private static string getKeyWordsOfCertainPrimaryGroup()
{
string output = "getKeyWordsOfCertainPrimaryGroup\n";
PrimaryGroups pg = new PrimaryGroups(primaryGroupId, "");
List <string> list = PrimaryGroupKeyWordsManager.getKeyWordsOfCertainPrimaryGroup(pg);
if (list == null)
{
output += "Error! getKeyWordsOfCertainPrimaryGroup失败，返回null。\n";
errorCount++;
return(output);
}
for (int i = 0; i < count; i++)
{
if (!list.Contains(keys[i]))
{
output += "Error! 未拿到keyWord为" + keys[i] + "的测试记录！\n";
}
else
{
output += "Ok! 成功拿到keyWord为" + keys[i] + "的测试记录！\n";
}
}
return(output);
}
private static string deleteRecord()
{
string output = "deleteRecord\n";
for (int i = 0; i < count; i++)
{
PrimaryGroupKeyWords kw = new PrimaryGroupKeyWords(primaryGroupId, keys[i]);
if (!PrimaryGroupKeyWordsManager.deleteRecord(kw))
{
output += "Error! 删除测试关键词（keyWord为\"" + keys[i] + "\"）失败！\n";
errorCount++;
}
else
{
output += "Ok! 删除测试关键词（keyWord为\"" + keys[i] + "\"）成功！\n";
}
}
return(output);
}
private static string addRecord()
{
string output = "addRecord\n";
Random r = new Random();
for (int i = 0; i < count; i++)
{
int no = r.Next(1000, 9000);
keys[i] = "测试样例" + no;
PrimaryGroupKeyWords kw = new PrimaryGroupKeyWords(primaryGroupId, keys[i]);
if (!PrimaryGroupKeyWordsManager.addRecord(kw))
{
output += "Error! 新增主分类（id为" + primaryGroupId + "）关键词\"" + keys[i] + "\"失败！返回false。\n";
errorCount++;
}
else
{
output += "Ok! 新增主分类（id为" + primaryGroupId + "）关键词\"" + keys[i] + "\"成功！返回true。\n";
}
}
return(output);
}
/**
* 输入：新主分类的名称，以及相应的关键词列表。各个关键词之间用空格分隔。
* 输出：返回所有未被分到任何主分类的文章列表
* 功能：1、新增一个主分类
* 2、在该主分类下增加一个“其他”子分类
* 2、调用NewsAssist.cs中getArticleListOfOthers( int userId , GroupNode gn )方法，通过将第二个参数置为null，拿到所有未被分到任何主分类的文章列表；
* 3、返回这些文章的列表，让管理员查看一下其中有没有可以直接被分到新的主分类去的文章。
*/
public static List <Article> addPrimaryGroup(string name, string keywordsList)
{
// 功能：1、新增一个主分类
PrimaryGroups pg = new PrimaryGroups();
pg.GroupName = name;
int gid = PrimaryGroupMananger.addRecord(pg);
PrimaryGroupKeyWords pgk = new PrimaryGroupKeyWords();
pgk.PrimaryGroupId = gid;
pgk.KeyWord = keywordsList;
bool addPGKsuccess = PrimaryGroupKeyWordsManager.addRecord(pgk);
if (addPGKsuccess)
{
addSecondaryGroup(gid, "其他", "");
}
// 功能：2、调用Articlemanager.cs 中的方法,返回所有未分类的文章
PrimaryGroups p = new PrimaryGroups();
p.GroupId = 0;
return(ArticleManager.getArticleListByPrimaryGroup(p));
}
/*
* 输入：一个Article的model实例以及其content字符串
* 输出：执行结果标示符（暂定如下：0表示成功；-1~-5表示不同类型的失败）
* 功能：
* 1、对传入参数标示的文章内容进行分词、词频统计、TF·IDF计算等
* 2、将文章总词数（article表的wordCount字段）更新到Article表中
* 3、将分词结果更新到GlobalParse数据表中，注意参数中的articleNumber字段值要设置好：如果GlobalParse表中尚无该词记录，则
* articleNumber设为1；若GlobalParse表中已有该词记录，则读取其articleNumber并加一再存进去
* 4、将分词结果更新到LocalParse数据表中，注意参数中的count字段值的设置要设置好：如果LocalParse表中尚无该词记录，则
* count设为1；若LocalParse表中已有该词记录，则读取其count并加一再存进去
* 5、对文章进行分类，并将分类结果更新到Article2Group数据表中。若没有任何类别适合该文章，则默认应该将其分到groupId为0的类别中，该类别即“分组失败”类别或“其他”类别（视用户需求而定）
* 6、返回执行结果。请实现者自己定义一下不同的返回值标示何种类型的失败
*
* 用途说明：该函数为私有函数，只被addArticleWrapper函数在将新文章添加到数据库后自动调用，用于解析新文章。
*
*/
private static void parseArticle(Article a, string content)
{
// 在这里先行数据库中读出 各个 primary group 的 关键词列表
List <List <string> > allGroupKeywordList = new List <List <string> >();
List <string[]> allGroup = PrimaryGroupMananger.getAllGroups();
foreach (string[] ag in allGroup)
{
PrimaryGroups g = new PrimaryGroups();
g.GroupId = Convert.ToInt32(ag[0]);
g.GroupName = ag[1];
allGroupKeywordList.Add(PrimaryGroupKeyWordsManager.getKeyWordsOfCertainPrimaryGroup(g));
}
List <String[]> parseList = stringParse(content);
a.WordCount = parseList.Count;
Dictionary <String, Int32> dic = new Dictionary <String, Int32>();
int wordMount = 0;
foreach (String[] tempString in parseList)
{
if (stop_list.Contains(tempString[0]))
{
continue;
}
if (!dic.ContainsKey(tempString[0]))
{
dic.Add(tempString[0], 1);
wordMount++;
}
else
{
int tempInt = dic[tempString[0]];
dic[tempString[0]] = tempInt + 1;
}
}
int articleID = a.ArticleId;
Dictionary <String, Double> tf_idf = new Dictionary <String, Double>();
foreach (KeyValuePair <String, Int32> keyPair in dic)
{
GlobalParse tempGP = new GlobalParse();
tempGP.ArticleNumber = 1; tempGP.WordContent = keyPair.Key; tempGP.Type = "q";
if (GlobalParseManager.addRecord(tempGP) == false)
{
int num = GlobalParseManager.selectRecordByWordContent(tempGP).ArticleNumber + 1;
tempGP.ArticleNumber = num;
GlobalParseManager.updateRecord(tempGP);
}
tempGP = GlobalParseManager.selectRecordByWordContent(tempGP);
LocalParse tempLP = new LocalParse();
tempLP.ArticleId = articleID; tempLP.WordContent = tempGP.WordContent; tempLP.Count = keyPair.Value;
tempLP.Type = "q";
LocalParseManager.addRecord(tempLP);
// Counting tf_idf
int tf_fenzi = tempLP.Count; int tf_fenmu = wordMount;
int total_document_number = ArticleManager.countArticleNum();
int document_number_with_word = tempGP.ArticleNumber;
double tf_idf_value = ((double)tf_fenzi) / tf_fenmu;// *Math.Log((total_document_number / document_number_with_word), Math.E);
tf_idf.Add(keyPair.Key, tf_idf_value);
}
//tf_idf.OrderByDescending(s => s.Value);
double[] answer = new double[allGroup.Count];
for (int i = 0; i < answer.Length; i++)
{
answer[i] = 0;
}
//Console.Write(tf_idf.ElementAt(0).Key);
//Console.WriteLine(other.Count);
for (int i = 0; i < answer.Length; i++)
{
List <string> list = allGroupKeywordList[i];
if (list == null)
{
continue;
}
foreach (String tempS in list)
{
if (tf_idf.ContainsKey(tempS))
{
answer[i] += tf_idf[tempS];
}
}
}
Article2Group a2g = new Article2Group();
a2g.ArticleId = a.ArticleId;
a2g.GroupId = max(answer);
Article2GroupManager.addRecord(a2g);
}
private static void addPrimaryGroup()
{
output += "addPrimaryGroup\n";
// add
string name = "sample123";
string keywords = "qawsed rftgyh";
ManagerAssist.addPrimaryGroup(name, keywords);
// check
List <string[]> list = PrimaryGroupMananger.getAllGroups();
bool found = false;
int groupId = -1;
foreach (string[] item in list)
{
if (item[1].Equals("sample123"))
{
found = true;
groupId = int.Parse(item[0]);
break;
}
}
if (!found)
{
output += "Error! addPrimaryGroup执行失败！主分类未加入数据库中。\n";
errorCount++;
}
else
{
output += "Ok! addPrimaryGroup执行成功！主分类已加入数据库中。\n";
}
PrimaryGroups pg = new PrimaryGroups(groupId, name);
List <string> list2 = PrimaryGroupKeyWordsManager.getKeyWordsOfCertainPrimaryGroup(pg);
if (list2.Count != 2)
{
output += "Error! PrimaryGroupKeyWords数目不对！应为2\n";
}
else
{
output += "Ok! PrimaryGroupKeyWords数目正确！应为2\n";
}
if (list2.First().Equals("qawsed") || list2.First().Equals("rftgyh"))
{
output += "Ok! PrimaryGroupKeyWords中对应的第一个关键词通过验证！\n";
}
else
{
output += "Error! PrimaryGroupKeyWords中对应的第一个关键词未通过验证！\n";
errorCount++;
}
List <Tag> tagL = TagManager.getAllTagsByCertainGroupId(groupId);
if (tagL == null || tagL.Count != 1)
{
output += "Error! addPrimaryGroup调用后没有将“其他”子分类插入tag表！\n";
errorCount++;
}
else
{
output += "Ok! addPrimaryGroup调用后成功将“其他”子分类插入tag表！\n";
}
// delete
foreach (string item in list2)
{
if (!PrimaryGroupKeyWordsManager.deleteRecord(new PrimaryGroupKeyWords(groupId, item)))
{
output += "Error! 删除测试关键词失败！\n";
errorCount++;
}
}
Tag t = new Tag();
t.TagId = tagL.First().TagId;
if (!TagManager.deleteTag(t))
{
output += "Error! 删除测试主分类下“其他”子分类失败！\n";
errorCount++;
}
PrimaryGroups temp = new PrimaryGroups();
temp.GroupId = groupId;
if (!PrimaryGroupMananger.deleteRecord(temp))
{
output += "Error! 删除测试主分类失败！\n";
errorCount++;
}
}
