我也一直在努力解决这个问题(我也在学习阶段)。
我发现最有效的方法是使用REST api,然后使用Linq to XML将结果解析为可以使用的自定义结果对象。请原谅长代码粘贴(不确定协议是否与此板上的代码一致),但我想你会想要全部的。其中大部分来自亚马逊自己的一个样本,但是这个网站没有组织起来,所以我很难再找到它。
// custom class for result item
public class AmazonBookSearchResult
{
public string ASIN;
public string Author;
public string Title;
public Uri DetailPageURL;
public Uri AllCustomerReviews;
}
// I call this with the search terms, returns an IEnumerable of results
public class ItemLookup
{
private const string MY_AWS_ACCESS_KEY_ID = "YOUR KEY";
private const string MY_AWS_SECRET_KEY = "YOUR KEY";
private const string DESTINATION = "ecs.amazonaws.com";
private const string NAMESPACE = "http://webservices.amazon.com/AWSECommerceService/2009-03-31";
public static IEnumerable<AmazonBookSearchResult> AmazonItemLookup(string category, string browseNode, string keyWords)
{
SignedRequestHelper helper = new SignedRequestHelper(MY_AWS_ACCESS_KEY_ID, MY_AWS_SECRET_KEY, DESTINATION);
String requestUrl;
IDictionary<string, string> r1 = new Dictionary<string, String>();
r1["Service"] = "AWSECommerceService";
r1["Version"] = "2009-03-31";
r1["Operation"] = "ItemSearch";
r1["ResponseGroup"] = "ItemAttributes";
r1["SearchIndex"] = category;
r1["Keywords"] = keyWords;
requestUrl = helper.Sign(r1);
XmlDocument xmlDoc = sendRequest(requestUrl);
XDocument doc = XDocument.Load(new XmlNodeReader(xmlDoc));
XNamespace ns = "http://webservices.amazon.com/AWSECommerceService/2009-03-31";
IEnumerable<AmazonBookSearchResult> result =
from items in doc.Element(ns + "ItemSearchResponse").Elements(ns + "Items").Elements(ns + "Item")
select new AmazonBookSearchResult
{
ASIN = (string)items.Element(ns + "ASIN")
};
int count = doc.Element(ns + "ItemSearchResponse").Elements(ns + "Items").Elements(ns + "Item").Count();
return result;
} // end Lookup()
// this is just taken from the amazon sample
private static XmlDocument sendRequest(string url)
{
try
{
WebRequest request = HttpWebRequest.Create(url);
WebResponse response = request.GetResponse();
XmlDocument doc = new XmlDocument();
doc.Load(response.GetResponseStream());
XmlNodeList errorMessageNodes = doc.GetElementsByTagName("Message", NAMESPACE);
if (errorMessageNodes != null && errorMessageNodes.Count > 0)
{
String message = errorMessageNodes.Item(0).InnerText;
throw new Exception("Error " + message);
}
return doc;
}
catch (Exception e)
{
System.Console.WriteLine("Caught Exception: " + e.Message);
System.Console.WriteLine("Stack Trace: " + e.StackTrace);
}
return null;
}// end FetchTitle()
}
class SignedRequestHelper
{
private string endPoint;
private string akid;
private byte[] secret;
private HMAC signer;
private const string REQUEST_URI = "/onca/xml";
private const string REQUEST_METHOD = "GET";
/*
* Use this constructor to create the object. The AWS credentials are available on
* http://aws.amazon.com
*
* The destination is the service end-point for your application:
* US: ecs.amazonaws.com
* JP: ecs.amazonaws.jp
* UK: ecs.amazonaws.co.uk
* DE: ecs.amazonaws.de
* FR: ecs.amazonaws.fr
* CA: ecs.amazonaws.ca
*/
public SignedRequestHelper(string awsAccessKeyId, string awsSecretKey, string destination)
{
this.endPoint = destination.ToLower();
this.akid = awsAccessKeyId;
this.secret = Encoding.UTF8.GetBytes(awsSecretKey);
this.signer = new HMACSHA256(this.secret);
}
/*
* Sign a request in the form of a Dictionary of name-value pairs.
*
* This method returns a complete URL to use. Modifying the returned URL
* in any way invalidates the signature and Amazon will reject the requests.
*/
public string Sign(IDictionary<string, string> request)
{
// Use a SortedDictionary to get the parameters in naturual byte order, as
// required by AWS.
ParamComparer pc = new ParamComparer();
SortedDictionary<string, string> sortedMap = new SortedDictionary<string, string>(request, pc);
// Add the AWSAccessKeyId and Timestamp to the requests.
sortedMap["AWSAccessKeyId"] = this.akid;
sortedMap["Timestamp"] = this.GetTimestamp();
// Get the canonical query string
string canonicalQS = this.ConstructCanonicalQueryString(sortedMap);
// Derive the bytes needs to be signed.
StringBuilder builder = new StringBuilder();
builder.Append(REQUEST_METHOD)
.Append("\n")
.Append(this.endPoint)
.Append("\n")
.Append(REQUEST_URI)
.Append("\n")
.Append(canonicalQS);
string stringToSign = builder.ToString();
byte[] toSign = Encoding.UTF8.GetBytes(stringToSign);
// Compute the signature and convert to Base64.
byte[] sigBytes = signer.ComputeHash(toSign);
string signature = Convert.ToBase64String(sigBytes);
// now construct the complete URL and return to caller.
StringBuilder qsBuilder = new StringBuilder();
qsBuilder.Append("http://")
.Append(this.endPoint)
.Append(REQUEST_URI)
.Append("?")
.Append(canonicalQS)
.Append("&Signature=")
.Append(this.PercentEncodeRfc3986(signature));
return qsBuilder.ToString();
}
/*
* Sign a request in the form of a query string.
*
* This method returns a complete URL to use. Modifying the returned URL
* in any way invalidates the signature and Amazon will reject the requests.
*/
public string Sign(string queryString)
{
IDictionary<string, string> request = this.CreateDictionary(queryString);
return this.Sign(request);
}
/*
* Current time in IS0 8601 format as required by Amazon
*/
private string GetTimestamp()
{
DateTime currentTime = DateTime.UtcNow;
string timestamp = currentTime.ToString("yyyy-MM-ddTHH:mm:ssZ");
return timestamp;
}
/*
* Percent-encode (URL Encode) according to RFC 3986 as required by Amazon.
*
* This is necessary because .NET's HttpUtility.UrlEncode does not encode
* according to the above standard. Also, .NET returns lower-case encoding
* by default and Amazon requires upper-case encoding.
*/
private string PercentEncodeRfc3986(string str)
{
str = HttpUtility.UrlEncode(str, System.Text.Encoding.UTF8);
str = str.Replace("'", "%27").Replace("(", "%28").Replace(")", "%29").Replace("*", "%2A").Replace("!", "%21").Replace("%7e", "~").Replace("+", "%20");
StringBuilder sbuilder = new StringBuilder(str);
for (int i = 0; i < sbuilder.Length; i++)
{
if (sbuilder[i] == '%')
{
if (Char.IsLetter(sbuilder[i + 1]) || Char.IsLetter(sbuilder[i + 2]))
{
sbuilder[i + 1] = Char.ToUpper(sbuilder[i + 1]);
sbuilder[i + 2] = Char.ToUpper(sbuilder[i + 2]);
}
}
}
return sbuilder.ToString();
}
/*
* Convert a query string to corresponding dictionary of name-value pairs.
*/
private IDictionary<string, string> CreateDictionary(string queryString)
{
Dictionary<string, string> map = new Dictionary<string, string>();
string[] requestParams = queryString.Split('&');
for (int i = 0; i < requestParams.Length; i++)
{
if (requestParams[i].Length < 1)
{
continue;
}
char[] sep = { '=' };
string[] param = requestParams[i].Split(sep, 2);
for (int j = 0; j < param.Length; j++)
{
param[j] = HttpUtility.UrlDecode(param[j], System.Text.Encoding.UTF8);
}
switch (param.Length)
{
case 1:
{
if (requestParams[i].Length >= 1)
{
if (requestParams[i].ToCharArray()[0] == '=')
{
map[""] = param[0];
}
else
{
map[param[0]] = "";
}
}
break;
}
case 2:
{
if (!string.IsNullOrEmpty(param[0]))
{
map[param[0]] = param[1];
}
}
break;
}
}
return map;
}
/*
* Consttuct the canonical query string from the sorted parameter map.
*/
private string ConstructCanonicalQueryString(SortedDictionary<string, string> sortedParamMap)
{
StringBuilder builder = new StringBuilder();
if (sortedParamMap.Count == 0)
{
builder.Append("");
return builder.ToString();
}
foreach (KeyValuePair<string, string> kvp in sortedParamMap)
{
builder.Append(this.PercentEncodeRfc3986(kvp.Key));
builder.Append("=");
builder.Append(this.PercentEncodeRfc3986(kvp.Value));
builder.Append("&");
}
string canonicalString = builder.ToString();
canonicalString = canonicalString.Substring(0, canonicalString.Length - 1);
return canonicalString;
}
}
/*
* To help the SortedDictionary order the name-value pairs in the correct way.
*/
class ParamComparer : IComparer<string>
{
public int Compare(string p1, string p2)
{
return string.CompareOrdinal(p1, p2);
}
}
