Blazor服务端复用数据访问接口实现文件导入数据库的问题

Hey there! Let's sort out that null reference exception you're hitting with IPermitData in your TextFileParser class. The core issue is that you're declaring IPermitData _db; but never assigning a valid instance to it—luckily, since you're already using dependency injection (DI) in your Blazor Server app, we can reuse that system to get the instance we need.

Step 1: Update TextFileParser to Use Constructor Injection #

Instead of declaring _db inside the method, we'll inject it directly through the class constructor. This lets the DI container automatically pass in the registered IPermitData instance when creating a TextFileParser. We'll also fix the async handling since InsertPermit is an asynchronous method.

Here's the revised TextFileParser code:

public class TextFileParser
{
    private readonly IPermitData _db;

    // Inject IPermitData via constructor
    public TextFileParser(IPermitData db)
    {
        _db = db;
    }

    // Change to async Task instead of void for proper async handling
    public async Task InsertTextFileData(string filePath)
    {
        string errMessage = string.Empty;
        FileInfo file = new FileInfo(filePath);
        
        if (!File.Exists(file.FullName))
        {
            errMessage = "File not found!";
            // Consider logging this error instead of just storing it
            return;
        }
        if (file.Length == 0) // Fixed your condition here (>=0 will always be true)
        {
            errMessage = "File has no data!";
            return;
        }

        string delimiter = "\t";
        using StreamReader fileReader = new StreamReader(filePath);
        int header = 1;

        while (!fileReader.EndOfStream)
        {
            string vText = fileReader.ReadLine();
            if (string.IsNullOrEmpty(vText)) continue; // Skip empty lines

            if (header == 1)
            {
                header++;
                // Optional: Validate header columns here if needed
                continue;
            }

            string[] vString = vText.Split(delimiter, StringSplitOptions.None);
            // Create a new PermitModel for each record to avoid reference issues
            PermitModel permitRecord = new PermitModel();
            
            // Map your columns (same as before)
            permitRecord.PD_ParcelID = vString[0];
            permitRecord.PD_Situs1 = vString[1];
            permitRecord.PD_Owner = vString[2];
            permitRecord.PD_Addr1 = vString[3];
            permitRecord.PD_Addr2 = vString[4];
            permitRecord.ADDR_3 = vString[5];
            permitRecord.PD_City = vString[6];
            permitRecord.PD_State = vString[7];
            permitRecord.PD_Zip = vString[8];
            permitRecord.Type_Construction = vString[9];
            permitRecord.Estimated_Cost = vString[10];
            permitRecord.Permit_Issue_Date = vString[11];
            permitRecord.Permit_Type = vString[12];
            permitRecord.Property_Type = vString[13];
            permitRecord.Permit_NO = vString[14];
            permitRecord.Completion_Date = vString[15];
            permitRecord.Percent_Complete = vString[16];
            permitRecord.Modified_Date = vString[17];
            permitRecord.Note = vString[18];
            permitRecord.Note_Date = vString[19];
            permitRecord.Submission_Date = vString[20];
            permitRecord.Submitter = vString[21];
            permitRecord.Submitter_Phone = vString[22];
            permitRecord.Submitter_Email = vString[23];
            permitRecord.Land_AV = vString[24];
            permitRecord.Impr_AV = vString[25];
            permitRecord.Sec_Location = vString[26];

            try
            {
                await _db.InsertPermit(permitRecord); // Use await for async method
            }
            catch (Exception ex)
            {
                // Log the exception instead of just continuing—helps with debugging
                // Example: Logger.LogError(ex, "Failed to insert permit record {ParcelId}", permitRecord.PD_ParcelID);
                continue;
            }
        }
    }
}

Key fixes here:

• Constructor injection for IPermitData ensures _db is always a valid instance
• Changed InsertTextFileData to async Task to properly handle the asynchronous InsertPermit method
• Fixed the file length check (original >=0 would never trigger the "no data" message)
• Created a new PermitModel per record to avoid accidental overwriting of data
• Added a check for empty lines to prevent index errors

Step 2: Register TextFileParser with the DI Container #

Now that TextFileParser has a dependency on IPermitData, you need to register it in your Startup.cs (or Program.cs if using .NET 6+) so the DI container can resolve its dependencies.

Add this line to your ConfigureServices method:

// Use the same lifetime as your IPermitData registration (likely Scoped)
services.AddScoped<TextFileParser>();

Step 3: Inject and Use TextFileParser in Your Razor Page #

Instead of injecting IPermitData directly if you're using the parser, inject TextFileParser and call its method:

@inject TextFileParser textFileParser

<!-- Your UI here, e.g., a file input or button to trigger processing -->
<button @onclick="ProcessSampleFile">Process File</button>

@code {
    private async Task ProcessSampleFile()
    {
        string filePath = "path/to/your/file.txt"; // Replace with actual path
        await textFileParser.InsertTextFileData(filePath);
        // Add success feedback here if needed
    }
}

Extra Notes #

• Async Best Practices: Avoid using void for async methods (except event handlers) because it can lead to unobserved exceptions. Using async Task ensures exceptions can be caught properly.
• Error Logging: Add a logger (inject ILogger<TextFileParser> via constructor) to log failed inserts—this will save you a lot of time debugging issues with bad data.
• File Path Handling: In Blazor Server, be careful with file paths—ensure the app has permission to read the file, and consider using a file upload component to let users select files instead of hardcoding paths.

内容的提问来源于stack exchange，提问作者Jason Garner