[Mypyspace-developer] SF.net SVN: mypyspace:[313] myspaceCrawler/trunk

[<gea...@us...>]

Revision: 313
          http://mypyspace.svn.sourceforge.net/mypyspace/?rev=313&view=rev
Author:   gearmonkey
Date:     2009-01-16 17:59:29 +0000 (Fri, 16 Jan 2009)

Log Message:
-----------
corrected the default base uri for rdf generated via mpsUser.

Modified Paths:
--------------
    musicGrabber/branches/webserv-branch/myspace2rdf.py
    myspaceCrawler/tags/0.8.1b_release/myspaceCrawler.py
    myspaceCrawler/trunk/mpsUser.py
    myspaceCrawler/trunk/myspaceuris.py

Modified: musicGrabber/branches/webserv-branch/myspace2rdf.py
===================================================================
--- musicGrabber/branches/webserv-branch/myspace2rdf.py	2009-01-16 17:24:24 UTC (rev 312)
+++ musicGrabber/branches/webserv-branch/myspace2rdf.py	2009-01-16 17:59:29 UTC (rev 313)
@@ -102,10 +102,8 @@
 				- should switch to check for genre tags instead???'''
 		
 		if self.page:
-			#############################################
-			# kludge set playr to always flase for now ##
-			#############################################
-			player = False #= scrapePage(self.page, [playerTag[0]], playerTag[1])
+		
+			player = scrapePage(self.page, [playerTag[0]], playerTag[1])
 			if player:
 				self.subject = mopy.mo.MusicArtist(dbtuneMyspace+'uid/'+str(self.uid))
 				#self.subjecttwo = mopy.foaf.Person('http://dbtune.org/myspace/uid/'+str(self.uid))

Modified: myspaceCrawler/tags/0.8.1b_release/myspaceCrawler.py
===================================================================
--- myspaceCrawler/tags/0.8.1b_release/myspaceCrawler.py	2009-01-16 17:24:24 UTC (rev 312)
+++ myspaceCrawler/tags/0.8.1b_release/myspaceCrawler.py	2009-01-16 17:59:29 UTC (rev 313)
@@ -32,7 +32,7 @@
 
 
 
-THREAD_CAP = 30 #maximum number of threads allowed to be firing at once
+THREAD_CAP = 10000 #maximum number of threads allowed to be firing at once
 THREAD_STALL_TIME = 30 #length of time in seconds to wait until the thread count is checked again
 LOG_FILENAME = "musicCrawler.log" #name of logger file (path set at commandline)
 

Modified: myspaceCrawler/trunk/mpsUser.py
===================================================================
--- myspaceCrawler/trunk/mpsUser.py	2009-01-16 17:24:24 UTC (rev 312)
+++ myspaceCrawler/trunk/mpsUser.py	2009-01-16 17:59:29 UTC (rev 313)
@@ -55,7 +55,7 @@
     
 	'''
 
-	def __init__(self, url, rdfprefix = dbtuneMyspace):
+	def __init__(self, url, rdfprefix = dbtuneMyspace + 'uid/'):
 		"""Initialization will set the source url, attempt to create a socket connection with the url and determine if this mpsUser is an artist.  If the user given is an artist, the initialization will also scrape the top Friends. rdfprefix is the uri base prepended to the uids of other myspace resources, by default it is set to the dbtune live service"""
 		self.source = url
 		self.uid = -1

Modified: myspaceCrawler/trunk/myspaceuris.py
===================================================================
--- myspaceCrawler/trunk/myspaceuris.py	2009-01-16 17:24:24 UTC (rev 312)
+++ myspaceCrawler/trunk/myspaceuris.py	2009-01-16 17:59:29 UTC (rev 313)
@@ -46,7 +46,6 @@
 
 
 #########################################################################################################
-
 # myspace uri for downloads  ----this has gotten a bit more complicated in the roll out of myspace's new media player
 # this xml file gives the songIDs, the songsIDs must be used individually to request another xml file that then contains the uri to the cached media
 #
@@ -75,7 +74,6 @@
 myspaceOwlURI = 'http://grasstunes.net/ontology/myspace.owl'
 dbtuneMyspace = 'http://dbtune.org/myspace/'
 
-
 countries = ['Afghanistan', 'Albania', 'Algeria', 'American Samoa','Andorra',
 			'Angola','Anguilla','Antarctica','Antigua and Barbuda','Argentina',
 			'Armenia','Aruba','Australia','Austria','Azerbaijan','Bahamas',


This was sent by the SourceForge.net collaborative development platform, the world's largest Open Source development site.