Plum Laravel Package

Getting Started

Plum is a lightweight, PSR-4-compatible data processing pipeline library for PHP. To begin:

• Install via Composer: composer require plumphp/plum
• Define a pipeline using Plum\Plum class — chain Readers (data sources), Processors (transformations), and Writers (destinations)
• First use case: read CSV → transform rows (e.g., filter uppercase fields) → write to JSON

use Plum\Plum\Plum;
use Plum\Plum\Reader\CSVReader;
use Plum\Plum\Processor\CallbackProcessor;
use Plum\Plum\Writer\JSONWriter;

$plum = new Plum();
$plum->readFrom(new CSVReader('input.csv'))
     ->processWith(new CallbackProcessor(function ($row) {
         $row['name'] = strtoupper($row['name'] ?? '');
         return $row;
     }))
     ->writeTo(new JSONWriter('output.json'));
$plum->run();

Start with the README — despite low activity, the API is stable and simple.

Implementation Patterns

• Modular Pipelines: Compose processors into reusable classes (e.g., UpperNameProcessor) for clarity and testability
• Batch Processing: Use BatchProcessor to handle array inputs efficiently, e.g., bulk sanitization
• Conditional Logic: Chain processors conditionally using ConditionalProcessor (e.g., skip validation for test data)
• Error Handling: Wrap writers/processors in ErrorLoggingProcessor to capture and log bad rows
• Lazy Streams: Prefer CallbackReader with generators for memory-efficient large-file handling
• Integration Tip: Pair with Laravel’s Storage facade — extend Writer/Reader to read/write from s3:// or local:// disks

Gotchas and Tips

• No Active Maintenance: As of 2015, no updates — verify compatibility with PHP 8+ (works via PSR-0 autoloading in older versions; add spl_autoload_register or use a bridge like composer-classmap-generator if needed)
• Processor Signatures: Processors must return the modified row (or null to skip) — forgetting to return causes silent data loss
• Memory Pitfall: JSONWriter loads entire dataset into memory — for large files, use NDJSONWriter (line-delimited JSON) instead
• Testing: Unit-test processors in isolation — they’re pure functions — and mock readers/writers in pipeline tests
• Extension Points: Implement custom ReaderInterface, ProcessorInterface, or WriterInterface — all have minimal contracts
• Debugging: Use DebugProcessor to dump rows mid-pipeline:

->processWith(new CallbackProcessor(function ($row) {
    // ...
}), new DebugProcessor()) // dumps each row to stderr  

— but remember to remove before production.