3.5.6 Release Notes#


  • [platform] The webshot service will skip the embedly strategy if no corresponding API key is configured. This key is configured via the embedly_key option in the webshot section. By default, no key is configured.

  • [Search] Extend query and get_item API to perform approximate search on longer text inputs (Concept Search: Perform search on a subset of important topics found within long input text).

  • Aggregate relevant communities per query in descending order of item count.

  • React item side panel with support for merged entities

  • Improve concept search functionality.

  • [Communities Widget] The widget can now display only communities of certain type, relevant to current search query

  • [Smart Answers] Extends the SquirroClient with the similar searches feature: get_similar_searches

  • Add ability to see queued batches for sources with low and high priority in Data tab

  • [Dashboard Edit] Add option to hide layers when widgets inside are empty

  • Allow libNLP-steps to share heavy models across Pipelines (one cache-client within the process). This is especially needed since each ML-workflow within the machinelearning service instantiates its own NLPRunner (and would duplicate models in memory)

  • Introducing the new concept search feature.

  • [Smart Answers] Introduces backend API for the similar searches functionality: find semantically similar searches in popular queries

  • [Item Detail] Improve label display, move to the top and fix toolbar

  • Create and activate workflow for document queries embedding during a project creation. It makes that embedding queries works out-of-the box.

  • Ability to enable activity logging in the JavaScript SDK.

  • This studio plugin contains all the assets, build files and add-in requirements to load a Squirro dashboard as an add-in for Microsoft Outlook.

  • Use new React search widget for global search; Added community search scope to global search.

  • config_handler retry

  • Include batches for sources with low and high priority in totals in Monitoring plugin

Bug Fixes#

  • [Query-Processing]: Make sure to always escape all tokens before mutating the query-string (via regex replace)

  • Make concept search (mlt) work together with legacy query_transformers

  • Fix error with fetching absolute urls for items without link.

  • Fixed styling of “add widget” dashboard button when navbar color is white. Fixed z-index issues with arrow button on items widget horizontal, Fixed widget container styling issues with “disable card styling”. Fixed don’t size difference between favorites and items widget.

  • Fixed a bug that caused infinite loop for tabs widget inside global search dashboard.

  • Fix the ensure PDF lookup when the filename contains URL quoted characters. This fixes the issue of converting those files even if they existed in PDF storage, delaying their display on the UI.

  • Fixes issues with card staying when “disable card styling” is enabled/disabled. Fixed check for is empty entities config option.

  • Fix libNLP content inside wheel

  • Prevent creation of /tmp/__pycache__/*.pyc files when loading Pipelet

  • Fix alignment of PDF display icons

  • Fix issue where the Pipeline Editor could not load when the machinelearning service was not able to respond to requests, and the pipeline workflows in the project included many ML model steps.

  • Fix Community Headlines Widget not passing community360 parameters correctly; Fixed a bug where ‘undefined’ was being added to the query in the community360 dashboard.

  • Fix similar searches widget header and empty image not showing

  • In Explore screen, after performing a search in PDF file, the keywords used to disappear

  • Added CentOS 8 missing pdfconversion service dependency

Breaking Changes#

  • Change a way parameters are defined for every step

Installation and Upgrade#

You will have to resolve at least the following config files when upgrading from Squirro 3.3.0

  • /etc/nginx/conf.d/ssl.inc

  • /etc/squirro/common.ini

For new installations, please follow the Installing Squirro on Linux instructions.

To upgrade an existing installation, please consult Upgrading Squirro.