星期五, 6月 19, 2015

Django Static Html Processing and Data Model Migration

Good Explain:

Django对静态文件的处理——部署阶段
http://blog.makto.me/post/2012-11-09/static-files-in-django-deployment

Django对静态文件的处理——开发阶段
http://blog.makto.me/post/2012-11-10/static-files-in-django-development

Django筆記(5) - 模型與資料庫
http://dokelung-blog.logdown.com/posts/220606-django-notes-5-model-and-database

星期六, 6月 06, 2015

Omnet++ INET UDPBasicApp生命週期觀察

Summary:

1. UDPBasicApp::initialize(int stage)
2. if (stage == INITSTAGE_LOCAL) {
   selfMsg = new cMessage("sendTimer");

2.1 selfMsg變數表「排程狀態」,用來當事件參數 OperationalBase::handleMessage(cMessage *message),做狀態機,控制用。

3. stage ++
4. 直到stage=12, i.e,  isInitializeStage(stage)==true, 執行     handleNodeStart(nullptr);
5.  handleNodeStart,     排程
selfMsg->setKind(START);
        scheduleAt(start, selfMsg);

以上是戴入omnet.ini就會執行的code

6. 至於真正的封包排程, schedule要等到按run才會執行, 此時 if (isOperational)為真
7. 事件發生,執行handleMessage, 會再呼叫  handleMessageWhenUp(message);
8. 會再呼叫processStart(), 正式app的啟動,
     case START:
                processStart();
                break;
9. processStart()呼叫 processSend();
10.  processSend()依設定的sendInterval參數, 送出封包後,再排程下次傳送時間點 
11. 重覆7-11, 執行
            case SEND:
                processSend();


                break;


-----------------------------------------------------------------
Details:

omnet++載入時omnet.ini會執行

0. stage=0, i.e, INITSTAGE_LOCAL
1. 一開始初始化階段,執行程式碼
   ApplicationBase::initialize(stage);
  if (stage == INITSTAGE_LOCAL) {
  EV_INFO << "hsuth: INITSTAGE_LOCAL"<< endl;
程式碼,進行初始化。
}

2. 每次跑完stage++
3. 在stage=12時, isInitializeStage(stage)會傳回 true, 此時會跑 handleNodeStart(nullptr);

void OperationalBase::initialize(int stage)
{
    if (isInitializeStage(stage)) {
        NodeStatus *nodeStatus = dynamic_cast(findContainingNode(this)->getSubmodule("status"));
        setOperational(!nodeStatus || nodeStatus->getState() == NodeStatus::UP);
        if (isOperational) {
            EV_INFO << "hsuth: OperationalBase::handleNodeStart"<< endl;
            handleNodeStart(nullptr);
        }
    }
}
會呼叫
bool UDPBasicApp::handleNodeStart(IDoneCallback *doneCallback)
{
    EV_INFO << "hsuth: UDPBasicApp::handleNodeStart"<< endl;
    simtime_t start = std::max(startTime, simTime());
    if ((stopTime < SIMTIME_ZERO) || (start < stopTime) || (start == stopTime && startTime == stopTime)) {
        selfMsg->setKind(START);
        scheduleAt(start, selfMsg);
    }
    return true;
}


當按run時,才會跑排程
排程,時間到呼叫handleMessage, 在
void OperationalBase::handleMessage(cMessage *message)
{
    if (isOperational)
        handleMessageWhenUp(message);
    else
        handleMessageWhenDown(message);
}
會呼叫 UDPBasicApp::handleMessageWhenUp

接著會跑到
void UDPBasicApp::handleMessageWhenUp(cMessage *msg)
{
    if (msg->isSelfMessage()) {
        ASSERT(msg == selfMsg);
        switch (selfMsg->getKind()) {
            case START:
                processStart();
                break;

執行
void UDPBasicApp::processStart()
{

if (!destAddresses.empty()) {
        selfMsg->setKind(SEND);
        processSend();
    }

}

執行

void UDPBasicApp::processSend()
{
    sendPacket();
    simtime_t d = simTime() + par("sendInterval").doubleValue();
    if (stopTime < SIMTIME_ZERO || d < stopTime) {
        selfMsg->setKind(SEND);
        scheduleAt(d, selfMsg);
    }
    else {
        selfMsg->setKind(STOP);
        scheduleAt(stopTime, selfMsg);
    }

}

重覆


void UDPBasicApp::handleMessageWhenUp(cMessage *msg)
{
    if (msg->isSelfMessage()) {
        ASSERT(msg == selfMsg);
        switch (selfMsg->getKind()) {
            case START:
                processStart();
                break;

            case SEND:
                processSend();
                break;

取得sendInterval,一直重覆,直到simulation end

void UDPBasicApp::processSend()
{
    sendPacket();
    simtime_t d = simTime() + par("sendInterval").doubleValue();
    if (stopTime < SIMTIME_ZERO || d < stopTime) {
        selfMsg->setKind(SEND);
        scheduleAt(d, selfMsg);
    }
    else {
        selfMsg->setKind(STOP);
        scheduleAt(stopTime, selfMsg);
    }
}

星期日, 5月 31, 2015

AngularJS Free Course

Reference:
Code School: Shaping up with angular.js
http://campus.codeschool.com/courses/shaping-up-with-angular-js/intro

Bootstrap Editor and Reference

Editor:

pingendo


Video:

week 1: Blog web design with pingendo
https://youtu.be/hCNRZnVoy0A

星期五, 5月 15, 2015

mac os 10.10.3 mail.app typing lag issue

Fix:

1. Open a message-editing window. Under the Edit menu, deselect all the options in the Spelling and Substitutions  sub-menus. Test.
2. Many problems with Mail are caused by iCloud Drive. If you use the service, open the iCloud preference pane, click Options... next to the iCloud Drive box, and in the sheet that opens, uncheck the box marked Mail. See whether there's an improvement.

System wide:  Go to System Preferences > Extensions > Actions and untick the "Markup" box

Mail only:  Customize the new-message toolbar (open a new message, right-click/control-click on the tool bar, select "Customize Toolbar" and add or subtract something - especially the "Markup" icon, but changing anything may work)

星期五, 4月 24, 2015

如何在Windows下安裝SciPy, matplotlib

在Windows下安裝SciPy, matplotlib 實在是很麻煩的事,上網查了一下,決定試用

http://conda.pydata.org/miniconda.html


Miniconda是Anaconda下的工具,Anaconda把很多在Windows需要編譯的package都編好了,直接安裝就可以,可以省很多事....相較之前的痛苦,一整個輕鬆愉快

把舊的python砍掉重練. 下戴miniconda後重新安裝....

安裝packeage方式如下:

$ conda install numpy

也可以產生像virtualenv的虛擬環境,解決dependency的問題

參閱https://gist.github.com/ccwang002/449159cc2a05b1011467

> conda create -n ngs python=2.7 pip
使用它很簡單就 activate ngsdeactivate, 細節可以看 conda 的說明文件。總之在這邊
> activate ngs
Activating environment ngs ...
[ngs]> conda install numpy scipy
...
Proceed ([y]/n)? 



在pycharm重設 interpter的路徑就好了:

http://unlikenoise.com/setup-pycharm-anaconda-python-windows/



Acer S7 鍵盤問題Solution

2015/5/13更新:
下戴zerofix後執行,此法更方便,不用設定半天....

https://dl.dropbox.com/u/26748522/zechofix.html

Reference:
http://community.acer.com/t5/Ultra-Thin/Acer-s7-keyboard-issue-with-repeating-characters/td-p/24407



Acer S7 的鍵盤常常會重覆前一個字元,例如: talkk, yearr...造成輸入文字時嚴重的問題。
上網查了一下,很多外國人都有反應這個問題,目前解法是:





基本上可以解決問題,缺點是按刪除鍵時不能一直壓著連續刪除,要改變使用習慣。

星期一, 4月 20, 2015

安裝Windows 版NumPy及SciPy

NumPy及SciPy在windows 安裝,都會有compile errors...

http://www.scipy.org/scipylib/download.html

要安裝的話,建議到sourceforge安裝預先compile的all-in-one版。

http://sourceforge.net/projects/numpy/files/

http://sourceforge.net/projects/scipy/files/

後記補充,由於這種方式安裝matplot等套件不方便,建議改用miniconda....

Python Text Classification using Naive Bayes and scikit-learn


Feature extraction (特徵擷取) [5]

CountVectorizer implements both tokenization (英文分詞) and occurrence counting(計算英文文字出現計數) in a single class:
>>>
>>> from sklearn.feature_extraction.text import CountVectorizer
This model has many parameters, however the default values are quite reasonable (please see the reference documentation for the details):
>>>
>>> vectorizer = CountVectorizer(min_df=1)
>>> vectorizer                     
CountVectorizer(analyzer=...'word', binary=False, decode_error=...'strict',
        dtype=<... 'numpy.int64'>, encoding=...'utf-8', input=...'content',
        lowercase=True, max_df=1.0, max_features=None, min_df=1,
        ngram_range=(1, 1), preprocessor=None, stop_words=None,
        strip_accents=None, token_pattern=...'(?u)\\b\\w\\w+\\b',
        tokenizer=None, vocabulary=None)
Let’s use it to tokenize and count the word occurrences of a minimalistic corpus of text documents:
說明: 
fit函式代表tokenize,加入到字典陣列vocabulary 
fit(raw_documents[, y])Learn a vocabulary dictionary of all tokens in the raw documents.
transform函式用來計計算英文文字出現計數
transform(raw_documents)Transform documents to document-term matrix.
fit_transform(raw_documents[, y])Learn the vocabulary dictionary and return term-document matrix.
>>>
>>> corpus = [
...     'This is the first document.',
...     'This is the second second document.',
...     'And the third one.',
...     'Is this the first document?',
... ]
>>> X = vectorizer.fit_transform(corpus)
>>> X                              
<4x9 matrix="" numpy.int64="" of="" sparse="" type="">'
    with 19 stored elements in Compressed Sparse ... format>
The default configuration tokenizes the string by extracting words of at least 2 letters. Each term found by the analyzer during the fit is assigned a unique integer index corresponding to a column in the resulting matrix. This interpretation of the columns can be retrieved as follows:
>>>
>>> vectorizer.get_feature_names() == (
...     ['and', 'document', 'first', 'is', 'one',
...      'second', 'the', 'third', 'this'])
True

>>> X.toarray()           
array([[0, 1, 1, 1, 0, 0, 1, 0, 1],
       [0, 1, 0, 1, 0, 2, 1, 0, 1],
       [1, 0, 0, 0, 1, 0, 1, 1, 0],
       [0, 1, 1, 1, 0, 0, 1, 0, 1]]...)

說明: 
經過特徵擷取後,可以利用get_feature_names()取得特徵索引字串字典,接著對映到結果計數陣列。
呼叫X.toarray()可以看出,每個文件,例[0, 1, 1, 1, 0, 0, 1, 0, 1]所對映到的英文字計數....
'This is the first document.' ->['and', 'document', 'first', 'is', 'one','second', 'the', 'third', 'this']


get_feature_names()Array mapping from feature integer indices to feature name



References:

1. Machine Learning Tutorial: The Naive Bayes Text Classifier

2. Naive Bayes

3. Working With Text Data — scikit-learn 0.16.1 documentation

4. Text Classification

5. Feature extraction

Videos:

高一下數學3-0引言01什麼是機率

高一下數學3-3A觀念01條件機率的概念


星期四, 4月 16, 2015

node js tcp socekt end and close different

Good Reference: http://maxogden.com/node-streams.html


end(): Half-closes the socket. i.e., it sends a FIN packet. It is possible the server will still send some data. only closes the writing stream of the socket, the remote host can keep his writing stream open and send you data.
destroy(): Ensures that no more I/O activity happens on this socket. Only necessary in case of errors (parse error or so).


Event: 'end'#
Emitted when the other end of the socket sends a FIN packet.

By default (allowHalfOpen == false) the socket will destroy its file descriptor once it has written out its pending write queue. However, by setting allowHalfOpen == true the socket will not automatically end() its side allowing the user to write arbitrary amounts of data, with the caveat that the user is required to end() their side now.

Event: 'error'#
Error object
Emitted when an error occurs. The 'close' event will be called directly following this event.

Event: 'close'#
had_error Boolean true if the socket had a transmission error
Emitted once the socket is fully closed. The argument had_error is a boolean which says if the socket was closed due to a transmission error.

星期日, 3月 15, 2015

Uniform Server, Ripple Chrome Extension, Nodejs, ExpressJS, Swig, D3JS, RickShaw.js, Socket.io

http://www.uniformserver.com/

The Uniform Server is a lightweight server solution for running a web server under the WindowsOS. Less than 10MiB, it includes the latest versions of Apache2, Perl5, PHP5, MySQL5, phpMyAdmin and more. No installation required! No registry dust! Just unpack and fire up

Ripple Chrome Extension
可以模擬各種行動裝置大小,方便做行動網頁設計。


Node Swig 樣版
http://paularmstrong.github.io/swig/
採用跟 python django同樣的模版語言,容易維護。

MangoDB
儲存資料

d3.js

Express JS

rickshaw.js
http://code.shutterstock.com/rickshaw/
利用d3來做各種線圖。


Socket.io
利用websocket來做browser即時通訊

Windows Terminal, Command Line Shell Suggestion

ConEmu is opensource console emulator with tabs, which represents multiple consoles and simple GUI applications as one customizable GUI window.

推薦,中文輸入記得進入Setting中,把monospace選項取消,中文字就可以正常移動。

星期一, 3月 09, 2015

apache2.conf vs site-avaliable virtualhost configuration 設定

Reference:

http://stackoverflow.com/questions/22579159/apache2-conf-vs-site-avaliable-virtualhost-configuration

結論:  site-available 裡的設定最後還是會被merge到apache2.conf。二種方式是一樣的,只是 site-available的方式比較方便管理多個網站。

星期二, 3月 03, 2015

星期六, 2月 28, 2015

su, sudo,sudo su

su
登入為root,需要輸入root的密碼,比較不安全。(fedora及大部份的distribution,建立系統時會建立root帳號及密碼)

sudo command
短暫取得root權限,執行後面指定的command命令。需要輸入自已己帳號的密碼,而不是root帳號的密碼。只有加入到sudoer檔案的帳號才可以執行sudo命令。優點,安全性較高,可避免登入到root帳號後,不小心下錯指令,導致大範圍失誤。( 在ubuntu系統中,預設不會建立root密碼啟用root登入功能。當需要執行root權限所需的命令時,都採用sudo command的方式來做。)

sudo su
同su功能,切換為root。即請sudo執行su命令,所不同的是輸入自己帳號密碼,而不是root密碼。原理當使用者輸入自己密碼後即取得root權限,此時再執行su,等同執行su root,就會切換成root身份。
為什麼要多此一舉? 因為在ubuntu系統中,預設不會建立root密碼啟用root登入功能。當需要執行root權限所需的命令時,都採用sudo command的方式來做。有時候要執行很多系統命令時,可以採用sudo su方式,切換到root身份後再下指令會比較方便。

WireShark 封包抓取學習好文章

Good Reference:

一站式学习Wireshark(一):Wireshark基本用法
https://community.emc.com/thread/194901

十大免费又好用的网络分析工具
https://community.emc.com/message/788944#788944

apache 權限設定好文章

Reference:
What permissions should my website files/folders have on a Linux webserver?

http://serverfault.com/questions/357108/what-permissions-should-my-website-files-folders-have-on-a-linux-webserver

星期三, 2月 25, 2015

iOS, Android, phonegap Push Notification

Good Reference:

iOS APNS 訊息推播 - Apple Push Notification Service 介紹
http://blog.toright.com/posts/2806/ios-apns-%E8%A8%8A%E6%81%AF%E6%8E%A8%E6%92%A5-apple-push-notification-service-%E4%BB%8B%E7%B4%B9.html

如何透過 PHP 發送 Apple Notification Push
http://blog.toright.com/posts/2846/%E5%A6%82%E4%BD%95%E9%80%8F%E9%81%8E-php-%E7%99%BC%E9%80%81-apple-notification-push.html


Android Push Notification推播通知訊息給Android客戶端
http://magiclen.org/android-push-notification/

Android GCM (Google Cloud Messaging):Server Side
http://blog.kenyang.net/2013/07/android-gcm-google-cloud.html

Android GCM (Google Cloud Messaging):Android Client Side
http://blog.kenyang.net/2013/07/android-gcm-google-cloud_1.html

Django and Facebook API

Good Reference:
Signing Up and Signing In: Users in Django with Django-AllAuth
https://speakerdeck.com/tedtieken/signing-up-and-signing-in-users-in-django-with-django-allauth

使用allauth套件對新手最方便

DjangoCon 2013 - Rapid prototyping and communicating with clients
http://www.slideshare.net/katychuang/kat-rapid-prototyping
django網站規劃好文章