Large-file processing

Large files

Design bounded-memory reading, import, progress, resume, and export workflows.

Processing path

Choose the workflow before tuning memory.

Start with the processing model, then add streaming, recovery, and output concerns only where the workload needs them.

Decide between normal sessions, streaming rows, chunks, database imports, and queue-backed processing.

Read forward-only, select source columns early, and keep shared strings off the PHP heap when required.

Report progress, collect row errors, persist manifests, and restart idempotently after interruption.

Write rows from iterables or database cursors without building the full workbook in memory.

Was this guide useful?Use GitHub issues for corrections, missing examples, or unclear behavior.

Open an issue