如何在C#中无需数据库将DAT文件转PDF并避免重复记录

Got it, let's tackle this problem step by step. You want to cut out the database middleman, convert DAT files straight to PDF in C#, and eliminate those duplicate records that were messing up your Crystal Reports output. Here's how I'd approach it:

Before even thinking about PDFs, let's get rid of duplicate records right when loading the DAT data into memory. First, define a model that matches your DAT file's structure (adjust fields to fit your actual data):

// Match this to your DAT file's columns
public class DatRecord
{
    public int RecordId { get; set; }
    public string CustomerName { get; set; }
    public decimal TransactionAmount { get; set; }
    // Add other fields as needed
}

Then write a method to load all DAT files, parse their contents, and deduplicate using a unique identifier (or combination of fields that define a unique record):

using System.IO;
using System.Linq;
using System.Collections.Generic;

public List<DatRecord> LoadAndDeduplicateDats(string[] datFilePaths)
{
    var allRawRecords = new List<DatRecord>();

    foreach (var filePath in datFilePaths)
    {
        var lines = File.ReadAllLines(filePath);
        // Skip header line if your DAT files have one; remove Skip(1) if not
        foreach (var line in lines.Skip(1))
        {
            // Replace ',' with your actual delimiter (e.g., '\t' for tab-separated)
            var fields = line.Split(',');
            var record = new DatRecord
            {
                RecordId = int.Parse(fields[0]),
                CustomerName = fields[1].Trim(),
                TransactionAmount = decimal.Parse(fields[2])
                // Map remaining fields here
            };
            allRawRecords.Add(record);
        }
    }

    // Deduplicate: Group by your unique key (use a composite key like new { r.RecordId, r.CustomerName } if needed)
    var deduplicatedRecords = allRawRecords
        .GroupBy(r => r.RecordId)
        .Select(group => group.First()) // Keep only the first instance of each unique record
        .ToList();

    return deduplicatedRecords;
}

You don't need a database to feed data into Crystal Reports—or you can even ditch Crystal Reports entirely for a lighter solution. Here are two options:

Option A: Keep Using Crystal Reports (With In-Memory Data) #

Crystal Reports supports binding to in-memory DataTables or lists, so you can skip inserting data into a database entirely. Convert your deduplicated list to a DataTable and bind it to your report:

using System.Data;
using System.Reflection;
using CrystalDecisions.CrystalReports.Engine;
using CrystalDecisions.Shared;

// Helper to convert list to DataTable for Crystal Reports
public DataTable ConvertListToDataTable<T>(List<T> data)
{
    var props = typeof(T).GetProperties();
    var dt = new DataTable();
    
    // Add columns matching your model properties
    foreach (var prop in props)
    {
        dt.Columns.Add(prop.Name, Nullable.GetUnderlyingType(prop.PropertyType) ?? prop.PropertyType);
    }
    
    // Populate rows
    foreach (var item in data)
    {
        var row = dt.NewRow();
        foreach (var prop in props)
        {
            row[prop.Name] = prop.GetValue(item) ?? DBNull.Value;
        }
        dt.Rows.Add(row);
    }
    return dt;
}

// Export to PDF using Crystal Reports
public void ExportCrystalReportToPdf(List<DatRecord> deduplicatedRecords, string outputPdfPath)
{
    var dataTable = ConvertListToDataTable(deduplicatedRecords);
    var report = new YourCustomCrystalReport(); // Replace with your actual report class
    
    report.SetDataSource(dataTable);
    
    // Configure export settings
    var exportOpts = report.ExportOptions;
    exportOpts.ExportFormatType = ExportFormatType.PortableDocFormat;
    exportOpts.DestinationOptions = new DiskFileDestinationOptions { DiskFileName = outputPdfPath };
    exportOpts.ExportDestinationType = ExportDestinationType.DiskFile;
    
    report.Export();
}

Option B: Use a Lightweight PDF Library (No Crystal Reports Dependency) #

If you want to avoid Crystal Reports' runtime dependencies, use a library like iTextSharp (or iText 7 for .NET Core) to build PDFs directly from your in-memory data:

First, add the iTextSharp NuGet package, then use this code:

using iTextSharp.text;
using iTextSharp.text.pdf;
using System.IO;

public void GeneratePdfFromRecords(List<DatRecord> records, string outputPath)
{
    using (var fs = new FileStream(outputPath, FileMode.Create))
    {
        var document = new Document(PageSize.A4, 20, 20, 20, 20);
        PdfWriter.GetInstance(document, fs);
        document.Open();
        
        // Add a table to hold your data
        var pdfTable = new PdfPTable(3); // Match column count to your data fields
        pdfTable.WidthPercentage = 100;
        
        // Add header cells
        pdfTable.AddCell(new Phrase("Record ID"));
        pdfTable.AddCell(new Phrase("Customer Name"));
        pdfTable.AddCell(new Phrase("Amount"));
        
        // Add data rows
        foreach (var record in records)
        {
            pdfTable.AddCell(new Phrase(record.RecordId.ToString()));
            pdfTable.AddCell(new Phrase(record.CustomerName));
            pdfTable.AddCell(new Phrase(record.TransactionAmount.ToString("C")));
        }
        
        document.Add(pdfTable);
        document.Close();
    }
}

• Define a solid unique key: Make sure the field/combination you use for grouping (e.g., RecordId) truly identifies unique records—using the wrong key will either leave duplicates or delete valid data.
• Validate DAT data: Add checks during parsing (e.g., ensure field counts match expected values, validate numeric data) to avoid corrupted records breaking your workflow.
• Test with sample data: Run your deduplication logic against a small set of DAT files with known duplicates to confirm it works as expected before scaling.

内容的提问来源于stack exchange，提问作者Vinoothna Y