2013
10.03

2PMMS(2-phase multiway merge sort),顾名思义,由2个阶段组成。第一个阶段将源文件分批次读到内存中,采用某种内部排序算法进行排序,然后将每次排序后的结果写到硬盘上,1个批次对应1个文件;第二个阶段将n个文件批量读到缓存里,从N个缓存序列里找到最小数,放到输出缓存里。每次放完后,要检查每个输入缓存是否读完,如果读完则从对应的文件里继续读数据到缓存,直到文件全部读完为止;并检查输出缓存是否已满,如满则输出到文件并清空输出缓存。

4个类:

(1)

public class Main {

/*
* 2 arguments
* first argument: inputfile ex: d:\\data\\input.dat
* second argument: output folder ex: d:\\data
*/
public static void main(String[] args) throws Exception {
    if(args.length!=2){
        System.out.println("arguments are not valid");
        return;
    }
    long startTime = System.currentTimeMillis();
    //read data, sort, output sorted data to files
    PhaseOne phaseOne = new PhaseOne();
    List<String> fileNameList = phaseOne.generateSortedFile(args[0], args[1]);
    //merge
    PhaseTwo phaseTwo = new PhaseTwo();
    phaseTwo.merge(fileNameList, args[1]);
    //delete temp files
    for(String fileName: fileNameList){
        new File(fileName).delete();
    }
    long finishTime = System.currentTimeMillis();
    double totalTime = (double)(finishTime - startTime) / 1000;
    System.out.println("total time: " + totalTime + " seconds");
}

}

(2)

import java.io.BufferedReader;
import java.io.BufferedWriter;
import java.io.FileReader;
import java.io.FileWriter;
import java.io.IOException;
import java.util.ArrayList;
import java.util.List;

public class PhaseOne {

private final int amountOnce = 550000;

public List<String> generateSortedFile(String inputFile, String outputFolder) {
    List<String> fileNameList = new ArrayList<String>();
    BufferedReader reader = null;
    try {
        reader = new BufferedReader(new FileReader(inputFile));
        String tempString = null;
        int line = 0;
        int[] array = new int[amountOnce];
        while ((tempString = reader.readLine()) != null) {
        //System.out.println("line " + line + ": " + tempString);
        array[line % amountOnce] = Integer.parseInt(tempString);
        if(line % amountOnce == amountOnce -1){
            String fileNO = (line/amountOnce)<10 ? "0"+ line/amountOnce : String.valueOf(line/amountOnce);
            String fileName = outputFolder+ "/tmp"+ fileNO+".dat";
            writeData(array, line%amountOnce+1, fileName,fileNameList);
        }
        line++;
    }
    String fileNO = (line/amountOnce)<10 ? "0"+ line/amountOnce : String.valueOf(line/amountOnce);
    String fileName = outputFolder+ "/tmp"+ fileNO+".dat";
    writeData(array, line%amountOnce, fileName, fileNameList);
    reader.close();
} catch (IOException e) {
    e.printStackTrace();
} finally {
    if (reader != null) {
    try {
        reader.close();
    }
    catch (IOException e1) {
    }
}
return fileNameList;
}
}

private void writeData(int[] array, int length, String fileName, List<String> fileNames) throws IOException{
    Quicksort.quickSort(array, length);
    BufferedWriter output = new BufferedWriter(new FileWriter(fileName));
    for(int i=0;i<length;i++){
        output.write(String.valueOf(array[i]));
        output.write("\r\n");
    }
    output.close();
    fileNames.add(fileName);
    array = new int[amountOnce];
}
}

(3)

import java.io.BufferedReader;
import java.io.BufferedWriter;
import java.io.FileOutputStream;
import java.io.FileReader;
import java.io.IOException;
import java.io.OutputStreamWriter;
import java.util.ArrayList;
import java.util.List;

public class PhaseTwo {

public static int inputBufferSize = 30000;

public static int outputBufferSize = 100000;

public static boolean finishFlag = false;

public void merge(List<String> fileNameList, String outputFolder) {
    int[] outputBuffer = new int[outputBufferSize];
    int[][] inputBuffers = new int[fileNameList.size()][inputBufferSize];
    List<BufferedReader> readers = new ArrayList<BufferedReader>();
    try {
        for(int i=0;i<fileNameList.size();i++){
            BufferedReader reader = new BufferedReader(new FileReader(fileNameList.get(i)));
            readers.add(reader);
            String tempString = null;
            int line = 0;
            while ((tempString = readers.get(i).readLine()) != null) {
            inputBuffers[i][line] = Integer.parseInt(tempString);
            line++;
            if(line >= inputBufferSize){
                break;
            }
        }
    }


    int[] pointers = new int[fileNameList.size()];

    int outputBufferIndex = 0;
    while(true){
        int min = findMin(inputBuffers,pointers, readers);
        outputBuffer[outputBufferIndex % outputBufferSize] = min;
        if(outputBufferIndex % outputBufferSize == outputBufferSize-1){
            outputData(outputBuffer, outputBufferSize, outputFolder+"/result.dat");
        }
        outputBufferIndex++;
        if(finishFlag){
            break;
        }
    }
    outputData(outputBuffer, outputBufferIndex%outputBufferSize, outputFolder+"/result.dat");

    for(int i=0;i<fileNameList.size();i++){
        readers.get(i).close();
    }
    System.out.println("------------------");

} catch (IOException e) {
    e.printStackTrace();
} finally {
    for(int i=0;i<fileNameList.size();i++){
    if (readers.get(i) != null) {
    try {
        readers.get(i).close();
    }
    catch (IOException e1) {
    }
}
}
}
}

private int findMin(int[][] arrays, int[] pointers, List<BufferedReader> readers) throws IOException{
    int min = 0;
    int index = 0;
    for(int i=0;i<arrays.length;i++){
        if(arrays[i][pointers[i]]>0){
            min = arrays[i][pointers[i]];
            index = i;
            break;
        }
    }

    for(int i=index+1;i<arrays.length;i++){
        if(arrays[i][pointers[i]]<=0){
            continue;
        }
        if(arrays[i][pointers[i]] < min){
            min = arrays[i][pointers[i]];
            index = i;
        }
    }
    pointers[index] = pointers[index]+1;

    for(int i=0;i<pointers.length;i++){
        if(pointers[i]>=inputBufferSize){
            arrays[i] = new int[inputBufferSize];
            String tempString = null;
            int line = 0;
            while ((tempString = readers.get(i).readLine()) != null) {
                arrays[i][line] = Integer.parseInt(tempString);
                line++;
                if(line >= inputBufferSize){
                    break;
                }
            }
            pointers[i]= 0;
        }
    } 

    //update finishFlag
    if(!finishFlag){
        boolean flag = true;
        for(int i=0;i<arrays.length;i++){
        if(arrays[i][pointers[i]]>0){
            flag = false;
            break;
        }
    }
    finishFlag = flag;
}
return min;
}

private void outputData(int[] outputBuffer, int length, String outputFile) throws IOException{
BufferedWriter output = new BufferedWriter(new OutputStreamWriter(new FileOutputStream(outputFile, true)));
for(int i=0;i<length;i++){
    StringBuffer sb = new StringBuffer();
    sb.append(outputBuffer[i]);
    sb.append("\r\n");
    output.write(sb.toString());
}
output.close();
outputBuffer = new int[outputBufferSize];
}

}

(4)

public class Quicksort {
    public static void quickSort(int[] data, int length) {
        recQuickSort(data, 0, length-1);
    }

private static void recQuickSort(int[] data,int left, int right) {
    if(right-left <= 0){
    return;
}
else {
    long pivot = data[right];
    int partition = partitionIt(data, left, right, pivot);
    recQuickSort(data, left, partition-1);
    recQuickSort(data, partition+1, right);
}
}

private static int partitionIt(int[] data, int left, int right, long pivot) {
    int leftPtr = left-1;
    int rightPtr = right;
    while(true) {
        while(data[++leftPtr] < pivot )
        ;
        while(rightPtr > 0 && data[--rightPtr] > pivot)
        ;
        if(leftPtr >= rightPtr){
        break;
    }
    else{
        swap(data, leftPtr, rightPtr);
    }
}
swap(data, leftPtr, right);
    return leftPtr;
}

private static void swap(int[] data, int dex1, int dex2) {
    int temp = data[dex1];
    data[dex1] = data[dex2];
    data[dex2] = temp;
}
}
2013
10.01

首先,我们要明确如何用LISP表示二叉树节点结构。一个节点由本身,左子树,右子树3部分组成。我们就可以用如下结构表示1个节点:

(node (left-subtree) (right-subtree))

比如,1个二叉树由3个节点组成,根为20, 左子树1个节点30, 右子树1个节点40。可以表示为:'(20 (30 () ()) (40 () ()))。

更复杂的情况,1个由7个节点组成的完全二叉树,从上到下从左到右,分别是1,2,3,4,5,6,7。则表示成 ‘(1 (2 (4 () ()) (5 () ())) (3 (6 () ()) (7 () ())))。

定义好二叉树的LISP表示方法后,就可以定义3种遍历算法了。

前序遍历:


(defun preorder(tree)
(if (or (null tree)
(not (listp tree))
(null (car tree))
(listp (car tree))
(not (= (length tree) 3))
(not (listp (car (cdr tree))))
(not (listp (car (cdr (cdr tree)))))
)
nil
(cons (car tree) (append (preorder (car (cdr tree))) (preorder (car (cdr (cdr tree))))))))

中序遍历:

(defun inorder(tree)
(if (or (null tree)
(not (listp tree))
(null (car tree))
(listp (car tree))
(not (= (length tree) 3))
(not (listp (car (cdr tree))))
(not (listp (car (cdr (cdr tree)))))
)
nil
(append (inorder (car (cdr tree))) (cons (car tree) '()) (inorder (car (cdr (cdr tree)))))))

后序遍历:

(defun postorder(tree)
(if (or (null tree)
(not (listp tree))
(null (car tree))
(listp (car tree))
(not (= (length tree) 3))
(not (listp (car (cdr tree))))
(not (listp (car (cdr (cdr tree)))))
)
nil
(append (postorder (car (cdr tree))) (postorder (car (cdr (cdr tree)))) (cons (car tree) '()))))
2013
09.06

SVN是一个很好的版本控制工具,可以进行协同开发,查看代码历史记录等等。

服务器已经安装了nginx,作为前端服务器正在使用中。nginx目前还不支持SVN以http形式访问,所以需要安装apache。nginx作为前端服务器,apache作为后端服务器。

1. 安装各种软件

sudo yum install httpd subversion  mod_dav_svn

2. 配置

新建svn 仓库:

mkdir /opt/svn/repos

chmod -R 777 /opt/svn/repos(一定要加-R,否则以后提交代码会有问题)

svnadmin create –fs-type fsfs /opt/svn/repos

添加项目:

mkdir -p /opt/svn/tmp/project1
cd /opt/svn/tmp/project1
mkdir branches
mkdir tags
mkdir trunk

svn import /opt/svn/tmp/ file:///opt/svn/repos/ –message “init”

配置apache:

vi /etc/httpd/conf/httpd.conf

增加:

LoadModule dav_svn_module modules/mod_dav_svn.so
LoadModule authz_svn_module modules/mod_authz_svn.so

<Location /repos>
DAV svn
SVNPath /opt/svn/repos
AuthType Basic
AuthName “svn repos”
AuthUserFile /etc/svn-auth-conf
AuthzSVNAccessFile /etc/svn-accesspolicy
Satisfy Any
Require valid-user
</Location>

nginx已经占用了80端口,修改apache端口:Listen 8180

配置nginx:

vi /etc/nginx/nginx.conf

增加:

server {
listen 80;
server_name svn.xxx.com;
location / {
proxy_pass http://127.0.0.1:8180;
}

增加用户校验:

htpasswd /etc/svn-auth-conf username,设定密码

新建文件/etc/svn-accesspolicy,文件内容如下:

[groups]
developers = username

[/]
@developers = rw

最后启动apache,重启nginx:

sudo service httpd start

sudo service nginx restart

打开浏览器,输入:http://svn.xxx.com/repos

2013
09.05

使用Eclipse Marketplace安装Maven插件

Maven是一个很好的项目管理工具。它的优点很多,比如可以简化依赖管理,不用手工逐个去获取、添加各种依赖包,而是在pom文件里添加几行配置就行。这样会节约大量时间和精力,因为我们依赖一个包,而这个包还可能依赖另外的包,以此类推,依赖的包会很多。如果采用最原始的方式,将这些包一一找到并添加到classpath下,这个工作会非常麻烦。幸好Maven帮我们把这些繁重的事情都做了。

过去我们在eclipse里添加maven插件,会通过菜单Help-Install New Software,指定maven插件的下载地址来进行安装。最新的eclipse推出了eclipse marketplace功能,其实就相当于apple的appstore,比以前安装插件的方式更方便了。试过以前2个maven插件地址,似乎都失效了,可能maven插件已经转移到eclipse marketplace里了吧。

下面介绍安装步骤:

1.安装Maven插件。

点击菜单Help——Eclipse Marketplace,在弹出对话框里find:右边的输入框输入“maven”,回车。在搜索结果里找到Maven Integration For Eclipse(Juno and newer),然后点击“Install”按钮。一路Next,接受安装协议,最后重启Eclipse。

2. 配置Maven插件。

点击菜单Window——Preferences,左边菜单找到Maven——Installations。我们看到右边显示已经安装了一个Embedded的Maven。我的电脑之前安装了一个Maven,我更喜欢用我安装的那个版本的Maven。这里点击右边的”Add”,添加之前已经安装好的Maven。

maven1

 

选择菜单Maven——User Settings,指定setting.xml文件的位置,以及本地Maven Repository的位置。

maven2

 

我们将一个Maven项目导入Eclipse中,选中项目名称,单击右键,Configure->Convert to Maven Project。再次选中项目,单击右键,这时右键菜单里会出现“Maven”选项。

2013
08.27

discuz删除一些版面后再访问之前的版面会有提示,但并没有返回我们期望的404状态,而是正常状态200。

google等搜索引擎已经收录了之前的网页,我们希望从google中删除这些已经不存在的内容。只有返回404,google才会认为这些网页已经不存在。我们需要对discuz做修改,如果版面或帖子已经删除,返回错误提示页面,并且状态是404。

最近,很多互联网公司和社会组织发起了寻找失踪儿童的公益活动。我们可以把自己的404页面设置成寻找失踪儿童页面。

以discuz X3为例,需要按照如下方式修改:

1. source/module/forum/forum_forumdisplay.php

注释掉:showmessage(‘forum_nonexistence’, NULL);

新增:dheader(“Location: /404.php”);

2. source/module/forum/forum_viewthread.php

和1一样。

3. 在网站根目录下新建文件404.php

4. 在template/default目录下新建1个目录,取名diy,然后在里面新建一个文件404.htm

代码见附件。upload

 

 


Hit Counter by http://yizhantech.com/