Fix Firestore replication ignoring initialCheckpoint

[calvinturbo]

This PR contains:

• a bugfix for the Firestore replication ignoring the initialCheckpoint variable, causing a new replication to replicate all documents

Describe the problem you have without this PR

• Firestore replication useless if the collection you wanna replicate has many documents

Todos

• adding documentation for the initialCheckpoint variable to https://rxdb.info/replication-firestore.html

Example how to 'build' the initialCheckpoint variable, something you could put in the docs:

 // 1. Query Firestore for the "boundary" document.
    let initialCheckpoint: any = null;
    try {
      const q = query(
          collection(this.firestore, 'collection'), // Your Firestore collection
          where('serverTimestamp', '>=', twoMonthsAgo), // Filter: serverTimestamp>= 2 months ago
          orderBy('serverTimestamp', 'asc'), // Order by createdAt ascending
          limit(1) // Get only the first (oldest) matching document
      );
      const querySnapshot = await getDocs(q);

      if (!querySnapshot.empty) {
        // 2. Construct the initialCheckpoint.
        const docData = querySnapshot.docs[0].data();
        initialCheckpoint = {
          id: querySnapshot.docs[0].id,  // Document ID
          serverTimestamp: docData['serverTimestamp'].toDate().toISOString(), // Server timestamp
        };
        console.log("Calculated initialCheckpoint:", initialCheckpoint);
      } else {
        // No documents within the last two months.  Handle as needed.
        // Option 1: Start with no checkpoint (sync all)
        // initialCheckpoint = null;
        // Option 2:  Don't start replication, show a message, etc.
        console.log("No documents found within the last two months.");
        return; // Or handle appropriately
      }


    } catch (error) {
      console.error("Error querying Firestore:", error);
      return; // Handle the error appropriately
    }

[pubkey]

Thank you for the fix. Can you add a test for this?