Skip to content

MIGRATION-FROM-FPDI.md

Latest commit

 

History

History
78 lines (61 loc) · 3.81 KB

MIGRATION-FROM-FPDI.md

File metadata and controls

78 lines (61 loc) · 3.81 KB
Raw

Migrating from FPDI

setasign/fpdi is free, but the moment you need to read PDFs that use compressed cross-reference streams (the default output of most modern producers) you need the commercial FPDI PDF-Parser add-on. dskripchenko/php-pdf reads classic xref, xref streams, object streams and hybrid files — plus encrypted input (RC4 / AES-128 / AES-256) and corrupt-xref recovery — under MIT.

Route 1 — the compat facade

The classic FPDI flow maps one-to-one:

// before                                   // after
use setasign\Fpdi\Fpdi;                     use Dskripchenko\PhpPdf\Compat\Fpdi;

$pdf = new Fpdi();                          $pdf = new Fpdi();
$count = $pdf->setSourceFile('in.pdf');     $count = $pdf->setSourceFile('in.pdf');
$tpl = $pdf->importPage(1);                 $tpl = $pdf->importPage(1);
$pdf->AddPage('', $pdf->getTemplateSize($tpl));
$pdf->useTemplate($tpl, x: 0, y: 0);        // same call
$pdf->Output('F', 'out.pdf');               // same call (both arg orders work)

Coordinates keep FPDF conventions: top-left origin, y downward, user units (mm by default; pt, cm, in supported in the constructor). useTemplate() scales proportionally when you pass only width or only height, exactly like FPDI.

Instead of FPDF's Cell()/SetFont() drawing API, the facade exposes the underlying objects — $pdf->page() returns the native Pdf\Page (text, images, form fields), $pdf->document() the native Pdf\Document (encryption, signing, metadata):

$pdf->page()->showText('Copy — not an original', 40, 40,
    \Dskripchenko\PhpPdf\Pdf\StandardFont::Helvetica, 9);

Route 2 — the native API

FPDI php-pdf
$pdf->setSourceFile($f) $src = ReaderDocument::fromBytes(file_get_contents($f)) (optional password argument)
(page count — return value) count($src->pages())
$tpl = $pdf->importPage($n) $form = PageImporter::intoDocument($doc, $src, $n - 1) — note 0-based index
$pdf->useTemplate($tpl, $x, $y, $w, $h) $page->useFormXObject($form, $x, $y, $w, $h) — PDF coordinates: bottom-left origin, points
$pdf->getTemplateSize($tpl) $form->bboxWidth() / $form->bboxHeight()
whole-file concatenation loop PdfMerger::create()->append(PdfSource::fromFile($a))->append(...)->toBytes()
stamp/watermark loop PdfMerger::create()->append($src)->stamp(PdfSource::fromFile($stamp), placement: Placement::fit())
encrypted source (unsupported without add-on) PdfSource::fromFile($f, password: '...') / ReaderDocument::fromBytes($bytes, '...')
xref-stream sources (commercial parser add-on) supported out of the box

Which native tool when:

  • PdfMerger — appending/reordering whole documents. Carries annotations, outlines and named destinations across (FPDI drops them).
  • PageImporter — FPDI-style: place an imported page as an XObject inside a freshly generated document, draw over or under it.

Gotchas

  • Coordinates: the facade keeps FPDF's top-left/mm conventions; the native API is PDF-native — bottom-left origin, points. Convert with y_pdf = pageHeight − y_mm × 72/25.4 − height_pt.
  • Page indexes: FPDI's importPage() is 1-based; the native PageImporter::intoDocument() is 0-based. The facade stays 1-based.
  • Annotations: like FPDI, importPage/PageImporter imports page content only. If you need links/outlines preserved, use PdfMerger.
  • adjustPageSize: instead of FPDI's flag, pass the template size to AddPage('', $pdf->getTemplateSize($tpl)) — explicit and equivalent.

The mappings above are exercised by tests/Compat/FpdiCompatTest.php.

Language: English · Русский · 中文 · Deutsch