You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@tika.apache.org by "Bob Paulin (JIRA)" <ji...@apache.org> on 2016/03/19 21:23:33 UTC
[jira] [Updated] (TIKA-1904) Tika 2.0 - Create Proxy Parser and
Detectors
[ https://issues.apache.org/jira/browse/TIKA-1904?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Bob Paulin updated TIKA-1904:
-----------------------------
Description:
There are several parsers and detectors that instantiate parsers and detectors that live in different modules in tika 2.0. As of now these modules have are dependent on other modules this includes:
tika-parser-office-module -> tika-parser-web-module, tika-parser-text-module, tika-parser-package-module
tika-parser-ebook-module -> tika-parser-text-module
tika-parser-journal-module -> tika-parser-pdf-module
May of these dependencies could be made optional by introducing the concept of proxy parser and detectors that would enable functionality if all the dependencies are included in the project but not throw a ClassNotFoundException if the dependent module was not include( ex. parse function would do nothing).
EX
Currently
ChmParser
{code}
private void parsePage(byte[] byteObject, ContentHandler xhtml) throws TikaException {// throws IOException
InputStream stream = null;
Metadata metadata = new Metadata();
HtmlParser htmlParser = new HtmlParser();
ContentHandler handler = new EmbeddedContentHandler(new BodyContentHandler(xhtml));// -1
ParseContext parser = new ParseContext();
try {
stream = new ByteArrayInputStream(byteObject);
htmlParser.parse(stream, handler, metadata, parser);
} catch (SAXException e) {
throw new RuntimeException(e);
} catch (IOException e) {
// Pushback overflow from tagsoup
}
}
{code}
Instead the HtmlParser could be Proxyed in the constructor
{code}
private final Parser htmlProxyParser;
public ChmParser() {
this.htmlProxyParser = new ParserProxy("org.apache.tika.parser.html.HtmlParser");
}
{code}
And
{code}
private void parsePage(byte[] byteObject, ContentHandler xhtml) throws TikaException {// throws IOException
InputStream stream = null;
Metadata metadata = new Metadata();
ContentHandler handler = new EmbeddedContentHandler(new BodyContentHandler(xhtml));// -1
ParseContext parser = new ParseContext();
try {
stream = new ByteArrayInputStream(byteObject);
htmlProxyParser.parse(stream, handler, metadata, parser);
} catch (SAXException e) {
throw new RuntimeException(e);
} catch (IOException e) {
// Pushback overflow from tagsoup
}
}
{code}
was:
There are several parsers and detectors that instantiate parsers and detectors that live in different modules in tika 2.0. As of now these modules have are dependent on other modules this includes:
tika-parser-office-module -> tika-parser-web-module, tika-parser-text-module, tika-parser-package-module
tika-parser-ebook-module -> tika-parser-text-module
tika-parser-journal-module -> tika-parser-pdf-module
May of these dependencies could be made optional by introducing the concept of proxy parser and detectors that would enable functionality if all the dependencies are included in the project but not throw a ClassNotFoundException if the dependent module was not include( ex. parse function would do nothing).
EX
Currently
ChmParser
{code}
private void parsePage(byte[] byteObject, ContentHandler xhtml) throws TikaException {// throws IOException
InputStream stream = null;
Metadata metadata = new Metadata();
HtmlParser htmlParser = new HtmlParser();
ContentHandler handler = new EmbeddedContentHandler(new BodyContentHandler(xhtml));// -1
ParseContext parser = new ParseContext();
try {
stream = new ByteArrayInputStream(byteObject);
htmlParser.parse(stream, handler, metadata, parser);
} catch (SAXException e) {
throw new RuntimeException(e);
} catch (IOException e) {
// Pushback overflow from tagsoup
}
}
{code}
Instead the HtmlParser could be Proxyed in the constructor
{code}
private final Parser htmlProxyParser;
public ChmParser() {
this.htmlProxyParser = new ProxyParser("org.apache.tika.parser.html.HtmlParser");
}
{code}
And
{code}
private void parsePage(byte[] byteObject, ContentHandler xhtml) throws TikaException {// throws IOException
InputStream stream = null;
Metadata metadata = new Metadata();
ContentHandler handler = new EmbeddedContentHandler(new BodyContentHandler(xhtml));// -1
ParseContext parser = new ParseContext();
try {
stream = new ByteArrayInputStream(byteObject);
htmlProxyParser.parse(stream, handler, metadata, parser);
} catch (SAXException e) {
throw new RuntimeException(e);
} catch (IOException e) {
// Pushback overflow from tagsoup
}
}
{code}
> Tika 2.0 - Create Proxy Parser and Detectors
> --------------------------------------------
>
> Key: TIKA-1904
> URL: https://issues.apache.org/jira/browse/TIKA-1904
> Project: Tika
> Issue Type: Improvement
> Affects Versions: 2.0
> Reporter: Bob Paulin
> Assignee: Bob Paulin
>
> There are several parsers and detectors that instantiate parsers and detectors that live in different modules in tika 2.0. As of now these modules have are dependent on other modules this includes:
> tika-parser-office-module -> tika-parser-web-module, tika-parser-text-module, tika-parser-package-module
> tika-parser-ebook-module -> tika-parser-text-module
> tika-parser-journal-module -> tika-parser-pdf-module
> May of these dependencies could be made optional by introducing the concept of proxy parser and detectors that would enable functionality if all the dependencies are included in the project but not throw a ClassNotFoundException if the dependent module was not include( ex. parse function would do nothing).
> EX
> Currently
> ChmParser
> {code}
> private void parsePage(byte[] byteObject, ContentHandler xhtml) throws TikaException {// throws IOException
> InputStream stream = null;
> Metadata metadata = new Metadata();
> HtmlParser htmlParser = new HtmlParser();
> ContentHandler handler = new EmbeddedContentHandler(new BodyContentHandler(xhtml));// -1
> ParseContext parser = new ParseContext();
> try {
> stream = new ByteArrayInputStream(byteObject);
> htmlParser.parse(stream, handler, metadata, parser);
> } catch (SAXException e) {
> throw new RuntimeException(e);
> } catch (IOException e) {
> // Pushback overflow from tagsoup
> }
> }
> {code}
> Instead the HtmlParser could be Proxyed in the constructor
> {code}
> private final Parser htmlProxyParser;
>
> public ChmParser() {
> this.htmlProxyParser = new ParserProxy("org.apache.tika.parser.html.HtmlParser");
> }
> {code}
> And
> {code}
> private void parsePage(byte[] byteObject, ContentHandler xhtml) throws TikaException {// throws IOException
> InputStream stream = null;
> Metadata metadata = new Metadata();
> ContentHandler handler = new EmbeddedContentHandler(new BodyContentHandler(xhtml));// -1
> ParseContext parser = new ParseContext();
> try {
> stream = new ByteArrayInputStream(byteObject);
> htmlProxyParser.parse(stream, handler, metadata, parser);
> } catch (SAXException e) {
> throw new RuntimeException(e);
> } catch (IOException e) {
> // Pushback overflow from tagsoup
> }
> }
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)