private static bool clean()
{
for (int i = 0; i < count; i++)
{
if (i < count / 2)
{
Article2Group a2g = new Article2Group(ids[i], primaryGroupId);
if (!Article2GroupManager.deleteRecord(a2g))
{
output += "Error! 将id为" + ids[i] + "的文章取消与id为" + primaryGroupId + "的主分类的关联时失败！测试无法继续进行。请先解决Article2GroupManager中的错误。\n";
errorCount++;
return(false);
}
}
else
{
Article2Group a2g = new Article2Group(ids[i], OtherGroupId);
if (!Article2GroupManager.deleteRecord(a2g))
{
output += "Error! 将id为" + ids[i] + "的文章取消与id为" + OtherGroupId + "的主分类的关联时失败！测试无法继续进行。请先解决Article2GroupManager中的错误。\n";
errorCount++;
return(false);
}
}
Article a = new Article(ids[i], titles[i], titles[i], "", DateTime.Now, 0, 0);
if (!ArticleManager.deleteRecord(a))
{
output += "Error! 删除测试文章\"" + titles[i] + "\"失败！测试无法继续进行。请先解决ArticleManager中的错误。\n";
errorCount++;
return(false);
}
}
return(true);
}
private static string getGroupIdsByArticle()
{
string output = "getGroupIdsByArticle\n";
for (int i = 0; i < count; i++)
{
Article a = new Article();
a.ArticleId = ids[i];
List <int> list = Article2GroupManager.getGroupIdsByArticle(a);
if (list == null)
{
output += "Error! 调用getGroupIdsByArticle失败！返回值为null。\n";
errorCount++;
continue;
}
if (!list.Contains(primaryGroupId))
{
output += "Error! 未找到关联（文章id为" + ids[i] + "，主分类id为" + primaryGroupId + "）！\n";
errorCount++;
}
else
{
output += "Ok! 找到关联（文章id为" + ids[i] + "，主分类id为" + primaryGroupId + "）！\n";
}
}
return(output);
}
private static string deleteRecord()
{
string output = "deleteRecord\n";
for (int i = 0; i < count; i++)
{
Article2Group a2g = new Article2Group(ids[i], primaryGroupId);
if (!Article2GroupManager.deleteRecord(a2g))
{
output += "Error! 删除关联（文章id为" + ids[i] + "，主分类id为" + primaryGroupId + "）失败！\n";
errorCount++;
}
else
{
output += "Ok! 删除关联（文章id为" + ids[i] + "，主分类id为" + primaryGroupId + "）成功！\n";
}
Article a = new Article();
a.ArticleId = ids[i];
if (!ArticleManager.deleteRecord(a))
{
output += "Error! 删除文章\"" + titles[i] + "\"失败！\n";
errorCount++;
}
else
{
output += "Ok! 删除文章\"" + titles[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);
titles[i] = "测试样例" + no;
Article a = new Article(no, titles[i], titles[i], "", DateTime.Now, 0, 0);
ids[i] = ArticleManager.addRecord(a);
if (ids[i] < 0)
{
output += "Error! 新增文章\"" + titles[i] + "\"失败！返回值为" + ids[i] + "\n";
errorCount++;
continue;
}
output += "Ok! 新增文章\"" + titles[i] + "\"成功！返回值为" + ids[i] + "\n";
Article2Group a2g = new Article2Group(ids[i], primaryGroupId);
if (!Article2GroupManager.addRecord(a2g))
{
output += "Error! 将id为" + ids[i] + "的文章关联到id为" + primaryGroupId + "的主分类时失败！\n";
errorCount++;
}
else
{
output += "Ok! 将id为" + ids[i] + "的文章关联到id为" + primaryGroupId + "的主分类时成功！\n";
}
}
return(output);
}
private static bool clean()
{
for (int i = 0; i < count; i++)
{
Article2Group a2g = new Article2Group(ids[i], primaryGroup[1].GroupId);
if (!Article2GroupManager.deleteRecord(a2g))
{
output += "Error! 删除关联（文章id为" + ids[i] + "，主分类id为" + primaryGroup[1].GroupId + "）失败！\n";
errorCount++;
}
Article a = new Article();
a.ArticleId = ids[i];
if (!ArticleManager.deleteRecord(a))
{
output += "Error! 删除文章\"" + titles[i] + "\"失败！\n";
errorCount++;
}
}
if (!PrimaryGroupMananger.deleteRecord(primaryGroup[0]))
{
output += "Error! 删除测试辅助主分类失败！\n";
errorCount++;
return(false);
}
if (!PrimaryGroupMananger.deleteRecord(primaryGroup[1]))
{
output += "Error! 删除测试辅助主分类失败！\n";
errorCount++;
return(false);
}
return(true);
}
private static string getArticleIdsByGroup()
{
string output = "getArticleIdsByGroup\n";
PrimaryGroups pg = new PrimaryGroups();
pg.GroupId = primaryGroupId;
List <int> list = Article2GroupManager.getArticleIdsByGroup(pg);
if (list == null)
{
output += "Error! 调用getArticleIdsByGroup失败，返回null。\n";
errorCount++;
return(output);
}
for (int i = 0; i < count; i++)
{
if (!list.Contains(ids[i]))
{
output += "Error! id为" + primaryGroupId + "的主分类中未找到id为" + ids[i] + "的文章！\n";
errorCount++;
}
else
{
output += "Ok! id为" + primaryGroupId + "的主分类中找到了id为" + ids[i] + "的文章！\n";
}
}
return(output);
}
private static bool prepare()
{
for (int i = 0; i < 2; i++)
{
primaryGroup[i] = new PrimaryGroups(0, "test1");
primaryGroup[i].GroupId = PrimaryGroupMananger.addRecord(primaryGroup[i]);
if (primaryGroup[i].GroupId < 0)
{
output += "Error! 新增测试辅助主分类\"test\"失败！返回值为" + primaryGroup[i].GroupId + "。测试无法继续进行。请先解决PrimaryGroupsManager中的错误。\n";
errorCount++;
return(false);
}
}
Random r = new Random();
for (int i = 0; i < count; i++)
{
int no = r.Next(1000, 9000);
titles[i] = "测试样例" + no;
Article a = new Article(no, titles[i], titles[i], "", DateTime.Now, 0, 0);
ids[i] = ArticleManager.addRecord(a);
if (ids[i] < 0)
{
output += "Error! 新增文章\"" + titles[i] + "\"失败！返回值为" + ids[i] + "。测试无法继续进行。请先解决Article2GroupManager中的错误。\n";
errorCount++;
return(false);
}
output += "Ok! 新增文章\"" + titles[i] + "\"成功！返回值为" + ids[i] + "\n";
Article2Group a2g = new Article2Group(ids[i], primaryGroup[0].GroupId);
if (!Article2GroupManager.addRecord(a2g))
{
output += "Error! 将id为" + ids[i] + "的文章关联到id为" + primaryGroup[0].GroupId + "的主分类时失败！测试无法继续进行。请先解决Article2GroupManager中的错误。\n";
errorCount++;
return(false);
}
}
return(true);
}
private static void changeGroupRelation()
{
output += "changeGroupRelation\n";
List <int> list0_pre = Article2GroupManager.getArticleIdsByGroup(primaryGroup[0]);
List <int> list1_pre = Article2GroupManager.getArticleIdsByGroup(primaryGroup[1]);
int list1_pre_len = 0;
if (list1_pre != null)
{
list1_pre_len = list1_pre.Count;
}
output += "调用changeGroupRelation前，primaryGroup[0]下有" + list0_pre.Count + "篇文章，primaryGroup[1]下有" + list1_pre_len + "篇文章。\n";
List <int> groupIds = new List <int>();
groupIds.Add(primaryGroup[0].GroupId);
groupIds.Add(primaryGroup[1].GroupId);
for (int i = 0; i < count; i++)
{
Article a = new Article();
a.ArticleId = ids[i];
ManagerAssist.changeGroupRelation(a, groupIds);
}
List <int> list0_post = Article2GroupManager.getArticleIdsByGroup(primaryGroup[0]);
List <int> list1_post = Article2GroupManager.getArticleIdsByGroup(primaryGroup[1]);
output += "调用changeGroupRelation前，primaryGroup[0]下有" + list0_post.Count + "篇文章，primaryGroup[1]下有" + list1_post.Count + "篇文章。\n";
if (list0_pre.Count - list1_pre_len > list0_post.Count - list1_post.Count)
{
output += "Ok! changeGroupRelation执行成功！\n";
}
else
{
output += "Ok! changeGroupRelation执行失败！\n";
errorCount++;
}
}
private static bool prepare()
{
Random r = new Random();
for (int i = 0; i < count; i++)
{
int no = r.Next(1000, 9000);
titles[i] = "测试样例" + no;
Article a = new Article(no, titles[i], titles[i], "", DateTime.Now, 0, 0);
ids[i] = ArticleManager.addRecord(a);
if (ids[i] < 0)
{
output += "Error! 新增测试文章\"" + titles[i] + "\"失败！返回值为" + ids[i] + "。测试无法继续进行。请先解决ArticleManager中的错误。\n";
errorCount++;
return(false);
}
if (i < count / 2)
{
Article2Group a2g = new Article2Group(ids[i], primaryGroupId);
if (!Article2GroupManager.addRecord(a2g))
{
output += "Error! 将id为" + ids[i] + "的文章关联到id为" + primaryGroupId + "的主分类时失败！测试无法继续进行。请先解决Article2GroupManager中的错误。\n";
errorCount++;
return(false);
}
}
else
{
Article2Group a2g = new Article2Group(ids[i], OtherGroupId);
if (!Article2GroupManager.addRecord(a2g))
{
output += "Error! 将id为" + ids[i] + "的文章关联到id为" + OtherGroupId + "的主分类时失败！测试无法继续进行。请先解决Article2GroupManager中的错误。\n";
errorCount++;
return(false);
}
}
}
return(true);
}
/*
* 输入：一个文章的Article实例，以及一个文章所属的groupId的列表
* 输出：成功返回true，失败返回false
* 功能：对指定文章进行分类，将其分到groupIds所包含的所有组别中去（即一篇文章可以被分到多个类别中）。
* 用途说明：管理员新增一篇文章到数据库中去以后，系统会自动将文章分类，并将结果显示在页面上，管理员若对自动
* 分类的结果不满意，可以手动调整，调整完毕后调用本函数修改文章的分类结果
*/
public static bool changeGroupRelation(Article a, List <int> groupIds)
{
bool result = true;
// 先删除旧的 groupIds
List <int> old_gids = Article2GroupManager.getGroupIdsByArticle(a); // 这里我认为这个传入的 Article 的 id 是已经赋值了的
foreach (int old_gid in old_gids)
{
Article2GroupManager.deleteRecord(new Article2Group(a.ArticleId, old_gid));
}
// 再插入新的 groupIds
foreach (int gid in groupIds)
{
if (Article2GroupManager.addRecord(new Article2Group(a.ArticleId, gid)) == false)
{
result = false; // 只要有一条记录插入不成功，则返回 false
}
}
return(result);
}
//获取该分类下的所有文章的article的列表（仅读取传入参数中的GroupId字段）
//返回article的列表List<Article>
public static List <Article> getArticleListByPrimaryGroup(PrimaryGroups g)
{
List <int> articlesByGroupId = Article2GroupManager.getArticleIdsByGroup(g);
return(getArticleListByArticleIdList(articlesByGroupId));
}
/*
* 输入：一个Article的model实例
* 输出：该文章所属的所有类别的groupId的列表
* 功能：查询指定文章所属的所有的组别的ID
* 用途说明：管理员调用addArticleWrapper函数将一篇新文章插入数据库后，得到返回的articleId，
* 便可用本函数查询该文章的自动分类结果（以组别ID的列表为形式）
*/
public static List <int> getGroupIdsByArticleWrapper(Article a)
{
return(Article2GroupManager.getGroupIdsByArticle(a));
}
/*
* 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);
}
