自动加载的陷阱 class_exists缺省情况下会触发autoload，如果你没注意到这点的话很可能会吃亏，演示代码： spl_autoload_register(function($name) { echo $name; }); class_exists('foo'); 你可以通过in_array($name, get_declared_classes())函数来判断是
自动加载的陷阱
class_exists缺省情况下会触发autoload，如果你没注意到这点的话很可能会吃亏，演示代码：
spl_autoload_register(function($name) { echo $name; }); class_exists('foo');
你可以通过in_array($name, get_declared_classes())函数来判断是否存在相关的class，这样不会触发autoload，不过稍显笨重，其实class_exists()函数本身可以不触发autoload，方法是第二个参数：class_exists('foo', false);，不过老实说，当初设计的时候缺省值是true实在是个错误的决定。
BTW：method_exists也要注意，不过它没有类似class_exists那样能关闭autoload的参数控制，这一点在手册里已经明确写出来了，需要注意：
Note: Using this function will use any registered autoloaders if the class is not already known.
所以如果你不想触发autoload，那么在使用method_exists之前，必须确保对应的类已经加载，否则就没戏了。
缓存代码的重复味道
缓存在Web程序里必不可少，最常见的形式如下：
<p><span>01</span><span>class</span><span> Foo </span><span>extends</span><span> DAO<br></span><span>02</span><span> {<br></span><span>03</span><span>public</span><span>function</span><span> find_by_a()<br></span><span>04</span><span> {<br></span><span>05</span><span>$result</span><span>=</span><span>$this</span><span>-></span><span>cache</span><span>-></span><span>get(</span><span>'</span><span>cache_a</span><span>'</span><span>);<br></span><span>06</span><span><br></span><span>07</span><span>if</span><span> (</span><span>!</span><span>$result</span><span>) {<br></span><span>08</span><span>$result</span><span>=</span><span>$this</span><span>-></span><span>db</span><span>-></span><span>getAll(</span><span>'</span><span>select ... from ... where a ...</span><span>'</span><span>);<br></span><span>09</span><span><br></span><span>10</span><span>$this</span><span>-></span><span>cache</span><span>-></span><span>set(</span><span>'</span><span>cache_a</span><span>'</span><span>,</span><span>$result</span><span>);<br></span><span>11</span><span> }<br></span><span>12</span><span><br></span><span>13</span><span>return</span><span>$result</span><span>;<br></span><span>14</span><span> }<br></span><span>15</span><span><br></span><span>16</span><span>public</span><span>function</span><span> find_by_b()<br></span><span>17</span><span> {<br></span><span>18</span><span>$result</span><span>=</span><span>$this</span><span>-></span><span>cache</span><span>-></span><span>get(</span><span>'</span><span>cache_b</span><span>'</span><span>);<br></span><span>19</span><span><br></span><span>20</span><span>if</span><span> (</span><span>!</span><span>$result</span><span>) {<br></span><span>21</span><span>$result</span><span>=</span><span>$this</span><span>-></span><span>db</span><span>-></span><span>getAll(</span><span>'</span><span>select ... from ... where b ...</span><span>'</span><span>);<br></span><span>22</span><span><br></span><span>23</span><span>$this</span><span>-></span><span>cache</span><span>-></span><span>set(</span><span>'</span><span>cache_b</span><span>'</span><span>,</span><span>$result</span><span>);<br></span><span>24</span><span> }<br></span><span>25</span><span><br></span><span>26</span><span>return</span><span>$result</span><span>;<br></span><span>27</span><span> }<br></span><span>28</span><span> }</span></p>这个代码很平常，实际情况中，多数人差不多都是这么写代码，先用某个键在缓存里取一下，如果没有就从数据库里实际查询一次，并且把结果缓存起来，这样的代码虽然不够健壮（没有捕捉可能存在的异常），不过本身并没有太大问题，但是若干个方法叠加起来，我们就能明显的感受到坏味道：重复！不说废话了哦，直接给出解决方案：
<p><span>01</span><span>abstract</span><span>class</span><span> DAO<br></span><span>02</span><span> {<br></span><span>03</span><span>public</span><span>function</span><span> getCache(</span><span>$key</span><span>,</span><span>$closure</span><span>)<br></span><span>04</span><span> {<br></span><span>05</span><span>$result</span><span>=</span><span>$this</span><span>-></span><span>cache</span><span>-></span><span>get(</span><span>$key</span><span>);<br></span><span>06</span><span><br></span><span>07</span><span>if</span><span> (</span><span>!</span><span>$result</span><span>) {<br></span><span>08</span><span>$result</span><span>=</span><span>$closure</span><span>();<br></span><span>09</span><span><br></span><span>10</span><span>$this</span><span>-></span><span>cache</span><span>-></span><span>set(</span><span>$key</span><span>,</span><span>$result</span><span>);<br></span><span>11</span><span> }<br></span><span>12</span><span><br></span><span>13</span><span>return</span><span>$result</span><span>;<br></span><span>14</span><span> }<br></span><span>15</span><span> }<br></span><span>16</span><span><br></span><span>17</span><span>class</span><span> Foo </span><span>extends</span><span> DAO<br></span><span>18</span><span> {<br></span><span>19</span><span>public</span><span>function</span><span> find_by_a()<br></span><span>20</span><span> {<br></span><span>21</span><span>return</span><span>$this</span><span>-></span><span>getCache(</span><span>'</span><span>cache_a</span><span>'</span><span>,</span><span>function</span><span>() {<br></span><span>22</span><span>return</span><span>$this</span><span>-></span><span>db</span><span>-></span><span>getAll(</span><span>'</span><span>select ... from ... where a ...</span><span>'</span><span>);<br></span><span>23</span><span> });<br></span><span>24</span><span> }<br></span><span>25</span><span><br></span><span>26</span><span>public</span><span>function</span><span> find_by_b()<br></span><span>27</span><span> {<br></span><span>28</span><span>return</span><span>$this</span><span>-></span><span>getCache(</span><span>'</span><span>cache_b</span><span>'</span><span>,</span><span>function</span><span>() {<br></span><span>29</span><span>return</span><span>$this</span><span>-></span><span>db</span><span>-></span><span>getAll(</span><span>'</span><span>select ... from ... where b ...</span><span>'</span><span>);<br></span><span>30</span><span> });<br></span><span>31</span><span> }<br></span><span>32</span><span> }</span></p>代码有点简陋，通过把非公共代码提取成一个closure，传递给getCache方法，从而消除了重复的坏味道。
