v1.19.2

1.19.2 - 2021-10-20

Fixed

• Error with library dependencies broke search capabilities, rolled back to known working versions

v1.19.1

1.19.1 - 2021-10-19

Added

• Support the DefaultAWSCredentialsProviderChain
  for passing in credentials to the S3ByteStorageService.

Fixed

• Cleaning up after a failed upload should no longer decrement the file + byte counts.
• Fix the broken preview after file deletion within a folder. #277
• Fix public spaces not displaying correctly if not logged in.

Changed

• Now building mongo-init and monitor docker containers with python 3.8
• Upgraded extractor parameters jsonform to version 2.2.5.

Removed

• Check image is now part of ncsa/checks

v1.19.0

1.19.0 - 2021-10-05

Important: This update requires a MongoDB update schema due to the new ability of showing summary statistics at the
space level. Make sure to start the application with -DMONGOUPDATE=1.

Fixed

• Adding dataset to space. Space list on dataset page would be empty - fixed error when no spaces would load. #274
• Typos "success" when returning status from API and "occurred" when logging to console.
• If a dataset had multiple folders the layout would be wrong.
• Collections created using api route are now indexed upon creation. #257

Added

• Mime type for geojson
• "when" parameter in a few GET API endpoints to enable pagination #266
• Show space statistics (bytes, users. etc) #119
• "id" in GET metadata.jsonld endpoints #278
• 'POST /api/files/uploadToDataset' now allows folder_id for uploading file to folder. #232

v1.18.1

1.18.1 - 2021-08-16

This release fixes a critical issue where invalid zip files could result in the files not being uploaded correctly. To check to see if you are affected, please use the following query:

db.uploads.find({"status": "CREATED", "contentType": "application/x-zip-compressed"}, {"author.fullName": 1, "author.email": 1, "filename": 1, "uploadDate": 1, "length": 1})

If any files are returned, you should check to see if these files affected and are missing from clowder.

Fixed

• When zip file is uploaded, it will parse the file to check if it is a valid zip file, this couuld result in files not stored in final storage space #264
• Updated swagger documentation
• Return 404 not found when calling file/dataset/space api endpoints with an invalid ID #251
• Line breaks in welcome message breaks swagger build #187

Changed

• Added more information when writing files to make sure files are written correctly
• Made cilogon group check debug message instead of error message

v1.18.0

1.18.0 - 2021-07-08

Added

• Added folder and folder id to API call GET /api/datasets/:id/files. #34
• Ability to queue archive / unarchive for full datasets.
• API status endpoint GET /api/status will now show what storage type is used and for superadmins will show more
  information about the backend storage.
• GET /api/files/bulkRemove now returns status of files deleted, not found, no permission, or errors.

Fixed

• When uploading a file, any extractors marked disabled at the space level would be ignored. #246
• RabbitMQ will not use connection if it does not exist.
• Previews returns 404 if preview is not found GET /api/previews/:id.
• Added index for comments, will speed up index creation.
• If using S3 storage in docker, it was not reflected correctly in the docker-compose file.
• Docker image for mongo-init now based on python:3.7-slim to reduce size.

v1.17.0

1.17.0 - 2021-04-29

Fixed

• Close channel after submitting events to RabbitMQMessageService.

Added

• Endpoint /api/datasets/createfrombag to ingest datasets in BagIt format. Includes basic dataset metadata, files,
  folders and technical metadata. Downloading datasets now includes extra Datacite and Clowder metadata.
• Endpoint /api/files/bulkRemove to delete multiple files in one call. #12
• Log an event each time that a user archives or unarchives a file.
• Persist name of message bus response queue, preventing status messages from getting lost after a reboot.

Changed

• Updated Sphinx dependencies due to security and changes in required packages.

v1.16.0

1.16.0 - 2021-03-31

Fixed

• Remove the RabbitMQ plugin from the docker version of clowder

Added

• Added a sort and order parameter to /api/search endpoint that supports date and numeric field sorting.
  If only order is specified, created date is used. String fields are not currently supported.
• Added a new /api/deleteindex admin endpoint that will queue an action to delete an Elasticsearch index (usually prior to a reindex).
• JMeter testing suite.

Changed

• Consolidated field names sent by the EventSinkService to maximize reuse.
• Add status column to files report to indicate if files are ARCHIVED, etc.
• Reworked auto-archival configuration options to make their meanings more clear.

v1.15.1

1.15.1 - 2021-03-12

Fixed

• Several views were throwing errors trying to access a None value in EventSinkService when a user was not logged in.
  Replaced get() with getOrElse().
• Consolidated field names sent by the EventSinkService to maximize reuse.
• Changed EventSinkService logging to debug to minimize chatter.
• Don't automatically create eventsink queue and bind it to eventsink exchange. Let clients do that so that we don't
  have a queue for the eventsink filling up if there are no consumers.

v1.15.0

1.15.0 - 2021-03-03

Added

• CSV/JSON previewer using Vega.
• Previewer for FBX files.
• created search option for filtering by upload/creation date of resource.
• EventSinkService to track user activity. All events are published to the message queue. Multiple consumers are
  available in event-sink-consumers.

Fixed

• Clowder will no longer offer a Download button for a file until it has been PROCESSED.
• When space created through api the creator was not added to space as admin #179.

Changed

• /api/me will now return some of the same information as response headers. Can be used by other services to single
  sign on when running on same host.
• RabbitMQPlugin has been split into ExtractorRoutingService and MessageService to isolate the rabbitmq code from
  the extraction code.

Removed

• the toolserver is no longer build as part of clowder since it is no longer maintained. We are working on a
  newer version that will be included in future versions of clowder.

v1.14.1

1.14.1 - 2021-02-02

• Google will no longer work as login provider, we are working on this issue #157.
• If non local accounts are used the count can be wrong. Use the fixcounts script to fix this.

Fixed

• Error logging in with Orcid due to changed URL. #91
• Fixed error in url for Twitter login.
• Users count was not correct if using anything else but local accounts. #136
• Files were not properly reindexed when the Move button was used to move a file into or out of a folder in a dataset.
• When adding a file to a dataset by URL, prioritize the URL content-type header over the file content type established
  by looking at the file name extension. #139
• Wrap words across lines to stay within interface elements. #160