User Exits and the Scanning Cycle

The following procedure outlines the scanning program. It shows when user exits occur in relation to other events in the program and lists some actions you can cause with the user exit. The diagram at the end of this topic summarizes the scanning actions and where user exits can occur.

NOTE: ONERRORcan occur at any time during the scanning process whenever a scanning error occurs.

NOTE: Field data areas that are not populated by a grid:
Usually ScanTools Plus fills the entire field with spaces (" ") before filling the used portions of the field with data from a grid. There are cases, however, where unused field data is not cleared with spaces. This typically occurs in the following situations:

• a sheet is scanned properly but a document order stop, edit stop, or user exit pauses the sheet on the bed
• a different form is then scanned which uses a different output record.

If your external data processing application requires that all unused field data be empty in all cases then you must use literal grids to fill the empty parts of the fields.

1. Record buffers, mark buffers, and document structures are initialized.

INIT. You can initialize buffers that will be used at later user exits, open any files you can need, or query for information from the user. This is called one time per batch, before any sheets are scanned.

2. The sheet is picked and scanned by the scanner.

ACQUIRE. This is called on every sheet, after it has been scanned and digitized, but before the form ID lookup has been done. You can alter the OMR image buffer, usually to add form ID marks.

3. Form ID marks and skew marks are checked.
4. The document is checked for booklet sequence and/or document sequence. If a form ID, skew mark, booklet or document sequence error is detected, scanning error processor takes over, and DOCCHK user exit is not called.

DOCCHK. You can suppress grid resolution. This is called on every sheet.

5. The sheet is resolved, including checking for the number of timing marks. If there is a timing mark error, the grid resolution results are discarded, error processor takes over, and RESOLVE user exit is not called.

RESOLVE. You can alter the results of grid resolution. This is called on every sheet.

6. The sheet is examined for edit failures.

PRESCORE. You can alter the contents of the data record before online scoring occurs, or suppress online scoring for a document. This is called one time per document, on the last sheet.

7. Online scoring occurs (only for the last sheet of a document, when a score profile is used during scanning).

XPRINT. You can alter the print string. This is called on every sheet.

8. Transport printing occurs.

DOCLIP At this time it is possible to enable users to do their image clipping after the next sheet is fed. This enables this clip time to be hidden in the feed for the next sheet. DOCLIP is not required for iNSIGHT 70/150, 5000i, iNSIGHT 20, or iNSIGHT 30 scanners. However, any user exit intended to run on both an OpScan iNSIGHT 70/150, 5000i, iNSIGHT 20, or iNSIGHT 30 scanner and an OpScan iNSIGHT 2/4 scanner with imaging should use DOCLIP to do its clipping.For Image Scanner Implementation Only. (

9. Continue loop, steps 2 - 8, until there are no more sheets in the document. At this point the document is completely scanned and stacked.

DOCDONE. You can alter the record from the scanned document before it is written to the output buffer, or reject the document and the record will not be written to disk. This is called one time per document; if it is a booklet, after the last sheet has been scanned.

10. If performing online conversion for Delimited ASCII (.dlm) or SPSS (Windows) formats, use PRECONVERTHDR to stop scanning after the first document has completed processing, but before the conversion header has been created. This is called one time, at the first sheet of the batch.
11. If performing online conversion for Delimited ASCII (.dlm) or SPSS (Windows) formats, use POSTCONVERTHDR to stop scanning after the first document has completed processing, after the conversion header has been created, but before it is written. This is called one time, at the first sheet of the batch.
12. If performing online conversion for all formats except Delimited ASCII (.dlm) or SPSS Windows, use PRECONVERT to suppress online conversion or to alter the contents of the data record before online conversion occurs. This is called one time per document.
13. For all formats except Delimited ASCII (.dlm) or SPSS Windows: online file conversion occurs (only for the last sheet of a document).

Use POSTCONVERT to prevent the converted output data record from being written for this document. This is called one time per document.

14. The record is written to the output buffer.
15. Continue loop, steps 2 through 11, until no more documents are the batch.

CLEANUP. You can do end-of-batch processing here.

Scanning errors can occur at almost any point in the scanning cycle listed above, which means that online edit correction is done asynchronously to the main flow. PRECORRECT occurs just before online edit correction, whenever a new record is ready to be presented in the Edit Correction dialog box.

See Scanner Errors and User Exit Sequence for information about how errors impacts the sequence of user exits.

The Scanning Cycle and the Occurrence of User Exits

This diagram summarizes the scanning cycle and the occurrence of user exits, except for PRECORRECT.