Remove unused image objects

问题:

I have PDF files that are being created with a composition tool to produce financial statements.

The PDF files are around the 5000 - 10000 pages per file using global image resources to maximise space efficiences.

These statements include marketing images. Many of them (about 3mb worth), not every particular statements uses all the images.

When I extract the PDF file using a tool that has been developed for this purpose (or if I use adobe acrobat just for testing purposes) - to extract a blank page at the start of the PDF file, the resulting extracted PDF is around the 3mb. Auditing the space usage sees that it is comprised of 3mb of images.

Using iTextSharp (latest 5.4.4) I have attempted to iterate through each page and copy to a writer calling reader.RemoveUnusedObjects. But this does not reduce the size.

I also found another example to use a pdfstamper and tried the same thing. Same result.

I've also tried setting maximum compression and SetFullCompression. Neither made any difference.

Can anyone give me any pointers for what I might do. I'm hoping I can do it as a simple exercise and not have to parse the objects in the PDF file and manually remove the unused ones.

Code Below:

iTextSharp.text.pdf.PdfReader reader = new iTextSharp.text.pdf.PdfReader(inputFile);

iTextSharp.text.Document document = new iTextSharp.text.Document(reader.GetPageSizeWithRotation(1));
// step 2: we create a writer that listens to the document
// step 3: we open the document

iTextSharp.text.pdf.PdfCopy pdfCpy = new iTextSharp.text.pdf.PdfCopy(document, new System.IO.FileStream(outputFile, System.IO.FileMode.Create));
document.Open();
iTextSharp.text.pdf.PdfContentByte cb = pdfCpy.DirectContent;
//pdfCpy.NewPage();
int objects = reader.RemoveUnusedObjects();
reader.RemoveFields();
reader.RemoveAnnotations();
// we retrieve the total number of pages
int numberofPages = reader.NumberOfPages;

int i = 0;
while (i < numberofPages)
{
    i++;
    document.SetPageSize(reader.GetPageSizeWithRotation(i));
    document.NewPage();

    iTextSharp.text.pdf.PdfImportedPage page = pdfCpy.GetImportedPage(reader, i);
    pdfCpy.SetFullCompression();
    reader.RemoveUnusedObjects();
    reader.RemoveFields();
    reader.RemoveAnnotations();
    int rotation = reader.GetPageRotation(i);
    if (rotation == 90 || rotation == 270)
    {
        cb.AddTemplate(page, 0, -1f, 1f, 0, 0, reader.GetPageSizeWithRotation(i).Height);
    }
    else
    {
        cb.AddTemplate(page, 1f, 0, 0, 1f, 0, 0);
    }
    pdfCpy.AddPage(page);

}
pdfCpy.NewPage();
pdfCpy.Add(new iTextSharp.text.Paragraph("This is added text"));

document.Close();
pdfCpy.CompressionLevel = iTextSharp.text.pdf.PdfStream.BEST_COMPRESSION;
pdfCpy.Close();
reader.Close();

Stamper example:

iTextSharp.text.pdf.PdfReader reader = new iTextSharp.text.pdf.PdfReader(inputFile);
using (FileStream fs = new FileStream(outputFile + ".2" , FileMode.Create))
{
    iTextSharp.text.pdf.PdfStamper stamper = new iTextSharp.text.pdf.PdfStamper(reader, fs, iTextSharp.text.pdf.PdfWriter.VERSION_1_5);
    iTextSharp.text.pdf.PdfWriter writer = stamper.Writer;
    writer.SetPdfVersion(iTextSharp.text.pdf.PdfWriter.PDF_VERSION_1_5);
    writer.CompressionLevel = iTextSharp.text.pdf.PdfStream.BEST_COMPRESSION;
    reader.RemoveFields();
    reader.RemoveUnusedObjects();
    stamper.Reader.RemoveUnusedObjects();

    stamper.SetFullCompression();
    stamper.Writer.SetFullCompression();
    stamper.Close();
}
reader.Close();

回答1:

Try using iTextSharp.text.pdf.PdfSmartCopy instead of PdfCopy.

For me it decreased a PDF with a size of ~43MB PDF to ~4MB.