introduce FederateSearchManager handling search heuristic to external systems via specific FederateSearchConnectors,
which provide the query() functionallity, the translation to YaCy schema .toYaCySchema() and the search() routine to deliver results to searchevents, which is generally implemented in Abstract connector.
The manager enforces now a min 15s delay between calls to external systems.
Besides the OpensearchConnector a SolrFederateSearchConnector is available. It uses a additional config file for fieldname translation.
default heuristicopensearch.conf:
- openbdb.com removed - seems not longer to deliver results
- config via solrconnector to datacite.org added (large technical library archive)
Intended for searches/research projects with not sufficient results from local and DHT selected remote target peers.
Function: the process checks newly created bookmarks for description starting with "query=..." and takes this to ask every peer for 20 search results and adds it to the local index in a background job.
link to start/stop the process added to /Bookmarks.html
This is intended for peers who want to participate in the P2P network but don't wish to load/fill-up their index with metadata of every received search result.
The DHT transfer is not effected by this option (and will work as usual, so that a peer disabling the new store to index switch still receives and holds the metadata according to DHT rules).
Downside for the local peer is that search speed will not improve if search terms are only avail. remote or by quick hits in local index.
To be able to improve the local index a Click-Servlet option was added additionally.
If switched on, all search result links point to this servlet, which forwards the users browser (by html header) to the desired page and feeds the page to the fulltext-index.
The servlet accepts a parameter defining the action to perform (see defaults/web.xml, index, crawl, crawllinks)
The option check-boxes are placed in ConfigPortal.html
stream the original to the search result thumb viewer. This has two
reasons:
- animated gifs cause 100% cpu and deadlocks in the jvm gif parser; a
known bug which is obviously not yet fixed
- animated gifs now appear in the search result also as animation
preferred over https. While this is a bad idea from the standpoint of
security it is more common applicable for environments where http and
https mix and for some domains https is not available. Then the
double-check is possible even if no postprocessing is performed.
parsing into individual pages and add them all using different URLs.
These constructed urls are generated from the source url with an
appended page=<pagenumber> attribute to the url get/post properties.
This will distinguish the different page entries. The search result list
will then replace the post parameter with a url anchor # mark which
causes that the original url is presented in the search result. These
URLs can be opened directly on the correct page using pdf.js which is
now built-in into firefox. That means: if you find a search hit on page
5 and click on the search result, firefox will open the pdf viewer and
shows page 5.
occurrences within the (web) page documents (not the document
last-modified!). This works only if the solr field dates_in_content_sxt
is enabled. A search request may then have the form "term on:<date>",
like
gift on:24.12.2014
gift on:2014/12/24
* on:2014/12/31
For the date format you may use any kind of human-readable date
representation(!yes!) - the on:<date> parser tries to identify language
and also knows event names, like:
bunny on:eastern
.. as long as the date term has no spaces inside (use a dot). Further
enhancement will be made to accept also strings encapsulated with
quotes.
notions within the fulltext of a document. This class attempts to
identify also dates given abbreviated or with missing year or described
with names for special days, like 'Halloween'. In case that a date has
no year given, the current year and following years are considered.
This process is therefore able to identify a large set of dates to a
document, either because there are several dates given in the document
or the date is ambiguous. Four new Solr fields are used to store the
parsing result:
dates_in_content_sxt:
if date expressions can be found in the content, these dates are listed
here in order of the appearances
dates_in_content_count_i:
the number of entries in dates_in_content_sxt
date_in_content_min_dt:
if dates_in_content_sxt is filled, this contains the oldest date from
the list of available dates
#date_in_content_max_dt:
if dates_in_content_sxt is filled, this contains the youngest date from
the list of available dates, that may also be possibly in the future
These fields are deactiviated by default because the evaluation of
regular expressions to detect the date is yet too CPU intensive. Maybe
future enhancements will cause that this is switched on by default.
The purpose of these fields is the creation of calendar-like search
facets, to be implemented next.
- new switch 'isFacet' which causes that the usage of the vocabulary for
search facets is enabled or disabled. This shall be used for large
vocabularies sind searched in solr are extremely slow if facets for a
large set of alternative terms are generated
- new option to disable auto-enrichment from synonyms
- new option to add synonyms from another column when importing from csv
- automatically recognize double-occurrences in synonyms and bundling
terms for such synonyms
- snapshots can now also be xml files which are extracted from the solr
index and stored as individual xml files in the snapshot directory along
the pdf and jpg images
- a transaction layer was placed above of the snapshot directory to
distinguish snapshots into 'inventory' and 'archive'. This may be used
to do transactions of index fragments using archived solr search results
between peers. This is currently unfinished, we need a protocol to move
snapshots from inventory to archive
- the SNAPSHOT directory was renamed to snapshot and contains now two
snapshot subdirectories: inventory and archive
- snapshots may now be generated by everyone, not only such peers
running on a server with xkhtml2pdf installed. The expert crawl starts
provides the option for snapshots to everyone. PDF snapshots are now
optional and the option is only shown if xkhtml2pdf is installed.
- the snapshot api now provides the request for historised xml files,
i.e. call:
http://localhost:8090/api/snapshot.xml?urlhash=Q3dQopFh1hyQ
The result of such xml files is identical with solr search results with
only one hit.
The pdf generation has been moved from the http loading process to the
solr document storage process. This may slow down the process a lot and
a different version of the process may be needed.
so viewed text and metadata (stored) info is similar
- to archive it, use request with profile to allow indexing (defaultglobaltext) and update index
(the resource is loaded, parsed anyway, so it's not a expensive operation)
Request: remove 2 unused init parameter
- number of anchors of the parent
- forkfactor sum of anchors of all ancestors
TimeoutRequests. The purpose is to test if YaCy runs better on VMs where
there is a limitation of concurrent processes; see
/proc/user_beancounters in row numproc; this value is limited and should
be low. Try to set timeoutrequests to keep this low. (works only after
restart)
be transcoded into jpg for image previews. To create such pdfs you must
do:
Add wkhtmltopdf and imagemagick to your OS, which you can do:
On a Mac download wkhtmltox-0.12.1_osx-cocoa-x86-64.pkg from
http://wkhtmltopdf.org/downloads.html and downloadh
ttp://cactuslab.com/imagemagick/assets/ImageMagick-6.8.9-9.pkg.zip
In Debian do "apt-get install wkhtmltopdf imagemagick"
Then check in /Settings_p.html?page=ProxyAccess: "Transparent Proxy" and
"Always Fresh" - this is used by wkhtmltopdf to fetch web pages using
the YaCy proxy. Using "Always Fresh" it is possible to get all pages
from the proxy cache.
Finally, you will see a new option when starting an expert web crawl.
You can set a maximum depth for crawling which should cause a pdf
generation. The resulting pdfs are then available in
DATA/HTCACHE/SNAPSHOTS/<host>.<port>/<depth>/<shard>/<urlhash>.<date>.pdf
concurrently. When a RWI search result is flushed into the result set,
id does Solr Queries (which replaced the old-style Metadata Queries) and
they are possibly running concurrently to a previously startet Solr
search. Both methods may block each other with IO. To enhance the speed,
they are now serialized. Because the Solr search results may result in
better results using the more advanced and configurable Ranking methods,
this result is preverred over the RWI search result. However, remote RWI
search results are still feeded concurrently into the search result as
well.
if file in DATA/SETTINGS it is loaded otherwise file in ./defaults is loaded
(if locale ./defaults/stopwords.xx doesn't exist take solr/lang/stopwords_xx.txt as default)
move yacy.stopwords, yacy.stopwords.de and yacy.badwords.example out of root directory to ./defaults directory
to always get fresh lists of documents. This is necessary since the
postprocessing changes the same documents which the
postprocessing-collection query selects.
hold a date for each URL to record when a url was first seen. This is
then used to overwrite the modification date for urls upon recrawl in
case that the first-seen date is before the latest document date. This
behaviour is necessary due to the common behaviour of content management
systems which attach always the current date to all documents. Using the
firstSeen database it is possible to approximate a real first document
creation date in case that the crawler starts frequently for the same
domain. As a result the search results ordered by date have a much
better quality and the usage of YaCy as search agent for latest news has
a better quality.
postprocessing the solr documents are now not completely retrieved.
instead, only fiels, needed for the postprocessing are extracted. When
Solr document are written, this is done using partial updates.
This increases postprocessing speed by about 50% for embedded Solr
configurations. For external Solr configurations the enhancement should
be much higher because the postprocessing with remote Solr is very slow.
When doing partial updates to a remote Solr, this method should perform
much better than before, it is expected that this is even much higher
than the increase with local Solr.
This is the first element of a new 'decoration' component which may hold
switches for different external appearance parameters.
The first switch in that context is decoration.audio (as usual in
yacy.init). This value is set to false by default, that means the audio
feedback element is switched off by default. To switch it on, set
decoration.audio = true (using /ConfigProperties_p.html). You will then
hear sounds for the following events:
- remote searches
- incoming dht transmissions
- new documents from the crawler
Sound clips are stored in htroot/env/soundclips/ which is done so
because a future implementation will read these files using the http
client and with configurable urls which will make it very easy for the
user to replace the given sounds with own sounds.
*) will try other ports if YaCy standard ports are not available
*) distinguish between internal and external port (not sure if this
works 100%)
Still to add: propery in config to enter own external port (in case of
manually configured NAT)
(this eventually can benefit image search by using mime only)
reduce redundant field assignment for Solrdocuments created from URIMetadataNode (URIMetadataNode = SolrDocument with partially assigned fields)
removed preferred IPv4 in start options and added a new field IP6 in
peer seeds which will contain one or more IPv6 addresses. Now every peer
has one or more IP addresses assigned, even several IPv6 addresses are
possible. The peer-ping process must check all given and possible IP
addresses for a backping and return the one IP which was successful when
pinging the peer. The ping-ing peer must be able to recognize which of
the given IPs are available for outside access of the peer and store
this accordingly. If only one IPv6 address is available and no IPv4,
then the IPv6 is stored in the old IP field of the seed DNA.
Many methods in Seed.java are now marked as @deprecated because they had
been used for a single IP only. There is still a large construction site
left in YaCy now where all these deprecated methods must be replaced
with new method calls. The 'extra'-IPs, used by cluster assignment had
been removed since that can be replaced with IPv6 usage in p2p clusters.
All clusters must now use IPv6 if they want an intranet-routing.
solr to the YaCy built-in solr search servlet. Its not complete and not
fully correct (there is still a utf8 encoding problem) but it is a way
to get easily requests forwarded through YaCy to an external Solr.
it is now possible to get the results in two steps:
- first retrieve all IDs as given for a query
- then retieve each document individually
This was necessary for very large result sets where a query may run for
hours and is possibly terminated by a solr-internal timeout. This occurs
regulary during postprocessing and therefore this commit may fix
unwanted postprocessing terminations.
hour. The table can be shown with
http://localhost:8090/Tables_p.html?table=stats
The entries have the following meaning:
aM: activeLastMonth
aW: activeLastWeek
aD: activeLastDay
aH: activeLastHour
cC: countConnected (Active Senior)
cD: countDisconnected (Passive Senior)
cP: countPotential (Junior)
cR: count of the RWI entries
cI: size of the index (number of documents)
The entry keys are abbreviated to reduce the space in the table as the
name is written again for every row.
This is the beginning of a 'yacystats' micro-alternative als built-in
function in YaCy. Graphics may follow after some time if enough test
data is available.
found a situation after crash (reboot) with existing running semaphore but YaCy not running.
Ping generated exception which finally deleted the conf file (during pre-read procedure)
- change to ping (catch exception solved it)
- additionally removed delete yacy.conf file (if needed we need to make a backup)
during document parsing; instead use the same references that would also
be written into the webgraph. That should cause that the webgraph and
the citation index express the exact same semantic.
attributes are attached to artificial constructed index.html files which
list directories. Such files are naturally rejected by the crawler and
should not appear in the error log because these files are part of the
construction of file crawlers and confuse users if they see them in the
error log.
- use CommonParams and DisMaxParams constants
- fix typo in get sort parameter
- getDocumentCountByParams redundant implementation and risk of not optimized call (row parameter unspecified) -> as only used from getCountByQuery removed from interface
see http://mantis.tokeek.de/view.php?id=437
test result (concurrency=7)
2000 docs = eom always
1000 docs = eom always
100 docs = eom never
chosen -> 200 docs (eom not encountered during test with 1GB mem setting)
formulated as edismax query but this was not set as query attribut. The
defType=edismax property needs a qf-field, so this was added as well. Do
not remove that field again! This fixes also a problem with title-unique
computation.
filled with the date, when the url is recognized as to be outdated. That
field was partly misinterpreted and the time interval was filled in. In
case that all the urls which are in the index shall be treated as
outdated, the field is filled now with Long.MAX_VALUE because then all
crawl dates are before that date and therefore outdated.
- with search modifier /heuristic a request is send to all configured opensearch target systems (old /heuristic/blekko modifier not longer valid)
- this allows to use opensearch heuristic on individual search request (in contrast to configuration HEURISTIC_OPENSEARCH=true which sends a osd request on all global searches
- the index.html searchoption text adjusted to be displayed only if option configured
- add Archive-It to predefined systems
attribute in the <a> tag for each crawl. This introduces a lot of
changes because it extends the usage of the AnchorURL Object type which
now also has a different toString method that the underlying
DigestURL.toString. It is therefore not advised to use .toString at all
for urls, just just toNormalform(false) instead.
with metadata retrieval from connectors directly. This should cause
better usage of the cache. Automatically increase the metadata cache if
more memory is available.
'cloud' was a bad idea. Changed also the accumulation process for peer
targets so that every dht chunk is not assigned the set of redundant
targets but they are assigned to redundant targets individually. This
enhances the granularity of the target accumulation and should enhance
the efficiency of the process. Finally the dht protocol client was
enriched with the ability to remove the 'accept remote index' flag from
peers or remove peers completely if they do not answer at all.
- author facet is based on omitted author_sxt field
- adjust to make author nav available on exist of author field but keep using author_sxt to construct the facet (why!?)
- add check for querymodifier author in searchevent
- Web servers may now deliver YaCy-specific http header field with a
title and keywords. The new http header fields are:
X-YaCy-Media-Title - to be used for media (image, audio, video) titles
X-YaCy-Media-Keywords - to be used for media (image, audio, video)
keywords
- both fields are written to document fields title and keywords and are
searched also during image search.
- to make the usage of arbitrary http header fields (including this new
fields) possible in the /api/push_p.json servlet, a new POST argument is
also introduced to push http header fields. The new POST attribute is
named "responseHeader-X" (where X is the counter). It is allowed to use
this attribute as multi-attribute several times, each can be filled with
a http header line.
- see /api/push_p.html for examples
Filter queries are not actually related to ranking, but user requests
have pointed out that specific boost queries to move results to the end
of the result list are not sufficient. Such boost filters may be better
executed as actual filter and therefore such a filter can now be
statically applied to every search request. A typical use could be the
expression "http_unique_b:true AND www_unique_b:true" which uses the
recently introduced fields http_unique_b and www_unique_b which are true
only for one of the alternatives with/without http(s) and with/without
prefix 'www.' in host names.
The collection field (can be filled i.e. in Crawl Start) can be used to
add categories to YaCy index entries. The usage of that field was
restricted to solr searches and post argument filters as implemented in
commit f7571386a3.
This commit extends collections to a full navigation option in the
standard YaCy search interface. The field is not active by default but
can be activated easily in the /ConfigSearchPage_p.html servlet (just
check the 'Collection' facet field). Collections can now be used for (at
least) two purposes:
- to provide search tenants (through post argument collection)
- to provide self-made category navigation
Search requests may now have (independently from switched on or off
collection facet) a "collection:<collection-name>" modifier attached;
firthermore collection names may use disjunctions using the '|' pipe
symbol. For example, this is a valid search request:
www collection:user|proxy
crawling to the YaCy indexer. Files are uploaded using POST multipart
requests; multiple file uploads are possible as well. Each file has
attached the file date and mime type which is used to get the right
parser for the submitted data. Also an url is submitted which is
assigned to the document.
The CrawlSwitchboard has a new option for default Crawl Profiles which
are assigned dynamically from the new push interface.
- unique-postprocessing was destroying results from other
postprocessings; removed cross-updates as they had been not necessary
- unique-postprocessing did not restrict on same protocol
- inefficient concurrent update cache was redesigned completely
- increased limits for concurrent blocking queues to prevent early
time-out
After import the conent is never accessed but may take up a lot of disk space,
also the getLoadedOAIServer (which lists the files in surrogate out) is not used.
Making the surrogate.out obsolete. Removed keeping of xmls after import.
