java - 使用 apache commons csv 按索引大小而不是 withFirstRecordAsHeader 读取 csv

Question

我正在尝试从 csv 中读取数据，如下所示:

id_number,last_name,first_name
123,Gibs,Walter
321,John,Chris
,,,,,,,,,,,,,,,,,
id_number,business_name
234,Abc Computers
417,Chris Inc

逗号行代表空白行(在 Excel 中)

为了清楚起见，下面是 csv 文件在 Excel 中的样子:

public class CsvFileParser {

    public static void customerMethod() {
        BufferedReader fileReader = null;
        CSVParser csvParser = null;

        try {
            fileReader = new BufferedReader(new FileReader("C:\\Users\\prassuha\\JavaProjects\\gs-maven-master\\complete\\src\\main\\java\\hello\\customer.csv"));
            csvParser = new CSVParser(fileReader,
                    CSVFormat.DEFAULT.withFirstRecordAsHeader().withIgnoreHeaderCase().withTrim());

            List<Customer> customers = new ArrayList<Customer>();

            Iterable<CSVRecord> csvRecords = csvParser.getRecords();

            for (CSVRecord csvRecord : csvRecords) {
                Customer customer = new Customer(
                        csvRecord.get("tracking_number"),
                        csvRecord.get("business_name"),
                );

                customers.add(customer);
            }

            for (Customer customer : customers) {
                System.out.println(customer);
            }

        } catch (Exception e) {
            System.out.println("Reading CSV Error!");
            e.printStackTrace();
        } finally {
            try {
                fileReader.close();
                csvParser.close();
            } catch (IOException e) {
                System.out.println("Closing fileReader/csvParser Error!");
                e.printStackTrace();
            }
        }
    }

    public static void personalMethod() {
        BufferedReader fileReader = null;
        CSVParser csvParser = null;

        try {
            fileReader = new BufferedReader(new FileReader("C:\\Users\\prassuha\\JavaProjects\\gs-maven-master\\complete\\src\\main\\java\\hello\\customer2.csv"));
            csvParser = new CSVParser(fileReader,
                    CSVFormat.DEFAULT.withFirstRecordAsHeader().withIgnoreHeaderCase().withTrim());

            List<Personal> personals = new ArrayList<Personal>();

            Iterable<CSVRecord> csvRecords = csvParser.getRecords();

            for (CSVRecord csvRecord : csvRecords) {
                Personal personal = new Personal(
                        csvRecord.get("tracking_number"),
                        csvRecord.get("full_name"),
                );
                personals.add(personal);
            }

            for (Personal personal : personals) {
                System.out.println(personal);
            }

        } catch (Exception e) {
            System.out.println("Reading CSV Error!");
            e.printStackTrace();
        } finally {
            try {
                fileReader.close();
                csvParser.close();
            } catch (IOException e) {
                System.out.println("Closing fileReader/csvParser Error!");
                e.printStackTrace();
            }
        }
    }

    public static void main(String[] args) {
        customerMethod();
        personalMethod();

    }
}

目前，我正在使用 withFirstRecordAsHeader() 但我想检查索引的大小，然后检查空行，然后再次检查索引的大小。我不确定这是否是正确的方法 - 如果有更好的方法来阅读本文，请分享。谢谢!

Answer 1

将整个文件加载到内存中，并将其拆分为“空白”行。然后分别解析每个部分。

下面的代码盲目地假设该文件有 2 个部分，第一个部分是人员，第二个部分是企业。您可能想检查这些假设。

CSVFormat csvFormat = CSVFormat.RFC4180
        .withFirstRecordAsHeader()
        .withIgnoreHeaderCase()
        .withTrim();

String fileName = "C:\\Users\\prassuha\\JavaProjects\\gs-maven-master\\complete\\src\\main\\java\\hello\\customer.csv";
String input = new String(Files.readAllBytes(Paths.get(fileName)), StandardCharsets.UTF_8);
String[] sections = input.split("(?m)^[,\\s]*\\R");

List<Person> persons = new ArrayList<>();
try (CSVParser csvParser = new CSVParser(new StringReader(sections[0]), csvFormat)) {
    for (CSVRecord csvRecord : csvParser) {
        String id_number = csvRecord.get("id_number");
        if (id_number.isEmpty())
            break; // end of person part of CSV data
        persons.add(new Person(
                id_number,
                csvRecord.get("last_name"),
                csvRecord.get("first_name")));
    }
}

List<Business> businesses = new ArrayList<>();
try (CSVParser csvParser = new CSVParser(new StringReader(sections[1]), csvFormat)) {
    for (CSVRecord csvRecord : csvParser) {
        String id_number = csvRecord.get("id_number");
        if (id_number.isEmpty())
            break; // end of business part of CSV data
        businesses.add(new Business(
                id_number,
                csvRecord.get("business_name")));
    }
}