[](https://github.com/DataShades/ckanext-search-tweaks/actions)
# ckanext-search-tweaks
Set of tools providing control over search results, sorting, etc.
## Requirements
Compatibility with core CKAN versions:
| CKAN version | Compatible? |
|-----------------|-------------|
| 2.8 and earlier | no |
| 2.9 | yes |
## Installation
To install ckanext-search-tweaks:
1. Activate your CKAN virtual environment, for example:
. /usr/lib/ckan/default/bin/activate
2. Install it on the virtualenv
pip install ckanext-search-tweaks
3. Add `search_tweaks` to the `ckan.plugins` setting in your CKAN
config file (by default the config file is located at
`/etc/ckan/default/ckan.ini`).
4. Restart CKAN.
## Usage
This extensions consists of multiple plugins. `search_tweaks` is the main
(major) one, that must be enabled all the time. And depending on the set of
secondary (minor) plugins, extra features and config options may be
available. Bellow are listed all the plugins with their side effects.
| Plugin | Functionality |
|-----------------------------------------------------------------|---------------------------------------------------------------------------------|
| [search_tweaks](#search_tweaks) | Allow all the other plugins to be enabled |
| [search_tweaks_query_relevance](#search_tweaks_query_relevance) | Promote datasets that were visited most frequently for the current search query |
| [search_tweaks_field_relevance](#search_tweaks_field_relevance) | Promote dataset depending on value of it's field |
| [search_tweaks_spellcheck](#search_tweaks_spellcheck) | Provides "Did you mean?" feature |
<!--
| [search_tweaks_advanced_search](#search_tweaks_advanced_search) | Basic configuration of ckanext-composite-search's search form |
-->
### <a id="search_tweaks"></a> search_tweaks
Provides base functionality and essential pieces of logic used by all the other
plugins. Must be enabled as long as at least one other plugin from this
extension is enabled.
- Switches search to `edismax` query parser if none was specified
- Enables `ckanext.search_tweaks.iterfaces.ISearchTweaks` interface with the
following methods:
def get_search_boost_fn(self, search_params: dict[str, Any]) -> Optional[str]:
"""Returns optional boost function that will be applied to the search query.
"""
return None
def get_extra_qf(self, search_params: dict[str, Any]) -> Optional[str]:
"""Return an additional fragment of the Solr's qf.
This fragment will be appended to the current qf
"""
return None
#### CLI
ckan search-tweaks -
Root of all the extension specific commands.
Every command from minor plugins is registered under this section.
#### Config settings
# Rewrite the default value of the qf parameter sent to Solr
# (optional, default: value of ckan.lib.search.query.QUERY_FIELDS).
ckanext.search_tweaks.common.qf = title^5 text
# Search by misspelled queries.
# (optional, default: false).
ckanext.search_tweaks.common.fuzzy_search.enabled = on
# Maximum number of misspelled letters. Possible values are 1 and 2.
# (optional, default: 1).
ckanext.search_tweaks.common.fuzzy_search.distance = 2
# Use `boost` instead of `bf` when `edismax` query parser is active
# (optional, default: true).
ckanext.search_tweaks.common.prefer_boost = no
# MinimumShouldMatch used in queries
# (optional, default: 1).
ckanext.search_tweaks.common.mm = 2<-1 5<80%
# Keep original query when using fuzzy search, e.g. "(hello~2) OR (hello)" if true
# (optional, default: true).
ckanext.search_tweaks.common.fuzzy_search.keep_original
---
### <a id="search_tweaks_query_relevance"></a> search_tweaks_query_relevance
Increase relevance of datasets for particular query depending on number of
direct visits of the dataset after running this search. I.e, if user searches
for `something` and then visits dataset **B** which is initially displayed in a
third row of search results, eventually this dataset will be displayed on the
second or even on the first row. This is implemented in three stages. On the
first stage, statistics collected and stored inside storage(redis, by default)
and then this statistics converted into numeric solr field via cronjob.
Finally, Solr's boost function that scales number of visits and improves score
for the given query is applied during search.
Following steps are required in order to configure this plugin:
- Add field that will store statistics to schema.xml(`query_relevance_` prefix
can be changed via config option):
<dynamicField name="query_relevance_*" type="int" indexed="true" stored="true"/>
- Configure a cronjob which will update search-index periodically:
0 0 * * * ckan search-index rebuild
#### CLI
relevance query align - remove old data from storage. Actual result of this command depends
on storage backend, that is controlled by config. At the momment, only `redis-daily` backend
is affected by this command - all records older than `query_relevance.daily.age` days are removed.
relevance query export - export statistics as CSV.
relevance query import - import statistics from CSV. Note, records that are already in storage but
are not listed in CSV won't be removed. It must be done manually
#### Config settings
# Which backend to use in order to collect information about dataset
# relevance for the particular search query. Possible values are:
# "redis-permanent", "redis-daily"
# (optional, default: redis-daily).
ckanext.search_tweaks.query_relevance.backend = redis-permanent
# How long(in days) information about dataset visits will be stored in order to
# update relevance of dataset in search query.
# (optional, default: 90).
ckanext.search_tweaks.query_relevance.daily.age = 90
# Solr boost function with $field placeholder that will be replaced by
# the correspoinding field name
# (optional, default: "scale(def($field,0),1,1.2)").
ckanext.search_tweaks.query_relevance.boost_function = recip($field,1,1000,1000)
# Prefix of the numeric field defined in Solr schema. This field will hold
# dataset's relevance for the given query.
# (optional, default: query_relevance_).
ckanext.search_tweaks.query_relevance.field_prefix = custom_score_
---
### <a id="search_tweaks_field_relevance"></a> search_tweaks_field_relevance
Increases the relevance of a dataset depending on value of its *numeric*
field. For now it's impossible to promote dataset using field with textual type.
No magic here either, this plugin allows you to specify Solr's boost function
that will be used during all the searches. One can achieve exactly the same
result using `ISearchTweaks.get_search_boost_fn`. But I expect this option to
be used often, so there is a possibility to update relevance without any extra
line of code.
#### Config settings
# Solr boost function for static numeric field
# (optional, default: None).
ckanext.search_tweaks.field_relevance.boost_function = pow(promoted_level,2)
# Field with dataset promotion level
# (optional, default: promotion_level).
ckanext.search_tweaks.field_relevance.blueprint.promotion.field_name = promotion
# Register pacakge promotion route
# (optional, default: False).
ckanext.search_tweaks.field_relevance.blueprint.promotion.enabled = true
#### Auth functions
search_tweaks_field_relevance_promote: access package promotion route. Calls `package_update` by default.
---
### <a id="search_tweaks_spellcheck"></a> search_tweaks_spellcheck
Exposes search suggestions from the Solr's spellcheck component to CKAN
templates. This plugin doesn't do much and mainly relies on the Solr's built-in
functionality. Thus you have to make a lot of changes inside Solr in order to
use it:
- `solrconfig.xml`. Configure spellcheck component. Search for `<searchComponent
name="spellcheck" class="solr.SpellCheckComponent">` section and add the
following item under it:
<lst name="spellchecker">
<str name="name">did_you_mean</str>
<str name="field">did_you_mean</str>
<str name="buildOnCommit">false</str>
</lst>
- Add cron job that will update suggestions dictionary periodically:
ckan search-tweaks spellcheck rebuild
- `solrconfig.xml`. Add spellcheck component to the search handler (`<requestHandler
name="/select" class="solr.SearchHandler">`):
<arr name="last-components">
<str>spellcheck</str>
</arr>
- Define spellcheck field in the schema. If you want to use an existing
field(`text` for example), change `<str name="field">did_you_mean</str>`
value inside `solrconfig.xml` to the name of the selected field instead.
<field name="did_you_mean" type="textgen" indexed="true" multiValued="true" />
- **Note:** skip if you've decided to use an existing field in the previous step.
<br/>
Copy meaningfull values into this field:
<copyField source="title" dest="did_you_mean"/>
<copyField source="notes" dest="did_you_mean"/>
<copyField source="res_name" dest="did_you_mean"/>
<copyField source="res_description" dest="did_you_mean"/>
<copyField source="extras_*" dest="did_you_mean"/>
After that you have to restart Solr service and rebuild search index:
ckan search-index rebuild
Now you can use `spellcheck_did_you_mean` template helper that returns better
search query when available instead of the current one. Consider including
`search_tweaks/did_you_mean.html` fragment under search form.
#### Config settings
# Do not show suggestions that have fewer results than current query
# (optional, default: true).
ckanext.search_tweaks.spellcheck.more_results_only = off
# How many different suggestions you expect to see for query
# (optional, default: 1).
ckanext.search_tweaks.spellcheck.max_suggestions = 3
#### CLI
spellcheck rebuild - rebuild/reload spellcheck dictionary.
---
<!--
### <a id="search_tweaks_advanced_search"></a> search_tweaks_advanced_search
Configure `ckanext-composite-search` for the basic usage. One need
`composite_search default_composite_search` plugins enabled in order to use
this plugin. It registers `advanced_search/search_form.html` snippet which can
be just used instead of `search_input` block of CKAN's
`snippets/search_form.html`. It can take a number of parameters, check its
content for details.
-->
## Developer installation
To install ckanext-search-tweaks for development, activate your CKAN virtualenv and
do:
git clone https://github.com/DataShades/ckanext-search-tweaks.git
cd ckanext-search-tweaks
python setup.py develop
pip install -r dev-requirements.txt
## Tests
Apart from the default configuration for CKAN testing, you have to create
`ckan_search_tweaks` Solr's core, replace its schema with
`ckanext/search_tweaks/tests/schema.xml` and make changes to `solrconfig.xml`
that are required by `search_tweaks_spellcheck`.
To run the tests, do:
pytest --ckan-ini=test.ini ckanext/search_tweaks/tests
## License
[AGPL](https://www.gnu.org/licenses/agpl-3.0.en.html)
Raw data
{
"_id": null,
"home_page": "https://github.com/DataShades/ckanext-search-tweaks",
"name": "ckanext-search-tweaks",
"maintainer": "",
"docs_url": null,
"requires_python": "",
"maintainer_email": "",
"keywords": "CKAN",
"author": "Sergey Motornyuk",
"author_email": "sergey.motornyuk@linkdigital.com.au",
"download_url": "https://files.pythonhosted.org/packages/c8/a6/820576bd4b6b2437f2a361ae9bce0876acff9b33750c909119104a871720/ckanext-search-tweaks-0.6.0.tar.gz",
"platform": null,
"description": "[](https://github.com/DataShades/ckanext-search-tweaks/actions)\n\n# ckanext-search-tweaks\n\nSet of tools providing control over search results, sorting, etc.\n\n## Requirements\n\nCompatibility with core CKAN versions:\n\n| CKAN version | Compatible? |\n|-----------------|-------------|\n| 2.8 and earlier | no |\n| 2.9 | yes |\n\n\n## Installation\n\nTo install ckanext-search-tweaks:\n\n1. Activate your CKAN virtual environment, for example:\n\n\t\t. /usr/lib/ckan/default/bin/activate\n\n2. Install it on the virtualenv\n\n\t\tpip install ckanext-search-tweaks\n\n3. Add `search_tweaks` to the `ckan.plugins` setting in your CKAN\n config file (by default the config file is located at\n `/etc/ckan/default/ckan.ini`).\n\n4. Restart CKAN.\n\n## Usage\n\nThis extensions consists of multiple plugins. `search_tweaks` is the main\n(major) one, that must be enabled all the time. And depending on the set of\nsecondary (minor) plugins, extra features and config options may be\navailable. Bellow are listed all the plugins with their side effects.\n\n| Plugin | Functionality |\n|-----------------------------------------------------------------|---------------------------------------------------------------------------------|\n| [search_tweaks](#search_tweaks) | Allow all the other plugins to be enabled |\n| [search_tweaks_query_relevance](#search_tweaks_query_relevance) | Promote datasets that were visited most frequently for the current search query |\n| [search_tweaks_field_relevance](#search_tweaks_field_relevance) | Promote dataset depending on value of it's field |\n| [search_tweaks_spellcheck](#search_tweaks_spellcheck) | Provides \"Did you mean?\" feature |\n<!--\n| [search_tweaks_advanced_search](#search_tweaks_advanced_search) | Basic configuration of ckanext-composite-search's search form |\n-->\n\n### <a id=\"search_tweaks\"></a> search_tweaks\n\nProvides base functionality and essential pieces of logic used by all the other\nplugins. Must be enabled as long as at least one other plugin from this\nextension is enabled.\n\n- Switches search to `edismax` query parser if none was specified\n- Enables `ckanext.search_tweaks.iterfaces.ISearchTweaks` interface with the\nfollowing methods:\n\n\t\tdef get_search_boost_fn(self, search_params: dict[str, Any]) -> Optional[str]:\n\t\t\t\"\"\"Returns optional boost function that will be applied to the search query.\n\t\t\t\"\"\"\n\t\t\treturn None\n\n\t\tdef get_extra_qf(self, search_params: dict[str, Any]) -> Optional[str]:\n\t\t\t\"\"\"Return an additional fragment of the Solr's qf.\n\t\t This fragment will be appended to the current qf\n\t\t\t\"\"\"\n\t\t\treturn None\n\n#### CLI\n\n\tckan search-tweaks -\n\t\tRoot of all the extension specific commands.\n\t\tEvery command from minor plugins is registered under this section.\n\n\n#### Config settings\n\n\t# Rewrite the default value of the qf parameter sent to Solr\n\t# (optional, default: value of ckan.lib.search.query.QUERY_FIELDS).\n\tckanext.search_tweaks.common.qf = title^5 text\n\n\t# Search by misspelled queries.\n\t# (optional, default: false).\n\tckanext.search_tweaks.common.fuzzy_search.enabled = on\n\n\t# Maximum number of misspelled letters. Possible values are 1 and 2.\n\t# (optional, default: 1).\n\tckanext.search_tweaks.common.fuzzy_search.distance = 2\n\n\t# Use `boost` instead of `bf` when `edismax` query parser is active\n\t# (optional, default: true).\n\tckanext.search_tweaks.common.prefer_boost = no\n\n\t# MinimumShouldMatch used in queries\n\t# (optional, default: 1).\n\tckanext.search_tweaks.common.mm = 2<-1 5<80%\n\n # Keep original query when using fuzzy search, e.g. \"(hello~2) OR (hello)\" if true\n # (optional, default: true).\n ckanext.search_tweaks.common.fuzzy_search.keep_original\n\n\n---\n\n### <a id=\"search_tweaks_query_relevance\"></a> search_tweaks_query_relevance\n\nIncrease relevance of datasets for particular query depending on number of\ndirect visits of the dataset after running this search. I.e, if user searches\nfor `something` and then visits dataset **B** which is initially displayed in a\nthird row of search results, eventually this dataset will be displayed on the\nsecond or even on the first row. This is implemented in three stages. On the\nfirst stage, statistics collected and stored inside storage(redis, by default)\nand then this statistics converted into numeric solr field via cronjob.\nFinally, Solr's boost function that scales number of visits and improves score\nfor the given query is applied during search.\n\nFollowing steps are required in order to configure this plugin:\n\n- Add field that will store statistics to schema.xml(`query_relevance_` prefix\n can be changed via config option):\n\n\t\t<dynamicField name=\"query_relevance_*\" type=\"int\" indexed=\"true\" stored=\"true\"/>\n\n- Configure a cronjob which will update search-index periodically:\n\n\t\t0 0 * * * ckan search-index rebuild\n\n#### CLI\n\n\trelevance query align - remove old data from storage. Actual result of this command depends\n\t\ton storage backend, that is controlled by config. At the momment, only `redis-daily` backend\n\t\tis affected by this command - all records older than `query_relevance.daily.age` days are removed.\n\n\trelevance query export - export statistics as CSV.\n\n\trelevance query import - import statistics from CSV. Note, records that are already in storage but\n\t\tare not listed in CSV won't be removed. It must be done manually\n\n\n#### Config settings\n\n\t# Which backend to use in order to collect information about dataset\n\t# relevance for the particular search query. Possible values are:\n\t# \"redis-permanent\", \"redis-daily\"\n\t# (optional, default: redis-daily).\n\tckanext.search_tweaks.query_relevance.backend = redis-permanent\n\n\t# How long(in days) information about dataset visits will be stored in order to\n\t# update relevance of dataset in search query.\n\t# (optional, default: 90).\n\tckanext.search_tweaks.query_relevance.daily.age = 90\n\n\t# Solr boost function with $field placeholder that will be replaced by\n\t# the correspoinding field name\n\t# (optional, default: \"scale(def($field,0),1,1.2)\").\n\tckanext.search_tweaks.query_relevance.boost_function = recip($field,1,1000,1000)\n\n\t# Prefix of the numeric field defined in Solr schema. This field will hold\n\t# dataset's relevance for the given query.\n\t# (optional, default: query_relevance_).\n\tckanext.search_tweaks.query_relevance.field_prefix = custom_score_\n\n---\n### <a id=\"search_tweaks_field_relevance\"></a> search_tweaks_field_relevance\n\nIncreases the relevance of a dataset depending on value of its *numeric*\nfield. For now it's impossible to promote dataset using field with textual type.\n\nNo magic here either, this plugin allows you to specify Solr's boost function\nthat will be used during all the searches. One can achieve exactly the same\nresult using `ISearchTweaks.get_search_boost_fn`. But I expect this option to\nbe used often, so there is a possibility to update relevance without any extra\nline of code.\n\n#### Config settings\n\n\t# Solr boost function for static numeric field\n\t# (optional, default: None).\n\tckanext.search_tweaks.field_relevance.boost_function = pow(promoted_level,2)\n\n\t# Field with dataset promotion level\n\t# (optional, default: promotion_level).\n\tckanext.search_tweaks.field_relevance.blueprint.promotion.field_name = promotion\n\n\t# Register pacakge promotion route\n\t# (optional, default: False).\n\tckanext.search_tweaks.field_relevance.blueprint.promotion.enabled = true\n\n#### Auth functions\n\n\tsearch_tweaks_field_relevance_promote: access package promotion route. Calls `package_update` by default.\n\n---\n\n### <a id=\"search_tweaks_spellcheck\"></a> search_tweaks_spellcheck\n\nExposes search suggestions from the Solr's spellcheck component to CKAN\ntemplates. This plugin doesn't do much and mainly relies on the Solr's built-in\nfunctionality. Thus you have to make a lot of changes inside Solr in order to\nuse it:\n\n- `solrconfig.xml`. Configure spellcheck component. Search for `<searchComponent\n name=\"spellcheck\" class=\"solr.SpellCheckComponent\">` section and add the\n following item under it:\n\n\t\t<lst name=\"spellchecker\">\n\t\t\t<str name=\"name\">did_you_mean</str>\n\t\t\t<str name=\"field\">did_you_mean</str>\n\t\t\t<str name=\"buildOnCommit\">false</str>\n\t\t</lst>\n\n- Add cron job that will update suggestions dictionary periodically:\n\n\t\tckan search-tweaks spellcheck rebuild\n\n- `solrconfig.xml`. Add spellcheck component to the search handler (`<requestHandler\n name=\"/select\" class=\"solr.SearchHandler\">`):\n\n\t\t<arr name=\"last-components\">\n\t\t\t<str>spellcheck</str>\n\t\t</arr>\n\n- Define spellcheck field in the schema. If you want to use an existing\n field(`text` for example), change `<str name=\"field\">did_you_mean</str>`\n value inside `solrconfig.xml` to the name of the selected field instead.\n\n\t\t<field name=\"did_you_mean\" type=\"textgen\" indexed=\"true\" multiValued=\"true\" />\n\n- **Note:** skip if you've decided to use an existing field in the previous step.\n <br/>\n Copy meaningfull values into this field:\n\n\t\t<copyField source=\"title\" dest=\"did_you_mean\"/>\n\t\t<copyField source=\"notes\" dest=\"did_you_mean\"/>\n\t\t<copyField source=\"res_name\" dest=\"did_you_mean\"/>\n\t\t<copyField source=\"res_description\" dest=\"did_you_mean\"/>\n\t\t<copyField source=\"extras_*\" dest=\"did_you_mean\"/>\n\nAfter that you have to restart Solr service and rebuild search index:\n\n\tckan search-index rebuild\n\nNow you can use `spellcheck_did_you_mean` template helper that returns better\nsearch query when available instead of the current one. Consider including\n`search_tweaks/did_you_mean.html` fragment under search form.\n\n#### Config settings\n\n\t# Do not show suggestions that have fewer results than current query\n\t# (optional, default: true).\n\tckanext.search_tweaks.spellcheck.more_results_only = off\n\n\t# How many different suggestions you expect to see for query\n\t# (optional, default: 1).\n\tckanext.search_tweaks.spellcheck.max_suggestions = 3\n\n#### CLI\n\n\tspellcheck rebuild - rebuild/reload spellcheck dictionary.\n\n---\n<!--\n### <a id=\"search_tweaks_advanced_search\"></a> search_tweaks_advanced_search\n\nConfigure `ckanext-composite-search` for the basic usage. One need\n`composite_search default_composite_search` plugins enabled in order to use\nthis plugin. It registers `advanced_search/search_form.html` snippet which can\nbe just used instead of `search_input` block of CKAN's\n`snippets/search_form.html`. It can take a number of parameters, check its\ncontent for details.\n-->\n\n## Developer installation\n\nTo install ckanext-search-tweaks for development, activate your CKAN virtualenv and\ndo:\n\n git clone https://github.com/DataShades/ckanext-search-tweaks.git\n cd ckanext-search-tweaks\n python setup.py develop\n pip install -r dev-requirements.txt\n\n\n## Tests\n\nApart from the default configuration for CKAN testing, you have to create\n`ckan_search_tweaks` Solr's core, replace its schema with\n`ckanext/search_tweaks/tests/schema.xml` and make changes to `solrconfig.xml`\nthat are required by `search_tweaks_spellcheck`.\n\nTo run the tests, do:\n\n pytest --ckan-ini=test.ini ckanext/search_tweaks/tests\n\n\n## License\n\n[AGPL](https://www.gnu.org/licenses/agpl-3.0.en.html)\n",
"bugtrack_url": null,
"license": "AGPL",
"summary": "",
"version": "0.6.0",
"project_urls": {
"Homepage": "https://github.com/DataShades/ckanext-search-tweaks"
},
"split_keywords": [
"ckan"
],
"urls": [
{
"comment_text": "",
"digests": {
"blake2b_256": "117d5baa91693b0653b832acf62a770ca8f304febcfbf7ea390dbda4187b2bd3",
"md5": "7fc2273690e11ccb74f9e8ea8982082d",
"sha256": "3f7960385e5c399655a0318f73645b32c072b89693dfab477707a694f1686adf"
},
"downloads": -1,
"filename": "ckanext_search_tweaks-0.6.0-py3-none-any.whl",
"has_sig": false,
"md5_digest": "7fc2273690e11ccb74f9e8ea8982082d",
"packagetype": "bdist_wheel",
"python_version": "py3",
"requires_python": null,
"size": 50122,
"upload_time": "2023-11-04T18:15:31",
"upload_time_iso_8601": "2023-11-04T18:15:31.699169Z",
"url": "https://files.pythonhosted.org/packages/11/7d/5baa91693b0653b832acf62a770ca8f304febcfbf7ea390dbda4187b2bd3/ckanext_search_tweaks-0.6.0-py3-none-any.whl",
"yanked": false,
"yanked_reason": null
},
{
"comment_text": "",
"digests": {
"blake2b_256": "c8a6820576bd4b6b2437f2a361ae9bce0876acff9b33750c909119104a871720",
"md5": "e535f8af893bbc281b4b4bf0b7de275b",
"sha256": "f674600abcf20cb8fe6871a25609a91d9b4228159726b1685df474ab87da232c"
},
"downloads": -1,
"filename": "ckanext-search-tweaks-0.6.0.tar.gz",
"has_sig": false,
"md5_digest": "e535f8af893bbc281b4b4bf0b7de275b",
"packagetype": "sdist",
"python_version": "source",
"requires_python": null,
"size": 43285,
"upload_time": "2023-11-04T18:15:34",
"upload_time_iso_8601": "2023-11-04T18:15:34.975116Z",
"url": "https://files.pythonhosted.org/packages/c8/a6/820576bd4b6b2437f2a361ae9bce0876acff9b33750c909119104a871720/ckanext-search-tweaks-0.6.0.tar.gz",
"yanked": false,
"yanked_reason": null
}
],
"upload_time": "2023-11-04 18:15:34",
"github": true,
"gitlab": false,
"bitbucket": false,
"codeberg": false,
"github_user": "DataShades",
"github_project": "ckanext-search-tweaks",
"travis_ci": false,
"coveralls": true,
"github_actions": true,
"requirements": [],
"lcname": "ckanext-search-tweaks"
}
JSON
