如何使用pdfjsLib检测PDF文件是否包含图片？

用pdfjsLib处理PDF文件，已成功提取文本，但卡了3天没找到检测PDF是否包含图片的方法，现有代码如下（stackItem为PDF文件URL），求解决思路：

const loadingTask = pdfjsLib.getDocument(stackItem);
loadingTask.promise
  .then(function (doc: { numPages: any; getMetadata: () => Promise<any>; getPage: (arg0: any) => Promise<any>; }) {
    const numPages = doc.numPages;
    console.log("# Document Loaded");
    console.log("Number of Pages: " + numPages);
    console.log();

    let lastPromise;
    lastPromise = doc.getMetadata().then(function (data: { info: any; metadata: { getAll: () => any; }; }) {
      console.log("# Metadata Is Loaded");
      console.log("## Info");
      console.log(JSON.stringify(data.info, null, 2));
      if (data.metadata) {
        console.log("## Metadata");
        console.log(JSON.stringify(data.metadata.getAll()),);
      }
    });

    const loadPage = function (pageNum: string) {
      return doc.getPage(pageNum).then(function (page: { getViewport: (arg0: { scale: number; }) => any; getTextContent: () => Promise<any>; cleanup: () => void; }) {
        console.log("# Page " + pageNum);
        const viewport = page.getViewport({ scale: 1.0 });
        console.log("Size: " + viewport.width + "x" + viewport.height);
        console.log();
        return page
          .getTextContent()
          .then(function (content: { items: any[]; }) {
            const strings = content.items.map(function (item: { str: any; }) {
              return item.str;
            });
            console.log("## Text Content");
            console.log(strings.join(" "));
            page.cleanup();
          })
          .then(function () {
            console.log();
          });
      });
    };
    // Loading of the first page will wait on metadata and subsequent loadings
    // will wait on the previous pages.
    for (let i = 1; i <= numPages; i++) {
      lastPromise = lastPromise.then(loadPage.bind(null, i));
    }
    return lastPromise;
  })
  .then(
    function () {
      console.log("# End of Document");
    },
    function (err: string) {
      console.error("Error: " + err);
    }
  );

PDF中的图片通常通过两种形式存在：作为XObject（外部对象）通过Do操作符调用，或是直接内嵌的图片通过BI/EI操作符包裹。我们可以通过解析页面的操作指令来检测图片，修改后的代码如下：

const loadingTask = pdfjsLib.getDocument(stackItem);
loadingTask.promise
  .then(function (doc: { numPages: any; getMetadata: () => Promise<any>; getPage: (arg0: any) => Promise<any>; }) {
    const numPages = doc.numPages;
    console.log("# Document Loaded");
    console.log("Number of Pages: " + numPages);
    console.log();

    let lastPromise;
    lastPromise = doc.getMetadata().then(function (data: { info: any; metadata: { getAll: () => any; }; }) {
      console.log("# Metadata Is Loaded");
      console.log("## Info");
      console.log(JSON.stringify(data.info, null, 2));
      if (data.metadata) {
        console.log("## Metadata");
        console.log(JSON.stringify(data.metadata.getAll()),);
      }
    });

    const loadPage = function (pageNum: string) {
      return doc.getPage(pageNum).then(function (page: { 
        getViewport: (arg0: { scale: number; }) => any; 
        getTextContent: () => Promise<any>; 
        getOperatorList: () => Promise<any>;
        cleanup: () => void; 
        getXObject: (arg0: string) => Promise<any>;
      }) {
        console.log("# Page " + pageNum);
        const viewport = page.getViewport({ scale: 1.0 });
        console.log("Size: " + viewport.width + "x" + viewport.height);
        console.log();

        // 检测当前页面是否包含图片
        const checkForImages = page.getOperatorList().then(async (opList) => {
          let hasImage = false;
          // 遍历所有页面操作指令
          for (const op of opList.ops) {
            const operator = op[0];
            // 检测内嵌图片操作符
            if (operator === "BI" || operator === "EI") {
              hasImage = true;
              break;
            }
            // 检测XObject调用并验证是否为图片类型
            if (operator === "Do") {
              const xObjectName = op[1];
              const xObject = await page.getXObject(xObjectName);
              if (xObject.type === "XObjectImage") {
                hasImage = true;
                break;
              }
            }
          }
          console.log(`## Page ${pageNum} has image: ${hasImage}`);
          return hasImage;
        });

        // 并行执行文本提取与图片检测，提升效率
        return Promise.all([
          page.getTextContent().then(function (content: { items: any[]; }) {
            const strings = content.items.map(function (item: { str: any; }) {
              return item.str;
            });
            console.log("## Text Content");
            console.log(strings.join(" "));
          }),
          checkForImages
        ]).finally(() => {
          page.cleanup();
          console.log();
        });
      });
    };
    // 按顺序加载页面
    for (let i = 1; i <= numPages; i++) {
      lastPromise = lastPromise.then(loadPage.bind(null, i));
    }
    return lastPromise;
  })
  .then(
    function () {
      console.log("# End of Document");
    },
    function (err: string) {
      console.error("Error: " + err);
    }
  );

关键说明 #

• page.getOperatorList()：获取页面的所有渲染操作指令，这是检测图片的核心入口
• 检测BI/EI：对应直接内嵌在页面中的图片资源
• 检测Do操作符+验证XObject类型：调用的XObject如果是XObjectImage，即为图片资源
• 用Promise.all并行处理文本提取和图片检测，避免串行等待浪费时间
• 用finally确保页面资源被及时清理，防止内存泄漏

内容的提问来源于stack exchange，提问作者tetar