boxtec
/
yacy_search_server
mirror of https://github.com/yacy/yacy_search_server
4
1
Fork 0
Code Issues Projects Releases Wiki Activity
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from 'a34f837592'
${ noResults }
yacy_search_server/source/net/yacy/search/schema/CollectionSchema.java

419 lines
32 KiB
Raw Normal View History Unescape

made the solr connection more generic
13 years ago
/**
introduced a second core named 'webgraph'. This core will hold the link structure, but is not filled yet. To have the opportunity of a second core, multi-core functionality had to be implemented to the deep-embedded solr: - migrated the solr_40 directory content to a subdirectory 'collection1'; the previously used default core is now called collection1 - added solr_40/webgraph subdirectory as second core - added a servlet configuration for the second core 'webgraph' in /IndexSchema_p.html - added instance handling as addition to solr connections: all solr connectors are now instances of an solr 'instance' object; this required a complete re-design of the solr embedding - migrated also caching and sharding ontop of new instance handling - migrated the search apis to handle now the access to a specific core, the default core named 'collection1' - migrated the remote solr search interface to access shards of cores; for the yacy remote search the default core is now called 'solr'; using the peer address as solr address - migrated the solr backup and restore process: old backups cannot be used after this migration! - redesign of solr instance handling in all methods which access the instances: they cannot hold copies of these instances any more; the must retrieve the actuall connection object every time they want to write to it (this solves also some bugs when switching the index/network) - added another schema 'solr.webgraph.schema', the old solr.keys.list is replaced by solr.collection.schema
12 years ago
* CollectionSchema
made the solr connection more generic
13 years ago
* Copyright 2011 by Michael Peter Christen
* First released 14.04.2011 at http://yacy.net
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this program in the file lgpl21.txt
* If not, see <http://www.gnu.org/licenses/>.
*/
introduced a second core named 'webgraph'. This core will hold the link structure, but is not filled yet. To have the opportunity of a second core, multi-core functionality had to be implemented to the deep-embedded solr: - migrated the solr_40 directory content to a subdirectory 'collection1'; the previously used default core is now called collection1 - added solr_40/webgraph subdirectory as second core - added a servlet configuration for the second core 'webgraph' in /IndexSchema_p.html - added instance handling as addition to solr connections: all solr connectors are now instances of an solr 'instance' object; this required a complete re-design of the solr embedding - migrated also caching and sharding ontop of new instance handling - migrated the search apis to handle now the access to a specific core, the default core named 'collection1' - migrated the remote solr search interface to access shards of cores; for the yacy remote search the default core is now called 'solr'; using the peer address as solr address - migrated the solr backup and restore process: old backups cannot be used after this migration! - redesign of solr instance handling in all methods which access the instances: they cannot hold copies of these instances any more; the must retrieve the actuall connection object every time they want to write to it (this solves also some bugs when switching the index/network) - added another schema 'solr.webgraph.schema', the old solr.keys.list is replaced by solr.collection.schema
12 years ago
package net.yacy.search.schema;
made the solr connection more generic
13 years ago
redesign of YaCySchema and SolrDoc handling
13 years ago
import java.util.Date;
import java.util.List;
introduced a second core named 'webgraph'. This core will hold the link structure, but is not filled yet. To have the opportunity of a second core, multi-core functionality had to be implemented to the deep-embedded solr: - migrated the solr_40 directory content to a subdirectory 'collection1'; the previously used default core is now called collection1 - added solr_40/webgraph subdirectory as second core - added a servlet configuration for the second core 'webgraph' in /IndexSchema_p.html - added instance handling as addition to solr connections: all solr connectors are now instances of an solr 'instance' object; this required a complete re-design of the solr embedding - migrated also caching and sharding ontop of new instance handling - migrated the search apis to handle now the access to a specific core, the default core named 'collection1' - migrated the remote solr search interface to access shards of cores; for the yacy remote search the default core is now called 'solr'; using the peer address as solr address - migrated the solr backup and restore process: old backups cannot be used after this migration! - redesign of solr instance handling in all methods which access the instances: they cannot hold copies of these instances any more; the must retrieve the actuall connection object every time they want to write to it (this solves also some bugs when switching the index/network) - added another schema 'solr.webgraph.schema', the old solr.keys.list is replaced by solr.collection.schema
12 years ago
import net.yacy.cora.federate.solr.SchemaDeclaration;
import net.yacy.cora.federate.solr.SolrType;
- added faceted drill-down for host and geolocation to solr queries - added a new geolocation field to index schema, the old values are migrated if possible
13 years ago
import org.apache.solr.common.SolrInputDocument;
introduced a second core named 'webgraph'. This core will hold the link structure, but is not filled yet. To have the opportunity of a second core, multi-core functionality had to be implemented to the deep-embedded solr: - migrated the solr_40 directory content to a subdirectory 'collection1'; the previously used default core is now called collection1 - added solr_40/webgraph subdirectory as second core - added a servlet configuration for the second core 'webgraph' in /IndexSchema_p.html - added instance handling as addition to solr connections: all solr connectors are now instances of an solr 'instance' object; this required a complete re-design of the solr embedding - migrated also caching and sharding ontop of new instance handling - migrated the search apis to handle now the access to a specific core, the default core named 'collection1' - migrated the remote solr search interface to access shards of cores; for the yacy remote search the default core is now called 'solr'; using the peer address as solr address - migrated the solr backup and restore process: old backups cannot be used after this migration! - redesign of solr instance handling in all methods which access the instances: they cannot hold copies of these instances any more; the must retrieve the actuall connection object every time they want to write to it (this solves also some bugs when switching the index/network) - added another schema 'solr.webgraph.schema', the old solr.keys.list is replaced by solr.collection.schema
12 years ago
public enum CollectionSchema implements SchemaDeclaration {
added a new fail type attribute for the index to distinguish two separate fail types: network fail and forced exclusion (i.e. by robots or forwarding rules).
12 years ago
sorted the solr schema into mandatory and optional fields; reduced number of used field to reduce solr index size
13 years ago
// mandatory
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
id(SolrType.string, true, true, false, false, false, "primary key of document, the URL hash **mandatory field**"),
replaced type of solr schema object sku of text_en_splitting_tight by string
12 years ago
sku(SolrType.string, true, true, false, true, true, "url of document"), // a 'sku' is a stock-keeping unit, a unique identifier and a default field in unmodified solr.
//sku(SolrType.text_en_splitting_tight, true, true, false, true, true, "url of document"), // a 'sku' is a stock-keeping unit, a unique identifier and a default field in unmodified solr.
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
last_modified(SolrType.date, true, true, false, false, false, "last-modified from http header"),
content_type(SolrType.string, true, true, true, false, false, "mime-type of document"),
added new fields http_unique_b and www_unique_b which can be used for ranking to prefer urls containing a www subdomain or using the https protocol
11 years ago
http_unique_b(SolrType.bool, true, true, false, false, false, "unique-field which is true when an url appears the first time. If the same url which was http then appears as https (or vice versa) then the field is false"),
www_unique_b(SolrType.bool, true, true, false, false, false, "unique-field which is true when an url appears the first time. If the same url within the subdomain www then appears without that subdomain (or vice versa) then the field is false"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
title(SolrType.text_general, true, true, true, false, true, "content of title tag"),
added new solr title_exact_signature_l and description_exact_signature_l to be able to identify unique title and unique description fields.
12 years ago
title_exact_signature_l(SolrType.num_long, true, true, false, false, false, "the 64 bit hash of the org.apache.solr.update.processor.Lookup3Signature of title, used to compute title_unique_b"),
fixed double content postprocessing
11 years ago
title_unique_b(SolrType.bool, true, true, false, false, false, "flag shows if title is unique within all indexable documents of the same host with status code 200; if yes and another document appears with same title, the unique-flag is set to false"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
host_id_s(SolrType.string, true, true, false, false, false, "id of the host, a 6-byte hash that is part of the document id"),// String hosthash();
md5_s(SolrType.string, true, true, false, false, false, "the md5 of the raw source"),// String md5();
exact_signature_l(SolrType.num_long, true, true, false, false, false, "the 64 bit hash of the org.apache.solr.update.processor.Lookup3Signature of text_t"),
exact_signature_unique_b(SolrType.bool, true, true, false, false, false, "flag shows if exact_signature_l is unique at the time of document creation, used for double-check during search"),
added two new fields, exact_signature_copycount_i and fuzzy_signature_copycount_i, which count the number of copies of non-unique documents and assigns this to each document. Thus, each document there is a number assigned which shows how many copies of this document exists. These fields are disabled by default.
12 years ago
exact_signature_copycount_i(SolrType.num_integer, true, true, false, false, false, "counter for the number of documents which are not unique (== count of not-unique-flagged documents + 1)"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
fuzzy_signature_l(SolrType.num_long, true, true, false, false, false, "64 bit of the Lookup3Signature from EnhancedTextProfileSignature of text_t"),
list more fields available for search and for ranking boosts
11 years ago
fuzzy_signature_text_t(SolrType.text_general, true, true, false, false, true, "intermediate data produced in EnhancedTextProfileSignature: a list of word frequencies"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
fuzzy_signature_unique_b(SolrType.bool, true, true, false, false, false, "flag shows if fuzzy_signature_l is unique at the time of document creation, used for double-check during search"),
added two new fields, exact_signature_copycount_i and fuzzy_signature_copycount_i, which count the number of copies of non-unique documents and assigns this to each document. Thus, each document there is a number assigned which shows how many copies of this document exists. These fields are disabled by default.
12 years ago
fuzzy_signature_copycount_i(SolrType.num_integer, true, true, false, false, false, "counter for the number of documents which are not unique (== count of not-unique-flagged documents + 1)"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
size_i(SolrType.num_integer, true, true, false, false, false, "the size of the raw source"),// int size();
re-declared some fields to be of type string rather than text which makes them more efficient and less large
12 years ago
failreason_s(SolrType.string, true, true, false, false, false, "fail reason if a page was not loaded. if the page was loaded then this field is empty"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
failtype_s(SolrType.string, true, true, false, false, false, "fail type if a page was not loaded. This field is either empty, 'excl' or 'fail'"),
httpstatus_i(SolrType.num_integer, true, true, false, false, false, "html status return code (i.e. \"200\" for ok), -1 if not loaded"),
httpstatus_redirect_s(SolrType.num_integer, true, true, false, false, false, "html status return code (i.e. \"200\" for ok), -1 if not loaded"),
added three new field for a better ranking: references_internal_i, references_external_i and references_exthosts_i. These can be used to count and evaluate the number of external links to every web page. An experimental ranking function can be i.e.: div(add(references_internal_i,product(references_external_i,references_exthosts_i)),add(clickdepth_i,1))
12 years ago
references_i(SolrType.num_integer, true, true, false, false, false, "number of unique http references, should be equal to references_internal_i + references_external_i"),
Added a citation reference computation for intra-domain link structures. While the values for the reference evaluation are computed, also a backlink-structure can be discovered and written to the index as well. The host browser has been extended to show such backlinks to each presented links. The host browser therefore can now show an information where an document is linked. The new citation reference is computed as likelyhood for a random click path with recursive usage of previously computed likelyhood. This process is repeated until the likelyhood converges to a specific number. This number is then normalized to a ranking value CRn, 0<=CRn<=1. The value CRn can therefore be used to rank popularity within intra-domain link structures.
12 years ago
references_internal_i(SolrType.num_integer, true, true, false, false, false, "number of unique http references from same host to referenced url"),
added three new field for a better ranking: references_internal_i, references_external_i and references_exthosts_i. These can be used to count and evaluate the number of external links to every web page. An experimental ranking function can be i.e.: div(add(references_internal_i,product(references_external_i,references_exthosts_i)),add(clickdepth_i,1))
12 years ago
references_external_i(SolrType.num_integer, true, true, false, false, false, "number of unique http references from external hosts"),
references_exthosts_i(SolrType.num_integer, true, true, false, false, false, "number of external hosts which provide http references"),
removed clickdepth_i field and related postprocessing. This information is now available in the crawldepth_i field which is identical to clickdepth_i because of a specific crawler strategy.
11 years ago
crawldepth_i(SolrType.num_integer, true, true, false, false, false, "crawl depth of web page according to the number of steps that the crawler did to get to this document; if the crawl was started at a root document, then this is equal to the clickdepth"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
process_sxt(SolrType.string, true, true, true, false, false, "needed (post-)processing steps on this metadata set"),
added the new field harvestkey_s to the collection index and the webgraph index which is temporary filled with the crawl profile key. This is used to select a set of documents for post-processing as soon as a crawl is finished. Now the postprocessing for a specific crawl is started when that specific crawl is finished and not at the end of all post-processing steps.
12 years ago
harvestkey_s(SolrType.string, true, true, false, false, false, "key from a harvest process (i.e. the crawl profile hash key) which is needed for near-realtime postprocessing. This shall be deleted as soon as postprocessing has been terminated."),
Preparations to produce a click depth attribute in the search index. This attribute can be used for ranking and for other purpose (demand by customer) The click depth is computed in two steps: - during indexing the current fill-state of the reverse link index is used to backtrack the current page to the root page. The length of that backtrack is the clickdepth. But this does not discover the shortest click depth. To get this, a second process to check again is needed - added a process tag that can be used to do operations on the existing index after a crawl; i.e. calculation the shortest clickpath. Added a field to control this operation but not a method to operate on this. - added a visualization of the clickpath length in the host browser
12 years ago
sorted the solr schema into mandatory and optional fields; reduced number of used field to reduce solr index size
13 years ago
// optional but recommended, part of index distribution
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
load_date_dt(SolrType.date, true, true, false, false, false, "time when resource was loaded"),
fresh_date_dt(SolrType.date, true, true, false, false, false, "date until resource shall be considered as fresh"),
Added a citation reference computation for intra-domain link structures. While the values for the reference evaluation are computed, also a backlink-structure can be discovered and written to the index as well. The host browser has been extended to show such backlinks to each presented links. The host browser therefore can now show an information where an document is linked. The new citation reference is computed as likelyhood for a random click path with recursive usage of previously computed likelyhood. This process is repeated until the likelyhood converges to a specific number. This number is then normalized to a ranking value CRn, 0<=CRn<=1. The value CRn can therefore be used to rank popularity within intra-domain link structures.
12 years ago
referrer_id_s(SolrType.string, true, true, false, false, false, "id of the referrer to this document, discovered during crawling"),// byte[] referrerHash();
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
publisher_t(SolrType.text_general, true, true, false, false, true, "the name of the publisher of the document"),// String dc_publisher();
language_s(SolrType.string, true, true, false, false, false, "the language used in the document"),// byte[] language();
audiolinkscount_i(SolrType.num_integer, true, true, false, false, false, "number of links to audio resources"),// int laudio();
videolinkscount_i(SolrType.num_integer, true, true, false, false, false, "number of links to video resources"),// int lvideo();
applinkscount_i(SolrType.num_integer, true, true, false, false, false, "number of links to application resources"),// int lapp();
- added faceted drill-down for host and geolocation to solr queries - added a new geolocation field to index schema, the old values are migrated if possible
13 years ago
sorted the solr schema into mandatory and optional fields; reduced number of used field to reduce solr index size
13 years ago
// optional but recommended
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
coordinate_p(SolrType.location, true, true, false, false, false, "point in degrees of latitude,longitude as declared in WSG84"),
coordinate_p_0_coordinate(SolrType.coordinate, true, true, false, false, false, "automatically created subfield, (latitude)"),
coordinate_p_1_coordinate(SolrType.coordinate, true, true, false, false, false, "automatically created subfield, (longitude)"),
ip_s(SolrType.string, true, true, false, false, false, "ip of host of url (after DNS lookup)"),
author(SolrType.text_general, true, true, false, false, true, "content of author-tag"),
author_sxt(SolrType.string, true, true, true, false, false, "content of author-tag as copy-field from author. This is used for facet generation"),
replaced the single-text description solr field with a multi-value description_txt text field
12 years ago
description_txt(SolrType.text_general, true, true, true, false, true, "content of description-tag(s)"),
added new solr title_exact_signature_l and description_exact_signature_l to be able to identify unique title and unique description fields.
12 years ago
description_exact_signature_l(SolrType.num_long, true, true, false, false, false, "the 64 bit hash of the org.apache.solr.update.processor.Lookup3Signature of description, used to compute description_unique_b"),
fixed double content postprocessing
11 years ago
description_unique_b(SolrType.bool, true, true, false, false, false, "flag shows if description is unique within all indexable documents of the same host with status code 200; if yes and another document appears with same description, the unique-flag is set to false"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
keywords(SolrType.text_general, true, true, false, false, true, "content of keywords tag; words are separated by space"),
charset_s(SolrType.string, true, true, false, false, false, "character encoding"),
wordcount_i(SolrType.num_integer, true, true, false, false, false, "number of words in visible area"),
added linkscount_i and linksnofollowcount_i to the default solr schema
11 years ago
linkscount_i(SolrType.num_integer, true, true, false, false, false, "number of all outgoing links; including linksnofollowcount_i"),
linksnofollowcount_i(SolrType.num_integer, true, true, false, false, false, "number of all outgoing inks with nofollow tag"),
inboundlinkscount_i(SolrType.num_integer, true, true, false, false, false, "number of outgoing inbound (to same domain) links; including inboundlinksnofollowcount_i"),
inboundlinksnofollowcount_i(SolrType.num_integer, true, true, false, false, false, "number of outgoing inbound (to same domain) links with nofollow tag"),
outboundlinkscount_i(SolrType.num_integer, true, true, false, false, false, "number of outgoing outbound (to other domain) links, including outboundlinksnofollowcount_i"),
outboundlinksnofollowcount_i(SolrType.num_integer, true, true, false, false, false, "number of outgoing outbound (to other domain) links with nofollow tag"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
imagescount_i(SolrType.num_integer, true, true, false, false, false, "number of images"),
responsetime_i(SolrType.num_integer, true, true, false, false, false, "response time of target server in milliseconds"),
text_t(SolrType.text_general, true, true, false, false, true, "all visible text"),
synonyms_sxt(SolrType.string, true, true, true, false, true, "additional synonyms to the words in the text"),
h1_txt(SolrType.text_general, true, true, true, false, true, "h1 header"),
h2_txt(SolrType.text_general, true, true, true, false, true, "h2 header"),
h3_txt(SolrType.text_general, true, true, true, false, true, "h3 header"),
h4_txt(SolrType.text_general, true, true, true, false, true, "h4 header"),
h5_txt(SolrType.text_general, true, true, true, false, true, "h5 header"),
h6_txt(SolrType.text_general, true, true, true, false, true, "h6 header"),
- added faceted drill-down for host and geolocation to solr queries - added a new geolocation field to index schema, the old values are migrated if possible
13 years ago
added a collection attribute to crawls and searches: - a solr field collection_sxt can be used to store a set of crawl tags - when this field is activated, a crawl tag can be assigned when crawls are started - the content of the collection field can be comma-separated, all of them are assigned to the documents when they are indexed as result of such a crawl start - a search result can be drilled down to a specific collection; this is currently only available in the solr interface and also in the gsa interface using the 'site' option - this adds a mandatory field for gsa queries (the google api demands that field all the time)
13 years ago
// optional values, not part of standard YaCy handling (but useful for external applications)
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
collection_sxt(SolrType.string, true, true, true, false, false, "tags that are attached to crawls/index generation to separate the search result into user-defined subsets"),
csscount_i(SolrType.num_integer, true, true, false, false, false, "number of entries in css_tag_txt and css_url_txt"),
re-declared some fields to be of type string rather than text which makes them more efficient and less large
12 years ago
css_tag_sxt(SolrType.string, true, true, true, false, false, "full css tag with normalized url"),
css_url_sxt(SolrType.string, true, true, true, false, false, "normalized urls within a css tag"),
scripts_sxt(SolrType.string, true, true, true, false, false, "normalized urls within a scripts tag"),
scriptscount_i(SolrType.num_integer, true, true, false, false, false, "number of entries in scripts_sxt"),
made the solr connection more generic
13 years ago
// encoded as binary value into an integer:
handle noarchive tag, skip writing page to cache http://mantis.tokeek.de/view.php?id=44
10 years ago
// bit 0: "all" contained in html header meta
// bit 1: "index" contained in html header meta
// bit 2: "follow" contained in html header meta
// bit 3: "noindex" contained in html header meta
// bit 4: "nofollow" contained in html header meta
// bit 5: "noarchive" contained in html header meta
// bit 8: "all" contained in http header X-Robots-Tag
// bit 9: "noindex" contained in http header X-Robots-Tag
// bit 10: "nofollow" contained in http header X-Robots-Tag
// bit 11: "noarchive" contained in http header X-Robots-Tag
// bit 12: "nosnippet" contained in http header X-Robots-Tag
// bit 13: "noodp" contained in http header X-Robots-Tag
// bit 14: "notranslate" contained in http header X-Robots-Tag
// bit 15: "noimageindex" contained in http header X-Robots-Tag
// bit 16: "unavailable_after" contained in http header X-Robots-Tag
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
robots_i(SolrType.num_integer, true, true, false, false, false, "content of <meta name=\"robots\" content=#content#> tag and the \"X-Robots-Tag\" HTTP property"),
metagenerator_t(SolrType.text_general, true, true, false, false, false, "content of <meta name=\"generator\" content=#content#> tag"),
inboundlinks_protocol_sxt(SolrType.string, true, true, true, false, false, "internal links, only the protocol"),
added the anchor text within web pages to the searcheable entities of a web page. This can be of benefit for the ranking if these fields are used for boosts.
11 years ago
inboundlinks_urlstub_sxt(SolrType.string, true, true, true, false, true, "internal links, the url only without the protocol"),
inboundlinks_anchortext_txt(SolrType.text_general, true, true, true, false, true, "internal links, the visible anchor text"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
outboundlinks_protocol_sxt(SolrType.string, true, true, true, false, false, "external links, only the protocol"),
added the anchor text within web pages to the searcheable entities of a web page. This can be of benefit for the ranking if these fields are used for boosts.
11 years ago
outboundlinks_urlstub_sxt(SolrType.string, true, true, true, false, true, "external links, the url only without the protocol"),
outboundlinks_anchortext_txt(SolrType.text_general, true, true, true, false, true, "external links, the visible anchor text"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
stub for better image indexing
12 years ago
images_text_t(SolrType.text_general, true, true, false, false, true, "all text/words appearing in image alt texts or the tokenized url"),
re-declared some fields to be of type string rather than text which makes them more efficient and less large
12 years ago
images_urlstub_sxt(SolrType.string, true, true, true, false, true, "all image links without the protocol and '://'"),
images_protocol_sxt(SolrType.string, true, true, true, false, false, "all image link protocols"),
turned images_alt_txt back to images_alt_sxt because it is not necessary to index the alt text. Indexed image Text is in images_text_t
12 years ago
images_alt_sxt(SolrType.string, true, true, true, false, true, "all image link alt tag"), // no need to index this; don't turn it into a txt field; use images_text_t instead
stub for better image indexing
12 years ago
images_height_val(SolrType.num_integer, true, true, true, false, false, "size of images:height"),
images_width_val(SolrType.num_integer, true, true, true, false, false, "size of images:width"),
images_pixel_val(SolrType.num_integer, true, true, true, false, false, "size of images as number of pixels (easier for a search restriction than with and height)"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
images_withalt_i(SolrType.num_integer, true, true, false, false, false, "number of image links with alt tag"),
htags_i(SolrType.num_integer, true, true, false, false, false, "binary pattern for the existance of h1..h6 headlines"),
canonical_s(SolrType.string, true, true, false, false, false, "url inside the canonical link element"),
canonical_equal_sku_b(SolrType.bool, true, true, false, false, false, "flag shows if the url in canonical_t is equal to sku"),
refresh_s(SolrType.string, true, true, false, false, false, "link from the url property inside the refresh link element"),
li_txt(SolrType.text_general, true, true, true, false, true, "all texts in <li> tags"),
licount_i(SolrType.num_integer, true, true, false, false, false, "number of <li> tags"),
bold_txt(SolrType.text_general, true, true, true, false, true, "all texts inside of <b> or <strong> tags. no doubles. listed in the order of number of occurrences in decreasing order"),
boldcount_i(SolrType.num_integer, true, true, false, false, false, "total number of occurrences of <b> or <strong>"),
italic_txt(SolrType.text_general, true, true, true, false, true, "all texts inside of <i> tags. no doubles. listed in the order of number of occurrences in decreasing order"),
italiccount_i(SolrType.num_integer, true, true, false, false, false, "total number of occurrences of <i>"),
underline_txt(SolrType.text_general, true, true, true, false, true, "all texts inside of <u> tags. no doubles. listed in the order of number of occurrences in decreasing order"),
underlinecount_i(SolrType.num_integer, true, true, false, false, false, "total number of occurrences of <u>"),
flash_b(SolrType.bool, true, true, false, false, false, "flag that shows if a swf file is linked"),
frames_sxt(SolrType.string, true, true, true, false, false, "list of all links to frames"),
framesscount_i(SolrType.num_integer, true, true, false, false, false, "number of frames_txt"),
iframes_sxt(SolrType.string, true, true, true, false, false, "list of all links to iframes"),
iframesscount_i(SolrType.num_integer, true, true, false, false, false, "number of iframes_txt"),
- added a solr type definition verifier - fixed type definition found by the verifier - added multivalue-string fields for solr with extension 'sxt' - added multivalue-integer fields for solr with extension 'val' - renamed some solr attributes from txt to sxt - changed solr query line to an explicit AND/OR structure - added a country code second level domain list to Domains class; with parser - added a host string parser to get domain class name, country-code second-level domain and subdomain out of it - removed old coordinate attributes
13 years ago
added new schema fields: hreflang_url_sxt and hreflang_cc_sxt for http://support.google.com/webmasters/bin/answer.py?hl=de&answer=189077 navigation_url_sxt and navigation_type_sxt for http://googlewebmastercentral.blogspot.de/2011/09/pagination-with-relnext-and-relprev.html publisher_url_s for http://support.google.com/plus/answer/1713826?hl=de all fields are disabled by default and not written to the index.
12 years ago
hreflang_url_sxt(SolrType.string, true, true, true, false, false, "url of the hreflang link tag, see http://support.google.com/webmasters/bin/answer.py?hl=de&answer=189077"),
hreflang_cc_sxt(SolrType.string, true, true, true, false, false, "country code of the hreflang link tag, see http://support.google.com/webmasters/bin/answer.py?hl=de&answer=189077"),
navigation_url_sxt(SolrType.string, true, true, true, false, false, "page navigation url, see http://googlewebmastercentral.blogspot.de/2011/09/pagination-with-relnext-and-relprev.html"),
navigation_type_sxt(SolrType.string, true, true, true, false, false, "page navigation rel property value, can contain one of {top,up,next,prev,first,last}"),
publisher_url_s(SolrType.string, true, true, false, false, false, "publisher url as defined in http://support.google.com/plus/answer/1713826?hl=de"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
url_protocol_s(SolrType.string, true, true, false, false, false, "the protocol of the url"),
list more fields available for search and for ranking boosts
11 years ago
url_file_name_s(SolrType.string, true, true, false, false, true, "the file name (which is the string after the last '/' and before the query part from '?' on) without the file extension"),
- Added new solr field url_file_name_tokens_t which stores the file name tokens. This can be used to enhance the ranking. - Added also a rating_i field as basis for later usage. - enhanced the tokenization process.
11 years ago
url_file_name_tokens_t(SolrType.text_general, true, true, false, false, true, "tokens generated from url_file_name_s which can be used for better matching and result boosting"),
list more fields available for search and for ranking boosts
11 years ago
url_file_ext_s(SolrType.string, true, true, false, false, true, "the file name extension"),
added new solr field url_paths_count_i which can be used to enhance the index browser and maybe also for ranking; possibly also for SEO-with-YaCy applications.
10 years ago
url_paths_count_i(SolrType.num_integer, true, true, false, false, false, "number of all path elements in the url hpath (see: http://www.ietf.org/rfc/rfc1738.txt) without the file name"),
added url_file_name_s in default collection schema for the file name without the file extension. This part of the file path is removed from the multi-field url_paths_sxt, which has now not the file name as last part of the path list. The same applies to the new fields source_file_name_s and target_file_name_s in the webgraph schema.
12 years ago
url_paths_sxt(SolrType.string, true, true, true, false, true, "all path elements in the url hpath (see: http://www.ietf.org/rfc/rfc1738.txt) without the file name"),
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
url_parameter_i(SolrType.num_integer, true, true, false, false, false, "number of key-value pairs in search part of the url"),
url_parameter_key_sxt(SolrType.string, true, true, true, false, false, "the keys from key-value pairs in the search part of the url"),
url_parameter_value_sxt(SolrType.string, true, true, true, false, false, "the values from key-value pairs in the search part of the url"),
url_chars_i(SolrType.num_integer, true, true, false, false, false, "number of all characters in the url == length of sku field"),
added new field for solr: url_paths_sxt url_parameter_i url_parameter_key_sxt url_parameter_value_sxt url_chars_i
13 years ago
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
host_s(SolrType.string, true, true, false, false, true, "host of the url"),
host_dnc_s(SolrType.string, true, true, false, false, true, "the Domain Class Name, either the TLD or a combination of ccSLD+TLD if a ccSLD is used."),
host_organization_s(SolrType.string, true, true, false, false, true, "either the second level domain or, if a ccSLD is used, the third level domain"),
host_organizationdnc_s(SolrType.string, true, true, false, false, true, "the organization and dnc concatenated with '.'"),
host_subdomain_s(SolrType.string, true, true, false, false, true, "the remaining part of the host without organizationdnc"),
added new field host_extent_i which, after a crawl and postprocessing, holds the number of documents for the host where the document is hosted. This is necessary for ranking and the norming of references per local host in the ranking computation.
12 years ago
host_extent_i(SolrType.num_integer, true, true, false, false, false, "number of documents from the same host; can be used to measure references_internal_i for likelihood computation"),
- added a solr type definition verifier - fixed type definition found by the verifier - added multivalue-string fields for solr with extension 'sxt' - added multivalue-integer fields for solr with extension 'val' - renamed some solr attributes from txt to sxt - changed solr query line to an explicit AND/OR structure - added a country code second level domain list to Domains class; with parser - added a host string parser to get domain class name, country-code second-level domain and subdomain out of it - removed old coordinate attributes
13 years ago
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
title_count_i(SolrType.num_integer, true, true, false, false, false, "number of titles (counting the 'title' field) in the document"),
title_chars_val(SolrType.num_integer, true, true, true, false, false, "number of characters for each title"),
title_words_val(SolrType.num_integer, true, true, true, false, false, "number of words in each title"),
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
description_count_i(SolrType.num_integer, true, true, false, false, false, "number of descriptions in the document. Its not counting the 'description' field since there is only one. But it counts the number of descriptions that appear in the document (if any)"),
description_chars_val(SolrType.num_integer, true, true, true, false, false, "number of characters for each description"),
description_words_val(SolrType.num_integer, true, true, true, false, false, "number of words in each description"),
added new field for solr: url_paths_sxt url_parameter_i url_parameter_key_sxt url_parameter_value_sxt url_chars_i
13 years ago
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
h1_i(SolrType.num_integer, true, true, false, false, false, "number of h1 header lines"),
h2_i(SolrType.num_integer, true, true, false, false, false, "number of h2 header lines"),
h3_i(SolrType.num_integer, true, true, false, false, false, "number of h3 header lines"),
h4_i(SolrType.num_integer, true, true, false, false, false, "number of h4 header lines"),
h5_i(SolrType.num_integer, true, true, false, false, false, "number of h5 header lines"),
h6_i(SolrType.num_integer, true, true, false, false, false, "number of h6 header lines"),
added Open Graph Metadata default fields, see http://ogp.me/ns#
12 years ago
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
schema_org_breadcrumb_i(SolrType.num_integer, true, true, false, false, false, "number of itemprop=\"breadcrumb\" appearances in div tags"),
opengraph_title_t(SolrType.text_general, true, true, false, false, true, "Open Graph Metadata from og:title metadata field, see http://ogp.me/ns#"),
opengraph_type_s(SolrType.text_general, true, true, false, false, false, "Open Graph Metadata from og:type metadata field, see http://ogp.me/ns#"),
opengraph_url_s(SolrType.text_general, true, true, false, false, false, "Open Graph Metadata from og:url metadata field, see http://ogp.me/ns#"),
opengraph_image_s(SolrType.text_general, true, true, false, false, false, "Open Graph Metadata from og:image metadata field, see http://ogp.me/ns#"),
added schema.org breadcrumb counter to parser and solr schema
12 years ago
Added a citation reference computation for intra-domain link structures. While the values for the reference evaluation are computed, also a backlink-structure can be discovered and written to the index as well. The host browser has been extended to show such backlinks to each presented links. The host browser therefore can now show an information where an document is linked. The new citation reference is computed as likelyhood for a random click path with recursive usage of previously computed likelyhood. This process is repeated until the likelyhood converges to a specific number. This number is then normalized to a ranking value CRn, 0<=CRn<=1. The value CRn can therefore be used to rank popularity within intra-domain link structures.
12 years ago
// link structure for ranking
cr_host_count_i(SolrType.num_integer, true, true, false, false, false, "the number of documents within a single host"),
cr_host_chance_d(SolrType.num_double, true, true, false, false, false, "the chance to click on this page when randomly clicking on links within on one host"),
cr_host_norm_i(SolrType.num_integer, true, true, false, false, false, "normalization of chance: 0 for lower halve of cr_host_count_i urls, 1 for 1/2 of the remaining and so on. the maximum number is 10"),
- Added new solr field url_file_name_tokens_t which stores the file name tokens. This can be used to enhance the ranking. - Added also a rating_i field as basis for later usage. - enhanced the tokenization process.
11 years ago
// custom rating; values to influence the ranking in combination with boost rules
rating_i(SolrType.num_integer, true, true, false, false, false, "custom rating; to be set with external rating information"),
- added a solr type definition verifier - fixed type definition found by the verifier - added multivalue-string fields for solr with extension 'sxt' - added multivalue-integer fields for solr with extension 'val' - renamed some solr attributes from txt to sxt - changed solr query line to an explicit AND/OR structure - added a country code second level domain list to Domains class; with parser - added a host string parser to get domain class name, country-code second-level domain and subdomain out of it - removed old coordinate attributes
13 years ago
// special values; can only be used if '_val' type is defined in schema file; this is not standard
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
bold_val(SolrType.num_integer, true, true, true, false, false, "number of occurrences of texts in bold_txt"),
italic_val(SolrType.num_integer, true, true, true, false, false, "number of occurrences of texts in italic_txt"),
underline_val(SolrType.num_integer, true, true, true, false, false, "number of occurrences of texts in underline_txt"),
ext_cms_txt(SolrType.text_general, true, true, true, false, false, "names of cms attributes; if several are recognized then they are listen in decreasing order of number of matching criterias"),
ext_cms_val(SolrType.num_integer, true, true, true, false, false, "number of attributes that count for a specific cms in ext_cms_txt"),
ext_ads_txt(SolrType.text_general, true, true, true, false, false, "names of ad-servers/ad-services"),
ext_ads_val(SolrType.num_integer, true, true, true, false, false, "number of attributes counts in ext_ads_txt"),
ext_community_txt(SolrType.text_general, true, true, true, false, false, "names of recognized community functions"),
ext_community_val(SolrType.num_integer, true, true, true, false, false, "number of attribute counts in attr_community"),
ext_maps_txt(SolrType.text_general, true, true, true, false, false, "names of map services"),
ext_maps_val(SolrType.num_integer, true, true, true, false, false, "number of attribute counts in ext_maps_txt"),
ext_tracker_txt(SolrType.text_general, true, true, true, false, false, "names of tracker server"),
ext_tracker_val(SolrType.num_integer, true, true, true, false, false, "number of attribute counts in ext_tracker_txt"),
ext_title_txt(SolrType.text_general, true, true, true, false, false, "names matching title expressions"),
added new solr schema fields which record the occurences of vocabulary matchings. These matches can be used for result boosting, i.e. if a document contains words from a specific vocabulary, boost it.
10 years ago
ext_title_val(SolrType.num_integer, true, true, true, false, false, "number of matching title expressions"),
vocabularies_sxt(SolrType.string, true, true, true, false, false, "collection of all vocabulary names that have a matcher in the document - use this to boost with vocabularies");
introduced a second core named 'webgraph'. This core will hold the link structure, but is not filled yet. To have the opportunity of a second core, multi-core functionality had to be implemented to the deep-embedded solr: - migrated the solr_40 directory content to a subdirectory 'collection1'; the previously used default core is now called collection1 - added solr_40/webgraph subdirectory as second core - added a servlet configuration for the second core 'webgraph' in /IndexSchema_p.html - added instance handling as addition to solr connections: all solr connectors are now instances of an solr 'instance' object; this required a complete re-design of the solr embedding - migrated also caching and sharding ontop of new instance handling - migrated the search apis to handle now the access to a specific core, the default core named 'collection1' - migrated the remote solr search interface to access shards of cores; for the yacy remote search the default core is now called 'solr'; using the peer address as solr address - migrated the solr backup and restore process: old backups cannot be used after this migration! - redesign of solr instance handling in all methods which access the instances: they cannot hold copies of these instances any more; the must retrieve the actuall connection object every time they want to write to it (this solves also some bugs when switching the index/network) - added another schema 'solr.webgraph.schema', the old solr.keys.list is replaced by solr.collection.schema
12 years ago
added new solr schema fields which record the occurences of vocabulary matchings. These matches can be used for result boosting, i.e. if a document contains words from a specific vocabulary, boost it.
10 years ago
public final static String CORE_NAME = "collection1"; // this was the default core name up to Solr 4.4.0. This default name was stored in CoreContainer.DEFAULT_DEFAULT_CORE_NAME but was removed in Solr 4.5.0
public final static String VOCABULARY_PREFIX = "vocabulary_"; // collects all terms that appear for each vocabulary
public final static String VOCABULARY_TERMS_SUFFIX = "_sxt"; // suffix for the term collector that start with VOCABULARY_PREFIX - middle part is vocabulary name
public final static String VOCABULARY_COUNT_SUFFIX = "_i"; // suffix for the term counter (>=1) that start with VOCABULARY_PREFIX - middle part is vocabulary name
public final static String VOCABULARY_LOGCOUNT_SUFFIX = "_log_i"; // log2(VOCABULARY_COUNT)] -- can be used for ranking boosts based on the number of occurrences
public final static String VOCABULARY_LOGCOUNTS_SUFFIX = "_log_val"; // all integers from [0 to log2(VOCABULARY_COUNT)] -- can be used for ranking boosts based on the number of occurrences
added a new fail type attribute for the index to distinguish two separate fail types: network fail and forced exclusion (i.e. by robots or forwarding rules).
12 years ago
Add possibility to set custom Solr field names for the YaCy default Solr attributes. - Changing the format of YaCy's solr.key.list while maintainig backward compatibility Federated index config screens adjusted accordingly - modified the Solr update request to use a 3 min Solr autocommit intervall
13 years ago
private String solrFieldName = null; // solr field name in custom solr schema, defaults to solcell schema field name (= same as this.name() )
private final SolrType type;
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
private final boolean indexed, stored, searchable, multiValued, omitNorms;
Add possibility to set custom Solr field names for the YaCy default Solr attributes. - Changing the format of YaCy's solr.key.list while maintainig backward compatibility Federated index config screens adjusted accordingly - modified the Solr update request to use a 3 min Solr autocommit intervall
13 years ago
private String comment;
made the solr connection more generic
13 years ago
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
private CollectionSchema(final SolrType type, final boolean indexed, final boolean stored, final boolean multiValued, final boolean omitNorms, final boolean searchable, final String comment) {
made the solr connection more generic
13 years ago
this.type = type;
this.indexed = indexed;
this.stored = stored;
- added a solr type definition verifier - fixed type definition found by the verifier - added multivalue-string fields for solr with extension 'sxt' - added multivalue-integer fields for solr with extension 'val' - renamed some solr attributes from txt to sxt - changed solr query line to an explicit AND/OR structure - added a country code second level domain list to Domains class; with parser - added a host string parser to get domain class name, country-code second-level domain and subdomain out of it - removed old coordinate attributes
13 years ago
this.multiValued = multiValued;
made the solr connection more generic
13 years ago
this.omitNorms = omitNorms;
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
this.searchable = searchable;
this.comment = comment;
added the new field harvestkey_s to the collection index and the webgraph index which is temporary filled with the crawl profile key. This is used to select a set of documents for post-processing as soon as a crawl is finished. Now the postprocessing for a specific crawl is started when that specific crawl is finished and not at the end of all post-processing steps.
12 years ago
// verify our naming scheme
String name = this.name();
int p = name.indexOf('_');
if (p > 0) {
String ext = name.substring(p + 1);
assert !ext.equals("i") || (type == SolrType.num_integer && !multiValued) : name;
assert !ext.equals("l") || (type == SolrType.num_long && !multiValued) : name;
assert !ext.equals("b") || (type == SolrType.bool && !multiValued) : name;
assert !ext.equals("s") || (type == SolrType.string && !multiValued) : name;
assert !ext.equals("sxt") || (type == SolrType.string && multiValued) : name;
assert !ext.equals("dt") || (type == SolrType.date && !multiValued) : name;
assert !ext.equals("t") || (type == SolrType.text_general && !multiValued) : name;
assert !ext.equals("coordinate") || (type == SolrType.coordinate && !multiValued) : name;
assert !ext.equals("txt") || (type == SolrType.text_general && multiValued) : name;
assert !ext.equals("val") || (type == SolrType.num_integer && multiValued) : name;
assert !ext.equals("d") || (type == SolrType.num_double && !multiValued) : name;
}
re-declared some fields to be of type string rather than text which makes them more efficient and less large
12 years ago
assert type.appropriateName(this) : "bad configuration: " + this.name();
made the solr connection more generic
13 years ago
}
introduced a second core named 'webgraph'. This core will hold the link structure, but is not filled yet. To have the opportunity of a second core, multi-core functionality had to be implemented to the deep-embedded solr: - migrated the solr_40 directory content to a subdirectory 'collection1'; the previously used default core is now called collection1 - added solr_40/webgraph subdirectory as second core - added a servlet configuration for the second core 'webgraph' in /IndexSchema_p.html - added instance handling as addition to solr connections: all solr connectors are now instances of an solr 'instance' object; this required a complete re-design of the solr embedding - migrated also caching and sharding ontop of new instance handling - migrated the search apis to handle now the access to a specific core, the default core named 'collection1' - migrated the remote solr search interface to access shards of cores; for the yacy remote search the default core is now called 'solr'; using the peer address as solr address - migrated the solr backup and restore process: old backups cannot be used after this migration! - redesign of solr instance handling in all methods which access the instances: they cannot hold copies of these instances any more; the must retrieve the actuall connection object every time they want to write to it (this solves also some bugs when switching the index/network) - added another schema 'solr.webgraph.schema', the old solr.keys.list is replaced by solr.collection.schema
12 years ago
Add possibility to set custom Solr field names for the YaCy default Solr attributes. - Changing the format of YaCy's solr.key.list while maintainig backward compatibility Federated index config screens adjusted accordingly - modified the Solr update request to use a 3 min Solr autocommit intervall
13 years ago
/**
* Returns the YaCy default or (if available) custom field name for Solr
* @return SolrFieldname String
*/
added solr field 'refresh_s' which stores the refresh url contained in the meta-refresh html header field.
13 years ago
@Override
Add possibility to set custom Solr field names for the YaCy default Solr attributes. - Changing the format of YaCy's solr.key.list while maintainig backward compatibility Federated index config screens adjusted accordingly - modified the Solr update request to use a 3 min Solr autocommit intervall
13 years ago
public final String getSolrFieldName() {
return (this.solrFieldName == null ? this.name() : this.solrFieldName);
}
/**
* Set a custom Solr field name (and converts it to lower case)
* @param theValue = the field name
*/
added missing @Override annotation
11 years ago
@Override
Add possibility to set custom Solr field names for the YaCy default Solr attributes. - Changing the format of YaCy's solr.key.list while maintainig backward compatibility Federated index config screens adjusted accordingly - modified the Solr update request to use a 3 min Solr autocommit intervall
13 years ago
public final void setSolrFieldName(String theValue) {
// make sure no empty string is assigned
if ( (theValue != null) && (!theValue.isEmpty()) ) {
this.solrFieldName = theValue.toLowerCase();
} else {
this.solrFieldName = null;
}
}
added solr field 'refresh_s' which stores the refresh url contained in the meta-refresh html header field.
13 years ago
@Override
made the solr connection more generic
13 years ago
public final SolrType getType() {
return this.type;
}
added solr field 'refresh_s' which stores the refresh url contained in the meta-refresh html header field.
13 years ago
@Override
made the solr connection more generic
13 years ago
public final boolean isIndexed() {
return this.indexed;
}
added solr field 'refresh_s' which stores the refresh url contained in the meta-refresh html header field.
13 years ago
@Override
made the solr connection more generic
13 years ago
public final boolean isStored() {
return this.stored;
}
added solr field 'refresh_s' which stores the refresh url contained in the meta-refresh html header field.
13 years ago
@Override
made the solr connection more generic
13 years ago
public final boolean isMultiValued() {
return this.multiValued;
}
added solr field 'refresh_s' which stores the refresh url contained in the meta-refresh html header field.
13 years ago
@Override
made the solr connection more generic
13 years ago
public final boolean isOmitNorms() {
return this.omitNorms;
}
changes in ranking computation - an existing ranking servlet for solr was extended. It is now possible to set boost values for fields, boost functions and boost queries. - The ranking can have different instances, but currently only the first one is used - added an abstraction layer for fields which can be used for search and those fields can be edited in the solr ranking configruation - the ranking value from solr within the field score is used to combine remote search requests, which all are created using the same locally defined boost values - reduced the number of fields which are used for search (makes it faster) - replaced some text fields by string fields (makes indexing faster) - removed classes which had no use - made a large number of experiments for a better ranking and created a temporary setting which prefers hits inside titles - adjusted also the RWI-based ranking computation to 'prefer title' - made special cases like for portal search where no post-processing and post-ranking is wanted: this keeps the original ranking order as done by Solr - fixed many bugs with old settings for ranking
12 years ago
@Override
public final boolean isSearchable() {
return this.searchable;
}
added solr field 'refresh_s' which stores the refresh url contained in the meta-refresh html header field.
13 years ago
@Override
made the solr connection more generic
13 years ago
public final String getComment() {
return this.comment;
- added faceted drill-down for host and geolocation to solr queries - added a new geolocation field to index schema, the old values are migrated if possible
13 years ago
}
redesign of YaCySchema and SolrDoc handling
13 years ago
added missing @Override annotation
11 years ago
@Override
redesign of YaCySchema and SolrDoc handling
13 years ago
public final void add(final SolrInputDocument doc, final String value) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
assert !this.isMultiValued();
redesign of YaCySchema and SolrDoc handling
13 years ago
doc.setField(this.getSolrFieldName(), value);
}
added missing @Override annotation
11 years ago
@Override
redesign of YaCySchema and SolrDoc handling
13 years ago
public final void add(final SolrInputDocument doc, final Date value) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
assert !this.isMultiValued();
added a feature to find similarities in documents. This uses an enhanced version of the Nutch/Solr TextProfileSignatue. As a result, a signature of the document is written to the solr search index. Additionally for each time when a signature is written, it is checked if the singature exists already in the index. If the signature does not exist, the document is marked as unique. The unique attribute can now be used to sort document lists and bring duplicates to the end of a result list. To enable this, a large portion of the search api to Solr had to be changed. This affected mainly caching of 'exists' searches to enhance the check for existing signatures and do this without actually doing a solr query. Because here the first time a long number is used as value in the Solr store, also the value naming in the YaCySchema had to be adopted and normalized. This caused that many files had to be changed.
12 years ago
assert this.type == SolrType.date;
redesign of YaCySchema and SolrDoc handling
13 years ago
doc.setField(this.getSolrFieldName(), value);
}
added missing @Override annotation
11 years ago
@Override
redesign of YaCySchema and SolrDoc handling
13 years ago
public final void add(final SolrInputDocument doc, final int value) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
assert !this.isMultiValued();
added a feature to find similarities in documents. This uses an enhanced version of the Nutch/Solr TextProfileSignatue. As a result, a signature of the document is written to the solr search index. Additionally for each time when a signature is written, it is checked if the singature exists already in the index. If the signature does not exist, the document is marked as unique. The unique attribute can now be used to sort document lists and bring duplicates to the end of a result list. To enable this, a large portion of the search api to Solr had to be changed. This affected mainly caching of 'exists' searches to enhance the check for existing signatures and do this without actually doing a solr query. Because here the first time a long number is used as value in the Solr store, also the value naming in the YaCySchema had to be adopted and normalized. This caused that many files had to be changed.
12 years ago
assert this.type == SolrType.num_integer;
redesign of YaCySchema and SolrDoc handling
13 years ago
doc.setField(this.getSolrFieldName(), value);
}
added missing @Override annotation
11 years ago
@Override
redesign of YaCySchema and SolrDoc handling
13 years ago
public final void add(final SolrInputDocument doc, final long value) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
assert !this.isMultiValued();
added a feature to find similarities in documents. This uses an enhanced version of the Nutch/Solr TextProfileSignatue. As a result, a signature of the document is written to the solr search index. Additionally for each time when a signature is written, it is checked if the singature exists already in the index. If the signature does not exist, the document is marked as unique. The unique attribute can now be used to sort document lists and bring duplicates to the end of a result list. To enable this, a large portion of the search api to Solr had to be changed. This affected mainly caching of 'exists' searches to enhance the check for existing signatures and do this without actually doing a solr query. Because here the first time a long number is used as value in the Solr store, also the value naming in the YaCySchema had to be adopted and normalized. This caused that many files had to be changed.
12 years ago
assert this.type == SolrType.num_long;
redesign of YaCySchema and SolrDoc handling
13 years ago
doc.setField(this.getSolrFieldName(), value);
}
added missing @Override annotation
11 years ago
@Override
redesign of YaCySchema and SolrDoc handling
13 years ago
public final void add(final SolrInputDocument doc, final String[] value) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
assert this.isMultiValued();
redesign of YaCySchema and SolrDoc handling
13 years ago
doc.setField(this.getSolrFieldName(), value);
}
added missing @Override annotation
11 years ago
@Override
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
public final void add(final SolrInputDocument doc, final Integer[] value) {
assert this.isMultiValued();
doc.setField(this.getSolrFieldName(), value);
}
added missing @Override annotation
11 years ago
@Override
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
public final void add(final SolrInputDocument doc, final List<?> value) {
assert this.isMultiValued();
if (value == null || value.size() == 0) {
added a feature to find similarities in documents. This uses an enhanced version of the Nutch/Solr TextProfileSignatue. As a result, a signature of the document is written to the solr search index. Additionally for each time when a signature is written, it is checked if the singature exists already in the index. If the signature does not exist, the document is marked as unique. The unique attribute can now be used to sort document lists and bring duplicates to the end of a result list. To enable this, a large portion of the search api to Solr had to be changed. This affected mainly caching of 'exists' searches to enhance the check for existing signatures and do this without actually doing a solr query. Because here the first time a long number is used as value in the Solr store, also the value naming in the YaCySchema had to be adopted and normalized. This caused that many files had to be changed.
12 years ago
if (this.type == SolrType.num_integer) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
doc.setField(this.getSolrFieldName(), new Integer[0]);
fix to collection schema; bug appeared for _txt fields with empty String as content
12 years ago
} else if (this.type == SolrType.string || this.type == SolrType.text_general) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
doc.setField(this.getSolrFieldName(), new String[0]);
} else {
fix to collection schema; bug appeared for _txt fields with empty String as content
12 years ago
assert false : "ADD(1): type is " + this.type.name();
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
doc.setField(this.getSolrFieldName(), new Object[0]);
}
return;
}
added a feature to find similarities in documents. This uses an enhanced version of the Nutch/Solr TextProfileSignatue. As a result, a signature of the document is written to the solr search index. Additionally for each time when a signature is written, it is checked if the singature exists already in the index. If the signature does not exist, the document is marked as unique. The unique attribute can now be used to sort document lists and bring duplicates to the end of a result list. To enable this, a large portion of the search api to Solr had to be changed. This affected mainly caching of 'exists' searches to enhance the check for existing signatures and do this without actually doing a solr query. Because here the first time a long number is used as value in the Solr store, also the value naming in the YaCySchema had to be adopted and normalized. This caused that many files had to be changed.
12 years ago
if (this.type == SolrType.num_integer) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
assert (value.iterator().next() instanceof Integer);
doc.setField(this.getSolrFieldName(), value.toArray(new Integer[value.size()]));
} else if (this.type == SolrType.string || this.type == SolrType.text_general) {
assert (value.iterator().next() instanceof String);
doc.setField(this.getSolrFieldName(), value.toArray(new String[value.size()]));
} else {
fix to collection schema; bug appeared for _txt fields with empty String as content
12 years ago
assert false : "ADD(2): type is " + this.type.name();
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
doc.setField(this.getSolrFieldName(), value.toArray(new Object[value.size()]));
}
redesign of YaCySchema and SolrDoc handling
13 years ago
}
added missing @Override annotation
11 years ago
@Override
redesign of YaCySchema and SolrDoc handling
13 years ago
public final void add(final SolrInputDocument doc, final float value) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
assert !this.isMultiValued();
redesign of YaCySchema and SolrDoc handling
13 years ago
doc.setField(this.getSolrFieldName(), value);
}
added missing @Override annotation
11 years ago
@Override
redesign of YaCySchema and SolrDoc handling
13 years ago
public final void add(final SolrInputDocument doc, final double value) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
assert !this.isMultiValued();
redesign of YaCySchema and SolrDoc handling
13 years ago
doc.setField(this.getSolrFieldName(), value);
}
added missing @Override annotation
11 years ago
@Override
redesign of YaCySchema and SolrDoc handling
13 years ago
public final void add(final SolrInputDocument doc, final boolean value) {
- added new solr fields: title_count_i, title_chars_val, title_words_val description_count_i, description_chars_val, description_words_val - added many asserts to ensure data type correctness from YaCy to Solr and vice versa - made many fixes according to new findings from these asserts (!)
13 years ago
assert !this.isMultiValued();
redesign of YaCySchema and SolrDoc handling
13 years ago
doc.setField(this.getSolrFieldName(), value);
}
made the solr connection more generic
13 years ago
}