// 实现IComparable接口的CompareTo()方法,
// obj: An object to compare with this instance
// 返回值 A 32-bit signed integer that indicates the relative order of the comparands. The return value has these meanings:
// Less than zero: This instance is less than obj.
// Zero: This instance is equal to obj.
// Greater than zero: This instance is greater than obj.
// 异常: ArgumentException,obj is not the same type as this instance.
public override int CompareTo(object obj)
{
DupLineItem item = (DupLineItem)obj;
// 小在前
return(String.Compare(this.Path, item.Path));
}
// 按照权值排序
public int CompareWeightTo(object obj)
{
DupLineItem item = (DupLineItem)obj;
int delta = this.Weight - item.Weight;
if (delta != 0)
{
return(-1 * delta); // 大在前
}
// 如权值相同，再按照路径排序
// 小在前
return(String.Compare(this.Path, item.Path));
}
// 按照差额排序。所谓差额就是权值和阈值的差额
public int CompareOverThresholdTo(object obj)
{
DupLineItem item = (DupLineItem)obj;
int over1 = this.Weight - this.Threshold;
int over2 = item.Weight - item.Threshold;
int delta = over1 - over2;
if (delta != 0)
{
return(-1 * delta); // 大在前
}
// 如差额相同，再按照路径排序
// 小在前
return(String.Compare(this.Path, item.Path));
}
// 使得可以按照多种风格排序
public override int Compare(long lPtr1, long lPtr2)
{
if (lPtr1 < 0 && lPtr2 < 0)
{
return(0);
}
else if (lPtr1 >= 0 && lPtr2 < 0)
{
return(1);
}
else if (lPtr1 < 0 && lPtr2 >= 0)
{
return(-1);
}
DupLineItem item1 = (DupLineItem)GetCompareItemByOffset(lPtr1);
DupLineItem item2 = (DupLineItem)GetCompareItemByOffset(lPtr2);
if (this.SortStyle == DupResultSetSortStyle.Path)
{
return(item1.CompareTo(item2));
}
else if (this.SortStyle == DupResultSetSortStyle.Weight)
{
return(item1.CompareWeightTo(item2));
}
else if (this.SortStyle == DupResultSetSortStyle.OverThreshold)
{
return(item1.CompareOverThresholdTo(item2));
}
else
{
Debug.Assert(false, "invalid sort style");
return(0);
}
}
// 针对一个from进行检索
// parameters:
// strExcludeBiblioRecPath 要排除掉的记录路径
// return:
// -1 error
// 0 not found
// 1 found
int SearchOneFrom(
// RmsChannelCollection Channels,
RmsChannel channel,
string strDbName,
string strFrom,
string strKey,
string strSearchStyle,
int nWeight,
int nThreshold,
long nMax,
string strExcludeBiblioRecPath,
out DupResultSet dupset,
out string strError)
{
strError = "";
dupset = null;
long lRet = 0;
if (strSearchStyle == "")
{
strSearchStyle = "exact";
}
string strQueryXml = "<target list='"
+ StringUtil.GetXmlStringSimple(strDbName + ":" + strFrom) // 2007/9/14
+ "'><item><word>"
+ StringUtil.GetXmlStringSimple(strKey)
+ "</word><match>" + strSearchStyle + "</match><relation>=</relation><dataType>string</dataType><maxCount>" + nMax.ToString() + "</maxCount></item><lang>zh</lang></target>";
string strSearchReason = "key='" + strKey + "', from='" + strFrom + "', weight=" + Convert.ToString(nWeight);
/*
* RmsChannel channel = Channels.GetChannel(this.WsUrl);
* if (channel == null)
* {
* strError = "get channel error";
* goto ERROR1;
* }
* */
Debug.Assert(channel != null, "");
lRet = channel.DoSearch(strQueryXml,
"dup",
"", // strOuputStyle
out strError);
if (lRet == -1)
{
goto ERROR1;
}
if (lRet == 0)
{
return(0); // not found
}
long lHitCount = lRet;
long lStart = 0;
long lPerCount = Math.Min(50, lHitCount);
List <string> aPath = null;
dupset = new DupResultSet();
dupset.Open(false, getTempFileName);
// 获得结果集，对逐个记录进行处理
for (; ;)
{
// TODO: 中间要可以中断
lRet = channel.DoGetSearchResult(
"dup", // strResultSetName
lStart,
lPerCount,
"zh",
null, // stop
out aPath,
out strError);
if (lRet == -1)
{
goto ERROR1;
}
if (lRet == 0)
{
strError = "未命中";
break; // ??
}
// TODO: 要判断 aPath.Count == 0 跳出循环。否则容易进入死循环
// 处理浏览结果
for (int i = 0; i < aPath.Count; i++)
{
string strPath = aPath[i];
// 忽略发起记录的路径
if (strPath == strExcludeBiblioRecPath)
{
continue;
}
DupLineItem item = new DupLineItem
{
Path = strPath,
Weight = nWeight,
Threshold = nThreshold,
Detail = BuildDetail(strDbName,
strFrom,
strKey,
strSearchStyle,
nWeight)
};
dupset.Add(item);
}
lStart += aPath.Count;
if (lStart >= lHitCount || lPerCount <= 0)
{
break;
}
}
return(1);
ERROR1:
return(-1);
}
// 获得查重检索命中结果
// parameters:
// lStart 返回命中结果集起始位置
// lCount 返回命中结果集的记录个数
// strBrowseInfoStyle 所返回的DupSearchResult中包含哪些信息
// "cols" 包含浏览列
// "excludecolsoflowthreshold" 不包含权值低于阈值的行的浏览列。要在同时包含cols时才起作用
// searchresults 包含记录信息的DupSearchResult数组
public LibraryServerResult GetDupSearchResult(
SessionInfo sessioninfo,
long lStart,
long lCount,
string strBrowseInfoStyle,
out DupSearchResult[] searchresults)
{
string strError = "";
searchresults = null;
int nRet = 0;
LibraryServerResult result = new LibraryServerResult();
RmsChannel channel = sessioninfo.Channels.GetChannel(this.WsUrl);
if (channel == null)
{
strError = "get channel error";
goto ERROR1;
}
DupResultSet dupset = sessioninfo.DupResultSet;
if (dupset == null)
{
strError = "查重结果集不存在";
goto ERROR1;
}
dupset.EnsureCreateIndex(getTempFileName);
int nCount = (int)lCount;
int nStart = (int)lStart;
if (nCount == -1)
{
nCount = (int)dupset.Count - nStart;
if (nCount < 0)
{
nCount = 0;
}
}
else
{
if (nCount > (int)dupset.Count - nStart)
{
nCount = (int)dupset.Count - nStart;
if (nCount < 0)
{
nCount = 0;
}
}
}
bool bDetail = (StringUtil.IsInList("detail", strBrowseInfoStyle));
bool bExcludeCols = (StringUtil.IsInList("excludecolsoflowthreshold", strBrowseInfoStyle) == true);
bool bCols = (StringUtil.IsInList("cols", strBrowseInfoStyle) == true);
List <string> pathlist = new List <string>();
List <DupSearchResult> results = new List <DupSearchResult>();
for (int i = 0; i < nCount; i++) // BUG nStart +
{
DupLineItem item = (DupLineItem)dupset[nStart + i]; // changed
DupSearchResult result_item = new DupSearchResult();
results.Add(result_item);
result_item.Path = item.Path;
result_item.Weight = item.Weight;
result_item.Threshold = item.Threshold;
if (bDetail)
{
result_item.Detail = item.Detail;
}
// paths[i] = item.Path;
if (bCols == true)
{
if (bExcludeCols == true && item.Weight < item.Threshold)
{
}
else
{
pathlist.Add(item.Path);
}
}
}
if (pathlist.Count > 0)
{
// string[] paths = new string[pathlist.Count];
string[] paths = StringUtil.FromListString(pathlist);
nRet = channel.GetBrowseRecords(paths,
"cols",
out ArrayList aRecord,
out strError);
if (nRet == -1)
{
strError = "GetBrowseRecords() error: " + strError;
goto ERROR1;
}
int j = 0;
for (int i = 0; i < results.Count; i++)
{
DupSearchResult result_item = results[i];
if (result_item.Path != pathlist[j])
{
continue;
}
string[] cols = (string[])aRecord[j];
results[i].Cols = cols; // style中不包含id
j++;
if (j >= pathlist.Count)
{
break;
}
}
}
searchresults = new DupSearchResult[results.Count];
results.CopyTo(searchresults);
result.Value = searchresults.Length;
return(result);
ERROR1:
result.Value = -1;
result.ErrorCode = ErrorCode.SystemError;
result.ErrorInfo = strError;
return(result);
}
// 针对一个from进行检索
// parameters:
// strExcludeBiblioRecPath 要排除掉的记录路径
// return:
// -1 error
// 0 not found
// 1 found
int SearchOneFrom(
// RmsChannelCollection Channels,
RmsChannel channel,
string strDbName,
string strFrom,
string strKey,
string strSearchStyle,
int nWeight,
int nThreshold,
long nMax,
string strExcludeBiblioRecPath,
out DupResultSet dupset,
out string strError)
{
strError = "";
dupset = null;
long lRet = 0;
if (strSearchStyle == "")
strSearchStyle = "exact";
string strQueryXml = "<target list='"
+ StringUtil.GetXmlStringSimple(strDbName + ":" + strFrom) // 2007/9/14
+ "'><item><word>"
+ StringUtil.GetXmlStringSimple(strKey)
+ "</word><match>" + strSearchStyle + "</match><relation>=</relation><dataType>string</dataType><maxCount>" + nMax.ToString() + "</maxCount></item><lang>zh</lang></target>";
string strSearchReason = "key='" + strKey + "', from='" + strFrom + "', weight=" + Convert.ToString(nWeight);
/*
RmsChannel channel = Channels.GetChannel(this.WsUrl);
if (channel == null)
{
strError = "get channel error";
goto ERROR1;
}
* */
Debug.Assert(channel != null, "");
lRet = channel.DoSearch(strQueryXml,
"dup",
"", // strOuputStyle
out strError);
if (lRet == -1)
goto ERROR1;
if (lRet == 0)
return 0; // not found
long lHitCount = lRet;
long lStart = 0;
long lPerCount = Math.Min(50, lHitCount);
List<string> aPath = null;
dupset = new DupResultSet();
dupset.Open(false, getTempFileName);
// 获得结果集，对逐个记录进行处理
for (; ; )
{
// TODO: 中间要可以中断
lRet = channel.DoGetSearchResult(
"dup", // strResultSetName
lStart,
lPerCount,
"zh",
null, // stop
out aPath,
out strError);
if (lRet == -1)
goto ERROR1;
if (lRet == 0)
{
strError = "未命中";
break; // ??
}
// TODO: 要判断 aPath.Count == 0 跳出循环。否则容易进入死循环
// 处理浏览结果
for (int i = 0; i < aPath.Count; i++)
{
string strPath = aPath[i];
// 忽略发起记录的路径
if (strPath == strExcludeBiblioRecPath)
continue;
DupLineItem item = new DupLineItem();
item.Path = strPath;
item.Weight = nWeight;
item.Threshold = nThreshold;
dupset.Add(item);
}
lStart += aPath.Count;
if (lStart >= lHitCount || lPerCount <= 0)
break;
}
return 1;
ERROR1:
return -1;
}
// 合并两个事项的 Detail 部分
static void MergeDetail(DupLineItem left, DupLineItem right)
{
left.Detail = left.Detail + " OR " + right.Detail;
}
