spark SQL读取不带表头的txt文件时，如果不传入schema信息，则会自动给列命名_c0、_c1等。而且也无法通过调用df.as()方法转换成dataset对象（甚至因为样例类的属性名称与df的列名不一致而抛出异常）。

这时候可以通过下面的方式添加schema

java 复制代码

// 定义schema
List<StructField> fields = new ArrayList<>();
fields.add(DataTypes.createStructField("word", DataTypes.StringType, true));
StructType schema = DataTypes.createStructType(fields);

Dataset<Row> dataFrame = sparkSession.createDataFrame(RowRDD, schema);// rdd -> dataframe

但是如果已经是dataframe对象则无法更新schema。所以我们需要在加载文件的时候通过调用schema()方法传入构造好的StructType对象以创建dataframe。例如：

java 复制代码

// 定义schema
List<StructField> fields = new ArrayList<>();
fields.add(DataTypes.createStructField("word", DataTypes.StringType, true));
fields.add(DataTypes.createStructField("cnt", DataTypes.StringType, true));
StructType schema = DataTypes.createStructType(fields);
Dataset<Row> citydf = reader.format("text")
        .option("delimiter", "\t")
        .option("header", true)
        .schema(schema)
        .csv("D:\project\sparkDemo\inputs\city_info.txt");

那么这时候就有问题了，如果需要加载的文件很多，全都要手动创建列表逐个添加字段会非常麻烦。

那么可以封装StructType对象的实例化方法，传入目标字段名称以及数据类型。字段名称以及数据类型可以通过样例类获取。

`StructType`对象的实例化方法

java 复制代码

package src.main.utils;

import java.util.*;
import java.util.stream.Collectors;
import java.util.stream.Stream;

import org.apache.spark.sql.types.DataType;
import org.apache.spark.sql.types.DataTypes;
import org.apache.spark.sql.types.StructField;
import org.apache.spark.sql.types.StructType;

import javax.activation.UnsupportedDataTypeException;

public class SchemaMaker {

    private LinkedHashMap<String, String> schemaMap = new LinkedHashMap<>();
    final private List<String> valueTypes = Stream.of("string", "integer", "double", "long").collect(Collectors.toList());
    List<StructField> fields = new ArrayList<>();
    public SchemaMaker(){
        this.fields.clear();
    }

    public SchemaMaker(ArrayList<ArrayList<String>> dataList){
        this.fields.clear();
        for (ArrayList<String> data : dataList) {
            int size = data.size();
            if (size != 2){
                throw new RuntimeException("每个数据必须为2个参数，第一个为字段名，第二个为字段类型");
            }
            String fieldName = data.get(0);
            String fieldType = getLowCase(data.get(1));

            if (checkType(fieldType)){
                this.schemaMap.put(fieldName, fieldType);
            }else {
                throw new RuntimeException("数据类型不符合预期" + this.valueTypes.toString());
            }
        }
    }

    public void add(String fieldName, String fieldType){
        String fieldtype = getLowCase(fieldType);
        if (checkType(fieldtype)){
            this.schemaMap.put(fieldName, fieldtype);
        }else {
            throw new RuntimeException("数据类型不符合预期" + this.valueTypes.toString());
        }
    }

    private String getLowCase(String s){
        return s.toLowerCase();
    }

    private boolean checkType(String typeValue){
        return this.valueTypes.contains(typeValue);
    }

    private DataType getDataType (String typeValue) throws UnsupportedDataTypeException {
        if (typeValue.equals("string")){
            return DataTypes.StringType;
        } else if (typeValue.equals("integer")) {
            return DataTypes.IntegerType;
        } else if (typeValue.equals("long")) {
            return DataTypes.LongType;
        } else if (typeValue.equals("double")) {
            return DataTypes.DoubleType;
        }else {
            throw new UnsupportedDataTypeException(typeValue);
        }
    }

    public StructType getStructType() throws UnsupportedDataTypeException {
        for (Map.Entry<String, String> schemaValue : schemaMap.entrySet()) {
            String fieldName = schemaValue.getKey();
            String fieldType = schemaValue.getValue();
            DataType fieldDataType = getDataType(fieldType);
            this.fields.add(DataTypes.createStructField(fieldName, fieldDataType, true));
        }
        
        return DataTypes.createStructType(this.fields);
    }

}

封装一层，通过传入的Object.class().getDeclaredFields()方法获取的字段信息构造StructType

java 复制代码

public static StructType getStructType(Field[] fields) throws UnsupportedDataTypeException {

    ArrayList<ArrayList<String>> lists = new ArrayList<>();
    for (Field field : fields) {
        String name = field.getName();
        AnnotatedType annotatedType = field.getAnnotatedType();
        String[] typeSplit = annotatedType.getType().getTypeName().split("\.");
        String type = typeSplit[typeSplit.length - 1];
        ArrayList<String> tmpList = new ArrayList<String>();
        tmpList.add(name);
        tmpList.add(type);
        lists.add(tmpList);
    }

    SchemaMaker schemaMaker = new SchemaMaker(lists);

    return schemaMaker.getStructType();
}

样例类的定义

java 复制代码

public static class City implements Serializable{
    private Long cityid;
    private String cityname;
    private String area;

    public City(Long cityid, String cityname, String area) {
        this.cityid = cityid;
        this.cityname = cityname;
        this.area = area;
    }

    public Long getCityid() {
        return cityid;
    }

    public void setCityid(Long cityid) {
        this.cityid = cityid;
    }

    public String getCityname() {
        return cityname;
    }

    public void setCityname(String cityname) {
        this.cityname = cityname;
    }

    public String getArea() {
        return area;
    }

    public void setArea(String area) {
        this.area = area;
    }
}

public static class Product implements Serializable{
    private Long productid;
    private String product;
    private String product_from;

    public Long getProductid() {
        return productid;
    }

    public void setProductid(Long productid) {
        this.productid = productid;
    }

    public String getProduct() {
        return product;
    }

    public void setProduct(String product) {
        this.product = product;
    }

    public String getProduct_from() {
        return product_from;
    }

    public void setProduct_from(String product_from) {
        this.product_from = product_from;
    }

    public Product(Long productid, String product, String product_from) {
        this.productid = productid;
        this.product = product;
        this.product_from = product_from;
    }
}

public static class UserVisitAction implements Serializable{
    private String date;
    private Long user_id;
    private String session_id;
    private Long page_id;
    private String action_time;
    private String search_keyword;
    private Long click_category_id;
    private Long click_product_id;
    private String order_category_ids;
    private String order_product_ids;
    private String pay_category_ids;
    private String pay_product_ids;
    private Long city_id;

    public String getDate() {
        return date;
    }

    public void setDate(String date) {
        this.date = date;
    }

    public Long getUser_id() {
        return user_id;
    }

    public void setUser_id(Long user_id) {
        this.user_id = user_id;
    }

    public String getSession_id() {
        return session_id;
    }

    public void setSession_id(String session_id) {
        this.session_id = session_id;
    }

    public Long getPage_id() {
        return page_id;
    }

    public void setPage_id(Long page_id) {
        this.page_id = page_id;
    }

    public String getAction_time() {
        return action_time;
    }

    public void setAction_time(String action_time) {
        this.action_time = action_time;
    }

    public String getSearch_keyword() {
        return search_keyword;
    }

    public void setSearch_keyword(String search_keyword) {
        this.search_keyword = search_keyword;
    }

    public Long getClick_category_id() {
        return click_category_id;
    }

    public void setClick_category_id(Long click_category_id) {
        this.click_category_id = click_category_id;
    }

    public Long getClick_product_id() {
        return click_product_id;
    }

    public void setClick_product_id(Long click_product_id) {
        this.click_product_id = click_product_id;
    }

    public String getOrder_category_ids() {
        return order_category_ids;
    }

    public void setOrder_category_ids(String order_category_ids) {
        this.order_category_ids = order_category_ids;
    }

    public String getOrder_product_ids() {
        return order_product_ids;
    }

    public void setOrder_product_ids(String order_product_ids) {
        this.order_product_ids = order_product_ids;
    }

    public String getPay_category_ids() {
        return pay_category_ids;
    }

    public void setPay_category_ids(String pay_category_ids) {
        this.pay_category_ids = pay_category_ids;
    }

    public String getPay_product_ids() {
        return pay_product_ids;
    }

    public void setPay_product_ids(String pay_product_ids) {
        this.pay_product_ids = pay_product_ids;
    }

    public Long getCity_id() {
        return city_id;
    }

    public void setCity_id(Long city_id) {
        this.city_id = city_id;
    }

    public UserVisitAction(String date, Long user_id, String session_id, Long page_id, String action_time, String search_keyword, Long click_category_id, Long click_product_id, String order_category_ids, String order_product_ids, String pay_category_ids, String pay_product_ids, Long city_id) {
        this.date = date;
        this.user_id = user_id;
        this.session_id = session_id;
        this.page_id = page_id;
        this.action_time = action_time;
        this.search_keyword = search_keyword;
        this.click_category_id = click_category_id;
        this.click_product_id = click_product_id;
        this.order_category_ids = order_category_ids;
        this.order_product_ids = order_product_ids;
        this.pay_category_ids = pay_category_ids;
        this.pay_product_ids = pay_product_ids;
        this.city_id = city_id;
    }
}

主程序部分

java 复制代码

        DataFrameReader reader = sparkSession.read();

        StructType citySchema = getStructType(City.class.getDeclaredFields());
        StructType productSchema = getStructType(Product.class.getDeclaredFields());
        StructType actionSchema = getStructType(UserVisitAction.class.getDeclaredFields());

        Dataset<Row> citydf = reader.format("text")
                .option("delimiter", "\t")
                .option("header", true)
                .schema(citySchema)
                .csv("D:\project\sparkDemo\inputs\city_info.txt");
        Dataset<Row> productdf = reader.format("text")
                .option("delimiter", "\t")
                .option("header", true)
                .schema(productSchema)
                .csv("D:\project\sparkDemo\inputs\product_info.txt");
        Dataset<Row> actiondf = reader.format("text")
                .option("delimiter", "\t")
                .option("header", true)
                .schema(actionSchema)
                .csv("D:\project\sparkDemo\inputs\user_visit_action.txt");

        Dataset<City> cityDataset = citydf.as(Encoders.bean(City.class));  // 转换为ds对象
//        cityDataset.show();

        citydf.write().format("jdbc").option("url", "jdbc:mysql://172.20.143.219:3306/test")
                .option("driver", "com.mysql.cj.jdbc.Driver").option("user", "root")
                .option("password", "mysql").option("dbtable", "city_info").mode("overwrite").save();
        productdf.write().format("jdbc").option("url", "jdbc:mysql://172.20.143.219:3306/test")
                .option("driver", "com.mysql.cj.jdbc.Driver").option("user", "root")
                .option("password", "mysql").option("dbtable", "product_info").mode("overwrite").save();
        actiondf.write().format("jdbc").option("url", "jdbc:mysql://172.20.143.219:3306/test")
                .option("driver", "com.mysql.cj.jdbc.Driver").option("user", "root")
                .option("password", "mysql").option("dbtable", "user_visit_action").mode("overwrite").save();

通过这个方法自定义了样例类之后可以进行批量读取与处理txt文件了。

PS:在缺乏文件信息的时候不要贸然加载文件，否则可能会造成严重的后果。

SparkSQL[java]读取不带表头的txt文件

StructType对象的实例化方法

样例类的定义

主程序部分

`StructType`对象的实例化方法