentSource, org.apache.xerces.xni.XMLDocumentHandler, org.cyberneko.html.HTMLComponent

public class HTMLPrinter
extends org.cyberneko.html.filters.DefaultFilter

This class reformats HTML documents in our DB to a format that can be used for the science. It was used for a syllable XBW paper.

Field Summary

protected static java.lang.String	AUGMENTATIONS
protected boolean	charNorm
protected int	depth
protected java.lang.String	enc
protected static java.lang.String	FILTERS
protected boolean	httpEqSeen
static java.lang.String	NOTIFY_CHAR_REFS
static java.lang.String	NOTIFY_HTML_BUILTIN_REFS
protected boolean	printChars
protected java.io.PrintWriter	pw
protected boolean	rootSeen

Fields inherited from class org.cyberneko.html.filters.DefaultFilter

fDocumentHandler, fDocumentSource

Constructor Summary

HTMLPrinter(java.io.OutputStream outputStream, java.lang.String encoding)

HTMLPrinter(java.io.Writer writer, java.lang.String encoding, boolean strictScript)

Method Summary

void	characters(org.apache.xerces.xni.XMLString text, org.apache.xerces.xni.Augmentations augs)
void	comment(org.apache.xerces.xni.XMLString text, org.apache.xerces.xni.Augmentations augs)
void	emptyElement(org.apache.xerces.xni.QName element, org.apache.xerces.xni.XMLAttributes attributes, org.apache.xerces.xni.Augmentations augs)
void	endElement(org.apache.xerces.xni.QName element, org.apache.xerces.xni.Augmentations augs)
void	endGeneralEntity(java.lang.String name, org.apache.xerces.xni.Augmentations augs)
static void	filter(java.lang.String systemId, byte[] file, int file_len, java.io.PrintStream ps, java.lang.String inpEnc, java.lang.String outEnc)
protected void	printAttributeValue(java.lang.String text)
protected void	printCharacters(org.apache.xerces.xni.XMLString text, boolean normalize)
protected void	printEndElement(org.apache.xerces.xni.QName element)
protected void	printEntity(java.lang.String name)
protected void	printStartElement(org.apache.xerces.xni.QName element, org.apache.xerces.xni.XMLAttributes attributes)
protected void	printStartElement(org.apache.xerces.xni.QName element, org.apache.xerces.xni.XMLAttributes attributes, boolean empty)
void	startDocument(org.apache.xerces.xni.XMLLocator locator, java.lang.String encoding, org.apache.xerces.xni.Augmentations augs)
void	startDocument(org.apache.xerces.xni.XMLLocator locator, java.lang.String encoding, org.apache.xerces.xni.NamespaceContext nscontext, org.apache.xerces.xni.Augmentations augs)
void	startElement(org.apache.xerces.xni.QName element, org.apache.xerces.xni.XMLAttributes attributes, org.apache.xerces.xni.Augmentations augs)
void	startGeneralEntity(java.lang.String name, org.apache.xerces.xni.XMLResourceIdentifier id, java.lang.String encoding, org.apache.xerces.xni.Augmentations augs)

Methods inherited from class org.cyberneko.html.filters.DefaultFilter

doctypeDecl, endCDATA, endDocument, endPrefixMapping, getDocumentHandler, getDocumentSource, getFeatureDefault, getPropertyDefault, getRecognizedFeatures, getRecognizedProperties, ignorableWhitespace, merge, processingInstruction, reset, setDocumentHandler, setDocumentSource, setFeature, setProperty, startCDATA, startPrefixMapping, textDecl, xmlDecl

Methods inherited from class java.lang.Object

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

Field Detail

NOTIFY_CHAR_REFS

public static final java.lang.String NOTIFY_CHAR_REFS

See Also:

Constant Field Values

NOTIFY_HTML_BUILTIN_REFS

public static final java.lang.String NOTIFY_HTML_BUILTIN_REFS

See Also:

Constant Field Values

AUGMENTATIONS

protected static final java.lang.String AUGMENTATIONS

See Also:

Constant Field Values

FILTERS

protected static final java.lang.String FILTERS

See Also:

Constant Field Values

enc

protected java.lang.String enc

pw

protected java.io.PrintWriter pw

rootSeen

protected boolean rootSeen

httpEqSeen

protected boolean httpEqSeen

depth

protected int depth

charNorm

protected boolean charNorm

printChars

protected boolean printChars

Constructor Detail

HTMLPrinter

public HTMLPrinter(java.io.OutputStream outputStream,
                   java.lang.String encoding)
            throws java.io.UnsupportedEncodingException

Throws:

java.io.UnsupportedEncodingException

HTMLPrinter

public HTMLPrinter(java.io.Writer writer,
                   java.lang.String encoding,
                   boolean strictScript)

Method Detail

startDocument

public void startDocument(org.apache.xerces.xni.XMLLocator locator,
                          java.lang.String encoding,
                          org.apache.xerces.xni.NamespaceContext nscontext,
                          org.apache.xerces.xni.Augmentations augs)
                   throws org.apache.xerces.xni.XNIException

Specified by:

startDocument in interface org.apache.xerces.xni.XMLDocumentHandler

Overrides:

startDocument in class org.cyberneko.html.filters.DefaultFilter

Throws:

org.apache.xerces.xni.XNIException

startDocument

public void startDocument(org.apache.xerces.xni.XMLLocator locator,
                          java.lang.String encoding,
                          org.apache.xerces.xni.Augmentations augs)
                   throws org.apache.xerces.xni.XNIException

Overrides:

startDocument in class org.cyberneko.html.filters.DefaultFilter

Throws:

org.apache.xerces.xni.XNIException

comment

public void comment(org.apache.xerces.xni.XMLString text,
                    org.apache.xerces.xni.Augmentations augs)
             throws org.apache.xerces.xni.XNIException

Specified by:

comment in interface org.apache.xerces.xni.XMLDocumentHandler

Overrides:

comment in class org.cyberneko.html.filters.DefaultFilter

Throws:

org.apache.xerces.xni.XNIException

startElement

public void startElement(org.apache.xerces.xni.QName element,
                         org.apache.xerces.xni.XMLAttributes attributes,
                         org.apache.xerces.xni.Augmentations augs)
                  throws org.apache.xerces.xni.XNIException

Specified by:

startElement in interface org.apache.xerces.xni.XMLDocumentHandler

Overrides:

startElement in class org.cyberneko.html.filters.DefaultFilter

Throws:

org.apache.xerces.xni.XNIException

emptyElement

public void emptyElement(org.apache.xerces.xni.QName element,
                         org.apache.xerces.xni.XMLAttributes attributes,
                         org.apache.xerces.xni.Augmentations augs)
                  throws org.apache.xerces.xni.XNIException

Specified by:

emptyElement in interface org.apache.xerces.xni.XMLDocumentHandler

Overrides:

emptyElement in class org.cyberneko.html.filters.DefaultFilter

Throws:

org.apache.xerces.xni.XNIException

characters

public void characters(org.apache.xerces.xni.XMLString text,
                       org.apache.xerces.xni.Augmentations augs)
                throws org.apache.xerces.xni.XNIException

Specified by:

characters in interface org.apache.xerces.xni.XMLDocumentHandler

Overrides:

characters in class org.cyberneko.html.filters.DefaultFilter

Throws:

org.apache.xerces.xni.XNIException

endElement

public void endElement(org.apache.xerces.xni.QName element,
                       org.apache.xerces.xni.Augmentations augs)
                throws org.apache.xerces.xni.XNIException

Specified by:

endElement in interface org.apache.xerces.xni.XMLDocumentHandler

Overrides:

endElement in class org.cyberneko.html.filters.DefaultFilter

Throws:

org.apache.xerces.xni.XNIException

startGeneralEntity

public void startGeneralEntity(java.lang.String name,
                               org.apache.xerces.xni.XMLResourceIdentifier id,
                               java.lang.String encoding,
                               org.apache.xerces.xni.Augmentations augs)
                        throws org.apache.xerces.xni.XNIException

Specified by:

startGeneralEntity in interface org.apache.xerces.xni.XMLDocumentHandler

Overrides:

startGeneralEntity in class org.cyberneko.html.filters.DefaultFilter

Throws:

org.apache.xerces.xni.XNIException

endGeneralEntity

public void endGeneralEntity(java.lang.String name,
                             org.apache.xerces.xni.Augmentations augs)
                      throws org.apache.xerces.xni.XNIException

Specified by:

endGeneralEntity in interface org.apache.xerces.xni.XMLDocumentHandler

Overrides:

endGeneralEntity in class org.cyberneko.html.filters.DefaultFilter

Throws:

org.apache.xerces.xni.XNIException

printAttributeValue

protected void printAttributeValue(java.lang.String text)

printCharacters

protected void printCharacters(org.apache.xerces.xni.XMLString text,
                               boolean normalize)

printStartElement

protected void printStartElement(org.apache.xerces.xni.QName element,
                                 org.apache.xerces.xni.XMLAttributes attributes)

printStartElement

protected void printStartElement(org.apache.xerces.xni.QName element,
                                 org.apache.xerces.xni.XMLAttributes attributes,
                                 boolean empty)

printEndElement

protected void printEndElement(org.apache.xerces.xni.QName element)

printEntity

protected void printEntity(java.lang.String name)

filter

public static void filter(java.lang.String systemId,
                          byte[] file,
                          int file_len,
                          java.io.PrintStream ps,
                          java.lang.String inpEnc,
                          java.lang.String outEnc)
                   throws java.lang.Exception

Throws:

java.lang.Exception