Firebase 分页的工作原理（续）

Question

我上周五从 分页持续问题提出了一个问题，因此 Firebase 文档有点令人困惑，所以我改进了我的代码，到 Firebase 分页文档。

这是我更新的代码：

async myForumQuestions(id: string, page = 1, pageSize =10): Promise<object> {
    const refDocs = admin.firestore().collection("forum").orderBy("timeStamp").where("createdby", "==", id);
    let lastDoc;

    // If it's not the first page, get the last document from the previous page
    if (page > 1) {
        const skip = (page - 1) * pageSize - 1;
        lastDoc = await refDocs.limit(skip).get().then((querySnapshot) => querySnapshot.docs[querySnapshot.docs.length - 1]);
    }

    const totalDocs = await refDocs.get().then((querySnapshot) => querySnapshot.size);
    const totalPages = Math.ceil(totalDocs / pageSize);

    // If there are no documents, return an empty list and set page and total_pages to 0
    if (totalDocs === 0) {
        return {
            data: [],
            pagination: {
                page: 0,
                total_pages: 0,
            },
        };
    }

    // Fetch the data with pagination
    let myQuestionData;
    if (lastDoc) {
        myQuestionData = await refDocs.startAfter(lastDoc).limit(pageSize).get().then((querySnapshot) => querySnapshot.docs.map((doc) => doc.data() as ForumInterface));
    } else {
        myQuestionData = await refDocs.limit(pageSize).get().then((querySnapshot) => querySnapshot.docs.map((doc) => doc.data() as ForumInterface));
    }

    await Promise.all(
        myQuestionData.map(async (question) => {
          if (question.quoteId != undefined) {
            const quotedSnapshot = await admin.firestore().collection("forum").doc(question.quoteId).get();
            if (quotedSnapshot.exists) {
              question.quoteData = quotedSnapshot.data() as ForumInterface;
            } else {
              question.quoteData = undefined;
            }
          }
        })
    );

    return {
        data: myQuestionData,
        pagination: {
            page: page,
            total_pages: totalPages,
        },
    };
}

Answer 1

所以根据上面的代码，他们 Firebase 将根据

skipped

文档收费，这是我试图避免的..

所以基于光标基本分页，我想我可以使用文档的ID并按

timestamp

排序。

这是我更新的代码..

 async allQuestions(cursor: string | null, pageSize =10): Promise<object> {
    const refDocs = admin.firestore().collection("forum").orderBy("timeStamp");
    let forumData;

    // If there's a cursor, start after the cursor
    if (cursor) {
      forumData = await refDocs.startAfter(cursor).limit(pageSize).get().then((querySnapshot) => querySnapshot.docs.map((doc) => doc.data() as ForumInterface));
    } else {
      forumData = await refDocs.limit(pageSize).get().then((querySnapshot) => querySnapshot.docs.map((doc) => doc.data() as ForumInterface));
    }

    if (forumData.length === 0) {
      return {
        data: forumData,
        cursor: null,
      };
    }

    await Promise.all(
        forumData.map(async (question) => {
          if (question.quoteId != undefined) {
            const quotedSnapshot = await admin.firestore().collection("forum").doc(question.quoteId).get();
            if (quotedSnapshot.exists) {
              question.quoteData = quotedSnapshot.data() as ForumInterface;
            } else {
              question.quoteData = undefined;
            }
          }
        })
    );

    // Get the ID of the last document to use as the next cursor
    let nextCursor = null;
    if (forumData.length === pageSize) {
      nextCursor = forumData[forumData.length - 1].id;
    }

    return {
      data: forumData,
      cursor: nextCursor,
    };
  }