/rput/bulk
POST /eas/rput/bulk
POST /eas/rput/bulkAccess token in the Authorization header:
Authorization: Bearer <token>.
If set, this setting avoids unique constraints error from duplicate reference. Instead, if a duplicate is encountered the response loads the exsting file and
marks it notes that in _duplicates.
If /rput/bulk is used and the request contains a new file and duplicates of it, all files will be carray the _duplicates marker in the response.
When adding the renditions job to the file worker queue a priority is used. Set 2 for interactive, 4 for highest and 0 for normal.
0The filename to store. The extension of the filename determines the routing for this file. It is the basis for the the permission check as well as the version production. If the filename is omitted, fylr will use the filename retrieved by the HEAD request which is performend to check the url. If skip_head_request is used, filename is mandatory.
If leave is set, the file will not be copied into the fylr storage. Use this to manage remote files which are not meant to be copied into fylr storage. Renditions for this file will still be produced by fylr.
If a version name is given, the upload is meant to be a manual rendition. As such, the id_parents must be given alongside with it. The referenced parent files must be original and have set their produce_versions to false. Version names must be unique for all manually uploaded versions.
A parent id must be set for manual rendition uploads when using version_names. The id must point to a file "original".
A unique file reference to identify the file. It can be used in file lookups when linking files to objects.
Custom produce config is explained in POST /api/v1/eas/{fileId}.
If set to true, fylr produces renditions for each file uploaded, if set to false, no versions are produced. If no values is given (no parameter set or less values than needed), true is assumed.
trueIf set to false, metadata for this file can be provived in the property metadata. This is only allowed if leave is set to true and produce_versions to false. This is used by fylr restore for the fastest mode to (test-)import files from another system.
trueThe URL to copy the file from. fylr is trying to connect to the URL using a HEAD request. If a non 200 response is received from that request, a GET request is performed (without reading the body).
If no filename is given, the request is used to determine the filename. If a filename without an extension is given, the request is used to determine the extension.
If a filename in the Content-Disposition header is found, fylr used that. If not, the last part of the URL is used to derive a filename. If that name has no extension, fylr parses the Content-Type header to determine a file extension from the provided MIME type. The Content-Type header also sets the MIME type of the file in the database.
The Content-Length header is used to set the filesize of the file. If leave is set, the Content-Length header with a filesize > 0 is required. If leave is not set, the Content-Length header can be missing. If it is there, it needs to be > 0 too.
The host portion of the URL is validated against the fylr.eas.rput.blockedHosts config before any outbound request is made. If the host (or any address it resolves to) matches an entry, the call is refused with a 400 error. The compiled-in default blocks loopback, link-local and private (RFC1918 / ULA) ranges; fylr's own internal services (backend callback URL, Elasticsearch nodes, execservers, backend / execserver listener addresses) are added automatically on their exact host:port. See fylr.example.yml for the entry syntax (IP / CIDR / hostname / single-label wildcard, optional :port) and how to override the list.
If set, fylr does not check the validity of the URL by performing the HEAD request to the remote server. If set, providing a filename is required. This can be used to speed up file imports.
The files were registered with fylr. The response carries
each file's descriptor in long format (with a duplicates
entry if skip_duplicate_reference was set and a match was
found).
The _id is created when a file is uploaded into the fylr server.
The eas_parent_id is the id of the file this file is derived from. When using /api/produce, the newly produced files will get the parent id set to the file they are produced from.
If set, this reference string is a unique string identifying the file. If the reference contains a sha244:<hash> or sha256:<hash>, the provided hash can be checked against the file's checksum using an action triggerable in /inspect/files. Other parts of the reference can be added but must be separated by :.
The UTC timestamp when this file was uploaded.
Best date is a compiled UTC timestamp from the technical metadata. fylr uses the first set date from technical_metadata.date_time_original, technical_metadata.create_date, date_uploaded.
The file size in bytes.
The status of the file. Known states are: pending, processing, sync, done, failed. Files and renditions can be accessed in status sync and done. Only files in state done can be exported.
The UTC timestamp when the last status change was last written to the database.
The compiled extension of the file. This is matched by parsing available recipes at the time of the metadata generation. Extensions are matched by longer length first. So for a file flower.webdvd.zip the extension matched would be .webdvd.zip and not .zip if there is an available recipe.
The compiled class of the file. The classes are set by the recipe. Standard recipes of fylr include image, audio, office, video. The fylr frontend uses the class to pick the general player to view the file. A class unknown is assigned if no matching recipe was found but the upload accepts unknown types.
A compiled concatenation of class, version and status, joined by .. This is used for aggregations of file status.
A compiled concatenation of class, version, joined by .. This is used for aggregations of files
Set to true if the file is an original. It is false if the file is a generated or uploaded rendition. Produced files are registered as originals underneath the original they are produced from.
A compiled version of the metadata for this file. For images this looks like this: JPG, 1920 x 1285 px, 569.5 kB.
Filename of the file as set during the upload. On Windows, this can contain drive letters and backslashes.
The filename as set by original_filepath minus the path.
The filename as set by original_filepath minus the path and the detected extension.
This may be set if the file is in the context of an object. If set, preferred indicates that this version is the preferred one in a list of other versions.
Custom name of the file. This is set in the relation to the object. When
retrieving the file via
/api/eas.
The bulk remote upload could not be processed. The code field
identifies the specific reason — typically EasInvalidURL for a
missing or unparseable url, EasEmptyFilename /
EasInvalidFilename (filename validation), or EasRputError
when an upstream URL is blocked by the SSRF list or is otherwise
unreachable. Unlike /eas/put and /eas/rput, this endpoint
has no read-only-mode guard and does not raise ReadOnlyMode.
No authenticated user (UserRequired). The upload-right check
runs before any remote URL is fetched, so an anonymous request
is rejected here.
The authenticated user holds no RIGHT_ASSET_UPLOAD anywhere
(and is neither system root nor system.poolmanager). The
code is RightRequired.
Last updated