A computer lets you make more mistakes faster than any other invention in human history, with the possible exceptions of handguns and tequila. - Mitch Ratcliffe

Haha, it's totally true ! And it makes me remember this meme when I was little bit drunk !

Example table

mysql> describe `test`;
+------------+-------------+------+-----+-------------------+----------------+
| Field      | Type        | Null | Key | Default           | Extra          |
+------------+-------------+------+-----+-------------------+----------------+
| id         | smallint(6) | NO   | PRI | NULL              | auto_increment |
| name       | varchar(50) | NO   |     | NULL              |                |
| created_at | timestamp   | YES  | MUL | CURRENT_TIMESTAMP |                |
+------------+-------------+------+-----+-------------------+----------------+
3 rows in set (0.00 sec)

Here is dump file of the table

DROP TABLE IF EXISTS `test`;

CREATE TABLE `test` (
  `id` smallint(6) NOT NULL AUTO_INCREMENT,
  `name` varchar(50) COLLATE utf8_unicode_ci NOT NULL,
  `created_at` timestamp NULL DEFAULT CURRENT_TIMESTAMP,
  PRIMARY KEY (`id`),
  KEY `created_at` (`created_at`)
) ENGINE=InnoDB DEFAULT CHARSET=utf8 COLLATE=utf8_unicode_ci;

LOCK TABLES `test` WRITE;
/*!40000 ALTER TABLE `test` DISABLE KEYS */;

INSERT INTO `test` (`id`, `name`, `created_at`)
VALUES
	(1,'abc','2017-02-16 17:28:59'),
	(2,' ABC','2017-02-16 17:29:14'),
	(3,'ABC     ','2017-02-16 17:29:21'),
	(4,'aBc','2017-02-16 17:29:31');

/*!40000 ALTER TABLE `test` ENABLE KEYS */;
UNLOCK TABLES;

Case Insensitive

When creating database, table and column, we have to set the default string COLLATION for them. And if we use collation end withs _ci, it means we ignore the Case Sensitive (or Case Insensitive).

Then "abc" == "ABC" or "abc" == "aBc" or "abc" = "Abc" or ... (X = Y <=> UPPER(X) == UPPER(Y))

mysql> select ("abc" = "ABC") as `case1`, ("abc" = "aBc") as `case2`, ("abc" = "Abc") as `case3`, ("abc" = "abcd") as `wrong`;
+-------+-------+-------+-------+
| case1 | case2 | case3 | wrong |
+-------+-------+-------+-------+
|     1 |     1 |     1 |     0 |
+-------+-------+-------+-------+
1 row in set (0.00 sec)

String Trimming

Check this out !

mysql> select `id`, CONCAT("'", `name`, "'") as `name_with_quote`, `created_at` from test;
+----+-----------------+---------------------+
| id | name_with_quote | created_at          |
+----+-----------------+---------------------+
|  1 | 'abc'           | 2017-02-16 17:28:59 |
|  2 | ' ABC'          | 2017-02-16 17:29:14 |
|  3 | 'ABC '          | 2017-02-16 17:29:21 |
|  4 | 'aBc'           | 2017-02-16 17:29:31 |
+----+-----------------+---------------------+
4 rows in set (0.00 sec)

mysql> select `id`, CONCAT("'", `name`, "'") as `name_with_quote`, `created_at` from test where `name` = 'abc';
+----+-----------------+---------------------+
| id | name_with_quote | created_at          |
+----+-----------------+---------------------+
|  1 | 'abc'           | 2017-02-16 17:28:59 |
|  3 | 'ABC '          | 2017-02-16 17:29:21 |
|  4 | 'aBc'           | 2017-02-16 17:29:31 |
+----+-----------------+---------------------+
3 rows in set (0.01 sec)

mysql> select `id`, CONCAT("'", `name`, "'") as `name_with_quote`, `created_at` from test where `name` = ' abc';
+----+-----------------+---------------------+
| id | name_with_quote | created_at          |
+----+-----------------+---------------------+
|  2 | ' ABC'          | 2017-02-16 17:29:14 |
+----+-----------------+---------------------+
1 row in set (0.00 sec)

BAAMMMMM !!!

MySQL do right trimming the string value before comparing.

So you must be becareful to trim value before storing to MySQL to make everything consistent.

Storing Emoji in MySQL

Because Emoji use utf-8 4 bytes, so we must use the utf8mb4 charset (utf8 multi-bytes 4).

It's safe when migrating charset from utf8 to utf8mb4 😎 , but not the reverse way 😅

Run this command, it will show top-10 frequently commands, explain shell

$ history | awk '{print $2}' | sort | uniq -c | sort -nr | head

Example result

2064 git
1284 ls
 826 cd
 700 ssh
 602 clear
 491 python
 473 exit
 341 vagrant
 242 export
 167 ping

• Ctrl + e : jump cursor to EOL
• Ctrl + a : jump cursor to BOL (beginning of line)
• Ctrl + u : delete all from cursor to BOL
• Ctrl + k : delete all from cursor to EOL
• Ctrl + r : search history, press again for next search
• Ctrl + l : clear shell screen
• Ctrl + c : terminate the command (sometimes have to press twice)
• Ctrl + z : suspend the command, back to shell. Run fg to resume the command

Today, I tried to debug many hours to find out why my POST request missing some data (specify _token hidden field). :disappointed:

I tried to config NGINX and PHPFPM max_post_size, client_max_body_size but they still gone. After 2-3 hours searching on Google, I found the link from PHP.net,
it has a config value about limiting max input vars (default = 1000), so it causes the problem about missing data.

So I changed max_input_vars = 9999 in my php.ini and everything works like a charm. :smiley:

At least, I had a luck cos it doesn't run my POST request when missing the CSRF token :grin: My data is save !!!

Command lsof

List all opened files, sockets, pipes

Eg:

• List processes are using port 80 (need root if port between 1-1023)

# sudo lsof -i:80

• List processes are using /bin/bash

# lsof /bin/bash

I found a handy tool for making a websocket right on your shell.
It is websocketd : http://websocketd.com/

Its phisolophy is so simple :

Just read incoming text from stdin and write outgoing text to stdout. Messaging is simple.

Read its docs, I follow the ez tutorial 10 second tutorial and see how cool does it work.
Let make something cool with this, we got a UNIX tool that print out to stdout in real-time with changes of end file.

Ok cool, change the command to

websocketd --port=8080 tail -f logs/web_access.log

Tada ! Open the webpage that listens the websocket you made and see the magic :D You can tracking access log of website or system in real-time way.

Docker is the fast and clean way to run Linux programs.

We can run a PHP project via PHP built-in server and Docker.

docker run -it -p 8080:8080 -v `pwd`:/code php:7 php -S 0.0.0.0:8080 -t /code/web /code/web/server.php

With server.php content is

<?php

$filename = __DIR__.preg_replace('#(\?.*)$#', '', $_SERVER['REQUEST_URI']);
if (php_sapi_name() === 'cli-server' && is_file($filename)) {
    return false;

}

// Run application below
$app = new Application();
$app->run();

When you have a object x and assign y = x, y will be a ref of x (it looks like pointer of C). So changing property of y means changing property of x.

Ex :

x = {"a": 1, "b": 2}
y = x
y['a'] = 100
print x['a'] # Result is 100

So if you want clone the value, use copy lib :

import copy
x = {"a": 1, "b": 2}
y = copy.deepcopy(x)
y['a'] = 100
print x['a'] # Result is 1

Can upload an file via FTP by curl tool with handy script :

curl -T file_need_to_upload ftp://hostname --user user:passwd

To create the n-to-m relationship in Eloquent, we create a table stand between 2 tables.
Eg:

We have the db schema :

In the model Post and Tag, we define a relation :

• App\Model\Post.php

public function tags() {
    return $this->belongsToMany('App\Model\Tag', 'post_tag', 'post_id', 'tag_id');
}

• App\Model\Tag.php

public function posts() {
    return $this->belongsToMany('App\Model\Post', 'post_tag', 'tag_id', 'post_id');
}

After that, we can use the relation to fetch the related data, like

$post = Post::firstOrFail(1);
$tags = $post->tags; // return the Collections class contains tags

or

$tag = Tag::where('name', '=', 'php')->first();
$posts_of_tag = $tag->posts; // return the Collections class contains posts

When updating or creating post, we have to sync with tags by

$tag_ids = [2, 3 ,4]; // Get tag ids from tag names
$post->tags()->sync($tag_ids);

unicode_thing = u"Xin chào mọi người"
a = '{}'.format(unicode_thing)

will cause the error UnicodeEncodeError: 'ascii' codec can't encode character u'\xe0' in position 6: ordinal not in range(128)

The solution is add u prefix the pattern (it means using unicode pattern) :

unicode_thing = u"Xin chào mọi người"
a = u'{}'.format(unicode_thing)

Reality, if using function on indexed field, you will broke indexing by accident.
Eg:

WHERE MONTH(`date`) = 11 AND YEAR(`date`) = 2015

Solution is transform the query to comparison query, like this :

WHERE `date` >= '2015-11-01' AND `date` < '2015-12-01'

• Open the preference of any Application by Cmd + ,.
• Press Cmd + ~ to go previous App when switching App on Cmd + Tab
• Copy screenshot to clipboard by Cmd + Ctl + Shift + 3
• Press Option when click to notification center is putting it on “Do no disturb” mode.

• To open directly the Console Dev Tools, press Cmd + Opt + J in MacOSX.
• To debug easily the any files of extension, open the url chrome-extension://<extension-id>/<file-name> in address bar.