使用 PDFBox 格式化数字

Question

I have a PDF file (that I cannot edit) with a table into which I can enter numbers. The lowest table cell automatically sums up the input. When I manually enter the numbers (using Acrobat Reader), they are formatted correctly and the sum works fine, but when using PDFBox, they are not, ie missing the thousands separator, and the sum is not calculated. I can calculate the sum and enter into the field though. This is all in German Locale, btw.

After I fill the PDF using PDFBox, other users might edit it using Acrobat Reader and input more numbers or edit them, so the sum has to work properly. Here is screenshot of what I mean:

Is there any way to tell the form fields to reformat their input to reflect the format that they have specified internally?

When I manually format my number, which I have as a Double, to the format of "###,##0.00", then the sum does not work anymore. When I manually change any of the input, the sum is recalculated and I get an error "The value entered does not match the format of the field". Unfortunately I can't share the file directly, because of confidentiality issues, but I could try to create one of my own using only the table, if needed...

Locale.setDefault(Locale.GERMAN);

File bbb = //obviously instantiated to the where the file is
InputStream in = new FileInputStream(bbb);
PDDocument doc = PDDocument.load(in);
PDAcroForm acro = doc.getDocumentCatalog().getAcroForm();

//using the following line messes up the sum
acro.getField("row1").setValue(new DecimalFormat("###,##0.00").format(1000));

//using the following line works (including sum) but no thousands separator
acro.getField("row1").setValue(new DecimalFormat("###,##0.00").format(1000).replaceAll("\\.", ""));

Answer 1

The problem is, Acrobat Forms, aside from their declarative layout (which is parseable and analyzable via PDFBox) can also have scripts written in JavaScript, which, for obvious reasons (such as lack of a complete PDF data model and interpreter) are not evaluated within PDFBox.

You can extract the scripts from the PDF (the form is a standard XML document and the scripts are in relevant script markers) and then try to mimic the behavior of the JavaScript within your Java code. Other than that, not much can be done.

Answer 2

As mentioned by Piotr Wilkin , the field format is written in JavaScript code. I have used the following code to extract the script markers that have the format:

String js = Optional.ofNullable(acroForm.getField(fieldName)).map(PDField::getCOSObject)
        // Additional-actions dictionary. Defining the actions to be taken in response to various trigger events.
        .map(d -> (COSDictionary) d.getDictionaryObject(COSName.AA))
        // F dictionary. A JavaScript action to be performed before the field is formatted to display its current value.
        .map(d -> (COSDictionary) d.getDictionaryObject(COSName.F))
        // JS string. A string or stream containing the JavaScript script to be executed.
        .map(d -> d.getString(COSName.JS))
        .orElse(null);

That will get you the JavaScript code that defines the format. However, it is then a matter of parsing it however you need and depending on what is the original field type.