qbo.Attachment

The Attachment modules provides:

- Document imaging with file storage on a SAN, in the cloud (Amazon), or any other file store wrapped with the IFileObject interface
- Mail merge capability via the IGenerate interface
- Web controls supporting multiple-document uploads, drag-and-drop HTML5 and image previewing
- Zip capability
- Remote printing
- Monitoring and processing of remote locations

Future enhancements planned include:

- Bar code generation and scanning
- Inbound faxing (with bar code scanning)
- Integration with document delivery services (both physical and electronic) via IDelivery interface

Behaviors:

- Single attachment upload control
- Multiple attachment upload control
- Photo control (with lazy loader capabilities)

Document Imaging: IFileObject Interface

The Attachment module's main task is managing files associated with a QBO application. Over the course of time, files uploaded to a QBO site have been stored on:

- A web server (but that was insufficient for load balancing situations and backups),
- A SAN (but we eventually ran out of SAN space for clients storing hundreds of large images per property inspection),
- On Amazon's S3 (where most of our client documents currently reside),
- On a data server (to allow SQL direct manipulation of data files for import),
- On a client FTP site,
- In a client imaging system, and
- On the public internet

The various use cases above led to the development of the IFileObject interface. When a user attempts to upload or download a file from a QBO system, the Attachment module loads a plug-in that implements the IFileObject interface, and is coded to read or write streams of data to some underlying repository. Current IFileObject plugins include:

- LocalFile (qbo.Attachment.FileObjects.LocalFile): reads and writes from a Windows file system using local paths
- UNCFile (qbo.Attachment.FileObjects.UNCFile): reads and writes from a Windows file system using UNC paths (including credentials to access other servers)
- S3FileObject (qbo.Attachment.Amazon.S3FileObject): reads and writes from Amazon's Simple Storage Service (S3)
- Internet (qbo.Attachment.FileObjects.Internet): reads (but does not write) from the public internet
- DistributedFile (qbo.Attachment.FileObjects.DistributedFile): writes to multiple locations
- WebDAV (currently a QBO 2.0 plugin only): reads and writes from any WebDAV compliant imaging system
- FTP (currently a QBO 2.0 plugin only): reads and writes from any FTP or FTPS site
- sFTP (currently a QBO 2.0 plugin only): reads and writes from any sFTP site

The IFileObject includes the following methods:

- Read(streamToWriteTo, relativePath): reads a file into a stream
- Write(streamToReadFrom, relativePath): writes a stream to a file
- List(relativePath, pattern, depth): lists files that exist in a file store
- Delete(relativePath): deletes a file
- Rename(oldPath, newPath): renames a file

Each QBO system will include a configuration file (Config/FileObject.config) that defines the plugins used for the system. Note that a given plugin may be used more than once.

</Permissions>

</FileObject>

</Permissions>

</FileObject>

</FileObjects>

</FileObject>

These low-level methods can be accessed from Attachment/FileObject.ashx directly. For example:

- FileObject.ashx/ReadFile?FileObject=AmazonS3&Path=/Loan/12345/MyLoanDoc.pdf
- FileObject.ashx/ListFiles?FileObject=AmazonS3&Path=/Loan/12345/&Pattern=*&Depth=0
- FileObject.ashx/DeleteFile?FileObject=AmazonS3&Path=/Loan/12345/MyLoanDoc.pdf

Normally, however, files are manipulated at the same time Attachment rows are manipulated. For example:

- When uploading a file, create an Attachment row, and stream the file to the appropriate location
- When downloading a file, ensure the user has access to the corresponding Attachment row
- When deleting an attachment, delete the underlying file as well

Attachment.ashx methods include:

- Upload: used for HTML and Flash file uploads
- UploadChunk: used for HTML5 uploads
- View?ID=1988: used to view the document for AttachmentID 1988
- Download?ID=1988: used to download the document for AttachmentID 1988
- Write?ID=1988:
  - writes the HttpContext.Request.InputStream to the file underlying AttachmentID 1988
- Write?Object=Loan&ObjectID=12345:
  - writes the InputStream to a new file, and creates a new Attachment under Loan 12345
- CopyFile?ID=1988&&ID=1989&FileObject=ClientABucket&PathURL=/{Object}/{ObjectID}/{Guid}.{Uri.Extension}:
  - copies a file to ClientABucket, using a path like /Loan/12345/e326f796-56d2-422b-b6f6-c568fbfb7fa2.pdf
  - the original file and Attachment record remain unchanged
- TransferFile?ID=1988&&ID=1989&FileObject=AmazonS3&PathURL=/{Uri.Host}/{Year}-{Month}/{Uri.FileName}:
  - copies the files underlying AttachmetnIDs 1988 and 1989 to AmazonS3
  - changes the PathURL to something like /images.sears.com/2012-05/MyOriginalFileName.jpg
  - modifies the Attachments to point to them
  - leaves the original files alone
- MoveFile?ID=1988&&ID=1989&FileObject=AmazonS3:
  - moves the files underlying AttachmetnIDs 1988 and 1989 from their current location to AmazonS3
  - leaves the PathURL alone
  - modifies the Attachments to point to them
  - deletes the original

Notes:

- See AttachmentObject.GetRelativePath() to see the string substitution you can use in PathURL