test.htm

this is very good for teacher

<base href="ftp://ftp.cs.pku.edu.cn/pub/games/../././incoming/document/rfc2616.pdf?search=key#SEC1"></base>
<A c="black    "href="../games/./././../software/../new_software/nasm.tar.gz">
</a>
<a href="a/a/a/a/a/a/b/b/b/b/b/../../../c/c/c/./././.././.././.././11/11/22/33/44/55/../.././xx/yy/zz?key=linux#XYZ   "color= yellow></a>
<base href="   http://www.pku.edu.cn/news/index.html     "></base>
<a color = red href="./.././.././.././../announce/././index.htm    "></a>
<a color=brown href=  "http://www.sohu.com/sports/index.htm"></a>
<A href=    mailto:[e-mail] other="some value"></a>
<a href  =     http://www.tsinghua.edu.cn></a>

<!-- The following is an illegal URI -->
<a href="./../student^/index.htm"></a>

<!-- The followings have grammar errors -->
<a href=http://www.sina.com.cn"></a>
<A href="http://www.cctv.com   ></a>
<a href=http://www.google.com</a>
<A href=http://www.yahoo.com
<a href="www.pku .edu.cn"></a>

<!-- The following is not right, but we tolerate it -->
<a href="http://www.kernel.org?abc#xyz" color="></a>

<!-- The one cannot be tolerated! -->
<a name= " href="http://www.fifa.com"></a>

<!-- This one is legal -->
<a href =></a>

<!-- Attribute names are case-insensitive. The following will been accpted -->
<a HReF=http://name:[e-mail]:8080></a>

<!-- changed the base to some strange value -->
<BasE href=aaa/.></base>
<!-- The following will result in an empty URI which is legal. -->
<A href="   ../../../.."></a>

<a href=hoho></a>
<a href=//a/b/c?name=xiehan#SEC1></a>
<a href=xyz>

<!-- This one is illegal -->
<a href=?></a>

<!-- This is right -->
<a href=/?></a>

<!-- This is not right -->
<a href=/ color></a>

<!-- "LINK" element is treated same as "A">
<link href="index.htm"></link>

<!-- how about this? -->
<a href="hoho^^" hREf="haha"></a>

<a href="http://a.b.c"></a>

<!-- Note! This URI is legal but the authority is a reg_name, i.e.,
     uri.authority_type = AT_REG_NAME, uri.reg_name = "user:passwd@123.222".
	 Because "123.456" is NOT legal host name. Such URI will be filtered
	 in our crawl program. Further more, parsing such URI consumes more time
	 than normal URI because reg_name's prefix conflicts with server.
	 Fortunately such URI is seldom found in pages, almost never. -->
<href=http://user:passwd@123.222/xyz></a>
<!-- The following URI's authority is a reg_name, not userinfo+host+port,
	because "$" is a illegal character in host name. -->
<a href="http://myinfo:[e-mail]$oft.com:8080/abc/xyz/index.htm?name#SEC1">
</a>
<!-- This is legal and authority is a server, i.e.,
	 uri.authority_tye = AT_SERVER, uri.userinfo = "user:passwd",
	 uri.host = "162.105.216.124", uri.port = "8080".
	 Because "162.105.216.124" is interpreted as an IPv4 address. -->
<a href="   http://user:passwd@162.105.216.124:8080/pub/incoming/index.htm?name=xiehan#SEC1

"></a>

<!-- IPv6 address are not allowed (RFC 2396) the following will be
	 interpreted as a reg_name -->
<a href="http://fff::123:78a/index.htm"></a>