opensemanticsearch / solr-php-ui Goto Github PK

Please do a git push / PR for app.js where the javascript was moved to by #19 from @rivimey

There's an XSS vulnerability in at least one view, i.e. src/templates/view.list.php.

Patch: https://github.com/honoki/solr-php-ui/commit/611ed5e114d67f38b5572bb1ce186a0b94f222a6

For performance issues if knowledge graph has large amounts of properties we should use a database, the index or a triplestore to get property/facet labels instead of config.facets.php within the new function get_label()

To be able to use URIs or RDF/SKOS labels with / in their label in taxonomy facets, change the default separator for values/paths in filter URL options of taxonomy facets (but let it for filepath facet)

After #35 missing the more button for entities under the snippets and more button for snippets doesn't work.

To check before new release.

Add option "Synonyms" in advanced search options to enable or disable to search for synonyms, too.

Technically like option "Stemming".

Any plans to release the source code here at Github?

After new Debian release, change dependency from former php5 package to new php package, so its not dependent on version any more and the same package name like for Ubuntu.

Add new facet content type group with user friendly groups and labels of similar content types like videos, audios, spreadsheets or text documents, even if file types of different software.

Start preview on first document of list beginning with document on the current page

Like now ocr text is used for snippets in result list if no content, use fallback to deskewed OCR or better merge all snippets from content and ocr and deskewed ocr.

Current only the first 50 results of facets are shown.

Sometimes you want to see more, so have to implement optional button to show more than 50 results of a facet by setting facet field limit up for this facet.

Move Tika processing metadata fields / values like
X-Parsed-By:
org.apache.tika.parser.DefaultParser
org.apache.tika.parser.mp4.MP4Parser

X-TIKA_content_handler:
ToTextContentHandler

X-TIKA_parse_time_millis:
283

From tab Metadata to tab ETL.

User interfaces for graph visualization parameters.

Maybe do graph visualization part by Django App or API or Javascript App instead of PHP

Seems after moving javascript to app.js the status is not set to loading by waiting_on()

Enable highlighting of matches in OCR text, since if no plain text in PDF the OCR is main content

So mixing of filters in flat and hierarchical view possible (important for graph view, which uses flat values and faceted search using hierarchical values)

Delete preview.php and set preview document link to tab preview with more functionality with id as search query.

Facet values / interactive filters with special chars don't work. Test if masked correctly for url and/or Solr query.

Exclude external libs d3js, foundation and tablesaw from code and add download script to build / documentation of git clone / downloadlinks to readme.

The network / graph view of named entities doen't work anymore.
Foundation upgrade broke the tabs, maybe JS issue because of this.

Upgrade tab classes to new Foundation standard.

Link knowledge graph URIs (ids) to view preview / knowledge graph view

Integrate leaflet.js to see search results including geo coordinates on an Openstreetmap map

Use same colors for facet filter headers like for printing same class of entities in snippets.

Use the colors of new designed entity printing, since more restrained.

PHP Notice: Undefined variable: preview_image in /usr/share/solr-php-ui/templates/view.preview.php on line 252

Is it possible to filter out words/phrases in analyze > named entities page? I have some common phrases in many documents and i dont want to show them in the graph.

Thanks in advance.

Optional preview tab and facets for ETL status / error messages of ETL plugins so admins or users have not to use more complicated Solr admin UI for that.

Change linkt to preview.php to new rich preview by index like in normal result lists.

Setup for autocomplete for Solr 7 defaults

In preview show annotations from Hypothesis and from Semantic Tagger in additional tab "Tags and Annotations", not only in tab "Meta".

Hi,

I have been installing the server on a Ubuntu box, was easy and running fast, thanks a lot for this. After indexing almost 1.300.000 documents search remained fast!

However I noticed that while I could make a search on words that were part of the file name, the text inside the document themselves (PDF for example) were not available for search.

I get this on the log:

Indexing new file: /XXXXX/docbookxsl_4_duplex.2.pdf
Exception while data enrichment with plugin enhance_extract_text_tika_server:
Exception while data enrichment with plugin enhance_pdf_ocr: content_type
Indexing new file: /XXXXX/Charte Guide Styles OCDE.pdf
Exception while data enrichment with plugin enhance_extract_text_tika_server:
Exception while data enrichment with plugin enhance_pdf_ocr: content_type

I don't know if that's relevant, nor how to fix.
I get results for searching for "Styles" but not for "lexicales" which is a word inside "Charte Guide Styles OCDE.pdf"

Note: the PDF are not images, so no OCR needed.

Thanks for your help.

Use facet config of admin UI / do not print entities of facets where snippets_enabled set to false

Use custom facet limit settings from new admin UI instead of default initial count for all facets.

Add UI element for choice of relevant facets/fields/properties for connections to disable visualization of (many) irrelevant connections/coocurences

Update default facet config with config from Open Semantic Search Apps, so new/additional facets for Named Entity Recognition and Hypothesis tags are configured out of the box before adding ontologies or facets in thesaurus or ontologies manager, too.

UI element to load more connections for an entity.

If solr-php-client should have problems with newer PHP developments or to use extended features, check migration to more active and current developed Solarium

For environments with many facets facets with activated option "Closed" shown closed in search UI, so their values are visible after opening it in the search UI.

The option is set as boolean variable "closed" in facet config by Open Semantic Search Apps UI.

Hey,

Hope you are all well !

Would be maybe easier and better to use one the following projects for the webui:

ReactJS UI packages:

• https://github.com/marmelab/admin-on-rest
• https://github.com/searchkit/searchkit (Elasticsearch)

Wrapped packages:

• https://github.com/aio-libs/aiohttp_admin/tree/master/demos (including admin-on-rest or ng-admin with python aiohttp server)
• https://github.com/marmelab/admin-on-rest-graphql-demo (NodeJS)

Do you have a docker-compose.yaml file to bootstrap all components of opensemanticsearch ?

Wish you an awesome day 1

Cheers,
Richard

Add Russian grammar config.

Move snippet generation from view.list.php to helpers.php, so we can reuse functions for video and image lists.

Hi guys,
OSS is very nice!
As a former web engineer and now solutions manager, I can say OSS is absolutely wonderful and a good alternative to the shut down Google Search Appliance.
So I adopted OSS in my company. It supports a lot of workflows.

As we operate part Windows/part Mac and not on Ubuntu Desktop, I implemented a few things extra for Browser Integration:

• PDF.js for PDF previews
• jQuery
• Fancybox 3 from fancyapps for browsing previews like the pdfs or Images in Preview Tab.
  -localexplorer: an already developed plugin for chrome, in order to open that file reference in a search result, locally in your PC, for editing purposes, rather than downloading a copy
  (In the command links below search result item, distinguishing between open, and open on PC)
  -localexplorer for Mac: I couldnt find such a plugin for Mac, so I wrote a simple AppleScript app doing the same.
  -Intro.js: a javascript highlighting-Get-started-with-Opensemantic Search Guide

Right now, I am integrating an ONLYOFFICE connector, so we do not depend on Microsoft (Online) anymore for previews and editing files like docx and xlsx.

Are you guys interested in having these feature in OSS‘s Solr PHP UI?
I could check it out, put my code in and then we can review the code in a a pull request to match existing code style and programming paradigmas?

Best regards

After merging #19 the list shows not $limit elements anymore, sometimes one sometimes more.

More results are there in the html. Seems that an html element (one that is hidden like the more sinppets or more entities) is opened or (not) closed on false place / to early / to often?

If Hypothesis annotation tools are configured, add a command link to "Annotate visual by Hypothesis" to each search result & preview.

Check if newest version of Tablesaw (table UIs) for better overview & easier scrolling for the table view is compatible with the used Zurb Foundation CSS / responsive layout and if yes integrate it.

Hi There,

Would it be possible to have the code available in tar.gz as well?

Thanks!

Move function filesize_formatted from index.php to helpers.php in templates.