Web Development on Coinidea's Blog

[LNMP] Nginx access_log Daily Rotation

Wed, 10 Apr 2019 02:30:15 +0000

Previously, I had a service using Nginx as a reverse proxy. Due to the high volume of traffic, I wanted to do some simple access statistics at the Nginx layer. I found that Nginx’s default access_log writes everything to a single file, which is very inconvenient. After some research, many people suggested using shell scripts to solve this problem. Finally, I found a better solution – Nginx supports the following configuration:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23


server {
 listen 8082;
 server_name your ip;

 if ($time_iso8601 ~ "^(\d{4})-(\d{2})-(\d{2})T(\d{2}):(\d{2}):(\d{2})"){
 set $year \$1;
 set $month \$2;
 set $day \$3;
 set $hour \$4;
 set $minutes \$5;
 set $seconds \$6;
 }

 access_log logs/8082_$year-$month-$day.log;
 error_log logs/8082.error;

 # Forward all requests
 location / {
 proxy_set_header Host $host;
 proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
 proxy_pass http://localhost:8082;
 }
}

[LNMP]Memory Optimization

Wed, 19 Dec 2018 03:01:19 +0000

Problem Description

Recently, my server ran into an issue where it would frequently kill my Tomcat process when memory ran low. Since the overall traffic isn’t very high, LNMP and Tomcat are currently running on the same machine, but this caused Tomcat-related Java applications to become unavailable.

1

free -m

Checking the memory usage, I found the operating system had only about 60MB of memory left – almost completely used up. This situation persisted for about a month, and it was a constant source of frustration. I considered adding more memory but thought it was too expensive. When resources are insufficient, there are generally two approaches: the first is to add resources, and the second is to optimize existing resources. For someone on a budget like me, the second approach is usually my first choice.

Initially, I thought Tomcat itself or some other process was consuming too much memory. After running:

1

top

I found that php-fpm processes were often the ones consuming the most memory.

To view the top 40 memory-consuming processes:

1

ps auxw | head -1; ps auxw | sort -rn -k4 | head -40

I found that php-fpm dominated the top 40 processes, with memory usage ranging from 2% to 7%. The problem was now quite clear – the culprit was php-fpm.

Solution

You can control the number of php-fpm child processes by configuring the pm.max_children property. First, open the php-fpm configuration file:

1

vim /etc/php-fpm.d/www.conf

pm.max_children defaults to 50. Each process uses 1%-2.5% of memory, which adds up to consuming most of the memory. Try reducing the value – I set mine to 25 (also referencing other blogs). At the same time, check the following two properties:

pm.max_spare_servers : This value represents the maximum number of idle processes. If idle processes exceed this value, they will be cleaned up.
pm.min_spare_servers : The minimum number of idle processes. If idle processes fall below this value, new child processes are created.

Neither of these values should exceed the pm.max_children value. Typically, pm.max_spare_servers is set to 60%-80% of the pm.max_children value.

Finally, restart php-fpm:

1

service php-fpm restart

Check the memory again:

1

free -m

I found there was about 600MB of free memory remaining. Even after starting Tomcat, there was still about 300MB to spare.

Reference: https://www.jb51.net/article/129528.htm

1
2

Hope this helps! If you need anything else, feel free to let me know.

[Web Development]HTML5 Geolocation Inaccuracy Issues

Thu, 04 Jan 2018 07:54:52 +0000

 1
 2
 3
 4
 5
 6
 7
 8
 9
10


if (navigator.geolocation) {
 navigator.geolocation.getCurrentPosition(getPosition, getPositionError, {
 // Indicates the browser should obtain high-accuracy position, default is false
 enableHighAccuracy: true,
 // Specifies the timeout for obtaining geolocation, default is unlimited, in milliseconds
 timeout: 5000,
 // Maximum cache age — when repeatedly obtaining geolocation, this parameter specifies how long before re-fetching the position
 maximumAge: 3000
 });
}

The above code is commonly seen in H5-related geolocation implementations. It uses the device’s own GPS for positioning, but through testing I found it was very inaccurate. The desktop version would directly throw errors.

On mobile, if you use Baidu’s API set, the location drifts significantly. Based on observation, it uses cell tower positioning.
I then thought — is Baidu really this bad? So I tried the native code above, and found it frequently errors out:
Error using Geolocation method: error.POSITION_UNAVAILABLE
After investigating for a long time, the desktop version just wouldn’t work. Finally I figured it out — key point:

GPS services now strongly prefer HTTPS. HTTP is increasingly unsupported. Chrome even explicitly states that geolocation requires HTTPS.

After testing again with HTTPS, the accuracy was incredible — within 10 meters.

[MySQL]Automatic MySQL Backup

Tue, 28 Nov 2017 02:59:06 +0000

Create the backup.sh Script

Create a new backup.sh file with the following content:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12


db_user="Your database user"
db_passwd="Your password"
db_name="Your database name"

# Backup file storage directory — you need to change this directory
backup_dir="/home/mysqlbackup"

# Date format for backup files (dd-mm-yyyy)
time="$(date +"%Y%m%d%H%M%S")"

# Execute the backup command
mysqldump -h localhost -u $db_user -p$db_passwd --all-databases --sock=/var/lib/mysql/mysql.sock > "$backup_dir/db_$time.sql"

Then, grant execution permissions and run the script:

1
2


chmod +x ./backup.sh
bash -x ./backup.sh

If you get a socket error, you need to specify sock. See: http://blog.csdn.net/seteor/article/details/18356079.

Create a Scheduled Task

Edit the crontab:

1

crontab -e

In the editor, add the following line:

1

00 23 * * * /home/mysqlbackup/backup.sh

This command will execute backup.sh every day at 23:00, backing up the database to /home/mysqlbackup.

View scheduled tasks:

1

crontab -l

[Push Notifications]Push Notifications Based on Umeng Service

Sun, 19 Nov 2017 11:48:10 +0000

When the server interacts with mobile clients, it often needs to push messages from the server to the user’s mobile device. Umeng provides such a push notification service — UPush.

Download the SDK code. Based on the demo, this article provides a simple ThinkPHP integration.

1
2
3
4
5
6


$appkey = "your app key";
$secret = "your app secret";
import('@.ORG.Notification');
$notification = new Notification($appkey, $secret);
$custom_data = array($key=>$value);
$notification->sendIOSBroadcast($your_title, $custom_data);

Notification.class.php is the demo.php from the SDK, encapsulated into ThinkPHP’s ORG folder.
If you have any questions, feel free to leave a comment below. I typically reply once a day.

[ThinkPHP]Table Locking Under Concurrency in ThinkPHP

Fri, 17 Nov 2017 03:10:04 +0000

Recently I needed to build a registration system that doesn’t require very high concurrency, but there’s still a possibility of multiple users registering at the same time. Since registration involves ordering, without row locking or table locking, the following situation can occur.
Example:
Currently the system only has user A registered with order 1, denoted as (A, 1).
At some moment t, users B, C, and D register simultaneously. The backend receives all three requests almost at the same time, queries the registration table, finds only (A, 1), and inserts (B, 2), (C, 2), (D, 2) into the table, resulting in:

1
2
3
4


(A, 1)
(B, 2)
(C, 2)
(D, 2)

My initial idea was: instead of recording the order number, I would record the insertion timestamp, which would give:

1
2
3
4


(A, ta)
(B, tb)
(C, tc)
(D, td)

ta, tb, tc, td are almost impossible to be equal.
However, this approach has three unsolvable issues:

There is a (very low) probability that timestamps could be equal
It’s not intuitive and requires sorting by timestamp
Registration has a capacity limit (e.g., 40 spots) and needs to immediately report whether the registration was successful

The current solution is to use database locking.
After researching online, ThinkPHP supports both pessimistic and optimistic locking. Given the low traffic of the target system, pessimistic locking is sufficient.

MyISAM can only lock tables, while InnoDB supports row-level locking. Table locking is sufficient for the target system.
The official documentation provides this table locking approach:

1

$User->lock(true)->save($data);// Use pessimistic locking

However, the target system needs to perform a series of operations, so the table locking code used is:

1
2
3
4


M()->query("lock tables yourtable write");
// TODO
// your code
M()->query("unlock tables");

The actual results worked quite well [real dates have been redacted]:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40


??-??-?? 10:00:00 1
??-??-?? 10:00:00 2
??-??-?? 10:00:00 3
??-??-?? 10:00:00 4
??-??-?? 10:00:01 5
??-??-?? 10:00:01 6
??-??-?? 10:00:01 7
??-??-?? 10:00:01 8
??-??-?? 10:00:02 9
??-??-?? 10:00:02 10
??-??-?? 10:00:02 11
??-??-?? 10:00:02 12
??-??-?? 10:00:02 13
??-??-?? 10:00:02 14
??-??-?? 10:00:03 15
??-??-?? 10:00:03 16
??-??-?? 10:00:03 17
??-??-?? 10:00:05 18
??-??-?? 10:00:06 19
??-??-?? 10:00:07 20
??-??-?? 10:00:08 21
??-??-?? 10:00:10 22
??-??-?? 10:00:15 23
??-??-?? 10:00:17 24
??-??-?? 10:00:19 25
??-??-?? 10:00:19 26
??-??-?? 10:00:24 27
??-??-?? 10:00:25 28
??-??-?? 10:00:34 29
??-??-?? 10:00:35 30
??-??-?? 10:00:38 31
??-??-?? 10:01:06 32
??-??-?? 10:01:11 33
??-??-?? 10:01:11 34
??-??-?? 10:01:17 35
??-??-?? 10:01:18 36
??-??-?? 10:02:27 37
??-??-?? 10:02:38 38
??-??-?? 10:02:39 39
??-??-?? 10:02:57 40

[MySQL] is marked as crashed and should be repaired

Fri, 28 Apr 2017 08:45:59 +0000

A website’s articles were not displaying. My best guess was that the database was down. After connecting, the database was accessible, but the articles table could not be opened, with the error:

Table is marked as crashed and should be repaired.

I was startled for a moment. Although I had scheduled backups, could the database really be broken? Or was there a security vulnerability?

Based on the literal meaning, I should repair the table.
After looking it up, Navicat happens to have this feature, so I tried to repair it.
Error:
mysql table is marked as crashed and last (automatic?) repair failed

The automatic repair failed.

Prepared to check the server logs
3.1 MySQL was running. Restarted it – normal, but the articles table still couldn’t be read.
3.2 Since the table uses the ISAM engine, I tried the method found online:

myisamchk -r /your/table/data/path.MYI

Still failed.
3.3 While about to check the logs, I noticed that the disk where MySQL data was stored had run out of space – 0 bytes remaining.

3.4 Obviously, after cleaning up and moving unnecessary files, I used Navicat to repair the table, and it succeeded on the first try.

It’s been a long time since I wrote a blog post. I need to keep at it.

[HTML5] Local Image Reading and Cropping with HTML5

Tue, 03 Jan 2017 08:18:50 +0000

You need to use HTML5’s FileReader. Here, image-file is an input type=file file browser. If you need to restrict it to images or camera input:
1

<input type="file" accept="image/*" capture="camera" id="image-file" name="image-file" />

When the user selects an image, bind a change event to the input file to automatically read the local file.

1
2
3
4
5
6
7


var image_file = document.getElementById("image-file");
if (typeof FileReader === 'undefined') {
 image.innerHTML = "Sorry, your browser does not support FileReader";
 image_file.setAttribute('disabled', 'disabled');
} else {
 image_file.addEventListener('change', readFile, false);
}

image is an img tag with the id image. The readFile function reads the image and automatically center-crops it to a width that adapts to the screen and a fixed height of 180px. The cropping process uses Canvas.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33


function readFile() {
 var file = this.files[0];
 if (!/image\/\w+/.test(file.type)) {
 alert("The file must be an image!");
 return false;
 }
 var reader = new FileReader();
 reader.readAsDataURL(file);
 reader.onload = function(e) {
 var image = document.getElementById("image");
 var img = new Image();
 img.src = this.result;
 img.onload = function() {
 var x = 0, y = 0, width = img.width, height = img.height;
 if (img.width / window.innerWidth > img.height / 180) {
 width = window.innerWidth * img.height / 180;
 x = (img.width - width) / 2;
 } else {
 height = 180 * img.width / window.innerWidth;
 y = (img.height - height) / 2;
 }

 var canvas = document.createElement('canvas');
 canvas.width = window.innerWidth;
 canvas.height = 180;
 var ctx = canvas.getContext('2d');
 ctx.drawImage(img, x, y, width, height, 0, 0, canvas.width, canvas.height);
 var data = canvas.toDataURL();

 image.innerHTML = '<img class="img-responsive" style="width:100%; height: 180px;" src="' + data + '" alt="" />';
 }
 }
}

Through these steps, the user can select an image, automatically read the local file, and then center-crop it to a width that adapts to the screen with a fixed height of 180px.

[MySQL] Removing Carriage Returns and Line Feeds from Fields

Sat, 31 Dec 2016 06:32:25 +0000

1

update table set item=replace(replace(item, char(10), ''), char(13), '') where condition;

table: table name
item: field name
condition: query condition

[JavaScript] Deep Copy in JavaScript

Thu, 08 Dec 2016 09:42:07 +0000

1

var obj1 = JSON.parse(JSON.stringify(obj2));

The above method uses JSON round-trip conversion for deep copying arrays or objects. It has some limitations but works for most cases. For details, see
https://segmentfault.com/a/1190000000501320

[PHP] Filtering All Punctuation with PHP Regular Expressions

Thu, 08 Dec 2016 09:40:15 +0000

 1
 2
 3
 4
 5
 6
 7
 8
 9
10


// Filter English punctuation
$word = preg_replace("/[[:punct:]]/i", " ", $word);

// Filter Chinese punctuation
mb_regex_encoding('utf-8');
$char = "。、！？：；﹑•＂…''""〝〞∕¦‖—　〈〉﹞﹝「」‹›〖〗】【»«』『〕〔》《﹐¸﹕︰﹔！¡？¿﹖﹌﹏﹋＇´ˊˋ―﹫︳︴¯＿￣﹢﹦﹤‐­˜﹟﹩﹠﹪﹡﹨﹍﹉﹎﹊ˇ︵︶︷︸︹︿﹀︺︽︾ˉ﹁﹂﹃﹄︻︼（）";
$word = mb_ereg_replace("[".$char."]", " ", $word, "UTF-8");

// Filter consecutive spaces
$word = preg_replace("/\s+/", " ", $word);

mb_ereg_replace is used for Chinese character replacement. When using regex, you don’t need to add / / delimiters.

[PHP] Large-Scale Data Import and Export with ThinkPHP, PHP, and MySQL

Thu, 08 Dec 2016 09:37:45 +0000

Large-Scale Data Import and Export

Note: “Large-scale” here refers to importing and exporting 100,000+ records from text files (including Office documents such as docx, xlsx).

MySQL has two very common table types: MyISAM and InnoDB. There are many comparisons of these two types online. The main difference is that InnoDB supports transactions and row-level locking, while MyISAM does not. MyISAM is relatively more efficient, and each has its own advantages and disadvantages.

Use Cases for MyISAM and InnoDB

MyISAM is suitable for:

Frequent COUNT calculations;
Infrequent inserts but frequent queries;
No transactions needed.

InnoDB is suitable for:

High reliability requirements or transaction support needed;
Frequent table updates and queries with high likelihood of table locking.

Differences Between MyISAM and InnoDB

MyISAM does not support advanced features like transactions, while InnoDB does.
MyISAM tables do not support foreign keys.
During write operations (insert, update, delete), MyISAM locks the entire table, while InnoDB locks only the affected rows.
When your database has heavy writes, updates, and deletions with relatively few queries, or when data integrity requirements are high, choose InnoDB. When your database is primarily read-heavy with relatively few updates and writes, and data integrity requirements are less strict, choose MyISAM, as its query performance is faster than InnoDB.

In summary, InnoDB is a good choice for this scenario.

Optimization Methods for Inserting Large Amounts of Data

When PHP needs to insert large amounts of data, inserting one record at a time results in severe I/O performance bottlenecks.

Demo Code:

1
2
3
4
5
6
7


$params = array('value' => '50');
set_time_limit(0);
echo date("H:i:s");
for($i = 0; $i < 2000000; $i++) {
 $connect_mysql->insert($params);
}
echo date("H:i:s");

There are two methods to solve this problem:

1. Using Transactions (InnoDB supported)

For example, you can commit a transaction every 10,000 or 100,000 records, which greatly reduces I/O.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12


echo date("H:i:s");
$connect_mysql->query('BEGIN');
$params = array('value' => '50');
for($i = 0; $i < 2000000; $i++) {
 $connect_mysql->insert($params);
 if($i % 100000 == 0) {
 $connect_mysql->query('COMMIT');
 $connect_mysql->query('BEGIN');
 }
}
$connect_mysql->query('COMMIT');
echo date("H:i:s");

2. SQL Statement Concatenation

This is a MySQL feature that uses optimized SQL statements: concatenate SQL statements using insert into table () values (),(),(),() and insert them all at once. If the string is too long, you need to configure MySQL by running:

1

SET GLOBAL max_allowed_packet = 2 * 1024 * 1024 * 10;

1
2
3
4
5
6


$sql = "INSERT INTO twenty_million (value) VALUES";
for($i = 0; $i < 2000000; $i++) {
 $sql .= "('50'),";
}
$sql = substr($sql, 0, strlen($sql) - 1);
$connect_mysql->query($sql);

The first method is more commonly used, while the second is mostly for test data.

By default, PHP has a memory limit of 128MB. When ThinkPHP exceeds this memory limit, you can either increase PHP’s default maximum memory or simply remove the APP_DEBUG constant definition in the entry file or set it to false. This will save a lot of memory. Of course, you can also use transaction mode to commit every n records. ThinkPHP’s addAll function can also be used for batch inserts, though it has quite a few bugs.

Starting and Committing Transactions in ThinkPHP

1
2


$table->startTrans();
$table->commit();

When working with Office files, PHPWord and PHPExcel are recommended.

[Web Development] Migrating DedeCMS to Discuz Portal

Fri, 18 Nov 2016 05:14:05 +0000

I recently needed to migrate from DedeCMS (dedecms) to Discuz Portal. The most important challenge was how to migrate categories and a large number of articles.

The database tables involved are:

dede_addonarticle - Article table
dede_arctype - Category table
dede_relation - Old-to-new category mapping table
pre_portal_article_content - Article content table
pre_portal_article_title - Article title table
pre_portal_article_count - Article view count table

1. Preprocess the old table, merge `typename`

1
2
3


UPDATE dede_addonarticle, dede_arctype
SET dede_addonarticle.typename = dede_arctype.typename
WHERE dede_arctype.id = dede_addonarticle.typeid;

2. Merge `click`, `title`, `writer`, `pubdate`, `keywords`, `description`

1
2
3
4
5
6
7
8
9


UPDATE dede_addonarticle, dede_archives
SET
 dede_addonarticle.click = dede_archives.click,
 dede_addonarticle.title = dede_archives.title,
 dede_addonarticle.writer = dede_archives.writer,
 dede_addonarticle.pubdate = dede_archives.pubdate,
 dede_addonarticle.keywords = dede_archives.keywords,
 dede_addonarticle.description = dede_archives.description
WHERE dede_archives.id = dede_addonarticle.aid;

3. Generate new IDs in `relation`

1
2
3


UPDATE dede_relation, pre_portal_category
SET dede_relation.catid = pre_portal_category.catid
WHERE dede_relation.new = pre_portal_category.catname;

4. Populate `catid` based on `typename` and `relation`

1
2
3


UPDATE dede_addonarticle, dede_relation
SET dede_addonarticle.catid = dede_relation.catid
WHERE dede_addonarticle.typename LIKE CONCAT("%", dede_relation.olds, "%");

5. Import into Discuz Portal Tables

Insert into `pre_portal_article_content` (Article content table)

1
2
3


INSERT INTO pre_portal_article_content(aid, title, content, dateline)
SELECT aid, title, body, pubdate AS dateline
FROM dede_addonarticle;

Insert into `pre_portal_article_title` (Article title table)

1
2
3


INSERT INTO pre_portal_article_title(aid, username, title, author, summary, dateline)
SELECT aid, 'admin', title, writer AS author, description AS summary, pubdate AS dateline
FROM dede_addonarticle;

Insert into `pre_portal_article_count` (Article view count table)

1
2
3


INSERT INTO pre_portal_article_count(aid, cid, viewnum)
SELECT aid, catid, click
FROM dede_addonarticle;

[Web Development] Postman Plugin

Fri, 18 Nov 2016 04:59:52 +0000

Recommending a Web Development Debugging Tool

It is especially useful for reading, modifying, and adding Cookie, Session, Header, GET, and POST data during network transmission. There are many blog posts about it online, so I won’t go into detail here.

Postman has three versions (as of November 2016):

Postman for Chrome
Postman for Mac
Postman for Linux (Beta)

The Chrome plugin is recommended. However, it’s not easy to download in China. Here is the official Google Chrome plugin download link as of November 2016:

https://chrome.google.com/webstore/detail/postman/fhbjgbiflinjbdggehcddcbncdddomop/related

And a very old version download: Postman-REST-Client_v0.8.4.14

Link: http://pan.baidu.com/s/1bo4nzQJ Password: h0w7

[PHP] call to undefined function mb_convert_encoding

Tue, 02 Aug 2016 03:37:01 +0000

Solution on Windows

In php.ini, enable the mbstring library: remove the semicolon before extension=php_mbstring.dll.

Solution on Linux

First, install the php-mbstring library:
1

yum install php-mbstring
Then, edit the /etc/php.ini file and add the following line:
1

extension=mbstring.so
Save and restart the httpd service:
1

service httpd restart

[IIS] File Formats (Fonts) Failing to Load on IIS Server

Thu, 28 Jul 2016 09:07:08 +0000

While deploying an IIS service recently, I noticed that some .ttf, .otf, .woff and other font files were not being loaded, even though js, css, images and other files loaded successfully. The reason is that IIS cannot serve files that don’t have a default MIME mapping.

The solution is simple – add the corresponding MIME type mappings. You can add them directly in the IIS panel under MIME Types. For specific type information, refer to the W3CSchool MIME Reference: http://www.w3school.com.cn/media/media_mimeref.asp

MIME mappings for font files:

1
2
3
4
5
6


<mimeMap fileExtension=".woff" mimeType="application/x-font-woff" />
<mimeMap fileExtension=".woff2" mimeType="application/x-font-woff2" />
<mimeMap fileExtension=".ttf" mimeType="application/x-font-truetype" />
<mimeMap fileExtension=".svg" mimeType="image/svg+xml" />
<mimeMap fileExtension=".otf" mimeType="application/x-font-opentype" />
<mimeMap fileExtension=".eot" mimeType="application/vnd.ms-fontobject" />

Font Subsetting for Web and Mobile

Fri, 22 Jul 2016 02:35:14 +0000

In web and mobile frontend development, developers often want to use non-system fonts, which are typically in TTF, OTF, and other formats.

English fonts are usually quite small, but Chinese fonts can range from 1-2 MB on the small end to over 10 MB for larger ones. This is unacceptable for web or mobile applications. This is where font subsetting comes in – you only need the specific characters used in a particular frontend interface, so you export just those characters from the font into a new TTF file, leaving out the unnecessary ones.

I highly recommend a Chinese font subsetting tool: TTF Font Library Subsetting and Conversion Tool

Download link: http://pan.baidu.com/s/1skI7FiP Password: qb27

Note that the target font to be subsetted must first be installed in the Windows system.

[Bootstrap] Triggering dropdown-toggle on Mouse Hover in Bootstrap

Fri, 27 May 2016 03:18:37 +0000

Using the `bootstrap-hover-dropdown` Plugin

1. Download `bootstrap-hover-dropdown`

Download link: http://www.bootcdn.cn/bootstrap-hover-dropdown/

Of course, you can also use the CDN version directly.

2. Include `bootstrap-hover-dropdown.js`

In your HTML file, make sure to include bootstrap-hover-dropdown.js. Typically, you can place it in a <script> tag at the bottom of the page:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20


<!DOCTYPE html>
<html lang="en">
<head>
 <meta charset="UTF-8">
 <meta name="viewport" content="width=device-width, initial-scale=1.0">
 <title>Bootstrap Hover Dropdown Example</title>
 <!-- Include Bootstrap CSS -->
 <link href="https://stackpath.bootstrapcdn.com/bootstrap/4.5.2/css/bootstrap.min.css" rel="stylesheet">
</head>
<body>
 <!-- Your content -->

 <!-- Include jQuery -->
 <script src="https://code.jquery.com/jquery-3.5.1.min.js"></script>
 <!-- Include Bootstrap JS -->
 <script src="https://stackpath.bootstrapcdn.com/bootstrap/4.5.2/js/bootstrap.min.js"></script>
 <!-- Include bootstrap-hover-dropdown.js -->
 <script src="path/to/bootstrap-hover-dropdown.js"></script>
</body>
</html>

Find the element with the dropdown-toggle class in your HTML file and modify it as follows:

1

<a href="#" class="dropdown-toggle" data-toggle="dropdown">Dropdown</a>

Change it to:

1

<a href="#" class="dropdown-toggle" data-hover="dropdown" data-close-others="true">Dropdown</a>

You can also add a parameter data-delay="1000", which sets the delay before the dropdown disappears (default is 500 milliseconds). For example:

1

<a href="#" class="dropdown-toggle" data-hover="dropdown" data-close-others="true" data-delay="1000">Dropdown</a>

4. Removing the Delay

If you want to completely remove the delay, you can directly modify the configuration in the bootstrap-hover-dropdown.js file, changing delay: 500 to delay: 0.

Find and modify the following code:

1

delay: 500

Change it to:

1

delay: 0

This will remove the dropdown menu delay.

[Node.js] Building an Image Server with UEditor + Node.js + SeaweedFS

Wed, 18 May 2016 03:16:01 +0000

1. UEditor + Node.js Image Upload

UEditor is a feature-rich open-source rich text editor by Baidu.
Download link: http://ueditor.baidu.com/website/download.html
It currently provides versions for PHP, ASP, .Net, and JSP. UEditor is primarily built with frontend HTML, CSS, and JS. The reason it is further divided by server-side language is, in my understanding, mainly for the image upload functionality which involves server interaction.
Among the many versions, there is no Node.js version. Below I will explain how to convert the PHP version of UEditor into a Node.js version.
After examining all requests in the PHP version, I found that the action parameter values include config (configuration file), uploadimage (image upload), listimage (online management), and catchimage (image capture). So we just need to rewrite these 4 request handlers to meet our needs.

1.1 Modify the `serverUrl` property in UEditor’s `ueditor.config.js`:

1

serverUrl: '/ue/uploads'

1.2 Rename `ueditor/php/config.json` to `config.js` and move it to the `ueditor` directory.

1.3 Next, write the four corresponding actions on the Node.js side. The image upload uses the `connect-busboy` middleware. Code as follows:

 1
 2
 3
 4
 5
 6
 7
 8
 9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104


// Image upload
var path = require('path');
var uploadsPath = path.resolve('public/uploads') + '/'; // Path to store images
var busboy = require('connect-busboy');
app.use(busboy({
 limits: {
 fileSize: 10 * 1024 * 1024 // 10MB
 }
}));
var action = {
 // Upload image
 uploadimage: function (req, res) {
 var fstream;
 req.pipe(req.busboy);
 req.busboy.on('file', function (fieldname, file, filename, encoding, mimetype) {
 var filesize = 0;
 var ext = path.extname(filename);
 var newFilename = (new Date() - 0) + ext;
 fstream = fs.createWriteStream(uploadsPath + newFilename);
 file.on('data', function (data) {
 filesize = data.length;
 });
 fstream.on('close', function () {
 console.log(JSON.stringify({
 "originalName": filename,
 "name": newFilename,
 "url": '/uploads/' + newFilename,
 "type": ext,
 "size": filesize,
 "state": "SUCCESS"
 }));
 res.send(JSON.stringify({
 "originalName": filename,
 "name": newFilename,
 "url": '/uploads/' + newFilename,
 "type": ext,
 "size": filesize,
 "state": "SUCCESS"
 }));
 });
 file.pipe(fstream);
 });
 },
 // Get configuration file
 config: function (req, res) {
 return res.redirect('/js/UEditor/config.js');
 },
 // Online management
 listimage: function (req, res) {
 fs.readdir(uploadsPath, function (err, files) {
 var total = 0, list = [];
 files.sort().splice(req.query.start, req.query.size).forEach(function (a, b) {
 /^.+\..+$/.test(a) &&
 list.push({
 url: '/uploads/' + a,
 mtime: new Date(fs.statSync(uploadsPath + a).mtime).getTime()
 });
 });
 total = list.length;
 res.json({state: total === 0 ? 'no match file' : 'SUCCESS', list: list, total: total, start: req.query.start});
 });
 },
 // Catch image (save image to server when pasting)
 catchimage: function (req, res) {
 var list = [];
 req.body.source.forEach(function (src, index) {
 http.get(src, function (_res) {
 var imagedata = '';
 _res.setEncoding('binary');
 _res.on('data', function (chunk) {
 imagedata += chunk
 });
 _res.on('end', function () {
 var pathname = url.parse(src).pathname;
 var original = pathname.match(/[^/]+\.\w+$/g)[0];
 var suffix = original.match(/[^\.]+$/)[0];
 var filename = Date.now() + '.' + suffix;
 var filepath = uploadsPath + 'catchimages/' + filename;
 fs.writeFile(filepath, imagedata, 'binary', function (err) {
 list.push({
 original: original,
 source: src,
 state: err ? "ERROR" : "SUCCESS",
 title: filename,
 url: '/uploads/catchimages/' + filename
 });
 })
 });
 })
 });
 var f = setInterval(function () {
 if (req.body.source.length === list.length) {
 clearInterval(f);
 res.json({state: "SUCCESS", list: list});
 }
 }, 50);
 }
};
app.get('/ue/uploads', function (req, res) {
 action[req.query.action](req, res);
});
app.post('/ue/uploads', function (req, res) {
 action[req.query.action](req, res);
});

The above mainly references this blog: http://www.xiaoboy.com/detail/1341545081.html

2. GoLang Installation and Configuration

In section 1, UEditor uploads images to the /public/uploads/ folder on the Node.js server. If you need an image server with high storage, portability, and scalability, you need to set up a dedicated server. This article uses the open-source distributed image (file) server seaweedfs.
GitHub link: https://github.com/chrislusf/seaweedfs
This server is written in GoLang, so you need to install and configure GoLang. Download link for China: http://www.golangtc.com/download – choose the appropriate package for your operating system. This article demonstrates installation on Windows 7 x64. After a straightforward installation, you need to configure the following environment variables:

1
2
3


GOROOT=D:\Go16
GOPATH=D:\ImageServer\seaweedfs
PATH=D:\Go16\bin

The GOPATH environment variable is very important – your own code must be placed in the directory configured by this variable for the Go compiler to find and compile it.

3. Compiling and Running seaweedfs

Clone the seaweedfs code from GitHub, then follow the makefile instructions or run the makefile directly. In the seaweedfs directory, run the following commands in the console:

1
2
3
4
5
6


go clean -i -v ./go/weed/

rm -f weed #for linux

go get -v -d ./go/weed
go build -v -o weed ./go/weed

Some dependencies may fail to download during go get and need to be downloaded manually and placed in the \seaweedfs\src directory.
Here is a recommended Go package download site: https://gopm.io/download?pkgname=golang.org/x/net
After building, a weed binary will be generated in the seaweedfs directory. You can also rename it to weed.exe. Now you can start weed. Using the local machine as an example:

1

weed master

1

weed volume --dir="./tmp/data1" --max=5 --mserver="localhost:9333" --port=8080 &

Following the GitHub instructions, uploading and downloading images works without issues. At this point, the basic image server setup is complete.

Modify the Node.js image upload code as follows:

 1
 2
 3
 4
 5
 6
 7
 8
 9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113


// Image upload
var path = require('path');
var multipart = require('connect-multiparty');
var multipartMiddleware = multipart();
function curlPostAssign(res, filepath, filename, filesize) {
 request.post(config.url_image + 'dir/assign',{},
 function (error, response, body) {
 if (!error && response.statusCode == 200) {
 assign = eval('(' + body + ')');
 var result = {};
 result.res = res;
 result.json = JSON.stringify({
 "originalName": filename,
 "name": assign.fid,
 "url": '/res?fid=' + assign.fid,
 "type": ext,
 "size": filesize,
 "state": "SUCCESS"
 });
 curlPostWrite(assign.fid, filepath, assign.publicUrl, result);
 var ext = path.extname(filename);
 } else {
 console.log("Image server assign error...")
 }
 });
}
function curlPostWrite(fid, tmp_file, publicUrl, result) {
 var files = [
 {urlKey: "file1", urlValue: tmp_file}
 ]
 var options = {
 host: publicUrl.split(":")[0],
 port: publicUrl.split(":")[1],
 method: "POST",
 path: "/" + fid
 }
 var req = http.request(options, function(res){
 res.on("data", function(chunk){
 })
 })

 req.on('error', function(e){
 console.log('problem with request:' + e.message);
 console.log(e);
 })
 postfile.postFile(files, req, result);
}
var action = {
 // Upload image
 uploadimage: function(req, res) {
 curlPostAssign(res, req.files.upfile.path, req.files.upfile.originalFilename, req.files.upfile.size);
 },
 // Get configuration file
 config: function (req, res) {
 return res.redirect('/js/UEditor/config.js');
 },
 // Online management
 listimage: function (req, res) {
 fs.readdir(uploadsPath, function (err, files) {
 var total = 0, list = [];
 files.sort().splice(req.query.start, req.query.size).forEach(function (a, b) {
 /^.+\..+$/.test(a) &&
 list.push({
 url: '/uploads/' + a,
 mtime: new Date(fs.statSync(uploadsPath + a).mtime).getTime()
 });
 });
 total = list.length;
 res.json({state: total === 0 ? 'no match file' : 'SUCCESS', list: list, total: total, start: req.query.start});
 });
 },
 // Catch image (save image to server when pasting)
 catchimage: function (req, res) {
 var list = [];
 req.body.source.forEach(function (src, index) {
 http.get(src, function (_res) {
 var imagedata = '';
 _res.setEncoding('binary');
 _res.on('data', function (chunk) {
 imagedata += chunk
 });
 _res.on('end', function () {
 var pathname = url.parse(src).pathname;
 var original = pathname.match(/[^/]+\.\w+$/g)[0];
 var suffix = original.match(/[^\.]+$/)[0];
 var filename = Date.now() + '.' + suffix;
 var filepath = uploadsPath + 'catchimages/' + filename;
 fs.writeFile(filepath, imagedata, 'binary', function (err) {
 list.push({
 original: original,
 source: src,
 state: err ? "ERROR" : "SUCCESS",
 title: filename,
 url: '/uploads/catchimages/' + filename
 });
 })
 });
 })
 });
 var f = setInterval(function () {
 if (req.body.source.length === list.length) {
 clearInterval(f);
 res.json({state: "SUCCESS", list: list});
 }
 }, 50);
 }
};
app.get('/ue/uploads', multipartMiddleware, function (req, res) {
 action[req.query.action](req, res);
});
app.post('/ue/uploads', multipartMiddleware, function (req, res) {
 action[req.query.action](req, res);
});

Running result:

[Spring] Creating a Spring Framework Project and Adding Your Own Controller in IntelliJ IDEA

Thu, 28 Apr 2016 07:24:58 +0000

As described in the previous blog post about creating a Maven project, create a Spring Maven project.

POM File Configuration:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62


<project xmlns="http://maven.apache.org/POM/4.0.0" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
 xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://www.w3.org/2001/XMLSchema-instance">
 <modelVersion>4.0.0</modelVersion>
 <groupId>com.coinidea</groupId>
 <artifactId>web</artifactId>
 <packaging>war</packaging>
 <version>1.0-SNAPSHOT</version>
 <name>hive.web Maven Webapp</name>
 <url>http://maven.apache.org</url>
 <properties>
 <spring.version>4.2.5.RELEASE</spring.version>
 </properties>
 <dependencies>
 <dependency>
 <groupId>junit</groupId>
 <artifactId>junit</artifactId>
 <version>3.8.1</version>
 <scope>test</scope>
 </dependency>
 <!--spring-->
 <dependency>
 <groupId>org.springframework</groupId>
 <artifactId>spring-context</artifactId>
 <version>${spring.version}</version>
 </dependency>
 <dependency>
 <groupId>org.springframework</groupId>
 <artifactId>spring-core</artifactId>
 <version>${spring.version}</version>
 </dependency>
 <dependency>
 <groupId>org.springframework</groupId>
 <artifactId>spring-web</artifactId>
 <version>${spring.version}</version>
 </dependency>
 <dependency>
 <groupId>org.springframework</groupId>
 <artifactId>spring-webmvc</artifactId>
 <version>${spring.version}</version>
 </dependency>
 <dependency>
 <groupId>org.springframework</groupId>
 <artifactId>spring-test</artifactId>
 <version>${spring.version}</version>
 </dependency>
 </dependencies>
 <build>
 <finalName>web</finalName>
 </build>
 <repositories>
 <repository>
 <id>people.apache.snapshots</id>
 <url>http://repository.apache.org/content/groups/snapshots-group/</url>
 <releases>
 <enabled>false</enabled>
 </releases>
 <snapshots>
 <enabled>true</enabled>
 </snapshots>
 </repository>
 </repositories>
</project>

Install Tomcat locally. If you don’t know how to install Tomcat, please search for instructions online.

Tomcat Configuration

Run -> Edit Configurations, click “+”, and select Tomcat Local.
In the “Server” tab, configure the local Tomcat path; in “Deployment”, select “+” -> Artifact.
After running, the default Hello World page will pop up.
Create your own Controller
web.xml configuration

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38


<?xml version="1.0" encoding="UTF-8"?>
<web-app xmlns="http://xmlns.jcp.org/xml/ns/javaee"
 xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
 xsi:schemaLocation="http://xmlns.jcp.org/xml/ns/javaee http://xmlns.jcp.org/xml/ns/javaee/web-app_3_1.xsd"
 version="3.1">
 <context-param>
 <param-name>contextConfigLocation</param-name>
 <param-value>/WEB-INF/applicationContext.xml</param-value>
 </context-param>
 <listener>
 <listener-class>org.springframework.web.context.ContextLoaderListener</listener-class>
 </listener>
 <servlet>
 <servlet-name>dispatcher</servlet-name>
 <servlet-class>org.springframework.web.servlet.DispatcherServlet</servlet-class>
 <load-on-startup>1</load-on-startup>
 </servlet>
 <servlet-mapping>
 <servlet-name>dispatcher</servlet-name>
 <url-pattern>/</url-pattern>
 </servlet-mapping>
 <servlet-mapping>
 <servlet-name>default</servlet-name>
 <url-pattern>*.css</url-pattern>
 </servlet-mapping>
 <servlet-mapping>
 <servlet-name>default</servlet-name>
 <url-pattern>*.gif</url-pattern>
 </servlet-mapping>
 <servlet-mapping>
 <servlet-name>default</servlet-name>
 <url-pattern>*.jpg</url-pattern>
 </servlet-mapping>
 <servlet-mapping>
 <servlet-name>default</servlet-name>
 <url-pattern>*.js</url-pattern>
 </servlet-mapping>
</web-app>

It is worth noting that the default servlet-mapping configuration above solves the Spring static file 404 issue.
All other URLs will go through the dispatcher configuration for URL mapping.
7. dispatcher-servlet.xml configuration

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21


<beans xmlns="http://www.springframework.org/schema/beans"
 xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
 xmlns:context="http://www.springframework.org/schema/context"
 xmlns:task="http://www.springframework.org/schema/task"
 xsi:schemaLocation="http://www.springframework.org/schema/beans
 http://www.springframework.org/schema/beans/spring-beans.xsd
 http://www.springframework.org/schema/context http://www.springframework.org/schema/context/spring-context.xsd
 http://www.springframework.org/schema/task http://www.springframework.org/schema/task/spring-task-3.0.xsd"
 default-lazy-init="false">
 <bean class="org.springframework.web.servlet.mvc.method.annotation.RequestMappingHandlerMapping"/>
 <bean class="org.springframework.web.servlet.mvc.method.annotation.RequestMappingHandlerAdapter"/>
 <!-- Enable Spring MVC annotations -->
 <context:annotation-config />
 <!-- Set the package where annotated classes are located -->
 <context:component-scan base-package="com.coinidea.controller"/>
 <!-- View resolver for page paths. prefix: prefix, suffix: suffix -->
 <bean class="org.springframework.web.servlet.view.InternalResourceViewResolver">
 <property name="prefix" value="/WEB-INF/views/"/>
 <property name="suffix" value=".jsp"/>
 </bean>
</beans>

This specifies that the JSP view folder is located at /WEB-INF/views/ and enables Spring’s annotation features. All Controllers will be looked up in the package com.coinidea.controller.
8. FirstHello Controller

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17


package com.cmri.web.controller;

import org.springframework.stereotype.Controller;
import org.springframework.web.bind.annotation.RequestMapping;
import org.springframework.web.servlet.ModelAndView;

@Controller
@RequestMapping("/first")
public class TestController {
 @RequestMapping("hello")
 public ModelAndView view() {
 ModelAndView mv = new ModelAndView();
 mv.addObject("message", "HelloWorld!");
 mv.setViewName("hello");
 return mv;
 }
}

hello.jsp in the views folder:

1
2
3
4
5
6
7
8
9


<%@ page language="java" contentType="text/html; charset=UTF-8" pageEncoding="UTF-8"%>
<html>
<head>
 <title>hello world</title>
</head>
<body>
${message}
</body>
</html>

When you visit http://localhost:8080/first/hello in the browser, it will display:

9. Java Web API Controller
When you need to respond to Ajax requests, add the following annotation to the method:

1
2
3
4
5


@RequestMapping("ajax")
@ResponseBody
public String ajax() {
 return "Hello world";
}

[SAE] SAE MySQL Cross-Application Authorization

Sat, 09 Apr 2016 07:50:12 +0000

As SAE (Sina App Engine) pricing continues to rise, cost savings become increasingly important. Starting March 2016, shared MySQL also began incurring charges, making it important for multiple applications to share a single MySQL instance.

Assume you have multiple applications (APP1, APP2, APP3): Only APP1 retains the shared MySQL, while APP2 and APP3 will use APP1’s MySQL.
Export APP2 and APP3’s MySQL databases as backups, then delete APP2 and APP3’s MySQL instances. This way, APP2 and APP3’s MySQL will be uninitialized and no longer incur charges.
Import APP2 and APP3’s database files (*.sql) into APP1’s MySQL. Before importing, make sure table names don’t conflict.
In APP1’s MySQL settings, click Cross-Application Authorization to select applications under your account or other accounts. Authorize APP2 and APP3 here.
Write a script to display APP1’s MySQL connection information for the authorized applications APP2 and APP3 to use (this will expose database credentials, so proceed with caution and delete the script promptly).

Related code:

1
2
3
4
5
6
7
8
9


<?php
header("Content-type:text/html;charset=utf-8");
echo "Username :".SAE_MYSQL_USER."<br>";
echo "Password :".SAE_MYSQL_PASS."<br>";
echo "Master Host:".SAE_MYSQL_HOST_M."<br>";
echo "Slave Host :".SAE_MYSQL_HOST_S."<br>";
echo "Port :".SAE_MYSQL_PORT."<br>";
echo "Database :".SAE_MYSQL_DB."<br>";
?>

Run the script: The page will display APP1’s MySQL information. Use this information in your other applications (such as APP2 and APP3) to connect to the database.

Notes

Security: Since the script exposes sensitive information such as database username and password, make sure to delete the script promptly after use to prevent information leakage.
Table name conflicts: When importing APP2 and APP3’s database files into APP1’s MySQL, ensure there are no table name conflicts. If conflicts exist, consider renaming tables before importing.
Authorization management: When performing cross-application authorization on the SAE management platform, only authorize necessary applications to avoid unnecessary permission exposure.

Through these steps, you can effectively reduce SAE MySQL costs while ensuring multiple applications can access the shared database normally.

[Node.js] Exception Handling in Node.js Express

Thu, 10 Mar 2016 09:30:18 +0000

Since Node.js is non-blocking, single-process, and single-threaded, once an exception is thrown, the entire service stops, making it very unstable. Here are the solutions:

Make the program robust: Wrap all potentially exception-prone areas with try { } catch(){ }.
Express error handling mechanism: As a commonly used Node.js framework, Express has its own error handling mechanism.

1
2
3
4
5
6


// Express' errorHandler
function errorHandler(err, req, res, next) {
 console.error(err.stack);
 res.status(500).send('Something broke!');
}
app.use(errorHandler);

Use the domain module: First install the domain module with npm install domain

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15


// Domain exception handling
const domain = require('domain');

app.use(function(req, res, next) {
 var reqDomain = domain.create();
 reqDomain.on('error', function(err) {
 console.log(err.stack);
 res.statusCode = 500;
 res.end(err.message + '\n');
 reqDomain.dispose();
 });
 reqDomain.add(req);
 reqDomain.add(res);
 reqDomain.run(next);
});

Use forever to start app.js: Forever acts as a daemon process for Node.js that can start, stop, and restart your app.

1

forever start app.js

Through these methods, you can improve the stability of Node.js services and prevent the entire service from crashing due to uncaught exceptions.

[Memcache] Testing Memcache Connection with PHP

Tue, 08 Mar 2016 12:42:07 +0000

1
2
3
4
5


<?php
$memcache_obj = memcache_connect('127.0.0.1', 11211);
$memcache_obj->add("name", "test");
echo $memcache_obj->get("name");
?>

If you encounter issues using this code, make sure of the following:

Is the Memcache service running: Ensure your Memcache service is running on 127.0.0.1 at port 11211.
Is the PHP Memcache extension installed and enabled: Make sure you have installed and enabled the PHP Memcache extension.

You can install the Memcache extension with the following command:

1

sudo apt-get install php-memcache

And enable the extension in your php.ini file:

1

extension=memcache.so

Restart your web server to apply changes:

1

sudo service apache2 restart

Or, if you’re using Nginx:

1

sudo service nginx restart

If you’re using memcached instead of memcache, make sure to use the correct extension and functions:

1
2
3
4
5
6


<?php
$memcached = new Memcached();
$memcached->addServer('127.0.0.1', 11211);
$memcached->set("name", "test");
echo $memcached->get("name");
?>

In this case, you need to install the php-memcached extension:

1

sudo apt-get install php-memcached

[Javascript] Calling the Printer from a Web Page

Tue, 01 Mar 2016 09:17:03 +0000

Some people reported that clicking doesn’t respond. I’ve tested this on both Chrome and IE and it works:

I think the likely reason is that jQuery was not included. I’ve now added jQuery from a CDN and updated the code to:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20


<html>
<head>
 <meta charset="utf-8">
 <script src="https://cdn.bootcss.com/jquery/3.3.1/jquery.min.js" type="text/javascript"></script>
 <title></title>
</head>
<body>
 <div id='div1'>Put the content to print here</div>
 <div id="div2">div2 content</div>
 <a href="javascript:printHTML('#div1')" target="_self">Print</a>
 <script language="javascript">
 function printHTML(page) {
 var bodyHTML = window.document.body.innerHTML;
 window.document.body.innerHTML = $(page).html();
 window.print();
 window.document.body.innerHTML = bodyHTML;
 }
 </script>
</body>
</html>

Original Javascript code, requires jQuery:

1
2
3
4
5
6


function printHTML(page) {
 var bodyHTML = window.document.body.innerHTML;
 window.document.body.innerHTML = $(page).html();
 window.print();
 window.document.body.innerHTML = bodyHTML;
}

[Javascript] Dynamic Loading and Hiding of UEditor

Fri, 26 Feb 2016 09:31:29 +0000

Based on UEditor, this implements dynamic loading and hiding of UEditor.

Github: https://github.com/hujiulin/UEditorComponent

1. Initial Page

2. Click the Left DIV

3. Click the Right DIV

4. Click an Empty Area on the Page

[Apache] Configuring Apache for Streaming Media

Tue, 05 Jan 2016 10:00:00 +0000

Apache Configuration

Download the mod_flvx.so and mod_h264_streaming.so modules.
Note: Make sure the x86/x64 versions of Apache and the modules match [the modules below are for Apache x86 2.2 and earlier].
Download link: http://pan.baidu.com/s/1dEzYX4d Password: jb0i

Edit apache\conf\httpd.conf, find LoadModule, and add the following at the end of (or near) the LoadModule section:

1
2
3
4


LoadModule flvx_module modules/mod_flvx.so
AddHandler flv-stream .flv
LoadModule h264_streaming_module modules/mod_h264_streaming.so
AddHandler h264-streaming.extensions .mp4

Restart Apache. If Apache fails to start, verify that the module versions match your Apache version.

Video Playback Example

Download ckplayer
Any web video player will work. This example uses ckplayer. Download link:
http://www.ckplayer.com/
Configuration guide:
http://www.ckplayer.com/tool/flashvars.htm
Place the ckplayer folder in the website root directory, and place the sample video example.flv in the website root directory.

Create video.html with the following code:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18


<html>
<meta charset="utf8">
<head>
<title>Streaming Video Playback</title>
</head>
<body>
<div id="a1"></div>
<script type="text/javascript" src="/ckplayer/ckplayer.js" charset="utf-8"></script>
<script type="text/javascript">
 var flashvars={
 f:'http://localhost/example.flv',
 c:0
 };
 var params={bgcolor:'#FFF',allowFullScreen:true,allowScriptAccess:'always',wmode:'transparent'};
 CKobject.embedSWF('/ckplayer/ckplayer.swf','a1','ckplayer_a1','600','400',flashvars,params);
</script>
</body>
</html>

Open a browser and visit http://localhost/video.html. You should be able to drag the video scrollbar.
Chrome result:

IE result:

—-Update 20160106—-
The previous test had issues because the FLV file was too small, so it was likely downloaded locally before playback.
The correct way to verify Apache streaming is:
Prepare a larger MP4/FLV file and replace the video URL above;
Open Chrome (recommended), and enter http://localhost/example.mp4?start=10;
If the video starts playing from the 10-second mark and supports seeking, the Apache configuration is successful.

The ckplayer mentioned above still requires waiting for the download to complete before sequential playback. It is recommended to use video.js instead, though it requires HTML5 browser support.
HTML code:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17


<html>
<head>
 <link href="/video-js/video-js.css" rel="stylesheet">
 <!-- If you'd like to support IE8 -->
 <script src="/video-js/ie8/videojs-ie8.min.js"></script>
</head>
<body>
 <video id="my-video" class="video-js" controls preload="auto" width="auto" height="auto" poster="http://vjs.zencdn.net/v/oceans.png" data-setup="{}">
 <source src="/video/example.mp4" type='video/mp4'>
 <p class="vjs-no-js">
 To view this video please enable JavaScript, and consider upgrading to a web browser that
 <a href="http://videojs.com/html5-video-support/" target="_blank">supports HTML5 video</a>
 </p>
 </video>
 <script src="/video-js/video.js"></script>
</body>
</html>

After testing, the video can be seeked to any position and played even before it is fully loaded. With this, Apache has achieved pseudo-streaming video playback.

[PHP] Encrypting PHP with ZendGuard

Tue, 29 Dec 2015 07:28:22 +0000

Downloading ZendGuard

Official download link: http://www.zend.com/en/products/guard/downloads#Windows

Registration is required before downloading. I downloaded Zend Guard 6.0. Simply double-click to run the installer.

The latest version is 7.0. If you want to download an earlier version, click “>Download here” and a list of older versions will appear below the current page.

Downloading PHP

Official download links: http://php.net/downloads.php | http://php.net/releases/

My current version: http://windows.php.net/download/#php-5.4

PHP typically releases two types of builds for each version: Thread Safe and Non Thread Safe.

Important note: ZendGuard only supports Non Thread Safe builds. So please download the Non Thread Safe version of PHP.

Configuring PHP – ZendLoader.dll

Download ZendLoader.dll

Official download link: http://www.zend.com/en/products/loader/downloads#Windows

Registration is required before downloading. Similar to ZendGuard, make sure to select the ZendLoader version that matches your PHP version.

Copy ZendLoader.dll to the PHP [PHP root directory]\ext\ folder.

Rename php.ini-development or php.ini-production to php.ini,
and add the following content to php.ini:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13


zend_extension="./ext/ZendLoader.dll"

; Enables loading encoded scripts. The default value is On
zend_loader.enable=1

; Disable license checks (for performance reasons)
zend_loader.disable_licensing=0

; The Obfuscation level supported by Zend Guard Loader. The levels are detailed in the official Zend Guard Documentation. 0 - no obfuscation is enabled
zend_loader.obfuscation_level_support=3

; Path to where licensed Zend products should look for the product license. For more information on how to create a license file, see the Zend Guard User Guide
zend_loader.license_path=

Creating PHP Code

Create an input folder, and inside it create test.php.

Code:

1
2
3
4
5


<?php
date_default_timezone_set("Asia/shanghai");
echo "Thanks ZendGuard!\n";
echo "Now datetime is ".date("Y-m-d h:i:s")."\n";
?>

The above code outputs:

Thanks ZendGuard

Now datetime is [current time].

Installing ZendGuard and Creating a Project

ZendGuard has a straightforward installer. After installation, launch it and you’ll see:

Create a new Zend Guard Project.
Set the output directory to the output folder.
Add the input folder as the input directory.
Select the PHP version.
Click
A new test.php is generated in the output folder. Opening it reveals garbled content (the encrypted code).

Running Source Code and Encrypted Code

Running the source code:

Running the encrypted code:

With this, you can now use ZendGuard to encrypt PHP code.

[PHP] Configuring PHPUnit for ThinkPHP

Thu, 10 Dec 2015 14:19:57 +0000

Version Details

PHP: 5.3

PHPUnit

ThinkPHP 3.1.3

IDE: PHPStorm 10 (recommended)

Configuring PHPUnit in PHPStorm

See the blog post:

http://blog.coinidea.com/web开发/php-1088.html

ThinkPHP Deployment

Official code download:

http://www.thinkphp.cn/down.html

Site initialization:

http://www.thinkphp.cn/info/60.html

Test Cases

In this example, the index.php in the root directory is configured as follows:

1
2
3
4
5
6
7


<?php
define('APP_NAME', 'example');
define('APP_PATH', '../example/');
define('APP_PHPUNIT', false);
define('APP_DEBUG', true);
require('../ThinkPHP/ThinkPHP.php');
?>

After the first visit, the following directory structure is generated:

Create a new folder named “Testcase” in the example site.

Testing the Model

Create HelloModel.class.php:

1
2
3
4
5
6
7
8
9


<?php
class HelloModel extends Model
{
 public function sayHello()
 {
 print 'Hello';
 return 'Hello';
 }
}

Create a Test.php file in the Test folder as the PHPUnit test file. Note that you need to require ThinkPHP to initialize the framework environment. Also, in Think.class.php, modify the

start() function by changing App::run() to !APP_PHPUNIT && App::run();

This distinguishes between site runtime and test case execution.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12


<?php
define('APP_NAME', 'example');
define('APP_PATH', './../../example/');
define('APP_PHPUNIT', true);
require('./../../ThinkPHP/ThinkPHP.php');
class TestSayHello extends PHPUnit_Framework_TestCase {

public function setUp() { }

public function tearDown(){ }

}

Add a test case in TestSayHello:

1
2
3
4
5


public function testHelloModel()
{
 $hello = D('Hello');
 $this->assertTrue($hello->sayHello('Hello') == 'Hello');
}

Testing the Action

Modify IndexAction.class.php as follows:

1
2
3
4
5
6
7
8
9


<?php
class IndexAction extends Action
{
 public function index()
 {
 $hello = D("Hello");
 return $hello->sayHello();
 }
}

Browser access to Index:

Add a test case in TestSayHello:

1
2
3
4
5


public function testHelloAction()
{
 $hello = new IndexAction();
 $this->assertTrue($hello->index() == 'Hello');
}

Results

Running Test.php produces the following result:

Tests passed. At this point, unit testing has been successfully added to ThinkPHP.

poisonbian 2016/05/10 14:57

I tried following this guide and found that the model couldn’t be used, and custom functions in the Common directory were not loaded. After some investigation, I think adding the APP_PHPUNIT check in Think.class.php is not ideal. It would be better to add it in App.class.php instead: !APP_PHPUNIT && App::exec();

[PHP] PHPUnit Installation, Configuration, and Examples

Fri, 04 Dec 2015 14:45:19 +0000

PHPUnit Installation and Sample Test

Many people online recommend installing via PEAR. I tried various methods from different blog posts, but all failed.
The method that worked for me is more straightforward:

Download link:

https://phpunit.de/index.html

The download is a .phar PHP extension. Download the PHPUnit version that matches your PHP version.

Follow the instructions at https://phpunit.de/manual/current/en/installation.html#installation.phar.windows step by step.

Download PHPUnit.phar to a folder, for example: C:\bin\
Add this folder C:\bin\ to the Path environment variable
In the command line, navigate to C:\bin and enter:
echo @php "%~dpophpunit.phar"%*>phpunit.cmd

In any command line location, enter:

1

phpunit --version

You should see:

At this point, PHPUnit is successfully installed.

Test Example:

SayHello.php

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11


<?php

class SayHello
{
 public function printHello()
 {
 echo 'Hello';
 return 'Hello';
 }
}
?>

TestSayHello.php

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17


<?php

require_once 'SayHello.php';

class TestSayHello extends PHPUnit_Framework_TestCase
{
 public function setUp() { }

 public function tearDown() { }

 public function testConnectionIsValid()
 {
 $hi = new SayHello();
 $this->assertTrue($hi->printHello() == 'Hello');
 }
}
?>

The SayHello class prints “Hello” to the screen and returns “Hello”.

TestSayHello verifies whether the SayHello class correctly returns the “Hello” string.

Run it in the command line, and the test passes.

Configuring PHPUnit in PHPStorm

Older versions of PHPStorm may not support PHPUnit well and will throw:

Warning: require_once(PHPUnit/Runner/Version.php)

See:

http://stackoverflow.com/questions/30278588/phpstorm-via-phpunit-phar-warning-require-oncephpunit-runner-version-php

This article uses PHPStorm 10. Download link:

https://www.jetbrains.com/phpstorm/?fromMenu

Open the two files from earlier

Open Settings: File => Settings

Configure PHP version:

Configure PHPUnit

Sample Test

Select TestSayHello and run it directly. The result:

Slow or Failed Grunt Downloads

Sun, 15 Nov 2015 03:37:01 +0000

Grunt is a front-end automation tool, and there are plenty of blog posts on the internet about how to install and use it:

http://www.cnblogs.com/wangfupeng1988/p/4561993.html

http://www.cnblogs.com/lhb25/archive/2013/01/24/grunt-for-javascript-project-a.html

Since Grunt installation depends on Node.js, this article mainly covers solutions for when “npm install grunt” downloads slowly or fails entirely. The fix is simple: specify the download registry by appending -registry "http://registry.npmjs.org" to the command, which changes the default from HTTPS to HTTP. Alternatively, Chinese internet companies like Taobao also provide registries that you can search for and use.

For example:

1

install grunt --save-dev -registry "http://registry.npmjs.org"

[Web Development] What Happens After a User Enters a URL?

Thu, 15 Oct 2015 15:24:04 +0000

When a user opens a browser and enters a URL:

The client checks whether there is an IP address cached for that URL. If not, it sends a request to the nearest DNS server for domain name resolution, querying up to the root DNS server if necessary. Once the query succeeds, the result is returned to the client.
The browser sends the request to the web server at the IP address corresponding to the URL.

This server may be a single standalone server, a lightweight load balancer, or a cluster.

If it is a load balancer or cluster, the servers internally synchronize based on timestamps to ensure the data is up to date; if not, synchronization is performed.
If the request involves application-level calls, the request is forwarded to the application server, which returns the result to the web server.
The web server executes programs written in the relevant dynamic languages, communicates with database servers and related files, and finally sends the resulting web page data back to the browser.
Based on transport protocols such as HTTP and HTTPS, the web server’s data is transmitted back to the browser, which then parses and renders the web page.

Finally, the user sees the web page.

[SEO] Search Engine Submission URLs

Thu, 10 Sep 2015 15:01:40 +0000

Website SEO is very important, and search engine optimization is a key part of it. The first step is to get your website indexed by search engines. Below are the submission URLs for major search engines:

Baidu Website Submission: http://zhanzhang.baidu.com/linksubmit/url

Google Website Submission: https://www.google.com/webmasters/tools/submit-url

360 Search Submission: http://info.so.360.cn/site_submit.html

Soso Website Submission: http://www.soso.com/help/usb/urlsubmit.shtml

Sogou Website Submission: http://www.sogou.com/feedback/urlfeedback.php

Bing Website Submission: http://www.bing.com/toolbox/submit-site-url

[JavaScript] Finding Prime Numbers Up to N

Sun, 26 Jul 2015 02:01:42 +0000

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21


<script type="text/javascript">
var arr = new Array();
var flag = new Array(1001);
for (var i = 1; i <= 1000; i++) {
 flag[i] = 0;
}
for (var i = 2; i <= 1000; i++) {
 if (flag[i] == 0) {
 flag[i] = 1;
 arr.push(i);
 for (var j = 1; j * i <= 1000; j++) {
 flag[j*i] = 1;
 }
 }
}
var ans = "2";
for (var i = 1; i < arr.length; i++) {
 ans = ans + "," + arr[i].toString();
}
alert(ans);
</script>

This uses the Sieve of Eratosthenes, which has good performance. Tested and working in Chrome. Hope it helps!

This was my answer on Baidu Zhidao (a Q&A platform).

[Yii] Defending Against CSRF and XSS in Yii [Part 2]

Tue, 21 Jul 2015 16:04:21 +0000

I previously wrote Defending Against CSRF and XSS in Yii [Part 1], but when it came to actual testing, there was a requirement to filter out special characters such as :, :, “, <, >, %, etc.

Yii’s CHtml::purifier actually only filters HTML, and CHtml::encode essentially calls htmlspecialchars:

htmlspecialchars only converts <, >, single quotes, double quotes, and &
htmlentities converts all HTML entities

As you can see, PHP’s built-in functions cannot filter all special characters. The approach I took was to create a new Filter class within the Yii framework and write a custom specialchar function, allowing you to replace characters as needed. For better performance, it can be defined as public static.

1
2
3
4
5
6
7


class Filter {
 public static function ReplaceSpecialChar($str) {
 $str=str_replace(":",$str); //TODO: replace other chars
 $str = htmlspecialchars($str);
 return $str;
 }
}

[Yii] Defending Against CSRF and XSS in Yii [Part 1]

Sun, 19 Jul 2015 02:59:44 +0000

CSRF stands for Cross-Site Request Forgery, and XSS stands for Cross-Site Scripting. To be honest, I only recently encountered these for the first time. After building a website with Yii and scanning it with testing software (e.g., IBM AppScan), two relatively common and serious issues were found: CSRF and XSS. So how do we fix them?

– CSRF

Reference: http://www.crarun.com/article-7.html

Yii supports CSRF configuration in its config file. Once configured, it embeds a hidden token with a hash key value in forms, and the server verifies whether each request is authorized through this token.

1
2
3
4
5
6


'components'=>array(
 'request'=>array(
 // Enable Yii Validate CSRF Token
 'enableCsrfValidation' => true,
 ),
),

In practice, because I was using an older version of Yii, the framework didn’t automatically create the input element in forms, so I had no choice but to create it manually by inserting the following code:

1

<input type="hidden" value="<?php echo Yii::app()->getRequest()->getCsrfToken(); ?>" name="YII_CSRF_TOKEN" />

Another issue arose: since CSRF wasn’t considered thoroughly early on, the entire system contained numerous AJAX POST requests, meaning lots of JS code had POST submissions. However, Yii’s view rendering doesn’t affect code inside .js files - even if you add the above code in a .js file, PHP won’t produce any output. So I came up with a workaround:

Add the following code to JS POST submissions:

1

$("input[name='YII_CSRF_TOKEN']").val()

Then add the hidden input in the PHP view. However, this also reveals a limitation - it doesn’t fully prevent CSRF. The optimal solution would be to embed authorization in each user’s session.

– XSS

When searching for solutions to this problem, a common question is: at which stage should string filtering be done - submission, storage, or display? I saw an answer on SegmentFault that said both are essential: one prevents malicious data from entering the database, and the other prevents it from being executed.

In my opinion, front-end user validation during submission is essential for good user experience. However, server-side validation is still a must, since programs can bypass the front end.

Should we also filter during display? I recommend yes - it adds an extra layer of safety to prevent attacks from taking effect.

How to do it? Yii provides CHtml and CHTMLPurifier for filtering. On GitHub, many people have open-sourced good front-end libraries for filtering - just search for them.

CoinJS - An Open Source Web Component Library

Fri, 03 Jul 2015 17:06:30 +0000

I’ve been building my own open-source web component library recently, keeping it as simple as possible for my own learning, while also sharing some insights with other beginners. GitHub download link: https://github.com/hujiulin/coinjs

- ProgressBar

- ScrollImage (Scrolling Images)

[Redis] Configuring Redis in Yii - Common Issues

Fri, 12 Jun 2015 06:24:58 +0000

Basic configuration:

In Yii’s main.php:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13


return array(
 ...
 'components'=>array(
 'redis_cache' => array (
 'class' => 'system.caching.CRedisCache',
 'hostname' =>'127.0.0.1',
 'port' => 6379,
 'password'=>'123456',
 'database'=>1
 ),
 ),
 ...
)

Notes:
The CRedisCache in the class field is the official Redis plugin;
The password field is required if a password has been set;
The database field specifies the corresponding database.

Usage:
$r_key = “key”;
Yii::app()->redis_cache->set($r_key, 99999);
echo Yii::app()->redis_cache->get($r_key);

However, you won’t find a key named “key” in the Redis database, because Yii’s Redis plugin applies MD5 hashing to keys by default.
By examining CCache, the parent class of CRedisCache, you’ll find that you need to declare the following two variables in CRedisCache.php:

1
2


public $hashKey = false;
public $keyPrefix = "";

This solves the problem.

Additionally, if you encounter garbled Chinese characters, it’s because Redis defaults to ANSI encoding. When connecting, use the following command:
./redis-cli –raw -h 127.0.0.1
to connect.

For official documentation, see: http://www.yiiframework.com/doc/api/1.1/CRedisCache/.

Online QR Code Generator

Sat, 30 May 2015 03:19:13 +0000

Sharing a simple online QR code generator I wrote. Click the link below:

Online QR Code Generator

[HTML] Differences and Connections Between DIV and Table Page Layouts

Sat, 23 May 2015 05:38:17 +0000

Differences and Connections Between DIV and Table Page Layouts

There is currently a debate over whether to use traditional Table or the newer DIV for web page design. Some say Table is better – faster to develop, easier to control, and better browser compatibility. Others believe DIV is better and represents the future trend, mainly for the following reasons:

DIV+CSS layouts save page code compared to Table layouts, and the code structure is cleaner and more readable.
DIV+CSS development is faster than Table with more precise layouts, though hand-written code increases noticeably. DIV+CSS layouts make website layout modifications simpler.
DIV+CSS layouts can adapt to future multi-client requirements.
DIV+CSS layouts save site storage space and bandwidth. These are all advantages of DIV.

With so many benefits of DIV, you might be tempted to learn it. DIV and Table each have their strengths, and in most cases they can be used interchangeably. I believe the correct standards-compliant design philosophy is: use DIV and other layout elements to create page design layouts, positioning, color blocks, images, etc. Use Table, UL, and similar elements to display data on the page. Because DIV does not behave like Table in IE, where the entire Table must be downloaded before any content is displayed (Firefox does not have this issue), using Table for layout is clearly not ideal, especially with large amounts of data – it is noticeably slower in IE. DIV handles this much better. Of course, DIV can also serve to organize data.

DIV for layout, Table for displaying data – this is the most fundamental design principle today.

Can a Table contain a DIV? And vice versa, can a DIV contain a Table?
Of course it can.
Does DIV serve the same purpose as Table?
DIV’s functionality is similar to Table, but DIV has better XML support and is more flexible to use, which is why it is recommended as the new approach for web page layout.
Which is faster, DIV or Table?
DIV is relatively faster. Web content display requires downloading a complete block of content before rendering. DIV layout consists of many independent, relatively small blocks, while Table layout typically involves multiple layers of nested tables forming larger download blocks.

Additionally, the difference between . and # in CSS:

. is called using class
# is called using id

[Discuz] Forum Invitation Codes

Sat, 23 May 2015 05:37:16 +0000

Whether the forum has invitation code registration enabled
Go to Admin Panel => Global => Registration & Access => Registration. Configure whether to enable invitation-based registration, the extended credits used for invitation rewards, the credit amounts for both the inviter and invitee, and whether successful invitation registrations automatically add each other as friends.
Edit administrator permissions
Users => User Groups => Others => Registration related
Purchase invitation codes and view them on the friends page

[nginx] Starting, Stopping, and Restarting Nginx on Linux

Sat, 23 May 2015 05:36:29 +0000

Starting Nginx

1

/usr/nginx/sbin/nginx

1

/usr/nginx/sbin/nginx -t

Check whether the configuration is correct.

Stopping Nginx

Stopping is done by sending signals to the nginx process (for more about signals, refer to Linux documentation).

Step 1: Find the nginx master process ID

1

ps -ef | grep nginx

Find the master process in the process list – its number is the master process ID.

Step 2: Send a signal

Graceful shutdown:
1

kill -QUIT master_process_id
Fast shutdown:
1

kill -TERM master_process_id
Forced shutdown:
1

pkill -9 nginx

Additionally, if the pid file path is configured in nginx.conf, that file contains the Nginx master process ID. If not specified, it is stored in nginx’s logs directory. With the pid file, you don’t need to look up the master process ID first and can send signals to Nginx directly:

1

kill -signal_type '/usr/nginx/logs/nginx.pid'

Graceful Restart

If you changed the configuration, do you need to stop Nginx and start it again? No – you can send a signal to Nginx for a graceful restart.

Graceful restart command:

1

kill -HUP master_process_id_or_pid_file_path

Or use:

1

/usr/nginx/sbin/nginx -s reload

Note: After modifying the configuration file, it’s best to check whether the modified configuration is correct first, to avoid errors after restart that could affect server stability. The command to check Nginx configuration:

1

nginx -t -c /usr/nginx/conf/nginx.conf

Or:

1

/usr/nginx/sbin/nginx -t

[HTML] Passing Values in HTML

Sat, 23 May 2015 05:35:22 +0000

In JavaScript, you can use the window.location.href property to get the full URL. This means we can pass parameters using GET mode, i.e., the form’s method="GET".

Below is a JavaScript function that retrieves values submitted via GET mode.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12


function Request(strName) {
 var strHref = window.document.location.href;
 var intPos = strHref.indexOf("?");
 var strRight = strHref.substr(intPos + 1);

 var arrTmp = strRight.split("&");
 for (var i = 0; i < arrTmp.length; i++) {
 var arrTemp = arrTmp[i].split("=");
 if (arrTemp[0].toUpperCase() == strName.toUpperCase()) return arrTemp[1];
 }
 return "";
}

Usage:
var id=Request("id");
document.write(id);

[Apache] Virtual Host Configuration for Multi-Domain Access to Local PHP Projects

Sat, 23 May 2015 05:34:23 +0000

First, modify the hosts file located at C:\WINDOWS\system32\drivers\etc. Open it with Notepad and add:
1 2

127.0.0.1 www.a.com 127.0.0.1 www.b.com
Open the file xampp\apache\conf\httpd.conf, search for “Include conf/extra/httpd-vhosts.conf”, and make sure there is no # comment symbol in front of it – this ensures the vhosts virtual host configuration file is included.

Once httpd-vhosts.conf is enabled, the default httpd.conf configuration becomes inactive (make sure virtual host configuration is also enabled in httpd-vhosts.conf, see step 3). All domain requests to this IP will be directed to the first virtual host in vhosts.conf.
In the virtual host configuration file xampp\apache\conf\extra\httpd-vhosts.conf:
Remove the ## before NameVirtualHost *:80 to enable vhosts.conf. The default httpd.conf configuration becomes inactive, and virtual host settings will only be configured in httpd-vhosts.conf.
1 2

DocumentRoot /xampp/htdocs/a ServerName www.a.com
1 2

DocumentRoot /xampp/htdocs/b ServerName www.b.com
After completing step 3, you will notice that accessing localhost redirects to the path configured for site a. This is explained in step 2 – once vhosts is enabled, the default httpd configuration becomes inactive, and the default access points to the first entry in vhosts. To fix this, you need to add the localhost directory configuration back.
1 2

DocumentRoot /xampp/htdocs/ ServerName localhost

With this, the XAMPP virtual host setup is complete. Now accessing localhost still shows the original XAMPP welcome page, accessing www.a.com points to the bound directory a, and accessing www.b.com points to the bound directory b.

[VPN] Configuration Guide

Sat, 23 May 2015 05:33:34 +0000

Setting Up on Windows XP:

Start – Control Panel – Administrative Tools – Services – Start four key services (Workstation, Telephony, Remote Access Connection Manager, Routing and Remote Access) and set them to start automatically – Control Panel – Network Connections – Incoming Connections – General tab (check: Virtual Private Network, allow others to establish a private connection to my computer via the Internet or other networks through “tunneling”) – Users tab (check which users can connect remotely via VPN) – Network tab (TCP/IP address assignment: manually specify an IP address in a different subnet from the server, e.g., if the server IP is in the 192.168.1.x range, enter an address in the 192.168.2.x range here).

Finding the server’s external IP address:

Direct dial-up – Start – Control Panel – Network Connections – Local Area Connection – Support – IP Address: something like “58.60._.133”.
Router dial-up – Enter the router’s IP address (local network gateway) in IE browser – Find the router’s IP address: something like “58.60._.133”.

Windows XP Client Connection Setup:

Start – Control Panel – Network Connections – Create a new network connection – Next – Connect to the network at my workplace – Virtual Private Network connection (V) – Give it a name: VPN – Do not dial the initial connection – Enter the server’s IP address or domain name, IP address like “58.60.*.133” – Finish.

Client Connection:

Start – Control Panel – Network Connections – Under Virtual Private Network you can see the VPN dial-up you just set up – Double-click to connect.
Once connected, your machine will be on the same local network as the server.

Common VPN Dial-up Errors:

VPN 800 error: The server’s external IP address is incorrect or DMZ is not configured.
VPN 628 error or VPN 721 error: The client and server’s external WAN port are the same. The two computers are on the same LAN under the same ADSL connection. In this case, you can enter the server’s (target host’s) internal IP address to test the connection.
VPN 733 error: TCP/IP CP reported an error. The server and client’s incoming IP addresses are in the same subnet. Set the server to manual assignment (Control Panel – Network Connections – Incoming Connections – Network tab – TCP/IP address assignment: manually specify an IP address in a different subnet from the server, e.g., if the server IP is in the 192.168.1.x range, enter an address in the 192.168.2.x range here).

Someone in the lab was around yesterday, and I tried it today – it works like a charm.

[MapService] How to Embed Maps in Your Own Web Pages

Sat, 23 May 2015 05:32:44 +0000

To embed a map in your own web page, the common methods can be summarized as follows:

The Simplest Method – Using Google Maps’ “Link” Feature

If you only need to display a specific area on your page, such as your company’s location, without adding any extra content like markers or polylines, this is the easiest way to embed Google Maps.

Go to the Google Maps homepage, navigate to the area you want to display, then click “Link” in the upper left corner of the map. An info box will appear with two input fields – copy the content from the second input field into your page and you’re done.

This embed code is essentially an iframe declaration. Although the Maps homepage provides a customization and preview feature, it only allows you to customize the map size. If needed, you can manually modify the iframe declaration for further customization, such as adding custom styles to the iframe.
The Most Lightweight Method – Using Google Static Maps

If you need to display a specific area and add some markers or polylines, but don’t care whether the map on your page is draggable, then static maps are what you need.

A static map means what you embed in your page is actually just a GIF image, dynamically fetched from Google via a URL. This differs from the “dynamic” maps we commonly use, and loading such a map is much faster than loading a full interactive map.

To use a static map on your page, simply use an img tag and set its src attribute to the Google Static Maps URL.

Here’s a simple static map URL:
1

http://ditu.google.cn/staticmap?center=39.970981,116.314108&zoom=16&size=400×400&key=YOUR_API_KEY
In this URL, you can edit parameters like center, zoom, and size to specify the map’s center point, zoom level, map dimensions, and more. Of course, you can also add parameters for markers and polylines. Don’t worry about memorizing all these parameters – there’s a static map customization wizard that guides you through a few simple steps to get the URL you need. If interested, you can explore the Google Static Maps API documentation.

Strictly speaking, Google Static Maps is also a type of Google Maps API, so using static maps requires your Google Maps API key. If you’ve already registered a Maps API key, you can simply reuse it.
The Most Flexible Method – Using the Google Maps API

If neither of the above methods meets your needs, try the Google Maps API. Although the Google Maps API has been divided into JavaScript, Flash, Earth, Static, and other versions, in my view, the JavaScript API is the foundation of the Google Maps API. So unless otherwise specified, “Google Maps API” here refers to the Google Maps JavaScript API. With this API, you can display your data on the map in any feasible way you like, and you can even make your own map look better than Google Maps itself.

To embed a map in your page using this API, the basic steps are:
1. Import the Maps API library using a JavaScript tag;
2. Define an element on the page to hold the map, typically a div tag with specified width and height;
3. In your JavaScript code: new GMap2(document.getElementById("your map container's id"));
4. Use classes provided by the API such as GMarker, GPolyline, etc. to customize the markers, polylines, and other elements on the map.
For a detailed guide on creating Google Maps, see the “Creating Maps with JavaScript Guide” above. However, to use this API proficiently, you need a certain level of JavaScript knowledge and hands-on ability. Additionally, it’s strongly recommended to read the Google Maps JavaScript API Developer Guide first, which can answer many common questions. Of course, you can also find the information you’re looking for in this documentation.

Beyond the methods described above, there are some less commonly used ways to embed maps in web pages, which won’t be covered here.

[HTML] Preventing Duplicate Form Submission on F5 Refresh

Sat, 23 May 2015 05:30:50 +0000

Page A has a form.
You submit this page and are redirected to page B.
On page B, pressing the refresh key will trigger a dialog asking to resend the form data.

This problem can be solved from the web development perspective.
Page A submits to your server page (ASP, JSP).
The server page processes the request but does not return page B to the user directly.
Instead, it returns a simple page C, such as a redirect page.

This immediately redirects to page B – in this case, main.jsp.

If you refresh on page B, the resend dialog will not appear.
If you press “Back” on page B, you will go back to page A.
The user will not notice the intermediate page C at all.

Going further, you can design a JavaScript dialog box that, once confirmed, will not pop up again regardless of whether the user refreshes or navigates back.
redirect.jsp

[JSP] Free JSP Hosting

Sat, 23 May 2015 05:29:00 +0000

When a JSP website is ready to be published, it can be a headache for individuals or students without a budget.
I encountered this problem too. After searching online, I found two helpful blog posts:
Recommending a free JSP hosting service: http://fireinwind.iteye.com/blog/705611

This blog post introduced me to two free JSP hosting options: jhost and EATJ.
Here are my thoughts on using these two free hosting services – free services inevitably have drawbacks. For example, jhost is extremely slow to access from the campus network, and uploading files takes forever. EATJ’s drawback is that it shuts down your server every six hours each day, and you have to restart it manually (shutdown times are 4:00, 10:00, 14:00, 18:00, and 24:00).

So I recommend using EATJ when you don’t have a budget in the early stages. The speed is decent and it works well enough – the only downside is having to restart the server. It is perfectly suitable for testing purposes.

As for jhost, it has a better user experience, designed with a desktop OS-style interface for easier management, but it is slow.

For detailed instructions on how to use these two hosting services, see the blog post:

jsp php enthusiasts’ good news: eatj and jhost

Note: Applying for jhost is somewhat complicated and requires an invitation code. You can search for “jhost invitation code” online.

After activating your account, log in to the hosting backend and activate your Website Hosting Service in the Control Panel.

Site policy: Regular users must log in and renew at least once every 3 days (click the Renew button in the Member Service Management section of the Control Panel) to confirm you are still using the service.
Additionally, the site periodically suspends the TOMCAT service for users who have not renewed within the specified time. The next renewal will restart the service. If you do not renew for over a month, your account will be deactivated.

You can extend your hosting usage period or redeem other services through site credits. Credits can be earned by referring users or purchased with cash. See the site’s help section for details.

Counting Valid Lines in a Textarea with JavaScript

Thu, 19 Mar 2015 12:30:42 +0000

I recently had a requirement to count the number of valid (non-empty) lines in a textarea on a page, limit the total number of lines, and display the remaining available lines. [Besides counting lines, you could also check for spaces, commas, semicolons, etc. – the approach is similar.]

My initial idea was: Write a regex to replace multiple consecutive newlines “\n” with a single “\n”: str.replace(/\n+/, "\n"); But this doesn’t handle cases where there are multiple empty lines with leading spaces.

Later: Simply split by “\n”, then count entries with length greater than 0 as valid lines.

Here’s the code:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33


<script type="text/javascript">
var maxTagCount = 100;
function checkLine(textArea) {
 var str = textArea.value;
 var lines = str.split('\n');
 var countLine = 0;
 for (var i = 0; i < lines.length; i++) {
 if (lines[i].length > 0) {
 countLine++;
 }
 }
 var tagElement = document.getElementById("tagTextarea");

 if (countLine > maxTagCount) {
 var varNewValue = "";
 var index = 0;
 for (var i = 0; i < lines.length; i++) {
 if (index >= maxTagCount) {
 break;
 }
 if (lines[i].length > 0) {
 index++;
 }
 varNewValue += lines[i];
 varNewValue += "\n";
 }
 textArea.value = varNewValue;
 stat_left.innerHTML = 0;
 } else {
 stat_left.innerHTML = maxTagCount - countLine;
 }
}
</script>

Web demo: countLines

Making Bootstrap Compatible with IE

Sat, 07 Mar 2015 07:24:29 +0000

The front-end of my previous websites was built entirely with Bootstrap.

I usually use Chrome, but one day I checked my website on IE6 running on XP, and it was a disaster. I had completely forgotten about browser compatibility. So I tested on:

Chrome, Opera, Firefox, IE 5.5-10 [using IE Tester]

The main issues were with IE, so I started looking for solutions:

http://www.jqueryba.com/1194.html

Add:

1
2
3


<!--[if IE]>
<script src="http://html5shiv.googlecode.com/svn/trunk/html5.js"></script>
<![endif]-->

Or directly add:

1
2
3
4
5


<!--[if IE]>
<script>
(function(){if(!/*@cc_on!@*/0)return;var e = "abbr,article,aside,audio,canvas,datalist,details,dialog,eventsource,figure,footer,header,hgroup,mark,menu,meter,nav,output,progress,section,time,video".split(','),i=e.length;while(i--){document.createElement(e[i])}})()
</script>
<![endif]-->

Then:

1

header, footer, article, section, nav, menu, hgroup {display: block;}

1
2
3
4
5
6
7
8
9


<script type="text/javascript">
document.createElement("section");
document.createElement("article");
document.createElement("footer");
document.createElement("header");
document.createElement("hgroup");
document.createElement("nav");
document.createElement("menu");
</script>

The results weren’t great though – many nav elements and alignments were off.

BSIE [Bootstrap IE] http://www.bootcss.com/p/bsie/ This is an open-source library built on top of Bootstrap. The demo looked great, but due to legacy reasons, my site was already built and making changes was too difficult, so I had to give up.
The final solution: just embed two JS files.

html5shiv.js + respond.js
Yes, just these two JS files. At least IE7 and above now display properly. There are still some minor issues like rounded corners, but the nav, grid layout, and overall appearance work well. Very satisfying.

CoinVote -- A Voting System

Thu, 19 Feb 2015 06:05:31 +0000

CoinVote is a flexible voting system. CoinVote was independently developed by me in 2015 and features the following:

Voting homepage
View votes
Vote listings
Candidate listings
Edit candidates

Demo | Details

CoinCMS -- A Portal System

Thu, 19 Feb 2015 05:54:17 +0000

CoinCMS is a clean and simple portal system. CoinCMS was independently developed by me in 2015 and features the following:

Portal homepage
Category listings
Article publishing
Admin panel

Demo | Details

WordPress Domain Configuration on SAE

Wed, 11 Feb 2015 09:12:06 +0000

I set up WordPress on SAE and bound a custom domain to the application: blog.coinidea.com. The link does redirect to the blog page, but all internal links still point to app-name.sinaapp.com. This means the custom domain doesn’t apply to WordPress’s internal articles and pages.

Solutions:

Option 1: Dashboard Settings

Go to the WordPress dashboard.
Click “Settings”.
Modify:

Option 2: Direct Database Modification [More aggressive]

Find the wp-options table.
Modify entries #1 and #37.

Blog Migration

Fri, 19 Dec 2014 02:42:52 +0000

As 2015 approaches, in order to become more focused and professional, I am migrating my blog from Sina:

http://blog.sina.com.cn/u/1731162620

to my personal site on SAE.

Many thanks to:

http://www.crifan.com/crifan_released_all/website/python/blogstowordpress/usage_example/

crifan’s blog migration tool.

Web Development on Coinidea's Blog

[LNMP] Nginx access_log Daily Rotation

[LNMP]Memory Optimization

Problem Description

Solution

[Web Development]HTML5 Geolocation Inaccuracy Issues

[MySQL]Automatic MySQL Backup

Create the backup.sh Script

Create a Scheduled Task

[Push Notifications]Push Notifications Based on Umeng Service

[ThinkPHP]Table Locking Under Concurrency in ThinkPHP

[MySQL] is marked as crashed and should be repaired

[HTML5] Local Image Reading and Cropping with HTML5

[MySQL] Removing Carriage Returns and Line Feeds from Fields

[JavaScript] Deep Copy in JavaScript

[PHP] Filtering All Punctuation with PHP Regular Expressions

[PHP] Large-Scale Data Import and Export with ThinkPHP, PHP, and MySQL

Large-Scale Data Import and Export

Differences Between MyISAM and InnoDB

Optimization Methods for Inserting Large Amounts of Data

Demo Code:

1. Using Transactions (InnoDB supported)

2. SQL Statement Concatenation

Starting and Committing Transactions in ThinkPHP

[Web Development] Migrating DedeCMS to Discuz Portal

I recently needed to migrate from DedeCMS (dedecms) to Discuz Portal. The most important challenge was how to migrate categories and a large number of articles.

1. Preprocess the old table, merge typename

2. Merge click, title, writer, pubdate, keywords, description

3. Generate new IDs in relation

4. Populate catid based on typename and relation

5. Import into Discuz Portal Tables

Insert into pre_portal_article_content (Article content table)

Insert into pre_portal_article_title (Article title table)

Insert into pre_portal_article_count (Article view count table)

[Web Development] Postman Plugin

Recommending a Web Development Debugging Tool

[PHP] call to undefined function mb_convert_encoding

Solution on Windows

Solution on Linux

[IIS] File Formats (Fonts) Failing to Load on IIS Server

Font Subsetting for Web and Mobile

[Bootstrap] Triggering dropdown-toggle on Mouse Hover in Bootstrap

Using the bootstrap-hover-dropdown Plugin

1. Download bootstrap-hover-dropdown

2. Include bootstrap-hover-dropdown.js

3. Modify the Dropdown in HTML

4. Removing the Delay

[Node.js] Building an Image Server with UEditor + Node.js + SeaweedFS

1. UEditor + Node.js Image Upload

1.1 Modify the serverUrl property in UEditor’s ueditor.config.js:

1.2 Rename ueditor/php/config.json to config.js and move it to the ueditor directory.

1.3 Next, write the four corresponding actions on the Node.js side. The image upload uses the connect-busboy middleware. Code as follows:

2. GoLang Installation and Configuration

3. Compiling and Running seaweedfs

[Spring] Creating a Spring Framework Project and Adding Your Own Controller in IntelliJ IDEA

POM File Configuration:

Tomcat Configuration

[SAE] SAE MySQL Cross-Application Authorization

Steps to Share a Single MySQL Across Multiple Applications

Notes

[Node.js] Exception Handling in Node.js Express

[Memcache] Testing Memcache Connection with PHP

[Javascript] Calling the Printer from a Web Page

[Javascript] Dynamic Loading and Hiding of UEditor

1. Initial Page

2. Click the Left DIV

3. Click the Right DIV

4. Click an Empty Area on the Page

[Apache] Configuring Apache for Streaming Media

Apache Configuration

Video Playback Example

[PHP] Encrypting PHP with ZendGuard

Downloading ZendGuard

Downloading PHP

Configuring PHP – ZendLoader.dll

Creating PHP Code

Installing ZendGuard and Creating a Project

Running Source Code and Encrypted Code

[PHP] Configuring PHPUnit for ThinkPHP

Version Details

1. Preprocess the old table, merge `typename`

2. Merge `click`, `title`, `writer`, `pubdate`, `keywords`, `description`

3. Generate new IDs in `relation`

4. Populate `catid` based on `typename` and `relation`

Insert into `pre_portal_article_content` (Article content table)

Insert into `pre_portal_article_title` (Article title table)

Insert into `pre_portal_article_count` (Article view count table)

Using the `bootstrap-hover-dropdown` Plugin

1. Download `bootstrap-hover-dropdown`

2. Include `bootstrap-hover-dropdown.js`

1.1 Modify the `serverUrl` property in UEditor’s `ueditor.config.js`:

1.2 Rename `ueditor/php/config.json` to `config.js` and move it to the `ueditor` directory.

1.3 Next, write the four corresponding actions on the Node.js side. The image upload uses the `connect-busboy` middleware. Code as follows: