具有特殊字符的 PDFBox 在 Windows 上工作正常，但字符在 Linux 上被其他字符替换

Question

我的要求是我有一个带有少量文本字段的 PDF Acroform 模板，并且基于登录用户，这些文本字段预先填充了用户名和城市并呈现给用户（注意用户不是手动输入数据到PDF 但我的后端代码正在这样做）。所以这些名称和城市字段给非英语字符带来了问题。奇怪的是这在我的 loca (windows 10) 上工作得很好。然而，当代码部署在 lunix 机器上时，那些非英文字符将被替换为其他一些字符。

我正在使用 pdfbox-app 版本 2.0.24

我的类路径中有这种字体 Arial_Narrow.ttf。（解决另一个问题 - java.lang.IllegalArgumentException：...在此字体的编码中不可用：WinAnsiEncoding）

我的假设是因为此 Arial_Narrow.ttf 字体具有 WinAnsiEncoding，它可以在 Windows 10 机器上运行，但在部署到 Lunix 机器上时会出现问题。

请帮忙！

package com.example.demo;

import org.apache.pdfbox.pdmodel.PDDocument;
import org.apache.pdfbox.pdmodel.PDResources;
import org.apache.pdfbox.pdmodel.font.PDType0Font;
import org.apache.pdfbox.pdmodel.interactive.form.PDAcroForm;
import org.apache.pdfbox.pdmodel.interactive.form.PDCheckBox;
import org.apache.pdfbox.pdmodel.interactive.form.PDField;
import org.apache.pdfbox.pdmodel.interactive.form.PDRadioButton;
import org.apache.pdfbox.pdmodel.interactive.form.PDTextField;

import java.io.ByteArrayOutputStream;
import java.io.IOException;
import java.util.Iterator;
import java.util.Map;
import java.util.Set;

public class PDFMailMergeUtil {

    /**
     * constants used to check or uncheck a checkbox in the PDF
     */
    private static final String CHECKBOX_VALUE_ON = "ON";
    private static final String CHECKBOX_VALUE_OFF = "OFF";

    /**
     * This method takes PDF Template and a map of PDF Form field names and their values,
     * and returns Actual PDF Document with appropriate values.
     * Consider templateContent as Class and returned data as Object of that cklass.
     *
     * @param templateContent PDF Template
     * @param pdf_fieldName_fieldValue_map PDF form field names and their values
     * @return Actual PDF document with filled values
     */
    public static byte[] createPdfDocumentFromPdfTemplate(byte[] templateContent, Map<String, String> pdf_fieldName_fieldValue_map) {

        byte[] mailMergedPDFContent = null;

        // validation
        if (templateContent == null) {
            System.out.println ("PDF Template Content is null.");
            return null;
        }

        ByteArrayOutputStream out = null;
        PDDocument pdDoc = null;

        try {
            pdDoc = PDDocument.load(templateContent);
            PDAcroForm pdAcroForm = pdDoc.getDocumentCatalog().getAcroForm();

            if (pdAcroForm == null) {
                System.out.println("No Form Field present in the PDF Template.");
            } else {
                PDType0Font font = PDType0Font.load(pdDoc, PDFMailMergeUtil.class.getResourceAsStream("/Arial_Narrow.ttf")); // I have this Arial_Narrow.ttf font in my resources folder so available in classpath

                PDResources res = pdAcroForm.getDefaultResources();

                String fontName = res.add(font).getName();
                String defaultAppearanceString = "/" + fontName + " 10 Tf 0 g";

                Iterator<PDField> fieldsIterator = pdAcroForm.getFieldIterator();
                while (fieldsIterator.hasNext()) {

                    PDField pdfield = fieldsIterator.next();

                    String formFieldName = pdfield.getFullyQualifiedName();
                    // check if fieldName-Value Map contains the form field name in template
                    if (!pdf_fieldName_fieldValue_map.containsKey(formFieldName)) {
                        continue;
                    }

                    // We are here - means the PDF Acro Form Field name is present in our name-value map
                    // get field value from map
                    String formFieldValue = pdf_fieldName_fieldValue_map.get(formFieldName);

                    if (pdfield instanceof PDTextField) {
                        // if the PDF Form field is a Text Field
                        ((PDTextField) pdfield).setDefaultAppearance(defaultAppearanceString);
                        pdfield.setValue(formFieldValue);
                        pdfield.setReadOnly(true);

                    } else if (pdfield instanceof PDRadioButton) {
                        PDRadioButton pdRadioButton = (PDRadioButton)pdfield;
                        // if the PDF Form field is a Radio Button
                        Set<String> allowedValues = pdRadioButton.getOnValues();
                        if (allowedValues != null && !allowedValues.isEmpty() && allowedValues.contains(formFieldValue)) {
                            pdfield.setValue(formFieldValue);
                        } else {
                            System.out.println("PDF Form Field with name '" + formFieldName + "' received value as '" + formFieldValue
                                    + "'. However allowed values for this field are " + allowedValues);
                        }
                        pdfield.setReadOnly(true);

                    } else if (pdfield instanceof PDCheckBox) {
                        // if the PDF Form field is a Checkbox
                        PDCheckBox pdCheckBox = (PDCheckBox)pdfield;
                        if (CHECKBOX_VALUE_ON.equalsIgnoreCase(formFieldValue)) {
                            pdCheckBox.check();
                        } else if (CHECKBOX_VALUE_OFF.equalsIgnoreCase(formFieldValue)) {
                            pdCheckBox.unCheck();
                        }
                        
                        pdCheckBox.setReadOnly(true);
                    }

                }

                // extract to output file byte[]
                out = new ByteArrayOutputStream();
                pdDoc.save(out);
                mailMergedPDFContent = out.toByteArray();
            }
        } catch (Exception e) {
            e.printStackTrace();
        } finally {
            // clear resources
            try {
                if (pdDoc != null) pdDoc.close();
                if (out != null) out.close();
            } catch (IOException e) {
                e.printStackTrace();
            }
        }

        return mailMergedPDFContent;
    }
}


Input to above utility class is

    byte[] template = Files.readAllBytes(Paths.get("SomePDFTemplate.pdf"));

    // set field name value map
    Map<String, String> map = new HashMap<>();
    map.put("Signing_Place", "İstanbul, Poznań, Łodź"); // these wierd characters not rendering properly when code runs on Lunix server
    map.put("Participant_Name","Test Präjakta");
    map.put("Radio_Button_Group","RB_Item_3"); // valid values are: [RB_Item_1, RB_Item_2, RB_Item_3]
    map.put("CB_Item_1","OFF");
    map.put("CB_Item_2","ON");

    // create document
    byte[] pdfDoc = PDFMailMergeUtil.createPdfDocumentFromPdfTemplate(template, map);

    // save document as .pdf
    try (FileOutputStream fos = new FileOutputStream("C:\data\Projects\demo\src\test\resources\Test_PROD.pdf")) {
        fos.write(pdfDoc);
    }

Answer 1

改变这个

PDType0Font font = PDType0Font.load(pdDoc, PDFMailMergeUtil.class.getResourceAsStream("/Arial_Narrow.ttf"));

至此

PDType0Font font = PDType0Font.load(pdDoc, PDFMailMergeUtil.class.getResourceAsStream("/Arial_Narrow.ttf"), false);

避免子集化。 IIRC 这是因为子集字体中的字体文件在您使用它时并不真正存在，因为您使用的对象是不同的 PDFont 对象。

具有特殊字符的 PDFBox 在 Windows 上工作正常，但字符在 Linux 上被其他字符替换

PDFBox with special characters working fine on Windows but characters getting replaced with other characters in Linux

java

unix

pdf-generation

pdfbox

acrofields