当从代码中调用web资源时，一个常见的任务是构建一个包含所有必要参数的查询字符串。虽然这绝不是火箭科学，但有一些漂亮的细节需要注意，例如，如果不是第一个参数，则添加&，对参数进行编码等。
实现它的代码非常简单，但有点乏味:
StringBuilder SB = new StringBuilder();
if (NeedsToAddParameter A)
{
SB.Append("A="); SB.Append(HttpUtility.UrlEncode("TheValueOfA"));
}
if (NeedsToAddParameter B)
{
if (SB.Length>0) SB.Append("&");
SB.Append("B="); SB.Append(HttpUtility.UrlEncode("TheValueOfB")); }
}
这是一个非常常见的任务，人们希望存在一个实用工具类，使其更加优雅和可读。扫描MSDN，我没有找到一个，这让我想到了以下问题:
你所知道的最优雅干净的方法是什么?
当前回答
这里有很多很好的答案，但对于使用现代c#的人来说，这可能是一个很好的实用程序类。
public class QueryParamBuilder
{
private readonly Dictionary<string, string> _fields = new();
public QueryParamBuilder Add(string key, string value)
{
_fields.Add(key, value);
return this;
}
public string Build()
{
return $"?{String.Join("&", _fields.Select(pair => $"{HttpUtility.UrlEncode(pair.Key)}={HttpUtility.UrlEncode(pair.Value)}"))}";
}
public static QueryParamBuilder New => new();
}
我在这里使用了一个内部Dictionary，因为字典在内部是可枚举的键值对，这使得遍历它们比NameValueCollection容易得多。
那么查询字符串本身就是一个简单的带有连接的插值字符串。
另外，我为构造函数提供了一个静态接口，使构造新的构造器非常容易，并且只允许一个公开的方法Add来添加新的查询参数值。最后，使用Build()终止链以实际获得最终字符串。
下面是它用法的一个例子
var queryString = QueryParamBuilder.New
.Add("id", "0123")
.Add("value2", 1234.ToString())
.Add("valueWithSpace","value with spa12!@#@!ce")
.Build();
结果和预期的一样
?id=0123&value2=1234&valueWithSpace=value+with+spa12!%40%23%40!ce
希望你们中的一些人会觉得这个很好很优雅。
其他回答
我将以下方法添加到PageBase类中。
protected void Redirect(string url)
{
Response.Redirect(url);
}
protected void Redirect(string url, NameValueCollection querystrings)
{
StringBuilder redirectUrl = new StringBuilder(url);
if (querystrings != null)
{
for (int index = 0; index < querystrings.Count; index++)
{
if (index == 0)
{
redirectUrl.Append("?");
}
redirectUrl.Append(querystrings.Keys[index]);
redirectUrl.Append("=");
redirectUrl.Append(HttpUtility.UrlEncode(querystrings[index]));
if (index < querystrings.Count - 1)
{
redirectUrl.Append("&");
}
}
}
this.Redirect(redirectUrl.ToString());
}
电话:
NameValueCollection querystrings = new NameValueCollection();
querystrings.Add("language", "en");
querystrings.Add("id", "134");
this.Redirect("http://www.mypage.com", querystrings);
我有一个扩展方法的Uri:
接受匿名对象:uri。WithQuery(new {name = "value"}) 接受字符串/字符串对的集合(例如Dictionary ' 2)。 接受字符串/对象对的集合(例如RouteValueDictionary)。 接受NameValueCollections。 按键对查询值进行排序，以便相同的值产生相同的uri。 每个键支持多个值，保持其原始顺序。
文档版本可以在这里找到。
扩展:
public static Uri WithQuery(this Uri uri, object values)
{
if (uri == null)
throw new ArgumentNullException(nameof(uri));
if (values != null)
{
var query = string.Join(
"&", from p in ParseQueryValues(values)
where !string.IsNullOrWhiteSpace(p.Key)
let k = HttpUtility.UrlEncode(p.Key.Trim())
let v = HttpUtility.UrlEncode(p.Value)
orderby k
select string.IsNullOrEmpty(v) ? k : $"{k}={v}");
if (query.Length != 0 || uri.Query.Length != 0)
uri = new UriBuilder(uri) { Query = query }.Uri;
}
return uri;
}
查询解析器:
private static IEnumerable<KeyValuePair<string, string>> ParseQueryValues(object values)
{
// Check if a name/value collection.
var nvc = values as NameValueCollection;
if (nvc != null)
return from key in nvc.AllKeys
from val in nvc.GetValues(key)
select new KeyValuePair<string, string>(key, val);
// Check if a string/string dictionary.
var ssd = values as IEnumerable<KeyValuePair<string, string>>;
if (ssd != null)
return ssd;
// Check if a string/object dictionary.
var sod = values as IEnumerable<KeyValuePair<string, object>>;
if (sod == null)
{
// Check if a non-generic dictionary.
var ngd = values as IDictionary;
if (ngd != null)
sod = ngd.Cast<dynamic>().ToDictionary<dynamic, string, object>(
p => p.Key.ToString(), p => p.Value as object);
// Convert object properties to dictionary.
if (sod == null)
sod = new RouteValueDictionary(values);
}
// Normalize and return the values.
return from pair in sod
from val in pair.Value as IEnumerable<string>
?? new[] { pair.Value?.ToString() }
select new KeyValuePair<string, string>(pair.Key, val);
}
下面是测试:
var uri = new Uri("https://stackoverflow.com/yo?oldKey=oldValue");
// Test with a string/string dictionary.
var q = uri.WithQuery(new Dictionary<string, string>
{
["k1"] = string.Empty,
["k2"] = null,
["k3"] = "v3"
});
Debug.Assert(q == new Uri(
"https://stackoverflow.com/yo?k1&k2&k3=v3"));
// Test with a string/object dictionary.
q = uri.WithQuery(new Dictionary<string, object>
{
["k1"] = "v1",
["k2"] = new[] { "v2a", "v2b" },
["k3"] = null
});
Debug.Assert(q == new Uri(
"https://stackoverflow.com/yo?k1=v1&k2=v2a&k2=v2b&k3"));
// Test with a name/value collection.
var nvc = new NameValueCollection()
{
["k1"] = string.Empty,
["k2"] = "v2a"
};
nvc.Add("k2", "v2b");
q = uri.WithQuery(nvc);
Debug.Assert(q == new Uri(
"https://stackoverflow.com/yo?k1&k2=v2a&k2=v2b"));
// Test with any dictionary.
q = uri.WithQuery(new Dictionary<int, HashSet<string>>
{
[1] = new HashSet<string> { "v1" },
[2] = new HashSet<string> { "v2a", "v2b" },
[3] = null
});
Debug.Assert(q == new Uri(
"https://stackoverflow.com/yo?1=v1&2=v2a&2=v2b&3"));
// Test with an anonymous object.
q = uri.WithQuery(new
{
k1 = "v1",
k2 = new[] { "v2a", "v2b" },
k3 = new List<string> { "v3" },
k4 = true,
k5 = null as Queue<string>
});
Debug.Assert(q == new Uri(
"https://stackoverflow.com/yo?k1=v1&k2=v2a&k2=v2b&k3=v3&k4=True&k5"));
// Keep existing query using a name/value collection.
nvc = HttpUtility.ParseQueryString(uri.Query);
nvc.Add("newKey", "newValue");
q = uri.WithQuery(nvc);
Debug.Assert(q == new Uri(
"https://stackoverflow.com/yo?newKey=newValue&oldKey=oldValue"));
// Merge two query objects using the RouteValueDictionary.
var an1 = new { k1 = "v1" };
var an2 = new { k2 = "v2" };
q = uri.WithQuery(
new RouteValueDictionary(an1).Concat(
new RouteValueDictionary(an2)));
Debug.Assert(q == new Uri(
"https://stackoverflow.com/yo?k1=v1&k2=v2"));
假设你想减少对其他程序集的依赖并保持简单，你可以这样做:
var sb = new System.Text.StringBuilder();
sb.Append("a=" + HttpUtility.UrlEncode("TheValueOfA") + "&");
sb.Append("b=" + HttpUtility.UrlEncode("TheValueOfB") + "&");
sb.Append("c=" + HttpUtility.UrlEncode("TheValueOfC") + "&");
sb.Append("d=" + HttpUtility.UrlEncode("TheValueOfD") + "&");
sb.Remove(sb.Length-1, 1); // Remove the final '&'
string result = sb.ToString();
这也适用于循环。最后一个&号的删除需要到循环之外。
注意，使用连接操作符是为了提高可读性。使用它的成本与使用StringBuilder的成本相比是最小的(我认为Jeff Atwood在这个主题上发表了一些文章)。
我的提供:
public static Uri AddQuery(this Uri uri, string name, string value)
{
// this actually returns HttpValueCollection : NameValueCollection
// which uses unicode compliant encoding on ToString()
var query = HttpUtility.ParseQueryString(uri.Query);
query.Add(name, value);
var uriBuilder = new UriBuilder(uri)
{
Query = query.ToString()
};
return uriBuilder.Uri;
}
用法:
var uri = new Uri("http://stackoverflow.com").AddQuery("such", "method")
.AddQuery("wow", "soFluent");
// http://stackoverflow.com?such=method&wow=soFluent
public string UrlQueryStr(object data)
{
if (data == null)
return string.Empty;
object val;
StringBuilder sb = new StringBuilder();
foreach (PropertyDescriptor prop in TypeDescriptor.GetProperties(data))
{
if ((val = prop.GetValue(data)) != null)
{
sb.AppendFormat("{0}{1}={2}", sb.Length == 0 ? '?' : '&',
HttpUtility.UrlEncode(prop.Name), HttpUtility.UrlEncode(val.ToString()));
}
}
return sb.ToString();
}
