随记： 整理出一个算法详解过程。 举个例子：
映射关系如下
a -> [b,c,d,e]
b->[c,d,e]
c->[d,e]
d->[e]
=================================================================
做一个入深度记录如下：
Key Value
e 4
d 3
c 2
b 1
a 0
====================================================================
此时我们需要判断深度值等于0的，我们插入到集合里
List result = new ArrayList<>();
result.add(0, a);
我们通过键值a，获得数组[b,c,d,e]处理（入深度）得到如下
e 3
d 2
c 1
b 0
再次操作
result.add(0,b);
反复操作如上步骤即可得到排序
public void topoSort(List params) { params.forEach(i->i.setLevel(0)); Map<String, TeamTemplateItem> totalMap = params.stream().collect(Collectors.toMap(TeamTemplateItem::getColumnName, o -> o, (k1, k2) -> k2));
Map<String, List<String>> formulas = new HashMap<>();
totalMap.forEach((k, v) -> {
if (StringUtils.isNotBlank(v.getRelationFormula())) {
formulas.put(k, Arrays.asList(v.getRelationFormula().split(",")));
}
});
Map<String, Integer> inDegree = new HashMap<>();
Map<String, List<String>> graph = new HashMap<>(formulas);
for (String node : formulas.keySet()) {
inDegree.putIfAbsent(node, 0);
}
for (String node : formulas.keySet()) {
for (String neighbor : formulas.get(node)) {
inDegree.put(neighbor, inDegree.getOrDefault(neighbor, 0) + 1);
}
}
Queue<String> queue = new LinkedList<>();
for (String node : inDegree.keySet()) {
if (inDegree.get(node) == 0) {
queue.offer(node);
}
}
List<String> result = new ArrayList<>();
while (!queue.isEmpty()) {
String node = queue.poll();
result.add(0, node); // Insert at the beginning of the list
for (String neighbor : graph.getOrDefault(node, new ArrayList<>())) {
inDegree.put(neighbor, inDegree.get(neighbor) - 1);
if (inDegree.get(neighbor) == 0) {
queue.offer(neighbor);
}
}
}
for (int i = 0; i < result.size(); i++) {
if (totalMap.containsKey(result.get(i))) {
totalMap.get(result.get(i)).setLevel(i);
}
}
}