Apache POI - 检索 .doc 文件中关键字之间的文本内容并有条件地呈现它

Question

我想在 .doc 文件中找到两个关键字之间的文本内容，并有条件地呈现该文本内容或隐藏它。例如：

Lorem Ipsum is simply dummy text ${if condition} of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s ${endif}

当我使用 Apache - POI 解析文档时，我希望能够以某种方式在文档中发现这些 blockquotes [=12] 之间的每一个内容=] ${endif} 并在我要生成的下一个文档中有条件地渲染它或不渲染它。

所以经过我解析后的上述文字应该有以下两种不同的形式：

1)如果满足条件

Lorem Ipsum is simply dummy text of the printing and typesetting industry. Lorem Ipsum has been the industry's standard dummy text ever since the 1500s

或

2) 如果不满足条件

Lorem Ipsum is simply dummy text

我尝试通过使用 XWPFParagraph 对象然后使用 XWPFRun 来做到这一点，但这不是可靠的方法，因为 运行 可以在中间随机分割在不可预测的条件下的单词。

您能否提出任何可靠的方法来实现我的用例？提前致谢。

Answer 1

以此为例（代码已测试）：

class ParagraphModifier {

    private final Pattern pIf = Pattern.compile("\$\{if\s+(\w+)\}");
    private final Pattern pEIf = Pattern.compile("\$\{endif\}");
    private final Function<String, Boolean> processor;

    public ParagraphModifier(Function<String, Boolean> processor) {
        this.processor = processor;
    }

    // Process

    static class Pair<K, V> {
        public K key;
        public V value;
        public Pair(K key, V value) {
            this.key = key;
            this.value = value;
        }
    }

    // 
    public static void cloneRun(XWPFRun clone, XWPFRun source) {
        CTRPr rPr = clone.getCTR().isSetRPr() ? clone.getCTR().getRPr() : clone.getCTR().addNewRPr();
        rPr.set(source.getCTR().getRPr());
        clone.setText(source.getText(0));
    }

    // Split runs in paragraph at a specific text offset and returns the run index
    int splitAtTextPosition(XWPFParagraph paragraph, int position) {
        List<XWPFRun> runs = paragraph.getRuns();
        int offset = 0;

        for (int i = 0; i < runs.size(); i++) {
            XWPFRun run = runs.get(i);
            String text = run.getText(0);
            int length = text.length();

            if (position >= (offset + length)) {
                offset += length;
                continue;
            }

            // Do split
            XWPFRun run2 = paragraph.insertNewRun(i + 1);
            cloneRun(run2, run);
            run.setText(text.substring(0, position - offset), 0);
            run2.setText(text.substring(position - offset), 0);
            return i + 1;
        }
        return -1;
    }

    String getParagraphText(XWPFParagraph paragraph) {
        StringBuilder sb = new StringBuilder("");
        for (XWPFRun run : paragraph.getRuns()) sb.append(run.getText(0));
        return sb.toString();
    }

    void removeRunsRange(XWPFParagraph paragraph, int from, int to) {
        int runs = paragraph.getRuns().size();
        to = Math.min(to, runs);
        for (int i = (to - 1); i >= from; i--) {
            paragraph.removeRun(i);
        }
    }

    Pair<Integer, String> extractToken(Pattern pattern, XWPFParagraph paragraph) {
        String text = getParagraphText(paragraph);
        Matcher matcher = pattern.matcher(text);

        if (matcher.find()) {
            int rStart = splitAtTextPosition(paragraph, matcher.start());
            int rEnd = splitAtTextPosition(paragraph, matcher.end());
            removeRunsRange(paragraph, rStart, rEnd);
            return new Pair<>(rStart, matcher.group());
        } else {
            return new Pair<>(-1, "");
        }
    }

    void applyParagraph(XWPFParagraph paragraph) {
        int lastIf = -1;

        while (true) {
            var tIf = extractToken(pIf, paragraph);
            if (tIf.key == -1) {
                break;
            }
            if (tIf.key < lastIf) {
                throw new IllegalStateException("If conditions can not be nested");
            }

            var tEIf = extractToken(pEIf, paragraph);
            if (tEIf.key == -1) {
                throw new IllegalStateException("If condition missing endif");
            }

            var m = pIf.matcher(tIf.value);
            var keep = m.find() && processor.apply(m.group(1));
            if (!keep) {
                removeRunsRange(paragraph, tIf.key, tEIf.key);
            }

            lastIf = tEIf.key;
        }
    }

    void apply(Iterable<XWPFParagraph> paragraphs) {
        for (XWPFParagraph p : paragraphs) {
            applyParagraph(p);
        }
    }

}

用法：

class Main {

    private static XWPFDocument loadDoc(String name) throws IOException, InvalidFormatException {
        String path = Main.class.getClassLoader().getResource(name).getPath();
        FileInputStream fis = new FileInputStream( path);
        return new XWPFDocument(OPCPackage.open(fis));
    }

    private static void saveDoc(String path, XWPFDocument doc) throws IOException {
        try (var fos = new FileOutputStream(path)) {
            doc.write(fos);
        }
    }

    public static void main (String[] args) throws Exception {
        var xdoc = loadDoc("test.docx");

        var pm = new ParagraphModifier(str -> str.toLowerCase().equals("true"));
        pm.apply(xdoc.getParagraphs());

        saveDoc("test.out.docx", xdoc);
    }
    
}

此解决方案不支持 ${if } 块跨越段落，如果嵌套，也不支持 Table 结构。扩展解决方案以支持它们应该很简单。