Reference table
File drops
Lists the file based interfaces Pensieve accepts, the formats and folders they use, and the schedule on which each drop is collected.
Pensieve accepts file-based interfaces for the systems that exchange data as files rather than live messages, and this page lists those interfaces, the formats and folders they use, and the schedule on which each drop is collected. orophin watches the nominated folders on the hospital network, collects each file as it lands, and carries it across the boundary for processing.
The file interfaces
A file interface names a folder the bridge watches and the format the files in it carry. The platform collects a file, holds it while it is scanned, then maps its contents into the record, so a file follows one controlled path from the folder to the record.
| Interface | Format | Collection |
|---|---|---|
| Row import | Comma-separated or delimited text with a header row | Collected as the file lands |
| Scheduled extract | Comma-separated or delimited text | Collected on a fixed interval an administrator sets |
| Document drop | PDF, image or office document | Collected as the file lands |
Imaging does not arrive on a file interface. A study travels from the hospital's imaging source over its own authenticated path to the Pensieve imaging store, so the pixel data stays off the watched folders and off the message channel.
Folders and formats
A file interface reads from a nominated share on the site network, arranged so a file moves through a small set of folders as it is handled. The bridge reads from the inbound folder, holds a file in the holding folder while it is scanned, and moves a collected file into the processed folder.
- nominated share
- inbounda file lands here
- holdinga file waits here while it is scanned
- processeda collected file moves here
A row import reads delimited text with a header row, and each row maps through the connection's field mapping into the record. Read how a row's columns are matched to record fields in mapping fields.
How a dropped file is collected
The bridge moves a file through a fixed sequence, and the holding stage is the load-bearing part, because a file becomes available to staff only once a scan has cleared it.
- 1
A file lands
A system or a member of staff places a file in the inbound folder of the nominated share.
- 2
The bridge collects it
The bridge collects the file and carries it across the boundary to the platform.
- 3
The file is scanned
The file waits in the holding folder while it is scanned, so it stays apart from the record until it is cleared.
- 4
The rows map
Once the scan clears the file, each row maps through the connection's field mapping and reaches the record.
- 5
The file is filed
The mapped file moves to the processed folder, and its collection is recorded in the upload trail.
A file waits for the scan
A collected file stays in the holding folder until a scan clears it, so an unscanned file waits apart from the record and reaches staff only once it is clean.
The collection schedule
A continuously watched folder is collected the moment a file lands in it, which suits a system that writes a file whenever it has one to send. A scheduled folder is collected on a fixed interval an administrator sets, which suits a system that writes an extract on its own clock, such as an overnight batch.
Both schedules run through the same holding and scanning path, so the collection schedule governs when a file is picked up rather than how it is handled once it is in hand.
The upload trail
Every collected file records an entry that names the connection, the moment of collection and the person who placed it, so staff read back every file they have sent. The entry holds the record of the crossing without the file's own patient content.
The file interface the file arrived on.
The moment the bridge took the file across the boundary.
The person or system that placed the file in the folder.
Whether the file cleared the scan and mapped into the record.
The trail is visible in the console. See how a connection and its files are watched in monitoring a connection.
Common confusions
When is a scheduled drop collected?
A scheduled folder is collected on a fixed interval an administrator sets, which suits a system that writes an extract on its own clock. A watched folder is collected the moment a file lands.
Can staff see a file straight away?
A collected file waits in the holding folder while a scan runs, so it reaches the record and staff once the scan has cleared it. The upload trail shows a file's progress through that path.
How does a row import reach the record?
Each row in a delimited file maps through the connection's field mapping into the record, so the columns a system writes are matched to record fields before anything is stored.
Read how a file interface is configured and watched in connections and credentials, and how a study reaches the viewer once it is in the imaging store in reading a study.