package mobvista.dmp.datasource.gender;
import com.google.common.collect.Sets;
import org.apache.commons.lang.StringUtils;
import org.apache.hadoop.conf.Configuration;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.io.LongWritable;
import org.apache.hadoop.io.Text;
import org.apache.hadoop.io.compress.GzipCodec;
import org.apache.hadoop.mapreduce.Job;
import org.apache.hadoop.mapreduce.Mapper;
import org.apache.hadoop.mapreduce.Reducer;
import org.apache.hadoop.mapreduce.lib.input.FileInputFormat;
import org.apache.hadoop.mapreduce.lib.output.FileOutputFormat;
import org.apache.hadoop.util.GenericOptionsParser;
import org.codehaus.jackson.map.ObjectMapper;
import org.codehaus.jackson.type.JavaType;
import java.io.IOException;
import java.text.ParseException;
import java.text.SimpleDateFormat;
import java.util.*;
public class ExtractDeviceMRV2 {
/**
* author:liushuai
*
* @throws InterruptedException
* @throws IOException
* @throws ClassNotFoundException desc : 对dm_install_list每天的数据根据device_id进行去重并合并package_name
*/
public static void main(String[] args) throws ClassNotFoundException, IOException, InterruptedException {
Configuration conf = new Configuration();
conf.set("mapreduce.map.speculative", "true");
conf.set("mapreduce.reduce.speculative", "true");
conf.set("mapreduce.task.io.sort.mb", "500");
conf.set("mapreduce.reduce.java.opts", "-Xmx1536m");
conf.set("mapreduce.reduce.memory.mb", "2048");
conf.set("mapreduce.reduce.shuffle.parallelcopies", "50");
String[] otherArgs = new GenericOptionsParser(conf, args).getRemainingArgs();
Job job = Job.getInstance(conf, "ExtractDeviceMR");
FileOutputFormat.setCompressOutput(job, true);
FileOutputFormat.setOutputCompressorClass(job, GzipCodec.class);
job.setJarByClass(ExtractDeviceMRV2.class);
job.setMapperClass(ExtractMapper.class);
job.setReducerClass(ExtractReducer.class);
job.setOutputKeyClass(Text.class);
job.setOutputValueClass(Text.class);
FileInputFormat.addInputPath(job, new Path(otherArgs[0]));
FileOutputFormat.setOutputPath(job, new Path(otherArgs[1]));
System.exit(job.waitForCompletion(true) ? 0 : 1);
}
public static class ExtractMapper extends Mapper<LongWritable, Text, Text, Text> {
ObjectMapper objectMapper = new ObjectMapper();
JavaType javaType = objectMapper.getTypeFactory()
.constructCollectionType(ArrayList.class, ExtractDeviceMould.class);
Text outKey = new Text();
Text outValue = new Text();
String time = "";
public void setup(Context context) {
try {
time = context.getConfiguration().get("time");
} catch (Exception e) {
System.exit(3);
}
}
public void map(LongWritable key, Text value, Context context) {
try {
String line = value.toString();
String[] fields = MRUtils.SPLITTER.split(line, -1);
if (fields.length != 4) {
return;
}
if (fields[3].equals("")) { //package
return;
}
String deviceId = fields[0];
String[] tmpDeviceId = deviceId.split("-", -1);
StringBuilder tmpDeviceId2 = new StringBuilder();
for (String add : tmpDeviceId) {
tmpDeviceId2.append(add);
}
if (tmpDeviceId.length == 1) {
return;
}
if (tmpDeviceId2.toString().matches("^0*$")) {
return;
}
if (deviceId.equals("")) {
return;
}
StringBuilder outputValue = new StringBuilder();
//使用jackson解析json数组
List<ExtractDeviceMould> packageList = objectMapper.readValue(fields[3], javaType);//[{"date":"2017-03-30","package_name":"com.myntra.android"}]
if (packageList.size() == 0) {//说明没有package_name,扔掉
return;
}
//在90天之内有过安装行为的数据保留，否则不保留
SimpleDateFormat sdf = new SimpleDateFormat("yyyy-MM-dd");
String day = time;
Date today = null;
try {
today = sdf.parse(day);
} catch (ParseException e) {
e.printStackTrace();
}
Calendar theCa = Calendar.getInstance();
theCa.setTime(today);
theCa.add(theCa.DATE, -90);//90天前
Date begin = theCa.getTime();//90天前
int num = 0;
for (ExtractDeviceMould deviceMould : packageList) {
if (deviceMould.getPackage_name().length() == 0 || deviceMould.getDate().length() == 0) {
continue;
}
try {
Date install = sdf.parse(deviceMould.getDate());//pkg安装的时间
if (install.getTime() >= begin.getTime()) {
num = 1;
break;
}
} catch (ParseException e) {
continue;
}
}
if (num == 0) {
return;
}
String updateDate = "";
SimpleDateFormat format = new SimpleDateFormat("yyyy-MM-dd");
for (ExtractDeviceMould deviceMould : packageList) {
if (!StringUtils.isNotBlank(updateDate) && StringUtils.isNotBlank(deviceMould.getDate())) {
updateDate = deviceMould.getDate();
} else {
try {
// 获取最近的活跃日期
if (format.parse(deviceMould.getDate()).after(format.parse(updateDate))) {
updateDate = deviceMould.getDate();
}
} catch (ParseException e) {
}
}
outputValue.append("#" + deviceMould.getPackage_name());//循环把package_name加到后面
}
if (outputValue.toString().equals("") || outputValue.toString().length() <= 1) {
return;
}
outKey.set(MRUtils.JOINER.join(fields[0], fields[1]));//device_id, device_type
outValue.set(mobvista.dmp.util.MRUtils.JOINER.join(outputValue.toString().substring(1), updateDate));
// pkg_name, updateDate
context.write(outKey, outValue);
} catch (Exception e) {
return;
}
}
}
public static class ExtractReducer extends Reducer<Text, Text, Text, Text> {
Text outKey = new Text();
Text outValue = new Text();
public void reduce(Text key, Iterable<Text> values, Context context) {
try {
String deviceId = MRUtils.SPLITTER.split(key.toString(), -1)[0];
String deviceType = MRUtils.SPLITTER.split(key.toString(), -1)[1];
StringBuilder packageName = new StringBuilder("");
Set<String> set = Sets.newHashSet();
String updateDate = "";
SimpleDateFormat format = new SimpleDateFormat("yyyy-MM-dd");
for (Text value : values) {
String[] arr = MRUtils.SPLITTER.split(value.toString(), -1);
String[] array = arr[0].split("#", -1);//以#分隔的pkg
for (String pkgName : array) {
if (pkgName.equals("") || pkgName.length() < 1) {
continue;
}
set.add(pkgName);
}
// 设备的最新活跃时间
if (!StringUtils.isNotBlank(updateDate) && StringUtils.isNotBlank(arr[1])) {
updateDate = arr[1];
} else {
try {
if (format.parse(arr[1]).after(format.parse(updateDate))) {
updateDate = arr[1];
}
} catch (ParseException e) {
}
}
}
if (set.size() < 2) {
return;
}
for (String pkg : set) {
if (packageName.toString().equals("")) {
packageName.append(pkg);
} else {
packageName.append("#" + pkg);
}
}
outKey.set(deviceId);
outValue.set(MRUtils.JOINER.join("B", packageName.toString(), deviceType, updateDate));
context.write(outKey, outValue);
} catch (Exception e) {
return;
}
}
}
}