- added statistics about database index caches in PerformanceMemory_p.html
- adoped many classes to use the new statistics
- added missing close statements
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@7018 6c8d7289-2bf4-0310-a012-ef5d649a1542
Here a new concept called 'search heuristics' is introduced. A heuristic is a kind of 'shortcut' to good results in IT, here for good search results. In this case it will be used to get a very transparent way to compare what YaCy is able to produce as search result and what g**gle produces as search result. Here is what your can do now:
- add the phrase 'heuristic:scroogle' to your search query, like 'oil spill heuristic:scroogle' and then a call to scroogle is made to get anonymous search results from g**gle.
- these results are _not_ taken as meta-search results, but are used to instantly feed a crawling and indexing process. This happens very fast, here 20 results from scroogle are taken and loaded all simultanously, parsed and indexed immediately and from the results of the parsed content the search result is feeded, along to the normal p2p search
- when new results from that heuristic (more to come) get part of the search results, then it is verified if such results are redundant to existing (they had been part of the normal YaCy search result anyway) or if they had been completely new to YaCy.
- in the search results the new search results from heuristics are marked with a 'H ++' and search results from heuristics that had been already found by YaCy are marked with a 'H ='. That means:
- you can now see YaCy and Scroogle search results in one result page but you also see that you would not have 'missed' the g**gle results when you would only have used YaCy.
- to make it short: YaCy now subsumes g**gle results. If you use only YaCy, you miss nothing.
to come: a configuration page that let you configure the usage of heuristics and get this feature by default.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@6944 6c8d7289-2bf4-0310-a012-ef5d649a1542
- removed usage of URL-Caches which could have been a memory leak
- removed unused classes and methods
- removed not necessary synchronizations
- added synchronization hacks where possible
- fine-tuned crawling speed to prevent IO of balancer
- fixed a bug in IODispatcher that may have caused that no merges were done
- reduced number of parameters in very often called methods (compare methods)
- reduced complexity of data structures of now massively used HandleSet class
- reduction of new String() and getBytes() usage / new methods to support this transition
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@6820 6c8d7289-2bf4-0310-a012-ef5d649a1542
The result should be a less usage of new String() and less memory usage (since a String-encapsulated byte[] has 40 bytes overhead)
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@6815 6c8d7289-2bf4-0310-a012-ef5d649a1542
This is a major change in the organization of indexes.
Please consider a back-up of your data before you run this update.
All existing index files will be moved and renamed to a new position.
With this change, it will be possible to maintain different indexes for different purposes and it will be possible to have a distinction between DHT-in and DHT-out specific indexes. Tenants may also have their own index, and it may be possible to have histories and back-ups of indexes. This is just the beginning, many servlets must be adopted after this change, but all functions that had been there should still work.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@6389 6c8d7289-2bf4-0310-a012-ef5d649a1542
- speed up deletion of containers when selscted from whole index
- correctly eliminate all references to unavailable URLs, not just the first encountered
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@6381 6c8d7289-2bf4-0310-a012-ef5d649a1542
high-performance search query situations as seen in yacy-metager integration showed deadlock situation caused by synchronization effects inside of sun.java code. It appears that the logger is not completely safe against deadlock situations in concurrent calls of the logger. One possible solution would be a outside-synchronization with 'synchronized' statements, but that would further apply blocking on all high-efficient methods that call the logger. It is much better to do a non-blocking hand-over of logging lines and work off log entries with a concurrent log writer. This also disconnects IO operations from logging, which can also cause IO operation when a log is written to a file. This commit not only moves the logger from kelondro to yacy.logging, it also inserts the concurrency methods to realize non-blocking logging.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@6078 6c8d7289-2bf4-0310-a012-ef5d649a1542
divided that class into three parts:
- the peers object is now hosted by the plasmaSwitchboard
- the crawler elements are now in a new class, crawler.CrawlerSwitchboard
- the index elements are core of the new segment data structure, which is a bundle of different indexes for the full text and (in the future) navigation indexes and the metadata store. The new class is now in kelondro.text.Segment
The refactoring is inspired by the roadmap to create index segments, the option to host different indexes on one peer.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@5990 6c8d7289-2bf4-0310-a012-ef5d649a1542
- added some debugging output to balancer to find a bug
- removed unused classes for index collection handling
- changed some default values for the process handling: more memory needed to prevent OOM
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@5856 6c8d7289-2bf4-0310-a012-ef5d649a1542
terms (words) are not any more retrieved by their word hash string, but by a byte[] containing the word hash.
this has strong advantages when RWIs are sorted in the ReferenceContainer Cache and compared with the sun.java TreeMap method, which needed getBytes() and new String() transformations before.
Many thousands of such conversions are now omitted every second, which increases the indexing speed by a factor of two.
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@5812 6c8d7289-2bf4-0310-a012-ef5d649a1542
- during index joins all word positions are maintained: better ranking for word distance possible; exact phrase match can be implemented soundly
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@5804 6c8d7289-2bf4-0310-a012-ef5d649a1542
This is a preparation to introduce other index tables as used now only for reverse text indexes. Next application of the reverse index is a citation index.
Moved to version 0.74
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@5777 6c8d7289-2bf4-0310-a012-ef5d649a1542
