The Watch Folder Service is a Windows service for converting files from "hot" folders. These are sometimes also called "drop" folders. The service will watch one or more folders at a time and convert any files dropped into those folders to the format specified for that folder. This gives you the freedom to do such tasks as watch two separate folders and create black and white TIFF images out of the files dropped in the first folder, and color TIFF images from files dropped in the other folder. This can be expanded to watch as many folders as needed.

The service is installed as part of the Document Conversion Service install and is configured to use the same user account as the PEERNET Document Conversion Service Monitor 1.0 specified during installation. This service requires a privileged user account to be able to access shared volumes and to allow for remote conversion.

This application is also provided as a Visual Studio project in C#.NET and demonstrates using PNDocConvQueueServiceLib from a service in a multithreaded environment.

High Performance Clustering and Failover (DCS 3.0.010)

Clustering allows more than one computer to process against the same group of files. This group of servers all working on the same set of files is called a cluster. This type of configuration allows for an increase in conversion speed and provides fail over support if a server in the cluster needs to be restarted. The other servers in the cluster will continue to convert files. This is explained in more detail in High Performance Clustering and Fail Over Conversion.

Large Volume Batch Conversion Using Clustering (DCS 3.0.010)

Extremely large folders of files can be processed efficiently using clustering on a single computer. The same technique that allows multiple computers to process from a shared location also allows a single computer to efficiently work its way through a large folder of documents. See Large Volume Batch Conversion Using Clustering for details.

Outlook and EML Message Archives (DCS 3.0.009, 3.0.029)

Each folder section can be configured to extract and convert all attachments in Outlook Message (*.msg) archives as well as the message itself. See Processing Outlook and EML Mail Messages and Attachments for full details.

Post Conversion Processing (DCS 3.0.010)

Each watch folder can optionally run a command at the end of the conversion process. Commands are run on each created file for successful conversions, and on the original source file in the case of a failed conversion. See Post-Conversion Processing for more information.

Folder Flattening and Unique File Names (DCS 3.0.019)

Each watch folder's default behavior is to maintain any folder structure found in the input folder when creating the converted files in the Output folder. If desired, this input folder structure can be ignored, or flattened, when creating files in the output folder. To prevent filename collision and overwrite issues, options to automatically create unique filenames using Globally Unique Identifiers (GUIDs) can be used. See Unique File Naming and Flat Folder Structures for an example.

Creating Done Files to Signal Completion

Added in version 3.0.025, this allows for the creation of a done file for each file converted, or a failed file if the conversion fails. The file lists the path to the input file that was converted, followed by a list of the files created. This file is not created until conversion is complete or failed and can be used to signal completed to other processes. See Creating Done Files to Signal Completion for more details.

Picking up Files in Sort Order

Added in version 3.0.027, this allows for the files be picked and converted in order by Name (default), DateModified or DateCreated and None. If the sorting mode is set to None then no ordering is applied to the list of files returned from the underlying file system. The sort order can be Ascending (default) or Descending. See Control Sort Order on File Pickup for more details.

Files in the root of the input folder are picked up and sorted first. Then, if enabled, any subfolders, in alphabetical order, are searched for files to convert. The file list for each subfolder is returned in the requested sort mode and added to the list of files to convert.

Sample Conversion Folders

The Watch Folder Service is pre-configured to create several example conversion folders for the following common conversion types and scenarios:

ConvertToTIFF

Creates 300 DPI Optimized TIFF images.

ConvertToFaxTIFF

Creates 204x196 DPI Monochrome faxable TIFF images.

ConvertToAdobePDF

Creates, where possible, vector (searchable) Adobe PDF files. If you need to keep the hyperlinks in your documents when creating PDF files, you'll want to use this folder.

ConvertToRasterPDF

Creates PDF files where each page is an image, similar to a scanned image. Good for archiving as the page content cannot be changed.

ConvertToJPG

Creates color JPG images at 300 DPI. One image is created for each page of each document.

LargeBatchTIFF

This folder is configured to allow dropping a large number of files at once into it's input folder. The files are then picked in small batches of up to 10 files until all files in the folder have been converted.

For a more in-depth explanation of converting an existing folder containing a very large number of files, look at Large Volume Batch Conversion with Watch Folder Service.

Clustered ConvertToTIFF

This folder is pre-configured for clustered processing using a shared folder that is created as part of the installation. It creates 300 DPI Optimized TIFF images. See High Performance Clustering and Fail Over Conversion.

Running the Watch Folder Service

Watch Folder Service Overview

Starting and Stopping the Watch Folder Service

Configure the Watch Folder Service

High Performance Clustering and Fail Over Conversion

Long Path Name Support

Skipping Files with the Passthrough Converter

Processing Outlook and EML Mail Messages and Attachments

Creating Done Files to Signal Completion

Control Sort Order on File Pickup

Post-Conversion Processing

Large Volume Batch Conversion Using Clustering

Large Volume Batch Conversion Using Synchronous File Pickup