- fixed broken downloads (flush was missing)
- different problem handling when download is corrupted
- different default values in yacy.init
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4669 6c8d7289-2bf4-0310-a012-ef5d649a1542
- removed all InputStream.available() because this does not work for files > 2GB
- iterator terminate when a IOException occurs
- added handling of non-executing index.add methods to enhance assert usage
- added index for file indexes > 2GB, to be used in new indexHeap
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4666 6c8d7289-2bf4-0310-a012-ef5d649a1542
storage and loading is 30 times faster! a cache of 100000 RWIs needed 180 seconds
to store and 100 seconds to restore; now the same cache needs only 6 seconds to store and
3 seconds to restore. The cache size has decreased now by 30% (95 MB instead of 150 MB).
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4634 6c8d7289-2bf4-0310-a012-ef5d649a1542
- concurrency for LURL-fetching: this can be done using a concurrent lookup into the separated url databases. Concurrency is possible because there is no IO during lookup. The more LURL-Tables are present, the better is the speedup. More CPUs will increase speed
- because a large number of LURL-lookups are made during crawling (for double-check), the LURL-Lookup speed enhancements enhances also crawling speed
- search speed also profits from LURL-lookup enhancement
- changed some flushing parameters in word index caching which should make better use of large word index caches and should speed up indexing
- removed flush chunksize parameter, because this was only useful for IO path enhancement feature which was removed some weeks ago to prevent blocking and deadlocks during search requests
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4628 6c8d7289-2bf4-0310-a012-ef5d649a1542
- the methods {parsing, semantic analysis (condensing), structure analysis (web structure)} in the serialized indexing path had been made concurrent.
- four BlockingQueues handle concurrency and hand-over of the indexing objects, the last object in the queue is stored into a blockingQueue of maximum size 1 to serialize the process for storage (which uses IO and therefore here should not be deserialized)
- a concurrency of (CPUs + 1) is default. Single-CPU users will profil from the change because large files cannot block the indexing process any more.
- removed the secondary indexing thread, which is superfluous now. Concurrency is default for all users.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4609 6c8d7289-2bf4-0310-a012-ef5d649a1542
- remote crawl receipts are now transmitted concurrently in separate threads (makes remove crawls much faster!)
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4605 6c8d7289-2bf4-0310-a012-ef5d649a1542
- refactoring of word/phrase handling: word abstraction from condenser becomes part of index element handling
- removed unused code parts from condenser
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4603 6c8d7289-2bf4-0310-a012-ef5d649a1542
this is another step to enable multiple, concurrent fulltext-indexes
- another try to make the yacy-httpc more stable
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4602 6c8d7289-2bf4-0310-a012-ef5d649a1542
danielr