Hadoop writing to a new file from mapper - java

I am trying to write a program that takes a huge data set and then run some queries on it using mapreduce. I have a code like this:
public static class MRMapper
extends Mapper<LongWritable, Text, Text, IntWritable>{
String output2="hdfs://master:9000/user/xxxx/indexln.txt";
FileSystem Phdfs =FileSystem.get(new Configuration());
Path fname1=new Path(output2);
BufferedWriter out=new BufferedWriter(new OutputStreamWriter(Phdfs.create(fname1,true)));
private final static IntWritable one = new IntWritable(1);
private Text word = new Text();
long max=0;
public void map(LongWritable key, Text value, Context context
) throws IOException, InterruptedException {
String binln = Long.toBinaryString(0x8000000000000000L | key).substring(1);
out2.write(binln+"\n");
out2.flush();
String line = value.toString();
String [] ST = line.split(",");
long val=Math.abs(Long.parseLong(ST[2]));
if (max < val){
max= val;
}
else{
word.set(line);
context.write(word, val);
}
}
}
What I am trying to do is to build an indexfile in the mapper. which would be used to access specific areas of the input file by the mappers. The mappers read a part of the input file based on the index and then print the part read and the number of lines read to the output. I am using one mapper with 9 reducers.
My question goes, is it possible to create/write to a file different from the output file in the map function and also, can a reducer read a file that is open in the mapper? If yes, Am i on the right path or totally wrong or maybe mapreduce is not the way for this? I apologize if this question sounds too noob but I'm actually a noob in hadoop. Trying to learn. thanks

Are you sure your are using a single mapper? Because Hadoop creates a number of mappers very close to the number of input splits (more details).
The concept of input split is very important as well: it means very big data files are splited into several chuncks, each chunck assigned to a mapper. Thus, unless you are totally sure only one mapper is being used, you wont be able to control which part of the file you are workin on, and you will not be able to control any kind of global index.
Being said that, by using a single mapper in MapReduce is the same than not using MapReduce at all :) Maybe the mistake is mine, and I'm assuming you have only one file to be analyzed, is that the case?
In the case you have several big data files the scenario changes, and it could make sense to create a single mapper for each file, but you will have to create your own InputSplit and override the isSplitable method by returning always false.

Related

Parsing a text file using java with multiple values per line to be extracted

I'm not going to lie I'm really bad at making regular expressions. I'm currently trying to parse a text file that is giving me a lot of issues. The goal is to extract the data between their respective "tags/titles". The file in question is a .qbo file laid out as follows personal information replaced with "DATA": The parts that I care about retrieving are between the "STMTTRM" and "/STMTTRM" tags as the rest I don't plan on putting in my database, but I figured it would help others see the file content I'm working with. I apologize for any confusion prior to this update.
FXHEADER:100
DATA:OFXSGML
VERSION:102
SECURITY:NONE
ENCODING:USASCII
CHARSET:1252
COMPRESSION:NONE
OLDFILEUID:NONE
NEWFILEUID:NONE
<OFX>
<SIGNONMSGSRSV1><SONRS>
<STATUS><CODE>0</CODE><SEVERITY>INFO</SEVERITY></STATUS>
<DTSERVER>20190917133617.000[-4:EDT]</DTSERVER>
<LANGUAGE>ENG</LANGUAGE>
<FI>
<ORG>DATA</ORG>
<FID>DATA</FID>
</FI>
<INTU.BID>DATA</INTU.BID>
<INTU.USERID>DATA</INTU.USERID>
</SONRS></SIGNONMSGSRSV1>
<BANKMSGSRSV1>
<STMTTRNRS>
<TRNUID>0</TRNUID>
<STATUS><CODE>0</CODE><SEVERITY>INFO</SEVERITY></STATUS>
<STMTRS>
<CURDEF>USD</CURDEF>
<BANKACCTFROM>
<BANKID>DATA</BANKID>
<ACCTID>DATA</ACCTID>
<ACCTTYPE>CHECKING</ACCTTYPE>
<NICKNAME>FREEDOM CHECKING</NICKNAME>
</BANKACCTFROM>
<BANKTRANLIST>
<DTSTART>20190717</DTSTART><DTEND>20190917</DTEND>
<STMTTRN><TRNTYPE>POS</TRNTYPE><DTPOSTED>20190717071500</DTPOSTED><TRNAMT>-5.81</TRNAMT><FITID>3893120190717WO</FITID><NAME>DATA</NAME><MEMO>POS Withdrawal</MEMO></STMTTRN>
<STMTTRN><TRNTYPE>DIRECTDEBIT</TRNTYPE><DTPOSTED>20190717085000</DTPOSTED><TRNAMT>-728.11</TRNAMT><FITID>4649920190717WE</FITID><NAME>CHASE CREDIT CRD</NAME><MEMO>DATA</MEMO></STMTTRN>
<STMTTRN><TRNTYPE>ATM</TRNTYPE><DTPOSTED>20190717160900</DTPOSTED><TRNAMT>-201.99</TRNAMT><FITID>6674020190717WA</FITID><NAME>DATA</NAME><MEMO>ATM Withdrawal</MEMO></STMTTRN>
</BANKTRANLIST>
<LEDGERBAL><BALAMT>2024.16</BALAMT><DTASOF>20190917133617.000[-4:EDT]</DTASOF></LEDGERBAL>
<AVAILBAL><BALAMT>2020.66</BALAMT><DTASOF>20190917133617.000[-4:EDT]</DTASOF></AVAILBAL>
</STMTRS>
</STMTTRNRS>
</BANKMSGSRSV1>
</OFX>
I want to be able to end with data that looks or acts like the following so that each row of data can easily be added to a database:
Example Parse
As David has already answered, It is good to parse the POS output XML using Java. If you are more interested about about regex to get all the information, you can use this regular expression.
<[^>]+>|\\n+
You can test in the following sites.
https://rubular.com/
https://www.regextester.com/
Given this is XML, I would do one of two things:
either use the Java DOM objects to marshall/unmarshall to/from Java objects (nodes and elements), or
use JAXB to achieve something similar but with better POJO representation.
Mkyong has tutorials for both. Try the dom parsing or jaxb. His tutorials are simple and easy to follow.
JAXB requires more work and dependencies. So try DOM first.
I would propose the following approach.
Read file line by line with Files:
final List<String> lines = Files.readAllLines(Paths.get("/path/to/file"));
At this point you would have all file line separated and ready to convert the string lines into something more useful. But you should create class beforehand.
Create a class for your data in line, something like:
public class STMTTRN {
private String TRNTYPE;
private String DTPOSTED;
...
...
//constructors
//getters and setters
}
Now when you have a data in each separate string and a class to hold the data, you can convert lines to objects with Jackson:
final XmlMapper xmlMapper = new XmlMapper();
final STMTTRN stmttrn = xmlMapper.readValue(lines[0], STMTTRN.class);
You may want to create a loop or make use of stream with a mapper and a collector to get the list of STMTTRN objects:
final List<STMTTRN> stmttrnData = lines.stream().map(this::mapLine).collect(Collectors.toList());
Where the mapper might be:
private STMTTRN mapLine(final String line) {
final XmlMapper xmlMapper = new XmlMapper();
try {
return xmlMapper.readValue(line, STMTTRN.class);
} catch (IOException e) {
throw new RuntimeException(e);
}
}

Using I/O stream to parse CSV file

I have a CSV file of US population data for every county in the US. I need to get each population from the 8th column of the file. I'm using a fileReader() and bufferedStream() and not sure how to use the split method to accomplish this. I know this isn't much information but I know that I'll be using my args[0] as the destination in my class.
I'm at a loss to where to being to be honest.
import java.io.FileReader;
public class Main {
public static void main(String[] args) {
BufferedReader() buff = new BufferedReader(new FileReader(args[0]));
String
}
try {
}
}
The output should be an integer of the total US population. Any help with pointing me in the right direction would be great.
Don't reinvent the wheel, don't parse CSV yourself: use a library. Even such a simple format as CSV has nuances: fields can be escaped with quotes or unescaped, the file can have or have not a header and so on. Besides that you have to test and maintain the code you've wrote. So writing less code and reusing libraries is good.
There are a plenty of libraries for CSV in Java:
Apache Commons CSV
OpenCSV
Super CSV
Univocity
flatpack
IMHO, the first two are the most popular.
Here is an example for Apache Commons CSV:
final Reader in = new FileReader("counties.csv");
final Iterable<CSVRecord> records = CSVFormat.DEFAULT.parse(in);
for (final CSVRecord record : records) { // Simply iterate over the records via foreach loop. All the parsing is handler for you
String populationString = record.get(7); // Indexes are zero-based
String populationString = record.get("population"); // Or, if your file has headers, you can just use them
… // Do whatever you want with the population
}
Look how easy it is! And it will be similar with other parsers.

Flink: How to pass a dynamic path while writing to files using writeFileAsText(path)?

Let's say I have a Stream with elements of type String. I want to write each element in the stream to a separate file in some folder. I'm using the following set up.
stream.writeAsText(path).setParallelism(1);
How do I make this path dynamic? I even tried adding System.nanotime() to the path to make it dynamic. But it still doesn't seem to work, everything gets written to a single file.
This sort of use case is explicitly supported in Flink by the Rolling File Sink with a custom bucketer, or the newer and prefered Streaming File Sink with a custom BucketAssigner and RollingPolicy.
Your problem is that DataStream.writeAsText() writes the entire content of the stream to the file at once, so you will only ever get a single file.
It looks like this will return a collection that you can use to output your strings as different files.
dataStream.flatMap(new FlatMapFunction<String, String>() {
#Override
public void flatMap(String value, Collector<String> out)
throws Exception {
for(String word: value.split(" ")){
out.collect(word);
}
}
});
Taken straight from the documentation here: https://ci.apache.org/projects/flink/flink-docs-release-1.2/dev/datastream_api.html

Using StAX to create index for XML for quick access

Is there a way to use StAX and JAX-B to create an index and then get quick access to an XML file?
I have a large XML file and I need to find information in it. This is used in a desktop application and so it should work on systems with few RAM.
So my idea is this: Create an index and then quickly access data from the large file.
I can't just split the file because it's an official federal database that I want to use unaltered.
Using a XMLStreamReader I can quickly find some element and then use JAXB for unmarshalling the element.
final XMLStreamReader r = xf.createXMLStreamReader(filename, new FileInputStream(filename));
final JAXBContext ucontext = JAXBContext.newInstance(Foo.class);
final Unmarshaller unmarshaller = ucontext.createUnmarshaller();
r.nextTag();
while (r.hasNext()) {
final int eventType = r.next();
if (eventType == XMLStreamConstants.START_ELEMENT && r.getLocalName().equals("foo")
&& Long.parseLong(r.getAttributeValue(null, "bla")) == bla
) {
// JAX-B works just fine:
final JAXBElement<Foo> foo = unmarshaller.unmarshal(r,Foo.class);
System.out.println(foo.getValue().getName());
// But how do I get the offset?
// cache.put(r.getAttributeValue(null, "id"), r.getCursor()); // ???
break;
}
}
But I can't get the offset. I'd like to use this to prepare an index:
(id of element) -> (offset in file)
Then I should be able use the offset to just unmarshall from there: Open file stream, skip that many bytes, unmarshall.
I can't find a library that does this. And I can't do it on my own without knowing the position of the file cursor. The javadoc clearly states that there is a cursor, but I can't find a way of accessing it.
Edit:
I'm just trying to offer a solution that will work on old hardware so people can actually use it. Not everyone can afford a new and powerful computer. Using StAX I can get the data in about 2 seconds, which is a bit long. But it doesn't require RAM. It requires 300 MB of RAM to just use JAX-B. Using some embedded db system would just be a lot of overhead for such a simple task. I'll use JAX-B anyway. Anything else would be useless for me since the wsimport-generated classes are already perfect. I just don't want to load 300 MB of objects when I only need a few.
I can't find a DB that just needs an XSD to create an in-memory DB, which doesn't use that much RAM. It's all made for servers or it's required to define a schema and map the XML. So I assume it just doesn't exist.
You could work with a generated XML parser using ANTLR4.
The Following works very well on a ~17GB Wikipedia dump /20170501/dewiki-20170501-pages-articles-multistream.xml.bz2 but I had to increase heap size using -xX6GB.
1. Get XML Grammar
cd /tmp
git clone https://github.com/antlr/grammars-v4
2. Generate Parser
cd /tmp/grammars-v4/xml/
mvn clean install
3. Copy Generated Java files to your Project
cp -r target/generated-sources/antlr4 /path/to/your/project/gen
4. Hook in with a Listener to collect character offsets
package stack43366566;
import java.util.ArrayList;
import java.util.List;
import org.antlr.v4.runtime.ANTLRFileStream;
import org.antlr.v4.runtime.CommonTokenStream;
import org.antlr.v4.runtime.tree.ParseTreeWalker;
import stack43366566.gen.XMLLexer;
import stack43366566.gen.XMLParser;
import stack43366566.gen.XMLParser.DocumentContext;
import stack43366566.gen.XMLParserBaseListener;
public class FindXmlOffset {
List<Integer> offsets = null;
String searchForElement = null;
public class MyXMLListener extends XMLParserBaseListener {
public void enterElement(XMLParser.ElementContext ctx) {
String name = ctx.Name().get(0).getText();
if (searchForElement.equals(name)) {
offsets.add(ctx.start.getStartIndex());
}
}
}
public List<Integer> createOffsets(String file, String elementName) {
searchForElement = elementName;
offsets = new ArrayList<>();
try {
XMLLexer lexer = new XMLLexer(new ANTLRFileStream(file));
CommonTokenStream tokens = new CommonTokenStream(lexer);
XMLParser parser = new XMLParser(tokens);
DocumentContext ctx = parser.document();
ParseTreeWalker walker = new ParseTreeWalker();
MyXMLListener listener = new MyXMLListener();
walker.walk(listener, ctx);
return offsets;
} catch (Exception e) {
throw new RuntimeException(e);
}
}
public static void main(String[] arg) {
System.out.println("Search for offsets.");
List<Integer> offsets = new FindXmlOffset().createOffsets("/tmp/dewiki-20170501-pages-articles-multistream.xml",
"page");
System.out.println("Offsets: " + offsets);
}
}
5. Result
Prints:
Offsets: [2441, 10854, 30257, 51419 ....
6. Read from Offset Position
To test the code I've written class that reads in each wikipedia page to a java object
#JacksonXmlRootElement
class Page {
public Page(){};
public String title;
}
using basically this code
private Page readPage(Integer offset, String filename) {
try (Reader in = new FileReader(filename)) {
in.skip(offset);
ObjectMapper mapper = new XmlMapper();
mapper.configure(DeserializationFeature.FAIL_ON_UNKNOWN_PROPERTIES, false);
Page object = mapper.readValue(in, Page.class);
return object;
} catch (Exception e) {
throw new RuntimeException(e);
}
}
Find complete example on github.
I just had to solve this problem, and spent way too much time figuring it out. Hopefully the next poor soul who comes looking for ideas can benefit from my suffering.
The first problem to contend with is that most XMLStreamReader implementations provide inaccurate results when you ask them for their current offsets. Woodstox however seems to be rock-solid in this regard.
The second problem is the actual type of offset you use. You have to use char offsets if you need to work with a multi-byte charset, which means the random-access retrieval from the file using the provided offsets is not going to be very efficient - you can't just set a pointer into the file at your offset and start reading, you have to read through until you get to the offset (that's what skip does under the covers in a Reader), then start extracting. If you're dealing with very large files, that means retrieval of content near the end of the file is too slow.
I ended up writing a FilterReader that keeps a buffer of byte offset to char offset mappings as the file is read. When we need to get the byte offset, we first ask Woodstox for the char offset, then get the custom reader to tell us the actual byte offset for the char offset. We can get the byte offset from the beginning and end of the element, giving us what we need to go in and surgically extract the element from the file by opening it as a RandomAccessFile, which means it's super fast at any point in the file.
I created a library for this, it's on GitHub and Maven Central. If you just want to get the important bits, the party trick is in the ByteTrackingReader.
Some people have commented about how this whole thing is a bad idea and why would you want to do it? XML is a transport mechanism, you should just import it to a DB and work with the data with more appropriate tools. For most cases this is true, but if you're building applications or integrations that communicate via XML, you need tooling to analyze and operate on the files that are exchanged. I get daily requests to verify feed contents, having the ability to quickly extract a specific set of items from a massive file and verify not only the contents, but the format itself is essential.
Anyhow, hopefully this can save someone a few hours, or at least get them closer to a solution.

MapReduce: How can I output key/value pair without newlines?

I am using a 0 reduce approach to my problem. I wish to preprocess data from one file and then to write it out as another file, but with no new lines and tab delimeters? How can I output my map job that has processed my data with the same file format it came in minus the preprocess.
That is, I have something like this:
Preprocess:
<TITLE> Herp derp </Title> I am a major general
Post Process:
Herp
Derp
I
am
a
major
general
What I want it to do is this:
Herp Derp I am a major general
I believe the issue is with this line of code:
job.setOutputFormatClass(TextOutputFormat.class);
However, when I tried, quite naively to do something like:
job.setOutputFormatClass(null);
It obviously would not work. Is there an format class that is provided that I can use to do this? If not, how could I write my own class to just output everything as I want? I am new to hadoop and map reduce.
I have included my map function below. I do not want to use reduce as it would sort between the map and reducer.
public void map(LongWritable key, Text value, Context context)
throws IOException, InterruptedException {
String line = value.toString();
StringTokenizer tokenizer = new StringTokenizer(line);
while (tokenizer.hasMoreTokens()) {
word.set(tokenizer.nextToken());
//Did preprocessing here, irrelevant to my problem
context.write(word, null);
}
}
Also, I have also googled this and read the apache hadoop api to see if I can gleam an answer.
On your mapper class, instead of parsing your line into individual words and writing them out, try sending the entire line to the
context.write(word, null);
That way it is keeping the entire string you are originally working with together, instead of sending out the line piece by piece.
So, cut your string apart for the preprocess work, then put it back together when you send it out with the context.write command.
If your mapper is writing multiple records containing the individual tokens from a single input line, then you will absolutely need a reducer to group those tokens back together into a single line for output. You can't do this without a reducer.

Categories