This installment covers two separate additions to a Uniface 10 customer-maintenance workflow: a custom CSV import that can be run as a dry run, and a backup attempt for the SQLite database file used by the application. The dry-run behavior belongs to that implementation; Uniface’s built-in data-copy facilities are not a general CSV preview or validation tool. The backup result also depends on the SQLite runtime and transaction path in use, so treat the reported file-copy fallback as a setup-specific observation, not a universal live-database backup recipe.
What the CSV dry run does—and what it does not
A dry run is part of the tutorial’s custom import workflow. It is intended to let the application process the import without committing the normal import changes, so a developer can inspect what the workflow finds before applying it. The available description does not include enough implementation detail to specify every validation rule, preview field, or guarantee that no side effect occurs. Confirm those behaviors in the code you are using rather than assuming that “dry run” has a fixed meaning in Uniface.
As an Amazon Associate I earn from qualifying purchases.
This is distinct from Uniface’s Data Copy Facility. Rocket’s documentation describes copying modeled entity data between databases and XML files through command-line and ProcScript entry points. It characterizes copying as a physical transformation and says the developer must ensure related data is copied correctly because referential-integrity constraints are not automatically considered: Uniface 10.4 Data Copy Facility. It does not describe a general-purpose CSV dry-run or validation feature.
Recommended Free Tools
The /cpy command is likewise a data-copy mechanism, not a substitute for designing the import’s preview and validation logic. Its source, target, and mapping locations depend on assignment settings. Rocket also cautions against using /cpy to export or import Repository definitions: Uniface /cpy reference.
#1 Best Overall
Check the form invocation before relying on a dry run
The tutorial’s reported behavior can depend on whether the import is invoked from the tools form or the customer form. In particular, unsaved edits in the customer form may be affected by the invocation path. Do not assume that a dry run preserves the current form buffer or that it is isolated from unsaved changes simply because it is not meant to commit imported records.
- Identify which form starts the import and how it passes the current customer state to the import logic.
- Check whether invoking that path saves, clears, reloads, or otherwise changes unsaved form edits.
- Test the dry run with representative input, including malformed rows and records related to existing customers, and verify both the preview and the state of the form and database afterward.
The tutorial’s description does not establish a complete validation contract or provide enough code to generalize the behavior to other Uniface applications. Treat those checks as part of adapting the workflow.
Rank #2
What Uniface establishes about the SQLite connection
Uniface’s assignment documentation identifies SLE as its SQLite connector and lists USYS$SLE_PARAMS as the connector-parameter variable: Uniface assignment reference: SQLite. That identifies the connector relationship; it does not establish which SQLite engine version or configuration is active in a particular installation.
Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallBefore choosing a backup method, determine the SQLite engine version used by the deployed application, how the SLE driver handles the relevant operation, whether the database is in a transaction, and how the database is journaled. The connector label alone is not enough to settle those runtime details.
Rank #3
VACUUM INTO and file copy: the reported result is setup-specific
The tutorial’s article result reports an attempt to use VACUUM INTO to make a backup and a fallback to copying the database file after a transaction-related error. In that reported path, a log message said the SLE driver could not VACUUM from within a transaction. This is an observed result for that setup, not proof that VACUUM INTO always fails through SLE or that copying a file is always safe.
| Approach | What is established here | What to verify |
|---|---|---|
VACUUM INTO |
The tutorial reports attempting it; its test path logged that the SLE driver could not VACUUM from within a transaction. | Confirm the deployed SQLite engine version, driver behavior, and transaction state for the exact call path. |
| File-copy fallback | The tutorial reports falling back to copying the SQLite file after that error. | Confirm that the copy is consistent for the database’s live-use and journaling conditions. The report does not establish universal safety. |
The tutorial’s filename includes a September 24, 2026 date, but that is a filename timestamp; it does not establish a publication date or independent certification of the backup method.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Make the backup workflow prove it can recover
A backup file’s existence is not enough to show that it is usable. Test the whole path in the environment you intend to support: create a backup under the same transaction and application conditions, then restore that copy to a separate location and verify that the restored database opens and contains the expected data. Keep the restore test separate from the live database so it cannot overwrite working customer data.
- Record the SQLite engine version, SLE configuration, transaction state, and journaling conditions relevant to the backup call.
- Check that the destination exists and that the application can report a clear success or failure instead of silently treating an attempted backup as complete.
- Restore a generated backup in a non-production location and verify the data needed by the application.
- Keep a separate copy of a verified backup if the workflow requires protection from loss of the application’s storage location.
Neither the /cpy facility nor the tutorial’s reported file-copy fallback establishes an online-backup guarantee for every SQLite runtime. Choose and validate the method against the actual application path rather than transferring a result from another setup.
Quick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
