Fetch Input

The Fetch Input screen is the batch automation of Processing → Extract and Process. It scans a directory of files or the BIP Print Queue, lists every candidate, and runs the same Extract → Process pipeline on every selected item.

The page applies regardless of source system — JD Edwards, SAP, NetSuite or a custom ERP — except for the BIP source, which is JD Edwards-specific.

The per-item semantics (mode resolution, validation, persistence, PA submission, BIP post-generation) are documented on Extract and Process. This page documents only what is specific to batch operation: the scan-then-select flow, the source modes, the Last Job Number incremental retrieval, and the aggregated results.

---

At a glance

---

Pipeline at a glance

The flow is two-step: a Scan call lists candidates without running any processing; the user then ticks the items to keep and clicks Process. The processing step iterates over the selection, applying the Extract and Process pipeline per item.

---

Processing Options

The first section configures how each selected item is processed. The semantics match Extract and Process — Processing; this page just exposes the same controls so the same options apply to every selected item.

Field	Description
Source (read-only display)	XML or UBL. Inferred from the chosen template's source property — set on the Documents page, not here. Selects which pipeline runs. See Process Document for the per-item semantics.
Mode (XML source only)	AUTO, SINGLE, BURST or UBL. See Process Document — Modes (XML source).
Mode (UBL only)	Process & Validate or Validate only.
Replace	Skip keeps existing invoices untouched; Overwrite re-imports them.
Send to PA	Use settings (default), Skip sending, or Force send (UBL only).

These settings apply uniformly to every selected item in the run. To process a subset with different options, run the page twice — once per option set.

---

Extract Options

The second section picks the source and any source-specific parameters.

Field	Description
Template	Required when Process As = XML. Selects the XSL pipeline applied to each item. Hidden in UBL mode (UBL files are picked up directly from dirInput/ubl/).
Source	BIP Queue (JD Edwards-specific) or Input Directory.

Source = BIP Queue

Field	Description
Language	Optional BIP language filter (e.g. FR).
Extract Mode	Extract Input (XML), Extract Output or Extract Both. See Extract BIP for the semantics.
Last Job Number	Pre-filled from global.lastBipJobNumber. The Scan call returns only jobs with jobNumber > Last Job Number. The field can be edited to re-scan a different range, but the global config is updated to the highest processed job number after each batch — incremental retrieval is the default.

Source = Input Directory

The Scan lists every .xml file present in:

• dirInput/<template>/ for Process As = XML;
• dirInput/ubl/ for Process As = UBL.

No additional parameters — every file in the directory is a candidate.

---

Scan and select

Click Scan to populate the candidate list. The list shows one row per candidate with a checkbox; rows are selected by default. Above the list:

Element	Description
Selection counter	N file(s) found in <directory> (Directory mode) or N new job(s) after #<lastJob> (BIP mode).
Select All / Deselect All	Mass toggles.

Each row carries the file basename (Directory mode) or the BIP job's base name (BIP mode). Untick rows to exclude them from the next Process call without altering the underlying directory or BIP queue.

Click Process (N) to run the selection. The button disables itself while the run is in progress and during the scan.

---

Results

After processing, the section displays an aggregated summary plus a per-item result list.

Summary bar

Metric	Description
Total	Number of items processed.
Succeeded	Items that finished with no ERROR / FATAL rows. WARNING rows do not count as failure.
Failed	Items that produced at least one blocking row.

Per-item rows

Each item appears as a row with a green ✓ (success) or red ✗ (failure) marker. Clicking a row expands the underlying log table (same columns as on the Process Document page: Severity / Module / Submodule / Message).

When the BIP source is used and processing succeeds, the Apply post-generation call runs after each item — exactly as on Extract and Process. The global lastBipJobNumber is also refreshed to the highest processed job number, so the next Scan returns only newer jobs.

---

Tips & best practices

• Use Fetch Input for unattended runs. The page is the batch counterpart to Extract and Process; the manual selection step makes it suitable for end-of-day batches or scheduled runs.
• Keep Last Job Number as the watermark. The default value is the last successfully processed job number — leaving it untouched is the supported way to do incremental retrieval. Lowering it manually re-scans older jobs (useful for replays).
• Scan first, process second. The two-step flow exists deliberately: a stale candidate list, an unintended template choice, or a wrong source mode shows up as zero / wrong rows in the candidate list before any side effect happens.
• Select All and Deselect All are top-of-list shortcuts. When the candidate list runs into hundreds of rows, mass-toggle then fine-tune is faster than ticking individually.
• Untick rather than delete. Removing the underlying file or BIP job to skip it is destructive; an untick on this page is reversible — the row reappears on the next Scan if the underlying source still has it.
• For BIP, Apply post-generation shifts the watermark. A successful processed job updates global.lastBipJobNumber automatically — nothing to maintain by hand.