如何计算 .csv 文件中的重复条目？

Question

我有一个格式如下的 .csv 文件：

ID,date,itemName
456,1-4-2020,Lemon
345,1-3-2020,Bacon
345,1-4-2020,Sausage
123,1-1-2020,Apple
123,1-2-2020,Pineapple
234,1-2-2020,Beer
345,1-4-2020,Cheese

我已经实现了遍历文件的算法，扫描第一个数字并按降序排序并生成新的输出：

123,1-1-2020,Apple
123,1-2-2020,Pineapple
234,1-2-2020,Beer
345,1-3-2020,Bacon
345,1-4-2020,Cheese
345,1-4-2020,Sausage
456,1-4-2020,Lemon

我的问题是，我如何实现我的算法来生成计算重复的第一个数字条目的输出并重新格式化它以使其看起来像这样......

123,1-1-2020,1,Apple
123,1-2-2020,1,Pineapple
234,1-2-2020,1,Beer
345,1-3-2020,1,Bacon
345,1-4-2020,2,Cheese,Sausage
456,1-4-2020,1,Lemon

...以便计算每个 ID 的出现次数，用次数表示，如果该 ID 的日期也相同，则将项目名称合并到同一行。下面是我的源代码（.csv 中的每一行都被制成一个名为“receipt”的对象，该对象具有 ID、日期和名称以及它们各自的 get() 方法）：

public class ReadFile {

    private static List<Receipt> readFile() {
        
        List<Receipt> receipts = new ArrayList<>();
        try {
            BufferedReader reader = new BufferedReader(new FileReader("dataset.csv"));

            // Move past the first title line
            reader.readLine();

            String line = reader.readLine();

            // Start reading from second line till EOF, split each string at ","
            while (line != null) {
                String[] attributes = line.split(",");
                Receipt attribute = getAttributes(attributes);
                receipts.add(attribute);
                line = reader.readLine();
            }
            reader.close();
        } catch (IOException e) {
            e.printStackTrace();
        }
        return receipts;
    }

    private static Receipt getAttributes(String[] attributes) {

        // Get ID located before the first ","
        long memberNumber = Long.parseLong(attributes[0]);

        // Get date located after the first ","
        String date = attributes[1];

        // Get name located after the second ","
        String name = attributes[2];

        return new Receipt(memberNumber, date, name);
    }

    // Parse the data into new file after sorting
    private static void parse(List<Receipt> receipts) {
        PrintWriter output = null;
        try {
            output = new PrintWriter("output.txt");
        } catch (FileNotFoundException e) {
            e.printStackTrace();
        }

        // For each receipts, assert the text output stream is not null, print line.
        for (Receipt p : receipts) {
            assert output != null;
            output.println(p.getMemberNumber() + "," + p.getDate() + "," + p.getName());
        }
        assert output != null;
        output.close();
    }

    // Main method, accept input file, sort and parse
    public static void main(String[] args) {

        List<Receipt> receipts = readFile();
        QuickSort q = new QuickSort();
        q.quickSort(receipts);
        parse(receipts);
    }
}

Answer 1

最简单的方法是使用地图。

文件中的数据示例。

String[] lines = {
"123,1-1-2020,Apple",
"123,1-2-2020,Pineapple",
"234,1-2-2020,Beer",
"345,1-3-2020,Bacon",
"345,1-4-2020,Cheese",
"345,1-4-2020,Sausage",
"456,1-4-2020,Lemon"};

• 创建地图
• 当您阅读这些行时，将它们拆分并使用compute 方法将它们添加到地图中。如果密钥（数字和日期）不存在，这将放入该行。否则，它只是将最后一项附加到现有条目。
• 文件不必排序，但值会在遇到时添加到末尾。

Map<String, String> map = new LinkedHashMap<>(); 
for (String line : lines) {
    String[] vals = line.split(",");

    // if v is null, add the line
    // if v exists, take the existing line and append the last value
    map.compute(vals[0]+vals[1], (k,v)->v == null ? line : v +","+vals[2]);
}

for (String line : map.values()) {
    String[] fields = line.split(",",3);
    int count = fields[2].split(",").length;
    System.out.printf("%s,%s,%s,%s%n", fields[0],fields[1],count,fields[2]);
}

对于这个示例运行打印

123,1-1-2020,1,Apple
123,1-2-2020,1,Pineapple
234,1-2-2020,1,Beer
345,1-3-2020,1,Bacon
345,1-4-2020,2,Cheese,Sausage
456,1-4-2020,1,Lemon