net.sourceforge.rssowl.controller.thread
Class FeedSearchManager

java.lang.Object
  net.sourceforge.rssowl.controller.thread.FeedSearchManager

public class FeedSearchManager

extends Object

The FeedSearchManager performs a search for RSS / RDF feeds on a certain topic. Each link that is in the result is checked for a valid feed. If the link is a newsfeed, a new line is added to the table of the FeedSearchDialog.

Version:

1.1.3

Author:

Benjamin Pasero

Field Summary

private static int	COLLECTOR_THREAD_COUNT Number of collector threads
(package private) int	collectorsDone Count the number of collector threads that have finished
private ExtendedThread[]	collectorThreads Collector threads to collect and parse links from the results of the search
private boolean	intensiveSearch If TRUE parse any link from the results.
private Vector	links List of links from the results
private static Vector	localeSearchList Domains that are allowed to use for search
private static int	MAX_CACHE_PER_RUN The maximum number of newsfeeds to cache per search
private boolean	nativeLangSearch If TRUE only display news that match the selected language
private static String	NEXT_IMAGE Token for next image
(package private) Vector	normalUrlObjects Synchronized list of normal URL objects
(package private) org.jdom.input.SAXBuilder[]	parser Create one parser for each Collector thread
(package private) Vector	preferredUrlObjects Synchronized list of preferred URL objects
(package private) Random	randomNumberGenerator Random Number Generator for sleep Time of Threads
private int	resultCounter Count the resulting Feeds
(package private) FeedSearchDialog	rssOwlFeedSearchDialog The opened rss feedsearch dialog
(package private) boolean	running Flag indicates the running threads
private static String[]	SEARCH_DELIMITER One of the words from this array is combined with the search of the user
(package private) boolean	searchDone Flag indicates when the search has finished
private Thread	searchThread Thread that performs the search
private String	searchUrl URL to get the results from
private static String	START_PARAM Token for start parameter
private String	topic Topic that the user searches for

Constructor Summary

FeedSearchManager(String topic, FeedSearchDialog rssOwlFeedSearchDialog, boolean intensiveSearch, boolean nativeLangSearch)
Instantiate a new FeedSearchManager

Method Summary

static void	addIntoLocaleSearchList(String value) Add a value to the locale search list
(package private) void	buildChannel(URL url, int parserNumber) Build RSS channel object from the given URl
private org.jdom.Document	buildDocument(URL xmlUrl, boolean forceDefaultEncoding, int parserNumber) Build the XML document from the given URL
private void	createCollectorThreads() Collector thread checks if URLs in "urlObj" are valid RSS / RDF feeds
private void	createSearchThread() Search thread performs the search for the topic
private String	createUrl() Create the URL to perform the search
private void	filterLink(String url) Filter the given URL.
private void	initXMLParsers() Init one SAX Parser for each collector thread
boolean	isRunning() Get the state of the search
(package private) void	performSearch() Perform a search on the search engine and add every link to the Stack, that was found.
void	startSearch() Start search- and collector threads
void	stopSearch() Interrupt the search- and collector threads

Methods inherited from class java.lang.Object

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

Field Detail

COLLECTOR_THREAD_COUNT

private static final int COLLECTOR_THREAD_COUNT

Number of collector threads

See Also:

Constant Field Values

localeSearchList

private static final Vector localeSearchList

Domains that are allowed to use for search

MAX_CACHE_PER_RUN

private static final int MAX_CACHE_PER_RUN

The maximum number of newsfeeds to cache per search

See Also:

Constant Field Values

NEXT_IMAGE

private static final String NEXT_IMAGE

Token for next image

See Also:

Constant Field Values

SEARCH_DELIMITER

private static final String[] SEARCH_DELIMITER

One of the words from this array is combined with the search of the user

START_PARAM

private static final String START_PARAM

Token for start parameter

See Also:

Constant Field Values

collectorThreads

private ExtendedThread[] collectorThreads

Collector threads to collect and parse links from the results of the search

intensiveSearch

private boolean intensiveSearch

If TRUE parse any link from the results. If FALSE parse .xml, .rss and .rdf

links

private Vector links

List of links from the results

nativeLangSearch

private boolean nativeLangSearch

If TRUE only display news that match the selected language

resultCounter

private int resultCounter

Count the resulting Feeds

searchThread

private Thread searchThread

Thread that performs the search

searchUrl

private String searchUrl

URL to get the results from

topic

private String topic

Topic that the user searches for

collectorsDone

int collectorsDone

Count the number of collector threads that have finished

normalUrlObjects

Vector normalUrlObjects

Synchronized list of normal URL objects

parser

org.jdom.input.SAXBuilder[] parser

Create one parser for each Collector thread

preferredUrlObjects

Vector preferredUrlObjects

Synchronized list of preferred URL objects

randomNumberGenerator

Random randomNumberGenerator

Random Number Generator for sleep Time of Threads

rssOwlFeedSearchDialog

FeedSearchDialog rssOwlFeedSearchDialog

The opened rss feedsearch dialog

running

boolean running

Flag indicates the running threads

searchDone

boolean searchDone

Flag indicates when the search has finished

Constructor Detail

FeedSearchManager

public FeedSearchManager(String topic,
                         FeedSearchDialog rssOwlFeedSearchDialog,
                         boolean intensiveSearch,
                         boolean nativeLangSearch)

Instantiate a new FeedSearchManager

Parameters:

topic - Topic to search for

rssOwlFeedSearchDialog - This dialog called the FeedSearchManager

intensiveSearch - TRUE if intensive search is performed

nativeLangSearch - TRUE if native lang search is performed

Method Detail

addIntoLocaleSearchList

public static void addIntoLocaleSearchList(String value)

Add a value to the locale search list

Parameters:

value - The value to add

isRunning

public boolean isRunning()

Get the state of the search

Returns:

boolean TRUE if search is running

startSearch

public void startSearch()

Start search- and collector threads

stopSearch

public void stopSearch()

Interrupt the search- and collector threads

buildDocument

private org.jdom.Document buildDocument(URL xmlUrl,
                                        boolean forceDefaultEncoding,
                                        int parserNumber)

Build the XML document from the given URL

Parameters:

xmlUrl - The XML Url object

forceDefaultEncoding - If set to TRUE the encoding is not being detected from the XMl document and the OS default encoding is used

parserNumber - The number of the index in the parser array to use

Returns:

Document The builded XML document from the URL or NULL on warning

createCollectorThreads

private void createCollectorThreads()

Collector thread checks if URLs in "urlObj" are valid RSS / RDF feeds

createSearchThread

private void createSearchThread()

Search thread performs the search for the topic

createUrl

private String createUrl()

Create the URL to perform the search

Returns:

String search URL

filterLink

private void filterLink(String url)

Filter the given URL. Do not add URLs which have a file name which marks them as binary file

Parameters:

url - Any URL to filter

initXMLParsers

private void initXMLParsers()

Init one SAX Parser for each collector thread

buildChannel

void buildChannel(URL url,
                  int parserNumber)

Build RSS channel object from the given URl

Parameters:

url - The URL of the newsfeed

parserNumber - The number of the index in the parser array to use

performSearch

void performSearch()
             throws IOException

Perform a search on the search engine and add every link to the Stack, that was found. Each link is first filtered, to see if it's not an image or zip for example.

Throws:

IOException - If connection fails