1.不要使用相对路径
常常会看到:
<ol class="dp-c"><li class="alt"><span><span class="keyword">require_once</span><span>(</span><span class="string">'../../lib/some_class.php'</span><span>); </span></span></li></ol>
该方法有很多缺点:
它首先查找指定的php包含路径, 然后查找当前目录.
因此会检查过多路径.
如果该脚本被另一目录的脚本包含, 它的基本目录变成了另一脚本所在的目录.
另一问题, 当定时任务运行该脚本, 它的上级目录可能就不是工作目录了.
因此最佳选择是使用绝对路径:
<ol class="dp-c"> <li class="alt"><span><span>define(</span><span class="string">'ROOT'</span><span> , </span><span class="string">'/var/www/project/'</span><span>); </span></span></li> <li> <span class="keyword">require_once</span><span>(ROOT . </span><span class="string">'../../lib/some_class.php'</span><span>); </span> </li> <li class="alt"><span> </span></li> <li> <span class="comment">//rest of the code</span><span> </span> </li> </ol>
我们定义了一个绝对路径, 值被写死了. 我们还可以改进它. 路径 /var/www/project 也可能会改变, 那么我们每次都要改变它吗? 不是的, 我们可以使用__FILE__常量, 如:
<ol class="dp-c"> <li class="alt"><span><span class="comment">//suppose your script is /var/www/project/index.php </span><span> </span></span></li> <li> <span class="comment">//Then __FILE__ will always have that full path. </span><span> </span> </li> <li class="alt"><span> </span></li> <li> <span>define(</span><span class="string">'ROOT'</span><span> , </span><span class="func">pathinfo</span><span>(</span><span class="keyword">__FILE__</span><span>, PATHINFO_DIRNAME)); </span> </li> <li class="alt"> <span class="keyword">require_once</span><span>(ROOT . </span><span class="string">'../../lib/some_class.php'</span><span>); </span> </li> <li><span> </span></li> <li class="alt"> <span class="comment">//rest of the code</span><span> </span> </li> </ol>
现在, 无论你移到哪个目录, 如移到一个外网的服务器上, 代码无须更改便可正确运行.
2. 不要直接使用 require, include, include_once, required_once
可以在脚本头部引入多个文件, 像类库, 工具文件和助手函数等, 如:
<ol class="dp-c"> <li class="alt"><span><span class="keyword">require_once</span><span>(</span><span class="string">'lib/Database.php'</span><span>); </span></span></li> <li> <span class="keyword">require_once</span><span>(</span><span class="string">'lib/Mail.php'</span><span>); </span> </li> <li class="alt"><span> </span></li> <li> <span class="keyword">require_once</span><span>(</span><span class="string">'helpers/utitlity_functions.php'</span><span>); </span> </li> </ol>
这种用法相当原始. 应该更灵活点. 应编写个助手函数包含文件. 例如:
<ol class="dp-c">
<li class="alt"><span><span class="keyword">function</span><span> load_class(</span><span class="vars">$class_name</span><span>) </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="comment">//path to the class file </span><span> </span>
</li>
<li>
<span> </span><span class="vars">$path</span><span> = ROOT . </span><span class="string">'/lib/'</span><span> . </span><span class="vars">$class_name</span><span> . </span><span class="string">'.php'</span><span>); </span>
</li>
<li class="alt">
<span> </span><span class="keyword">require_once</span><span>( </span><span class="vars">$path</span><span> ); </span>
</li>
<li><span>} </span></li>
<li class="alt"><span> </span></li>
<li>
<span>load_class(</span><span class="string">'Database'</span><span>); </span>
</li>
<li class="alt">
<span>load_class(</span><span class="string">'Mail'</span><span>); </span>
</li>
</ol>有什么不一样吗? 该代码更具可读性.
將来你可以按需扩展该函数, 如:
<ol class="dp-c">
<li class="alt"><span><span class="keyword">function</span><span> load_class(</span><span class="vars">$class_name</span><span>) </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="comment">//path to the class file </span><span> </span>
</li>
<li>
<span> </span><span class="vars">$path</span><span> = ROOT . </span><span class="string">'/lib/'</span><span> . </span><span class="vars">$class_name</span><span> . </span><span class="string">'.php'</span><span>); </span>
</li>
<li class="alt"><span> </span></li>
<li>
<span> </span><span class="keyword">if</span><span>(</span><span class="func">file_exists</span><span>(</span><span class="vars">$path</span><span>)) </span>
</li>
<li class="alt"><span> { </span></li>
<li>
<span> </span><span class="keyword">require_once</span><span>( </span><span class="vars">$path</span><span> ); </span>
</li>
<li class="alt"><span> } </span></li>
<li><span>} </span></li>
</ol>还可做得更多:
为同样文件查找多个目录
能很容易的改变放置类文件的目录, 无须在代码各处一一修改
可使用类似的函数加载文件, 如html内容.
3. 为应用保留调试代码
在开发环境中, 我们打印数据库查询语句, 转存有问题的变量值, 而一旦问题解决, 我们注释或删除它们. 然而更好的做法是保留调试代码.
在开发环境中, 你可以:
<ol class="dp-c">
<li class="alt"><span><span>define(</span><span class="string">'ENVIRONMENT'</span><span> , </span><span class="string">'development'</span><span>); </span></span></li>
<li><span> </span></li>
<li class="alt">
<span class="keyword">if</span><span>(! </span><span class="vars">$db</span><span>->query( </span><span class="vars">$query</span><span> ) </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="keyword">if</span><span>(ENVIRONMENT == </span><span class="string">'development'</span><span>) </span>
</li>
<li><span> { </span></li>
<li class="alt">
<span> </span><span class="func">echo</span><span> </span><span class="string">"$query failed"</span><span>; </span>
</li>
<li><span> } </span></li>
<li class="alt">
<span> </span><span class="keyword">else</span><span> </span>
</li>
<li><span> { </span></li>
<li class="alt">
<span> </span><span class="func">echo</span><span> </span><span class="string">"Database error. Please contact administrator"</span><span>; </span>
</li>
<li><span> } </span></li>
<li class="alt"><span>} </span></li>
</ol>在服务器中, 你可以:
<ol class="dp-c">
<li class="alt"><span><span>define(</span><span class="string">'ENVIRONMENT'</span><span> , </span><span class="string">'production'</span><span>); </span></span></li>
<li><span> </span></li>
<li class="alt">
<span class="keyword">if</span><span>(! </span><span class="vars">$db</span><span>->query( </span><span class="vars">$query</span><span> ) </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="keyword">if</span><span>(ENVIRONMENT == </span><span class="string">'development'</span><span>) </span>
</li>
<li><span> { </span></li>
<li class="alt">
<span> </span><span class="func">echo</span><span> </span><span class="string">"$query failed"</span><span>; </span>
</li>
<li><span> } </span></li>
<li class="alt">
<span> </span><span class="keyword">else</span><span> </span>
</li>
<li><span> { </span></li>
<li class="alt">
<span> </span><span class="func">echo</span><span> </span><span class="string">"Database error. Please contact administrator"</span><span>; </span>
</li>
<li><span> } </span></li>
<li class="alt"><span>} </span></li>
</ol>4. 使用可跨平台的函数执行命令
system, exec, passthru, shell_exec 这4个函数可用于执行系统命令. 每个的行为都有细微差别. 问题在于, 当在共享主机中, 某些函数可能被选择性的禁用. 大多数新手趋于每次首先检查哪个函数可用, 然而再使用它.
更好的方案是封成函数一个可跨平台的函数.
<ol class="dp-c">
<li class="alt"><span><span class="comment">/** </span> </span></li>
<li><span><span class="comment"> Method to execute a command in the terminal </span> </span></li>
<li class="alt"><span><span class="comment"> Uses : </span> </span></li>
<li><span> </span></li>
<li class="alt"><span><span class="comment"> 1. system </span> </span></li>
<li><span><span class="comment"> 2. passthru </span> </span></li>
<li class="alt"><span><span class="comment"> 3. exec </span> </span></li>
<li><span><span class="comment"> 4. shell_exec </span> </span></li>
<li class="alt"><span> </span></li>
<li><span><span class="comment">*/</span><span> </span></span></li>
<li class="alt">
<span class="keyword">function</span><span> terminal(</span><span class="vars">$command</span><span>) </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="comment">//system </span><span> </span>
</li>
<li>
<span> </span><span class="keyword">if</span><span>(function_exists(</span><span class="string">'system'</span><span>)) </span>
</li>
<li class="alt"><span> { </span></li>
<li><span> ob_start(); </span></li>
<li class="alt">
<span> system(</span><span class="vars">$command</span><span> , </span><span class="vars">$return_var</span><span>); </span>
</li>
<li>
<span> </span><span class="vars">$output</span><span> = ob_get_contents(); </span>
</li>
<li class="alt"><span> ob_end_clean(); </span></li>
<li><span> } </span></li>
<li class="alt">
<span> </span><span class="comment">//passthru </span><span> </span>
</li>
<li>
<span> </span><span class="keyword">else</span><span> </span><span class="keyword">if</span><span>(function_exists(</span><span class="string">'passthru'</span><span>)) </span>
</li>
<li class="alt"><span> { </span></li>
<li><span> ob_start(); </span></li>
<li class="alt">
<span> </span><span class="func">passthru</span><span>(</span><span class="vars">$command</span><span> , </span><span class="vars">$return_var</span><span>); </span>
</li>
<li>
<span> </span><span class="vars">$output</span><span> = ob_get_contents(); </span>
</li>
<li class="alt"><span> ob_end_clean(); </span></li>
<li><span> } </span></li>
<li class="alt"><span> </span></li>
<li>
<span> </span><span class="comment">//exec </span><span> </span>
</li>
<li class="alt">
<span> </span><span class="keyword">else</span><span> </span><span class="keyword">if</span><span>(function_exists(</span><span class="string">'exec'</span><span>)) </span>
</li>
<li><span> { </span></li>
<li class="alt">
<span> </span><span class="func">exec</span><span>(</span><span class="vars">$command</span><span> , </span><span class="vars">$output</span><span> , </span><span class="vars">$return_var</span><span>); </span>
</li>
<li>
<span> </span><span class="vars">$output</span><span> = implode(</span><span class="string">"n"</span><span> , </span><span class="vars">$output</span><span>); </span>
</li>
<li class="alt"><span> } </span></li>
<li><span> </span></li>
<li class="alt">
<span> </span><span class="comment">//shell_exec </span><span> </span>
</li>
<li>
<span> </span><span class="keyword">else</span><span> </span><span class="keyword">if</span><span>(function_exists(</span><span class="string">'shell_exec'</span><span>)) </span>
</li>
<li class="alt"><span> { </span></li>
<li>
<span> </span><span class="vars">$output</span><span> = shell_exec(</span><span class="vars">$command</span><span>) ; </span>
</li>
<li class="alt"><span> } </span></li>
<li><span> </span></li>
<li class="alt">
<span> </span><span class="keyword">else</span><span> </span>
</li>
<li><span> { </span></li>
<li class="alt">
<span> </span><span class="vars">$output</span><span> = </span><span class="string">'Command execution not possible on this system'</span><span>; </span>
</li>
<li>
<span> </span><span class="vars">$return_var</span><span> = 1; </span>
</li>
<li class="alt"><span> } </span></li>
<li><span> </span></li>
<li class="alt">
<span> </span><span class="keyword">return</span><span> </span><span class="keyword">array</span><span>(</span><span class="string">'output'</span><span> => </span><span class="vars">$output</span><span> , </span><span class="string">'status'</span><span> => </span><span class="vars">$return_var</span><span>); </span>
</li>
<li><span>} </span></li>
<li class="alt"><span> </span></li>
<li>
<span>terminal(</span><span class="string">'ls'</span><span>); </span>
</li>
</ol>上面的函数將运行shell命令, 只要有一个系统函数可用, 这保持了代码的一致性.
5. 灵活编写函数
<ol class="dp-c">
<li class="alt"><span><span class="keyword">function</span><span> add_to_cart(</span><span class="vars">$item_id</span><span> , </span><span class="vars">$qty</span><span>) </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="vars">$_SESSION</span><span>[</span><span class="string">'cart'</span><span>][</span><span class="string">'item_id'</span><span>] = </span><span class="vars">$qty</span><span>; </span>
</li>
<li><span>} </span></li>
<li class="alt"><span> </span></li>
<li>
<span>add_to_cart( </span><span class="string">'IPHONE3'</span><span> , 2 ); </span>
</li>
</ol>使用上面的函数添加单个项目. 而当添加项列表的时候,你要创建另一个函数吗? 不用, 只要稍加留意不同类型的参数, 就会更灵活. 如:
<ol class="dp-c">
<li class="alt"><span><span class="keyword">function</span><span> add_to_cart(</span><span class="vars">$item_id</span><span> , </span><span class="vars">$qty</span><span>) </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="keyword">if</span><span>(!</span><span class="func">is_array</span><span>(</span><span class="vars">$item_id</span><span>)) </span>
</li>
<li><span> { </span></li>
<li class="alt">
<span> </span><span class="vars">$_SESSION</span><span>[</span><span class="string">'cart'</span><span>][</span><span class="string">'item_id'</span><span>] = </span><span class="vars">$qty</span><span>; </span>
</li>
<li><span> } </span></li>
<li class="alt"><span> </span></li>
<li>
<span> </span><span class="keyword">else</span><span> </span>
</li>
<li class="alt"><span> { </span></li>
<li>
<span> </span><span class="keyword">foreach</span><span>(</span><span class="vars">$item_id</span><span> </span><span class="keyword">as</span><span> </span><span class="vars">$i_id</span><span> => </span><span class="vars">$qty</span><span>) </span>
</li>
<li class="alt"><span> { </span></li>
<li>
<span> </span><span class="vars">$_SESSION</span><span>[</span><span class="string">'cart'</span><span>][</span><span class="string">'i_id'</span><span>] = </span><span class="vars">$qty</span><span>; </span>
</li>
<li class="alt"><span> } </span></li>
<li><span> } </span></li>
<li class="alt"><span>} </span></li>
<li><span> </span></li>
<li class="alt">
<span>add_to_cart( </span><span class="string">'IPHONE3'</span><span> , 2 ); </span>
</li>
<li>
<span>add_to_cart( </span><span class="keyword">array</span><span>(</span><span class="string">'IPHONE3'</span><span> => 2 , </span><span class="string">'IPAD'</span><span> => 5) ); </span>
</li>
</ol>现在, 同个函数可以处理不同类型的输入参数了. 可以参照上面的例子重构你的多处代码, 使其更智能.
6. 有意忽略php关闭标签
我很想知道为什么这么多关于php建议的博客文章都没提到这点.
<ol class="dp-c"> <li class="alt"><span><span><?php </span></span></span></li> <li><span> </span></li> <li class="alt"> <span class="func">echo</span><span> </span><span class="string">"Hello"</span><span>; </span> </li> <li><span> </span></li> <li class="alt"> <span class="comment">//Now dont close this tag</span><span> </span> </li> </ol>
这將节约你很多时间. 我们举个例子:
一个 super_class.php 文件
<ol class="dp-c">
<li class="alt"><span><span><?php </span></span></span></li>
<li>
<span class="keyword">class</span><span> super_class </span>
</li>
<li class="alt"><span>{ </span></li>
<li>
<span> </span><span class="keyword">function</span><span> super_function() </span>
</li>
<li class="alt"><span> { </span></li>
<li>
<span> </span><span class="comment">//super code </span><span> </span>
</li>
<li class="alt"><span> } </span></li>
<li><span>} </span></li>
<li class="alt"><span>?> </span></li>
<li>
<span class="comment">//super extra character after the closing tag</span><span> </span>
</li>
</ol>index.php
<ol class="dp-c"> <li class="alt"><span><span class="keyword">require_once</span><span>(</span><span class="string">'super_class.php'</span><span>); </span></span></li> <li><span> </span></li> <li class="alt"> <span class="comment">//echo an image or pdf , or set the cookies or session data</span><span> </span> </li> </ol>
这样, 你將会得到一个 Headers already send error. 为什么? 因为 “super extra character” 已经被输出了. 现在你得开始调试啦. 这会花费大量时间寻找 super extra 的位置.
因此, 养成省略关闭符的习惯:
<ol class="dp-c">
<li class="alt"><span><span><?php </span></span></span></li>
<li>
<span class="keyword">class</span><span> super_class </span>
</li>
<li class="alt"><span>{ </span></li>
<li>
<span> </span><span class="keyword">function</span><span> super_function() </span>
</li>
<li class="alt"><span> { </span></li>
<li>
<span> </span><span class="comment">//super code </span><span> </span>
</li>
<li class="alt"><span> } </span></li>
<li><span>} </span></li>
<li class="alt"><span> </span></li>
<li>
<span class="comment">//No closing tag</span><span> </span>
</li>
</ol>这会更好.
7. 在某地方收集所有输入, 一次输出给浏览器
这称为输出缓冲, 假如说你已在不同的函数输出内容:
<ol class="dp-c">
<li class="alt"><span><span class="keyword">function</span><span> print_header() </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="func">echo</span><span> </span><span class="string">"<div>Site Log and Login links</div>"</span><span>; </span>
</li>
<li><span>} </span></li>
<li class="alt"><span> </span></li>
<li>
<span class="keyword">function</span><span> print_footer() </span>
</li>
<li class="alt"><span>{ </span></li>
<li>
<span> </span><span class="func">echo</span><span> </span><span class="string">"<div>Site was made by me</div>"</span><span>; </span>
</li>
<li class="alt"><span>} </span></li>
<li><span> </span></li>
<li class="alt"><span>print_header(); </span></li>
<li>
<span class="keyword">for</span><span>(</span><span class="vars">$i</span><span> = 0 ; </span><span class="vars">$i</span><span> <span class="vars">$i</span><span>++) </span></span>
</li>
<li class="alt"><span>{ </span></li>
<li>
<span> </span><span class="func">echo</span><span> "I is : </span><span class="vars">$i</span><span> <br>'; </span>
</li>
<li class="alt"><span>} </span></li>
<li><span>print_footer(); </span></li>
</ol><ol class="dp-c">
<li class="alt"><span><span class="keyword">function</span><span> print_header() </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="vars">$o</span><span> = </span><span class="string">"<div>Site Log and Login links</div>"</span><span>; </span>
</li>
<li>
<span> </span><span class="keyword">return</span><span> </span><span class="vars">$o</span><span>; </span>
</li>
<li class="alt"><span>} </span></li>
<li><span> </span></li>
<li class="alt">
<span class="keyword">function</span><span> print_footer() </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="vars">$o</span><span> = </span><span class="string">"<div>Site was made by me</div>"</span><span>; </span>
</li>
<li>
<span> </span><span class="keyword">return</span><span> </span><span class="vars">$o</span><span>; </span>
</li>
<li class="alt"><span>} </span></li>
<li><span> </span></li>
<li class="alt">
<span class="func">echo</span><span> print_header(); </span>
</li>
<li>
<span class="keyword">for</span><span>(</span><span class="vars">$i</span><span> = 0 ; </span><span class="vars">$i</span><span> <span class="vars">$i</span><span>++) </span></span>
</li>
<li class="alt"><span>{ </span></li>
<li>
<span> </span><span class="func">echo</span><span> "I is : </span><span class="vars">$i</span><span> <br>'; </span>
</li>
<li class="alt"><span>} </span></li>
<li>
<span class="func">echo</span><span> print_footer(); </span>
</li>
</ol>为什么需要输出缓冲:
>>可以在发送给浏览器前更改输出. 如 str_replaces 函数或可能是 preg_replaces 或添加些监控/调试的html内容.
>>输出给浏览器的同时又做php的处理很糟糕. 你应该看到过有些站点的侧边栏或中间出现错误信息. 知道为什么会发生吗? 因为处理和输出混合了.
8. 发送正确的mime类型头信息, 如果输出非html内容的话.
输出一些xml.
<ol class="dp-c"> <li class="alt"><span><span class="vars">$xml</span><span> = </span><span class="string">'<?xml version="1.0" encoding="utf-8" standalone="yes"?>'</span><span>; </span></span></li> <li> <span class="vars">$xml</span><span> = "<response> </response></span> </li> <li class="alt"><span> <code>0</code> </span></li> <li><span>"; </span></li> <li class="alt"><span> </span></li> <li> <span class="comment">//Send xml data </span><span> </span> </li> <li class="alt"> <span class="func">echo</span><span> </span><span class="vars">$xml</span><span>; </span> </li> </ol>
工作得不错. 但需要一些改进.
<ol class="dp-c"> <li class="alt"><span><span class="vars">$xml</span><span> = </span><span class="string">'<?xml version="1.0" encoding="utf-8" standalone="yes"?>'</span><span>; </span></span></li> <li> <span class="vars">$xml</span><span> = "<response> </response></span> </li> <li class="alt"><span> <code>0</code> </span></li> <li><span>"; </span></li> <li class="alt"><span> </span></li> <li> <span class="comment">//Send xml data </span><span> </span> </li> <li class="alt"> <span>header(</span><span class="string">"content-type: text/xml"</span><span>); </span> </li> <li> <span class="func">echo</span><span> </span><span class="vars">$xml</span><span>; </span> </li> </ol>
注意header行. 该行告知浏览器发送的是xml类型的内容. 所以浏览器能正确的处理. 很多的javascript库也依赖头信息.
类似的有 javascript , css, jpg image, png image:
JavaScript
<ol class="dp-c"> <li class="alt"><span><span>header(</span><span class="string">"content-type: application/x-javascript"</span><span>); </span></span></li> <li> <span class="func">echo</span><span> </span><span class="string">"var a = 10"</span><span>; </span> </li> </ol>
CSS
<ol class="dp-c">
<li class="alt"><span><span>header(</span><span class="string">"content-type: text/css"</span><span>); </span></span></li>
<li>
<span class="func">echo</span><span> </span><span class="string">"#div id { background:#000; }"</span><span>; </span>
</li>
</ol>9. 为mysql连接设置正确的字符编码
曾经遇到过在mysql表中设置了unicode/utf-8编码, phpadmin也能正确显示, 但当你获取内容并在页面输出的时候,会出现乱码. 这里的问题出在mysql连接的字符编码.
<ol class="dp-c">
<li class="alt"><span><span class="comment">//Attempt to connect to database </span><span> </span></span></li>
<li>
<span class="vars">$c</span><span> = mysqli_connect(</span><span class="vars">$this</span><span>->host , </span><span class="vars">$this</span><span>->username, </span><span class="vars">$this</span><span>->password); </span>
</li>
<li class="alt"><span> </span></li>
<li>
<span class="comment">//Check connection validity </span><span> </span>
</li>
<li class="alt">
<span class="keyword">if</span><span> (!</span><span class="vars">$c</span><span>) </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="keyword">die</span><span> (</span><span class="string">"Could not connect to the database host: <br>"</span><span>. mysqli_connect_error()); </span>
</li>
<li><span>} </span></li>
<li class="alt"><span> </span></li>
<li>
<span class="comment">//Set the character set of the connection </span><span> </span>
</li>
<li class="alt">
<span class="keyword">if</span><span>(!mysqli_set_charset ( </span><span class="vars">$c</span><span> , </span><span class="string">'UTF8'</span><span> )) </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="keyword">die</span><span>(</span><span class="string">'mysqli_set_charset() failed'</span><span>); </span>
</li>
<li><span>} </span></li>
</ol>一旦连接数据库, 最好设置连接的 characterset. 你的应用如果要支持多语言, 这么做是必须的.
10. 使用 htmlentities 设置正确的编码选项
php5.4前, 字符的默认编码是ISO-8859-1, 不能直接输出如À â等.
<ol class="dp-c"><li class="alt"><span><span class="vars">$value</span><span> = htmlentities(</span><span class="vars">$this</span><span>->value , ENT_QUOTES , CHARSET); </span></span></li></ol>
php5.4以后, 默认编码为UTF-8, 这將解决很多问题. 但如果你的应用是多语言的, 仍然要留意编码问题,.
11. 不要在应用中使用gzip压缩输出, 让apache处理
考虑过使用 ob_gzhandler 吗? 不要那样做. 毫无意义. php只应用来编写应用. 不应操心服务器和浏览器的数据传输优化问题.
使用apache的mod_gzip/mod_deflate 模块压缩内容.
12. 使用json_encode输出动态javascript内容
时常会用php输出动态javascript内容:
<ol class="dp-c">
<li class="alt"><span><span class="vars">$images</span><span> = </span><span class="keyword">array</span><span>( </span></span></li>
<li>
<span> </span><span class="string">'myself.png'</span><span> , </span><span class="string">'friends.png'</span><span> , </span><span class="string">'colleagues.png'</span><span> </span>
</li>
<li class="alt"><span>); </span></li>
<li><span> </span></li>
<li class="alt">
<span class="vars">$js_code</span><span> = </span><span class="string">''</span><span>; </span>
</li>
<li><span> </span></li>
<li class="alt">
<span class="keyword">foreach</span><span>(</span><span class="vars">$images</span><span> </span><span class="keyword">as</span><span> </span><span class="vars">$image</span><span>) </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span class="vars">$js_code</span><span> .= </span><span class="string">"'$image' ,"</span><span>; </span>
</li>
<li><span>} </span></li>
<li class="alt"><span> </span></li>
<li>
<span class="vars">$js_code</span><span> = </span><span class="string">'var images = ['</span><span> . </span><span class="vars">$js_code</span><span> . </span><span class="string">']; '</span><span>; </span>
</li>
<li class="alt"><span> </span></li>
<li>
<span class="func">echo</span><span> </span><span class="vars">$js_code</span><span>; </span>
</li>
<li class="alt"><span> </span></li>
<li>
<span class="comment">//Output is var images = ['myself.png' ,'friends.png' ,'colleagues.png' ,];</span><span> </span>
</li>
</ol>更聪明的做法, 使用 json_encode:
<ol class="dp-c"> <li class="alt"><span><span class="vars">$images</span><span> = </span><span class="keyword">array</span><span>( </span></span></li> <li> <span> </span><span class="string">'myself.png'</span><span> , </span><span class="string">'friends.png'</span><span> , </span><span class="string">'colleagues.png'</span><span> </span> </li> <li class="alt"><span>); </span></li> <li><span> </span></li> <li class="alt"> <span class="vars">$js_code</span><span> = </span><span class="string">'var images = '</span><span> . json_encode(</span><span class="vars">$images</span><span>); </span> </li> <li><span> </span></li> <li class="alt"> <span class="func">echo</span><span> </span><span class="vars">$js_code</span><span>; </span> </li> <li><span> </span></li> <li class="alt"> <span class="comment">//Output is : var images = ["myself.png","friends.png","colleagues.png"]</span><span> </span> </li> </ol>
优雅乎?
13. 写文件前, 检查目录写权限
写或保存文件前, 确保目录是可写的, 假如不可写, 输出错误信息. 这会节约你很多调试时间. linux系统中, 需要处理权限, 目录权限不当会导致很多很多的问题, 文件也有可能无法读取等等.
确保你的应用足够智能, 输出某些重要信息.
<ol class="dp-c"> <li class="alt"><span><span class="vars">$contents</span><span> = </span><span class="string">"All the content"</span><span>; </span></span></li> <li> <span class="vars">$file_path</span><span> = </span><span class="string">"/var/www/project/content.txt"</span><span>; </span> </li> <li class="alt"><span> </span></li> <li> <span class="func">file_put_contents</span><span>(</span><span class="vars">$file_path</span><span> , </span><span class="vars">$contents</span><span>); </span> </li> </ol>
这大体上正确. 但有些间接的问题. file_put_contents 可能会由于几个原因失败:
>>父目录不存在
>>目录存在, 但不可写
>>文件被写锁住?
所以写文件前做明确的检查更好.
<ol class="dp-c">
<li class="alt"><span><span class="vars">$contents</span><span> = </span><span class="string">"All the content"</span><span>; </span></span></li>
<li>
<span class="vars">$dir</span><span> = </span><span class="string">'/var/www/project'</span><span>; </span>
</li>
<li class="alt">
<span class="vars">$file_path</span><span> = </span><span class="vars">$dir</span><span> . </span><span class="string">"/content.txt"</span><span>; </span>
</li>
<li><span> </span></li>
<li class="alt">
<span class="keyword">if</span><span>(</span><span class="func">is_writable</span><span>(</span><span class="vars">$dir</span><span>)) </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="func">file_put_contents</span><span>(</span><span class="vars">$file_path</span><span> , </span><span class="vars">$contents</span><span>); </span>
</li>
<li><span>} </span></li>
<li class="alt">
<span class="keyword">else</span><span> </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="keyword">die</span><span>(</span><span class="string">"Directory $dir is not writable, or does not exist. Please check"</span><span>); </span>
</li>
<li><span>} </span></li>
</ol>这么做后, 你会得到一个文件在何处写及为什么失败的明确信息.
14. 更改应用创建的文件权限
在 linux环境中, 权限问题可能会浪费你很多时间. 从今往后, 无论何时, 当你创建一些文件后, 确保使用chmod设置正确权限. 否则的话, 可能文件先是由"php"用户创建, 但你用其它的用户登录工作, 系统將会拒绝访问或打开文件, 你不得不奋力获取root权限, 更改文件的权限等等.
<ol class="dp-c"> <li class="alt"><span><span class="comment">// Read and write for owner, read for everybody else </span><span> </span></span></li> <li> <span class="func">chmod</span><span>(</span><span class="string">"/somedir/somefile"</span><span>, 0644); </span> </li> <li class="alt"><span> </span></li> <li> <span class="comment">// Everything for owner, read and execute for others </span><span> </span> </li> <li class="alt"> <span class="func">chmod</span><span>(</span><span class="string">"/somedir/somefile"</span><span>, 0755); </span> </li> </ol>
15. 不要依赖submit按钮值来检查表单提交行为
<ol class="dp-c">
<li class="alt"><span><span class="keyword">if</span><span>(</span><span class="vars">$_POST</span><span>[</span><span class="string">'submit'</span><span>] == </span><span class="string">'Save'</span><span>) </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="comment">//Save the things </span><span> </span>
</li>
<li><span>} </span></li>
</ol>上面大多数情况正确, 除了应用是多语言的. 'Save' 可能代表其它含义. 你怎么区分它们呢. 因此, 不要依赖于submit按钮的值.
<ol class="dp-c">
<li class="alt"><span><span class="keyword">if</span><span>( </span><span class="vars">$_SERVER</span><span>[</span><span class="string">'REQUEST_METHOD'</span><span>] == </span><span class="string">'POST'</span><span> </span><span class="keyword">and</span><span> isset(</span><span class="vars">$_POST</span><span>[</span><span class="string">'submit'</span><span>]) ) </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="comment">//Save the things </span><span> </span>
</li>
<li><span>} </span></li>
</ol>现在你从submit按钮值中解脱出来了.
16. 为函数内总具有相同值的变量定义成静态变量
<ol class="dp-c">
<li class="alt"><span><span class="comment">//Delay for some time </span><span> </span></span></li>
<li>
<span class="keyword">function</span><span> delay() </span>
</li>
<li class="alt"><span>{ </span></li>
<li>
<span> </span><span class="vars">$sync_delay</span><span> = get_option(</span><span class="string">'sync_delay'</span><span>); </span>
</li>
<li class="alt"><span> </span></li>
<li>
<span> </span><span class="func">echo</span><span> </span><span class="string">"<br>Delaying for $sync_delay seconds..."</span><span>; </span>
</li>
<li class="alt">
<span> sleep(</span><span class="vars">$sync_delay</span><span>); </span>
</li>
<li>
<span> </span><span class="func">echo</span><span> </span><span class="string">"Done <br>"</span><span>; </span>
</li>
<li class="alt"><span>} </span></li>
</ol>用静态变量取代:
<ol class="dp-c">
<li class="alt"><span><span class="comment">//Delay for some time </span><span> </span></span></li>
<li>
<span class="keyword">function</span><span> delay() </span>
</li>
<li class="alt"><span>{ </span></li>
<li>
<span> </span><span class="keyword">static</span><span> </span><span class="vars">$sync_delay</span><span> = null; </span>
</li>
<li class="alt"><span> </span></li>
<li>
<span> </span><span class="keyword">if</span><span>(</span><span class="vars">$sync_delay</span><span> == null) </span>
</li>
<li class="alt"><span> { </span></li>
<li>
<span> </span><span class="vars">$sync_delay</span><span> = get_option(</span><span class="string">'sync_delay'</span><span>); </span>
</li>
<li class="alt"><span> } </span></li>
<li><span> </span></li>
<li class="alt">
<span> </span><span class="func">echo</span><span> </span><span class="string">"<br>Delaying for $sync_delay seconds..."</span><span>; </span>
</li>
<li>
<span> sleep(</span><span class="vars">$sync_delay</span><span>); </span>
</li>
<li class="alt">
<span> </span><span class="func">echo</span><span> </span><span class="string">"Done <br>"</span><span>; </span>
</li>
<li><span>} </span></li>
</ol>17. 不要直接使用 $_SESSION 变量
<ol class="dp-c"> <li class="alt"><span><span class="vars">$_SESSION</span><span>[</span><span class="string">'username'</span><span>] = </span><span class="vars">$username</span><span>; </span></span></li> <li> <span class="vars">$username</span><span> = </span><span class="vars">$_SESSION</span><span>[</span><span class="string">'username'</span><span>]; </span> </li> </ol>
这会导致某些问题. 如果在同个域名中运行了多个应用, session 变量可能会冲突. 两个不同的应用可能使用同一个session key. 例如, 一个前端门户, 和一个后台管理系统使用同一域名.
从现在开始, 使用应用相关的key和一个包装函数:
<ol class="dp-c">
<li class="alt"><span><span>define(</span><span class="string">'APP_ID'</span><span> , </span><span class="string">'abc_corp_ecommerce'</span><span>); </span></span></li>
<li><span> </span></li>
<li class="alt">
<span class="comment">//Function to get a session variable </span><span> </span>
</li>
<li>
<span class="keyword">function</span><span> session_get(</span><span class="vars">$key</span><span>) </span>
</li>
<li class="alt"><span>{ </span></li>
<li>
<span> </span><span class="vars">$k</span><span> = APP_ID . </span><span class="string">'.'</span><span> . </span><span class="vars">$key</span><span>; </span>
</li>
<li class="alt"><span> </span></li>
<li>
<span> </span><span class="keyword">if</span><span>(isset(</span><span class="vars">$_SESSION</span><span>[</span><span class="vars">$k</span><span>])) </span>
</li>
<li class="alt"><span> { </span></li>
<li>
<span> </span><span class="keyword">return</span><span> </span><span class="vars">$_SESSION</span><span>[</span><span class="vars">$k</span><span>]; </span>
</li>
<li class="alt"><span> } </span></li>
<li><span> </span></li>
<li class="alt">
<span> </span><span class="keyword">return</span><span> false; </span>
</li>
<li><span>} </span></li>
<li class="alt"><span> </span></li>
<li>
<span class="comment">//Function set the session variable </span><span> </span>
</li>
<li class="alt">
<span class="keyword">function</span><span> session_set(</span><span class="vars">$key</span><span> , </span><span class="vars">$value</span><span>) </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="vars">$k</span><span> = APP_ID . </span><span class="string">'.'</span><span> . </span><span class="vars">$key</span><span>; </span>
</li>
<li>
<span> </span><span class="vars">$_SESSION</span><span>[</span><span class="vars">$k</span><span>] = </span><span class="vars">$value</span><span>; </span>
</li>
<li class="alt"><span> </span></li>
<li>
<span> </span><span class="keyword">return</span><span> true; </span>
</li>
<li class="alt"><span>} </span></li>
</ol>18. 將工具函数封装到类中
假如你在某文件中定义了很多工具函数:
<ol class="dp-c">
<li class="alt"><span><span class="keyword">function</span><span> utility_a() </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="comment">//This function does a utility thing like string processing </span><span> </span>
</li>
<li><span>} </span></li>
<li class="alt"><span> </span></li>
<li>
<span class="keyword">function</span><span> utility_b() </span>
</li>
<li class="alt"><span>{ </span></li>
<li>
<span> </span><span class="comment">//This function does nother utility thing like database processing </span><span> </span>
</li>
<li class="alt"><span>} </span></li>
<li><span> </span></li>
<li class="alt">
<span class="keyword">function</span><span> utility_c() </span>
</li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="comment">//This function is ... </span><span> </span>
</li>
<li><span>} </span></li>
</ol>这些函数的使用分散到应用各处. 你可能想將他们封装到某个类中:
<ol class="dp-c">
<li class="alt"><span><span class="keyword">class</span><span> Utility </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="keyword">public</span><span> </span><span class="keyword">static</span><span> </span><span class="keyword">function</span><span> utility_a() </span>
</li>
<li><span> { </span></li>
<li class="alt"><span> </span></li>
<li><span> } </span></li>
<li class="alt"><span> </span></li>
<li>
<span> </span><span class="keyword">public</span><span> </span><span class="keyword">static</span><span> </span><span class="keyword">function</span><span> utility_b() </span>
</li>
<li class="alt"><span> { </span></li>
<li><span> </span></li>
<li class="alt"><span> } </span></li>
<li><span> </span></li>
<li class="alt">
<span> </span><span class="keyword">public</span><span> </span><span class="keyword">static</span><span> </span><span class="keyword">function</span><span> utility_c() </span>
</li>
<li><span> { </span></li>
<li class="alt"><span> </span></li>
<li><span> } </span></li>
<li class="alt"><span>} </span></li>
<li><span> </span></li>
<li class="alt">
<span class="comment">//and call them as </span><span> </span>
</li>
<li><span> </span></li>
<li class="alt">
<span class="vars">$a</span><span> = Utility::utility_a(); </span>
</li>
<li>
<span class="vars">$b</span><span> = Utility::utility_b(); </span>
</li>
</ol>显而易见的好处是, 如果php内建有同名的函数, 这样可以避免冲突.
另一种看法是, 你可以在同个应用中为同个类维护多个版本, 而不导致冲突. 这是封装的基本好处, 无它.
19. Bunch of silly tips
>>使用echo取代print
>>使用str_replace取代preg_replace, 除非你绝对需要
>>不要使用 short tag
>>简单字符串用单引号取代双引号
>>head重定向后记得使用exit
>>不要在循环中调用函数
>>isset比strlen快
>>始中如一的格式化代码
>>不要删除循环或者if-else的括号
不要这样写代码:
<ol class="dp-c"><li class="alt"><span><span><span><span class="string">"color:#333333;font-family:''Helvetica, Arial, sans-serif'';"</span><span>></span><span class="keyword">if</span><span>(</span><span class="vars">$a</span><span> == true) </span><span class="vars">$a_count</span><span>++;</span> </span></span></span></li></ol>
这绝对WASTE.
写成:
<ol class="dp-c">
<li class="alt"><span><span><span><span class="string">"color:#333333;font-family:''Helvetica, Arial, sans-serif'';"</span><span>></span><span class="keyword">if</span><span>(</span><span class="vars">$a</span><span> == true) </span></span></span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="vars">$a_count</span><span>++; </span>
</li>
<li>
<span>}</span> </li>
</ol>不要尝试省略一些语法来缩短代码. 而是让你的逻辑简短.
>>使用有高亮语法显示的文本编辑器. 高亮语法能让你减少错误.
20. 使用array_map快速处理数组
比如说你想 trim 数组中的所有元素. 新手可能会:
<ol class="dp-c">
<li class="alt"><span><span class="keyword">foreach</span><span>(</span><span class="vars">$arr</span><span> </span><span class="keyword">as</span><span> </span><span class="vars">$c</span><span> => </span><span class="vars">$v</span><span>) </span></span></li>
<li><span>{ </span></li>
<li class="alt">
<span> </span><span class="vars">$arr</span><span>[</span><span class="vars">$c</span><span>] = trim(</span><span class="vars">$v</span><span>); </span>
</li>
<li><span>} </span></li>
</ol>但使用 array_map 更简单:
<ol class="dp-c"><li class="alt"><span><span class="vars">$arr</span><span> = </span><span class="func">array_map</span><span>(</span><span class="string">'trim'</span><span> , </span><span class="vars">$arr</span><span>); </span></span></li></ol>
这会为$arr数组的每个元素都申请调用trim. 另一个类似的函数是 array_walk. 请查阅文档学习更多技巧.
21. 使用 php filter 验证数据
你肯定曾使用过正则表达式验证 email , ip地址等. 是的,每个人都这么使用. 现在, 我们想做不同的尝试, 称为filter.
php的filter扩展提供了简单的方式验证和检查输入.
22. 强制类型检查
<ol class="dp-c"> <li class="alt"><span><span class="vars">$amount</span><span> = </span><span class="func">intval</span><span>( </span><span class="vars">$_GET</span><span>[</span><span class="string">'amount'</span><span>] ); </span></span></li> <li> <span class="vars">$rate</span><span> = (int) </span><span class="vars">$_GET</span><span>[</span><span class="string">'rate'</span><span>]; </span> </li> </ol>
这是个好习惯.
23. 如果需要,使用profiler如xdebug
如果你使用php开发大型的应用, php承担了很多运算量, 速度会是一个很重要的指标. 使用profile帮助优化代码. 可使用
xdebug和webgrid.
24. 小心处理大数组
对于大的数组和字符串, 必须小心处理. 常见错误是发生数组拷贝导致内存溢出,抛出Fat
<ol class="dp-c"> <li class="alt"><span><span class="vars">$db_records_in_array_format</span><span>; </span><span class="comment">//This is a big array holding 1000 rows from a table each having 20 columns , every row is atleast 100 bytes , so total 1000 * 20 * 100 = 2MB </span><span> </span></span></li> <li><span> </span></li> <li class="alt"> <span class="vars">$cc</span><span> = </span><span class="vars">$db_records_in_array_format</span><span>; </span><span class="comment">//2MB more </span><span> </span> </li> <li><span> </span></li> <li class="alt"> <span>some_function(</span><span class="vars">$cc</span><span>); </span><span class="comment">//Another 2MB ?</span><span> </span> </li> </ol>
当导入或导出csv文件时, 常常会这么做.
不要认为上面的代码会经常因内存限制导致脚本崩溃. 对于小的变量是没问题的, 但处理大数组的时候就必须避免.
确保通过引用传递, 或存储在类变量中:
<ol class="dp-c"> <li class="alt"><span><span class="vars">$a</span><span> = get_large_array(); </span></span></li> <li> <span>pass_to_function(&</span><span class="vars">$a</span><span>); </span> </li> </ol>
这么做后,向函数传递变量引用(而不是拷贝数组). 查看
