分包

src/jy/pretreated/division.go

@@ -585,198 +585,164 @@ func divisionPackageChild(blockPackage *map[string]*util.BlockPackage, content,
 	//	util.Debug(con)
 	//	util.Debug(pkg)
 	//分包前面添加换行
-	//log.Println(con)
-	bools := make(map[string]bool)
-	for k, v := range pkg {
+	appendWarpIndex := []int{}
+	for _, v := range pkg {
 		//如果文本内容以识别出来的分包标识结尾，不是分包
 		if len(pkg) == 1 && strings.HasSuffix(con, v[0]) {
 			return false, ""
 		}
 		//
-		is := regexp.MustCompile(v[0] + "[:：]*").FindAllString(con, -1)
+		is := regexp.MustCompile(v[0]+"[:：]*").FindAllStringIndex(con, -1)
+		for _, sv := range is {
+			appendWarpIndex = append(appendWarpIndex, sv[0])
+		}
+	}
+	appendWarpIndex = getPkgIndex(appendWarpIndex)
+	conTemp := ""
+	for k, v := range appendWarpIndex {
+		if k == 0 {
+			conTemp += con[:v] + "\n"
+		} else {
+			conTemp += "\n" + con[appendWarpIndex[k-1]:v]
+		}
+		if k == len(appendWarpIndex)-1 {
+			conTemp += "\n" + con[v:]
+		}
+	}
+	con = conTemp
+	con = replSerial.ReplaceAllString(con, "\n")
+	con = regMoreWrap.ReplaceAllString(con, "\n")
+	//log.Println(con)
+	//根据分包，找索引位置
+	indexMap := map[int]int{}
+	indexKeyStringMap := map[int]string{}
+	indexKeyIntMap := map[int]int{}
+	indexs := []int{}
+	startEndMap := map[int]int{}
+	pkgIndexMap := map[string][]int{}
+	indexPkgMap := map[int]string{}
+	//遍历分包，把kv在包前面的移动到包后面
+	for _, v := range pkg {
+		pgflag := v[0] + "[:：]*"
+		is := regexp.MustCompile(pgflag).FindAllStringIndex(con, -1)
 		for _, sv := range is {
-			newBpkg := &util.BlockPackage{
-				Origin:   sv,
-				Text:     con,
-				Index:    k,
-				Accuracy: false,
+			indexMap[sv[0]] = sv[1]
+			indexs = append(indexs, sv[0])
+			pkgIndexMap[v[0]] = append(pkgIndexMap[v[0]], sv[0])
+			indexPkgMap[sv[0]] = v[0]
+		}
+		//key在包前面，并且在一行的开头
+		keys := regexp.MustCompile("([\r\n]|^)([\u4e00-\u9fa5]{2,30}?([（(].{1,8}?[)）])?[:：\\s\u3000\u2003\u00a0]+.*?)"+pgflag).FindAllStringSubmatchIndex(con, -1)
+		if len(keys) == 0 {
+			//key在包前面，并且key以冒号结尾
+			keys = regexp.MustCompile("()([\u4e00-\u9fa5]{2,30}?([（(].{1,8}?[)）])?[:：]+[\\s\u3000\u2003\u00a0]*[\r\n])"+pgflag).FindAllStringSubmatchIndex(con, -1)
+		}
+		if len(keys) == 0 {
+			keys = regexp.MustCompile("()注[:：]([\u4e00-\u9fa5]{2,8}?([（(].{1,8}?[)）])?[\\s\u3000\u2003\u00a0]*[\r\n])"+pgflag).FindAllStringSubmatchIndex(con, -1)
+		}
+		for _, key := range keys {
+			startEndMap[key[5]] = key[4]
+			//
+			headkey := con[key[4]:key[5]]
+			headkey = regReplAllSpace.ReplaceAllString(headkey, "")
+			if !regDivision.MatchString(headkey) {
+				headkey += "："
 			}
-			if (*blockPackage)[k] == nil && !bools[sv] {
-				bools[sv] = true
-				con = strings.ReplaceAll(con, sv, "\n")
-				//log.Println(k, con)
-				kvAll := GetKVAll(con, title, nil, 4)
-				newBpkg.ColonKV = kvAll
-				newBpkg.SpaceKV = SspacekvEntity.Entrance(con, "", nil)
-				(*blockPackage)[k] = newBpkg
-			} else if (*blockPackage)[k].ColonKV != nil {
-				kvAll := GetKVAll(con, title, nil, 4)
-				MergeKvTags((*blockPackage)[k].ColonKV.KvTags, kvAll.KvTags)
+			headkey = moreColonReg.ReplaceAllString(headkey, "：")
+			colonIndexs := regDivision.FindAllStringIndex(headkey, -1)
+			if len(colonIndexs) > 1 {
+				headkey = headkey[colonIndexs[len(colonIndexs)-2][1]:colonIndexs[len(colonIndexs)-1][1]]
 			}
-			for kk, vv := range (*blockPackage)[k].ColonKV.KvTags {
-				for _, vvvv := range vv {
-					//log.Println(kk,vv,kkk,vvvv.Value,vvvv.Key)
-					//if kk == "项目编号" && (*blockPackage)[k].Origin == "" {
-					//	(*blockPackage)[k].Origin = strings.TrimRightFunc(vvvv.Value, func(r rune) bool {
-					//		return r == 65289 || r == 41 || r == 46 || r == 12290
-					//	})
-					//	break
-					//} else
-					if kk == "项目名称" && (*blockPackage)[k].Name == "" {
-						(*blockPackage)[k].Name = vvvv.Value
-						break
-					} else if kk == "预算" && (*blockPackage)[k].Budget == 0 {
-						(*blockPackage)[k].Budget = qutil.Float64All(vvvv.Value)
-						break
-					} else if kk == "中标单位" && (*blockPackage)[k].Winner == "" {
-						(*blockPackage)[k].Winner = vvvv.Value
-						break
-					} else if kk == "标段类型" && (*blockPackage)[k].Type == "" {
-						(*blockPackage)[k].Type = vvvv.Value
-						break
-					} else if kk == "中标金额" && (*blockPackage)[k].Bidamount == 0 {
-						(*blockPackage)[k].Bidamount = qutil.Float64All(vvvv.Value)
-						break
-					}
+			indexKeyStringMap[key[5]] = headkey
+			indexKeyIntMap[key[5]] = key[1]
+		}
+	}
+	indexs = getPkgIndex(indexs)
+	for ik, iv := range indexs {
+		if indexKeyStringMap[iv] != "" {
+			continue
+		}
+		if indexKeyIntMap[iv] == indexMap[iv] {
+			continue
+		}
+		if ik > 0 {
+			indexKeyStringMap[iv] = indexKeyStringMap[indexs[ik-1]]
+		}
+	}
+	//
+	//获取截取标识
+	surplusText, maxWarpCount, indexTextMap, indexWarpMap := interceptText(indexs, indexPkgMap, pkgIndexMap, startEndMap, con)
+	//查找分包内容，分kv
+	for _, iv := range indexs {
+		text := indexTextMap[iv]
+		//
+		warpIndex := regSpliteSegment.FindAllStringIndex(text, -1)
+		if len(indexWarpMap) > 0 {
+			maxWarpCount = indexWarpMap[iv]
+		}
+		if maxWarpCount > 0 && len(warpIndex) >= 5 && len(warpIndex) > maxWarpCount {
+			textTemp := text
+			text = textTemp[:warpIndex[maxWarpCount-1][1]]
+			surplusText += textTemp[warpIndex[maxWarpCount-1][0]:]
+		}
+		for bk, bv := range pkg {
+			//判断分包如果在这段文字里面，该段文字就属于该包的
+			if !strings.HasPrefix(text, bv[0]) {
+				continue
+			}
+			index := util.PackageNumberConvert(bk)
+			//去掉前缀,空格必须要加，分kv的时候要用
+			text = regexp.MustCompile(bv[0]+"[:：]*").ReplaceAllString(text, "")
+			headKey := ""
+			if indexKeyStringMap[iv] != "" {
+				//if !filterPkgTitleKey.MatchString(indexKeyStringMap[iv]) {
+				headKey = indexKeyStringMap[iv]
+				text = indexKeyStringMap[iv] + "  " + text
+				//}
+				for _, pkgIndexMap_v := range pkgIndexMap[bv[0]] {
+					delete(indexKeyStringMap, pkgIndexMap_v)
 				}
-
+			}
+			//如果一块中有多个相同的包，合并到一个
+			if (*blockPackage)[index] != nil {
+				//合并文本
+				(*blockPackage)[index].Text += "\n" + text
+				//合并冒号kv
+				colonJobKv := GetKVAll(strings.TrimLeft(text, headKey), "", nil, 1)
+				if headKey != "" {
+					kvAgain := GetKVAll(text, "", nil, 4)
+					MergeKvTags(colonJobKv.KvTags, kvAgain.KvTags)
+				}
+				MergeKvTags((*blockPackage)[index].ColonKV.KvTags, colonJobKv.KvTags)
+				//合并空格kv
+				spaceJobKv := SspacekvEntity.Entrance(text, "", nil)
+				MergeKvTags((*blockPackage)[index].SpaceKV.KvTags, spaceJobKv.KvTags)
+			} else {
+				newBpkg := &util.BlockPackage{
+					Origin:   bk,
+					Text:     text,
+					Index:    index,
+					Type:     bv[1],
+					Accuracy: accuracy,
+				}
+				finalKv := GetKVAll(strings.TrimLeft(text, headKey), "", nil, 4)
+				if headKey != "" {
+					kvAgain := GetKVAll(text, "", nil, 4)
+					MergeKvTags(finalKv.KvTags, kvAgain.KvTags)
+				}
+				newBpkg.ColonKV = finalKv
+				newBpkg.SpaceKV = SspacekvEntity.Entrance(text, "", nil)
+				(*blockPackage)[index] = newBpkg
 			}
 		}
 	}
-	//log.Println(con)
-	//con = replSerial.ReplaceAllString(con, "\n")
-	//con = regMoreWrap.ReplaceAllString(con, "\n")
-	////util.Debug(con)
-	////根据分包，找索引位置
-	//indexMap := map[int]int{}
-	//indexKeyStringMap := map[int]string{}
-	//indexKeyIntMap := map[int]int{}
-	//indexs := []int{}
-	//startEndMap := map[int]int{}
-	//pkgIndexMap := map[string][]int{}
-	//indexPkgMap := map[int]string{}
-	////遍历分包，把kv在包前面的移动到包后面
-	//for _, v := range pkg {
-	//	pgflag := v[0] + "[:：]*"
-	//	is := regexp.MustCompile(pgflag).FindAllStringIndex(con, -1)
-	//	for _, sv := range is {
-	//		indexMap[sv[0]] = sv[1]
-	//		indexs = append(indexs, sv[0])
-	//		pkgIndexMap[v[0]] = append(pkgIndexMap[v[0]], sv[0])
-	//		indexPkgMap[sv[0]] = v[0]
-	//	}
-	//	//key在包前面，并且在一行的开头
-	//	keys := regexp.MustCompile("([\r\n]|^)([\u4e00-\u9fa5]{2,30}?([（(].{1,8}?[)）])?[:：\\s\u3000\u2003\u00a0]+.*?)"+pgflag).FindAllStringSubmatchIndex(con, -1)
-	//	if len(keys) == 0 {
-	//		//key在包前面，并且key以冒号结尾
-	//		keys = regexp.MustCompile("()([\u4e00-\u9fa5]{2,30}?([（(].{1,8}?[)）])?[:：]+[\\s\u3000\u2003\u00a0]*[\r\n])"+pgflag).FindAllStringSubmatchIndex(con, -1)
-	//	}
-	//	if len(keys) == 0 {
-	//		keys = regexp.MustCompile("()注[:：]([\u4e00-\u9fa5]{2,8}?([（(].{1,8}?[)）])?[\\s\u3000\u2003\u00a0]*[\r\n])"+pgflag).FindAllStringSubmatchIndex(con, -1)
-	//	}
-	//	for _, key := range keys {
-	//		startEndMap[key[5]] = key[4]
-	//		//
-	//		headkey := con[key[4]:key[5]]
-	//		headkey = regReplAllSpace.ReplaceAllString(headkey, "")
-	//		if !regDivision.MatchString(headkey) {
-	//			headkey += "："
-	//		}
-	//		headkey = moreColonReg.ReplaceAllString(headkey, "：")
-	//		colonIndexs := regDivision.FindAllStringIndex(headkey, -1)
-	//		if len(colonIndexs) > 1 {
-	//			headkey = headkey[colonIndexs[len(colonIndexs)-2][1]:colonIndexs[len(colonIndexs)-1][1]]
-	//		}
-	//		indexKeyStringMap[key[5]] = headkey
-	//		indexKeyIntMap[key[5]] = key[1]
-	//	}
-	//}
-	//indexs = getPkgIndex(indexs)
-	//for ik, iv := range indexs {
-	//	if indexKeyStringMap[iv] != "" {
-	//		continue
-	//	}
-	//	if indexKeyIntMap[iv] == indexMap[iv] {
-	//		continue
-	//	}
-	//	if ik > 0 {
-	//		indexKeyStringMap[iv] = indexKeyStringMap[indexs[ik-1]]
-	//	}
-	//}
-	////
-	////获取截取标识
-	//surplusText, maxWarpCount, indexTextMap, indexWarpMap := interceptText(indexs, indexPkgMap, pkgIndexMap, startEndMap, con)
-	////查找分包内容，分kv
-	//for _, iv := range indexs {
-	//	text := indexTextMap[iv]
-	//	//
-	//	warpIndex := regSpliteSegment.FindAllStringIndex(text, -1)
-	//	if len(indexWarpMap) > 0 {
-	//		maxWarpCount = indexWarpMap[iv]
-	//	}
-	//	if maxWarpCount > 0 && len(warpIndex) >= 5 && len(warpIndex) > maxWarpCount {
-	//		textTemp := text
-	//		text = textTemp[:warpIndex[maxWarpCount-1][1]]
-	//		surplusText += textTemp[warpIndex[maxWarpCount-1][0]:]
-	//	}
-	//	for bk, bv := range pkg {
-	//		//判断分包如果在这段文字里面，该段文字就属于该包的
-	//		if !strings.HasPrefix(text, bv[0]) {
-	//			continue
-	//		}
-	//		index := util.PackageNumberConvert(bk)
-	//		//去掉前缀,空格必须要加，分kv的时候要用
-	//		text = regexp.MustCompile(bv[0]+"[:：]*").ReplaceAllString(text, "")
-	//		headKey := ""
-	//		if indexKeyStringMap[iv] != "" {
-	//			//if !filterPkgTitleKey.MatchString(indexKeyStringMap[iv]) {
-	//			headKey = indexKeyStringMap[iv]
-	//			text = indexKeyStringMap[iv] + "  " + text
-	//			//}
-	//			for _, pkgIndexMap_v := range pkgIndexMap[bv[0]] {
-	//				delete(indexKeyStringMap, pkgIndexMap_v)
-	//			}
-	//		}
-	//		//如果一块中有多个相同的包，合并到一个
-	//		if (*blockPackage)[index] != nil {
-	//			//合并文本
-	//			(*blockPackage)[index].Text += "\n" + text
-	//			//合并冒号kv
-	//			colonJobKv := GetKVAll(strings.TrimLeft(text, headKey), "", nil, 1)
-	//			if headKey != "" {
-	//				kvAgain := GetKVAll(text, "", nil, 4)
-	//				MergeKvTags(colonJobKv.KvTags, kvAgain.KvTags)
-	//			}
-	//			MergeKvTags((*blockPackage)[index].ColonKV.KvTags, colonJobKv.KvTags)
-	//			//合并空格kv
-	//			spaceJobKv := SspacekvEntity.Entrance(text, "", nil)
-	//			MergeKvTags((*blockPackage)[index].SpaceKV.KvTags, spaceJobKv.KvTags)
-	//		} else {
-	//			newBpkg := &util.BlockPackage{
-	//				Origin:   bk,
-	//				Text:     text,
-	//				Index:    index,
-	//				Type:     bv[1],
-	//				Accuracy: accuracy,
-	//			}
-	//			finalKv := GetKVAll(strings.TrimLeft(text, headKey), "", nil, 4)
-	//			if headKey != "" {
-	//				kvAgain := GetKVAll(text, "", nil, 4)
-	//				MergeKvTags(finalKv.KvTags, kvAgain.KvTags)
-	//			}
-	//			newBpkg.ColonKV = finalKv
-	//			newBpkg.SpaceKV = SspacekvEntity.Entrance(text, "", nil)
-	//			(*blockPackage)[index] = newBpkg
-	//		}
-	//	}
-	//}
 	//中标人排序
-	if !strings.Contains(title, "招标") && isFindWinnerOrder && blockPackage != nil && len(*blockPackage) > 0 {
+	if isFindWinnerOrder && blockPackage != nil && len(*blockPackage) > 0 {
 		for _, v := range *blockPackage {
 			v.WinnerOrder = winnerOrderEntity.Find(v.Text, true, 2)
 		}
 	}
-	return true, con
+	return true, surplusText
 }
 func getPkgIndex(indexs []int) []int {
 	sort.Ints(indexs)