正确格式化这些用户代理
我有一个包含用户代理数组的缓存文件。
当我打印出每个用户代理时，它看起来是这样的：
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\).*Chrome/6\..*Safari/.*$@
@^Mozilla/5\.0 \(X11; U; Linux.*; .*; rv\:1\.9\..*\) Gecko/.* Firefox/.* Maemo Browser 1\.7\..*$@
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\).*Chrome/7\..*Safari/.*$@
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\).*Chrome/8\..*Safari/.*$@
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\) .*Iron/6\..* Safari/.*$@
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\) .*Iron/7\..* Safari/.*$@
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\).*Chrome/9\..*Safari/.*$@
ect ect....
现在看起来这些用户代理已经被转义了，所以我已经stripslashes()它们了。
我还注意到这里有一些无效字符（@、^、.*$@），所以我 str_replaced( ) 他们什么都没有。
最终结果：
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko).Chrome/6..Safari/
Mozilla/5.0 (X11; U; Linux.; .; rv:1.9..) Gecko/. Firefox/. Maemo Browser 1.7.
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko).Chrome/7..Safari/
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko).Chrome/8..Safari/
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko) .Iron/6.. Safari/
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko) .Iron/7.. Safari/
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko).Chrome/9..Safari/
我计划将这些用户代理插入数据库（已经充满了数十万个），我担心它们在某些方面仍然无效，而我没有注意到。
如果我确实将这些插入数据库并且它们无效，那么我将不得不删除整个数据库，然后重新开始...
所以我不想冒任何风险...
你们能发现什么问题吗使用这些用户代理字符串？
:)
I have a cache file with an array of user agents.
When i print each user agent out, this is what it looks like:
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\).*Chrome/6\..*Safari/.*$@
@^Mozilla/5\.0 \(X11; U; Linux.*; .*; rv\:1\.9\..*\) Gecko/.* Firefox/.* Maemo Browser 1\.7\..*$@
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\).*Chrome/7\..*Safari/.*$@
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\).*Chrome/8\..*Safari/.*$@
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\) .*Iron/6\..* Safari/.*$@
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\) .*Iron/7\..* Safari/.*$@
@^Mozilla/5\.0 \(.*Linux i686.*\) AppleWebKit/.* \(KHTML, like Gecko\).*Chrome/9\..*Safari/.*$@
ect ect....
Now it looks like these user-agents have been escaped, so I've stripslashes() them.
I've also noticed some invalid characters here (@,^,.*$@), so I've str_replaced() them with nothing.
And the end result:
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko).Chrome/6..Safari/
Mozilla/5.0 (X11; U; Linux.; .; rv:1.9..) Gecko/. Firefox/. Maemo Browser 1.7.
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko).Chrome/7..Safari/
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko).Chrome/8..Safari/
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko) .Iron/6.. Safari/
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko) .Iron/7.. Safari/
Mozilla/5.0 (.Linux i686.) AppleWebKit/. (KHTML, like Gecko).Chrome/9..Safari/
I'm planning to insert these user-agents into a database (already full of hundreds of thousands), and I'm worried they're still invalid in some way that I haven't noticed.
If I do insert these into the DB and they're invalid, I would then have to delete the whole DB, and start all over...
So I don't want to take any risks...
Can you guys spot anything wrong with these user-agent strings?
:)
如果你对这篇内容有疑问，欢迎到本站社区发帖提问 参与讨论，获取更多帮助，或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱，如果其他用户或者作者回复了您的评论，将不能在第一时间通知您！
发布评论
评论（1）
这些是正则表达式，而不是无效字符。如果您只是将字符串与用户代理字符串进行匹配，它们将不会匹配。它们旨在作为正则表达式进行匹配。
These are regular expressions, not invalid characters. If you simply match the strings against the user agent string, they will not match. They are meant to be matched as regular expressions.