{"id":42,"date":"2013-10-03T01:41:10","date_gmt":"2013-10-02T17:41:10","guid":{"rendered":"http:\/\/www.lixiaodong.com\/?p=42"},"modified":"2014-02-03T07:07:54","modified_gmt":"2014-02-02T23:07:54","slug":"2pmms2-phase-multiway-merge-sort%e7%ae%97%e6%b3%95java%e7%89%88","status":"publish","type":"post","link":"https:\/\/www.lixiaodong.com\/?p=42","title":{"rendered":"2PMMS(2-phase multiway merge sort)\u7b97\u6cd5java\u5b9e\u73b0"},"content":{"rendered":"<p>2PMMS(2-phase multiway merge sort)\uff0c\u987e\u540d\u601d\u4e49\uff0c\u75312\u4e2a\u9636\u6bb5\u7ec4\u6210\u3002\u7b2c\u4e00\u4e2a\u9636\u6bb5\u5c06\u6e90\u6587\u4ef6\u5206\u6279\u6b21\u8bfb\u5230\u5185\u5b58\u4e2d\uff0c\u91c7\u7528\u67d0\u79cd\u5185\u90e8\u6392\u5e8f\u7b97\u6cd5\u8fdb\u884c\u6392\u5e8f\uff0c\u7136\u540e\u5c06\u6bcf\u6b21\u6392\u5e8f\u540e\u7684\u7ed3\u679c\u5199\u5230\u786c\u76d8\u4e0a\uff0c1\u4e2a\u6279\u6b21\u5bf9\u5e941\u4e2a\u6587\u4ef6\uff1b\u7b2c\u4e8c\u4e2a\u9636\u6bb5\u5c06n\u4e2a\u6587\u4ef6\u6279\u91cf\u8bfb\u5230\u7f13\u5b58\u91cc\uff0c\u4eceN\u4e2a\u7f13\u5b58\u5e8f\u5217\u91cc\u627e\u5230\u6700\u5c0f\u6570\uff0c\u653e\u5230\u8f93\u51fa\u7f13\u5b58\u91cc\u3002\u6bcf\u6b21\u653e\u5b8c\u540e\uff0c\u8981\u68c0\u67e5\u6bcf\u4e2a\u8f93\u5165\u7f13\u5b58\u662f\u5426\u8bfb\u5b8c\uff0c\u5982\u679c\u8bfb\u5b8c\u5219\u4ece\u5bf9\u5e94\u7684\u6587\u4ef6\u91cc\u7ee7\u7eed\u8bfb\u6570\u636e\u5230\u7f13\u5b58\uff0c\u76f4\u5230\u6587\u4ef6\u5168\u90e8\u8bfb\u5b8c\u4e3a\u6b62\uff1b\u5e76\u68c0\u67e5\u8f93\u51fa\u7f13\u5b58\u662f\u5426\u5df2\u6ee1\uff0c\u5982\u6ee1\u5219\u8f93\u51fa\u5230\u6587\u4ef6\u5e76\u6e05\u7a7a\u8f93\u51fa\u7f13\u5b58\u3002<\/p>\n<p>4\u4e2a\u7c7b\uff1a<\/p>\n<p>(1)<br \/>\n[code lang=&#8221;java&#8221;]<br \/>\npublic class Main {<\/p>\n<p>\/*<br \/>\n* 2 arguments<br \/>\n* first argument: inputfile ex: d:\\\\data\\\\input.dat<br \/>\n* second argument: output folder ex: d:\\\\data<br \/>\n*\/<br \/>\npublic static void main(String[] args) throws Exception {<br \/>\n    if(args.length!=2){<br \/>\n        System.out.println(&quot;arguments are not valid&quot;);<br \/>\n        return;<br \/>\n    }<br \/>\n    long startTime = System.currentTimeMillis();<br \/>\n    \/\/read data, sort, output sorted data to files<br \/>\n    PhaseOne phaseOne = new PhaseOne();<br \/>\n    List&lt;String&gt; fileNameList = phaseOne.generateSortedFile(args[0], args[1]);<br \/>\n    \/\/merge<br \/>\n    PhaseTwo phaseTwo = new PhaseTwo();<br \/>\n    phaseTwo.merge(fileNameList, args[1]);<br \/>\n    \/\/delete temp files<br \/>\n    for(String fileName: fileNameList){<br \/>\n        new File(fileName).delete();<br \/>\n    }<br \/>\n    long finishTime = System.currentTimeMillis();<br \/>\n    double totalTime = (double)(finishTime &#8211; startTime) \/ 1000;<br \/>\n    System.out.println(&quot;total time: &quot; + totalTime + &quot; seconds&quot;);<br \/>\n}<\/p>\n<p>}<br \/>\n[\/code]<br \/>\n(2)<br \/>\n[code lang=&#8221;java&#8221;]<br \/>\nimport java.io.BufferedReader;<br \/>\nimport java.io.BufferedWriter;<br \/>\nimport java.io.FileReader;<br \/>\nimport java.io.FileWriter;<br \/>\nimport java.io.IOException;<br \/>\nimport java.util.ArrayList;<br \/>\nimport java.util.List;<\/p>\n<p>public class PhaseOne {<\/p>\n<p>private final int amountOnce = 550000;<\/p>\n<p>public List&lt;String&gt; generateSortedFile(String inputFile, String outputFolder) {<br \/>\n    List&lt;String&gt; fileNameList = new ArrayList&lt;String&gt;();<br \/>\n    BufferedReader reader = null;<br \/>\n    try {<br \/>\n        reader = new BufferedReader(new FileReader(inputFile));<br \/>\n        String tempString = null;<br \/>\n        int line = 0;<br \/>\n        int[] array = new int[amountOnce];<br \/>\n        while ((tempString = reader.readLine()) != null) {<br \/>\n        \/\/System.out.println(&quot;line &quot; + line + &quot;: &quot; + tempString);<br \/>\n        array[line % amountOnce] = Integer.parseInt(tempString);<br \/>\n        if(line % amountOnce == amountOnce -1){<br \/>\n            String fileNO = (line\/amountOnce)&lt;10 ? &quot;0&quot;+ line\/amountOnce : String.valueOf(line\/amountOnce);<br \/>\n            String fileName = outputFolder+ &quot;\/tmp&quot;+ fileNO+&quot;.dat&quot;;<br \/>\n            writeData(array, line%amountOnce+1, fileName,fileNameList);<br \/>\n        }<br \/>\n        line++;<br \/>\n    }<br \/>\n    String fileNO = (line\/amountOnce)&lt;10 ? &quot;0&quot;+ line\/amountOnce : String.valueOf(line\/amountOnce);<br \/>\n    String fileName = outputFolder+ &quot;\/tmp&quot;+ fileNO+&quot;.dat&quot;;<br \/>\n    writeData(array, line%amountOnce, fileName, fileNameList);<br \/>\n    reader.close();<br \/>\n} catch (IOException e) {<br \/>\n    e.printStackTrace();<br \/>\n} finally {<br \/>\n    if (reader != null) {<br \/>\n    try {<br \/>\n        reader.close();<br \/>\n    }<br \/>\n    catch (IOException e1) {<br \/>\n    }<br \/>\n}<br \/>\nreturn fileNameList;<br \/>\n}<br \/>\n}<\/p>\n<p>private void writeData(int[] array, int length, String fileName, List&lt;String&gt; fileNames) throws IOException{<br \/>\n    Quicksort.quickSort(array, length);<br \/>\n    BufferedWriter output = new BufferedWriter(new FileWriter(fileName));<br \/>\n    for(int i=0;i&lt;length;i++){<br \/>\n        output.write(String.valueOf(array[i]));<br \/>\n        output.write(&quot;\\r\\n&quot;);<br \/>\n    }<br \/>\n    output.close();<br \/>\n    fileNames.add(fileName);<br \/>\n    array = new int[amountOnce];<br \/>\n}<br \/>\n}<br \/>\n[\/code]<br \/>\n(3)<br \/>\n[code lang=&#8221;java&#8221;]<br \/>\nimport java.io.BufferedReader;<br \/>\nimport java.io.BufferedWriter;<br \/>\nimport java.io.FileOutputStream;<br \/>\nimport java.io.FileReader;<br \/>\nimport java.io.IOException;<br \/>\nimport java.io.OutputStreamWriter;<br \/>\nimport java.util.ArrayList;<br \/>\nimport java.util.List;<\/p>\n<p>public class PhaseTwo {<\/p>\n<p>public static int inputBufferSize = 30000;<\/p>\n<p>public static int outputBufferSize = 100000;<\/p>\n<p>public static boolean finishFlag = false;<\/p>\n<p>public void merge(List&lt;String&gt; fileNameList, String outputFolder) {<br \/>\n    int[] outputBuffer = new int[outputBufferSize];<br \/>\n    int[][] inputBuffers = new int[fileNameList.size()][inputBufferSize];<br \/>\n    List&lt;BufferedReader&gt; readers = new ArrayList&lt;BufferedReader&gt;();<br \/>\n    try {<br \/>\n        for(int i=0;i&lt;fileNameList.size();i++){<br \/>\n            BufferedReader reader = new BufferedReader(new FileReader(fileNameList.get(i)));<br \/>\n            readers.add(reader);<br \/>\n            String tempString = null;<br \/>\n            int line = 0;<br \/>\n            while ((tempString = readers.get(i).readLine()) != null) {<br \/>\n            inputBuffers[i][line] = Integer.parseInt(tempString);<br \/>\n            line++;<br \/>\n            if(line &gt;= inputBufferSize){<br \/>\n                break;<br \/>\n            }<br \/>\n        }<br \/>\n    }<\/p>\n<p>    int[] pointers = new int[fileNameList.size()];<\/p>\n<p>    int outputBufferIndex = 0;<br \/>\n    while(true){<br \/>\n        int min = findMin(inputBuffers,pointers, readers);<br \/>\n        outputBuffer[outputBufferIndex % outputBufferSize] = min;<br \/>\n        if(outputBufferIndex % outputBufferSize == outputBufferSize-1){<br \/>\n            outputData(outputBuffer, outputBufferSize, outputFolder+&quot;\/result.dat&quot;);<br \/>\n        }<br \/>\n        outputBufferIndex++;<br \/>\n        if(finishFlag){<br \/>\n            break;<br \/>\n        }<br \/>\n    }<br \/>\n    outputData(outputBuffer, outputBufferIndex%outputBufferSize, outputFolder+&quot;\/result.dat&quot;);<\/p>\n<p>    for(int i=0;i&lt;fileNameList.size();i++){<br \/>\n        readers.get(i).close();<br \/>\n    }<br \/>\n    System.out.println(&quot;&#8212;&#8212;&#8212;&#8212;&#8212;&#8212;&quot;);<\/p>\n<p>} catch (IOException e) {<br \/>\n    e.printStackTrace();<br \/>\n} finally {<br \/>\n    for(int i=0;i&lt;fileNameList.size();i++){<br \/>\n    if (readers.get(i) != null) {<br \/>\n    try {<br \/>\n        readers.get(i).close();<br \/>\n    }<br \/>\n    catch (IOException e1) {<br \/>\n    }<br \/>\n}<br \/>\n}<br \/>\n}<br \/>\n}<\/p>\n<p>private int findMin(int[][] arrays, int[] pointers, List&lt;BufferedReader&gt; readers) throws IOException{<br \/>\n    int min = 0;<br \/>\n    int index = 0;<br \/>\n    for(int i=0;i&lt;arrays.length;i++){<br \/>\n        if(arrays[i][pointers[i]]&gt;0){<br \/>\n            min = arrays[i][pointers[i]];<br \/>\n            index = i;<br \/>\n            break;<br \/>\n        }<br \/>\n    }<\/p>\n<p>    for(int i=index+1;i&lt;arrays.length;i++){<br \/>\n        if(arrays[i][pointers[i]]&lt;=0){<br \/>\n            continue;<br \/>\n        }<br \/>\n        if(arrays[i][pointers[i]] &lt; min){<br \/>\n            min = arrays[i][pointers[i]];<br \/>\n            index = i;<br \/>\n        }<br \/>\n    }<br \/>\n    pointers[index] = pointers[index]+1;<\/p>\n<p>    for(int i=0;i&lt;pointers.length;i++){<br \/>\n        if(pointers[i]&gt;=inputBufferSize){<br \/>\n            arrays[i] = new int[inputBufferSize];<br \/>\n            String tempString = null;<br \/>\n            int line = 0;<br \/>\n            while ((tempString = readers.get(i).readLine()) != null) {<br \/>\n                arrays[i][line] = Integer.parseInt(tempString);<br \/>\n                line++;<br \/>\n                if(line &gt;= inputBufferSize){<br \/>\n                    break;<br \/>\n                }<br \/>\n            }<br \/>\n            pointers[i]= 0;<br \/>\n        }<br \/>\n    } <\/p>\n<p>    \/\/update finishFlag<br \/>\n    if(!finishFlag){<br \/>\n        boolean flag = true;<br \/>\n        for(int i=0;i&lt;arrays.length;i++){<br \/>\n        if(arrays[i][pointers[i]]&gt;0){<br \/>\n            flag = false;<br \/>\n            break;<br \/>\n        }<br \/>\n    }<br \/>\n    finishFlag = flag;<br \/>\n}<br \/>\nreturn min;<br \/>\n}<\/p>\n<p>private void outputData(int[] outputBuffer, int length, String outputFile) throws IOException{<br \/>\nBufferedWriter output = new BufferedWriter(new OutputStreamWriter(new FileOutputStream(outputFile, true)));<br \/>\nfor(int i=0;i&lt;length;i++){<br \/>\n    StringBuffer sb = new StringBuffer();<br \/>\n    sb.append(outputBuffer[i]);<br \/>\n    sb.append(&quot;\\r\\n&quot;);<br \/>\n    output.write(sb.toString());<br \/>\n}<br \/>\noutput.close();<br \/>\noutputBuffer = new int[outputBufferSize];<br \/>\n}<\/p>\n<p>}<br \/>\n[\/code]<br \/>\n(4)<br \/>\n[code lang=&#8221;java&#8221;]<br \/>\npublic class Quicksort {<br \/>\n    public static void quickSort(int[] data, int length) {<br \/>\n        recQuickSort(data, 0, length-1);<br \/>\n    }<\/p>\n<p>private static void recQuickSort(int[] data,int left, int right) {<br \/>\n    if(right-left &lt;= 0){<br \/>\n    return;<br \/>\n}<br \/>\nelse {<br \/>\n    long pivot = data[right];<br \/>\n    int partition = partitionIt(data, left, right, pivot);<br \/>\n    recQuickSort(data, left, partition-1);<br \/>\n    recQuickSort(data, partition+1, right);<br \/>\n}<br \/>\n}<\/p>\n<p>private static int partitionIt(int[] data, int left, int right, long pivot) {<br \/>\n    int leftPtr = left-1;<br \/>\n    int rightPtr = right;<br \/>\n    while(true) {<br \/>\n        while(data[++leftPtr] &lt; pivot )<br \/>\n        ;<br \/>\n        while(rightPtr &gt; 0 &amp;&amp; data[&#8211;rightPtr] &gt; pivot)<br \/>\n        ;<br \/>\n        if(leftPtr &gt;= rightPtr){<br \/>\n        break;<br \/>\n    }<br \/>\n    else{<br \/>\n        swap(data, leftPtr, rightPtr);<br \/>\n    }<br \/>\n}<br \/>\nswap(data, leftPtr, right);<br \/>\n    return leftPtr;<br \/>\n}<\/p>\n<p>private static void swap(int[] data, int dex1, int dex2) {<br \/>\n    int temp = data[dex1];<br \/>\n    data[dex1] = data[dex2];<br \/>\n    data[dex2] = temp;<br \/>\n}<br \/>\n}<br \/>\n[\/code]<\/p>\n","protected":false},"excerpt":{"rendered":"<p>2PMMS(2-phase multiway merge sort)\uff0c\u987e\u540d\u601d\u4e49\uff0c\u75312\u4e2a\u9636\u6bb5\u7ec4\u6210\u3002\u7b2c\u4e00\u4e2a\u9636\u6bb5\u5c06\u6e90 [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[8],"tags":[7,42],"class_list":["post-42","post","type-post","status-publish","format-standard","hentry","category-8","tag-2pmms","tag-java"],"_links":{"self":[{"href":"https:\/\/www.lixiaodong.com\/index.php?rest_route=\/wp\/v2\/posts\/42","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.lixiaodong.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.lixiaodong.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.lixiaodong.com\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.lixiaodong.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=42"}],"version-history":[{"count":7,"href":"https:\/\/www.lixiaodong.com\/index.php?rest_route=\/wp\/v2\/posts\/42\/revisions"}],"predecessor-version":[{"id":129,"href":"https:\/\/www.lixiaodong.com\/index.php?rest_route=\/wp\/v2\/posts\/42\/revisions\/129"}],"wp:attachment":[{"href":"https:\/\/www.lixiaodong.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=42"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.lixiaodong.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=42"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.lixiaodong.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=42"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}