ContentExtractor中的computeInfo函数会出现StackOverflowError

[yanpeng]

比如处理这个url http://www.suixian.gov.cn/news/News_View.asp?NewsID=20939 时会出现上述错误，我已经将其修改为非递归版本了，请问可以将代码提交上来吗？

[xiaozizero]

修改的代码可以贴在这里啊

[yanpeng]

    // 计算叶子节点信息
    private CountInfo computeLeafNodeInfo(Node node)
    {
        if (node instanceof TextNode)
        {
            TextNode tn = (TextNode) node;
            CountInfo countInfo = new CountInfo();
            String text = tn.text();
            int len = text.length();
            countInfo.textCount = len;
            countInfo.leafList.add(len);
            return countInfo;
        }
        else
        {
            return new CountInfo();
        }
    }

    protected CountInfo computeInfo(Node node)
    {
        if (node instanceof Element)
        {
            Deque<Node> stack = new ArrayDeque<Node>();
            Deque<Node> queue = new ArrayDeque<Node>();
            Set<Node> accessedNodes = new HashSet<Node>();
            Map<Node, CountInfo> nodeInfoMap = new HashMap<Node, CountInfo>();

            stack.addFirst(node);
            while (!stack.isEmpty())
            {
                Node headerNode = stack.getFirst();
                // 如果是非叶子节点添加至已访问节点集合
                // 并且将它的孩子以逆序入栈
                if ((headerNode instanceof Element) && !accessedNodes.contains(headerNode))
                {
                    accessedNodes.add(headerNode);
                    for (int i = headerNode.childNodeSize() - 1; i >= 0; --i)
                    {
                        stack.addFirst(headerNode.childNode(i));
                    }
                }
                else // 对于叶子节点和已经访问过的非叶子节点则入队列
                {
                    queue.addLast(stack.removeFirst());
                }
            }

            while (!queue.isEmpty())
            {
                Node headerNode = queue.removeFirst();
                if (headerNode instanceof Element)
                {
                    Element tag = (Element) headerNode;
                    CountInfo countInfo = new CountInfo();

                    for (Node childNode : headerNode.childNodes()) 
                    {
                        CountInfo childCountInfo = nodeInfoMap.get(childNode);
                        countInfo.textCount += childCountInfo.textCount;
                        countInfo.linkTextCount += childCountInfo.linkTextCount;
                        countInfo.tagCount += childCountInfo.tagCount;
                        countInfo.linkTagCount += childCountInfo.linkTagCount;
                        countInfo.leafList.addAll(childCountInfo.leafList);
                        countInfo.densitySum += childCountInfo.density;
                        countInfo.pCount += childCountInfo.pCount;
                    }

                    countInfo.tagCount++;
                    String tagName = tag.tagName();
                    if (tagName.equals("a"))
                    {
                        countInfo.linkTextCount = countInfo.textCount;
                        countInfo.linkTagCount++;
                    }
                    else if (tagName.equals("p"))
                    {
                        countInfo.pCount++;
                    }

                    int pureLen = countInfo.textCount - countInfo.linkTextCount;
                    int len = countInfo.tagCount - countInfo.linkTagCount;
                    if (pureLen == 0 || len == 0)
                    {
                        countInfo.density = 0;
                    }
                    else
                    {
                        countInfo.density = (pureLen + 0.0) / len;
                    }

                    infoMap.put(tag, countInfo);
                    nodeInfoMap.put(headerNode, countInfo);
                }
                else
                {
                    nodeInfoMap.put(headerNode, computeLeafNodeInfo(headerNode));
                }
            }

            return nodeInfoMap.get(node);
        }
        else
        {
            return computeLeafNodeInfo(node);
        }
    }

[hujunxianligong]

等有空我看下，最近比较忙不好意思 yanpeng <notifications@github.com> 于2019年11月29日周五 上午11:47写道：

…

`// 计算叶子节点信息 private CountInfo computeLeafNodeInfo(Node node) { if (node instanceof TextNode) { TextNode tn = (TextNode) node; CountInfo countInfo = new CountInfo(); String text = tn.text(); int len = text.length(); countInfo.textCount = len; countInfo.leafList.add(len); return countInfo; } else { return new CountInfo(); } } protected CountInfo computeInfo(Node node) { if (node instanceof Element) { Deque<Node> stack = new ArrayDeque<Node>(); Deque<Node> queue = new ArrayDeque<Node>(); Set<Node> accessedNodes = new HashSet<Node>(); Map<Node, CountInfo> nodeInfoMap = new HashMap<Node, CountInfo>(); stack.addFirst(node); while (!stack.isEmpty()) { Node headerNode = stack.getFirst(); // 如果是非叶子节点添加至已访问节点集合 // 并且将它的孩子以逆序入栈 if ((headerNode instanceof Element) && !accessedNodes.contains(headerNode)) { accessedNodes.add(headerNode); for (int i = headerNode.childNodeSize() - 1; i >= 0; --i) { stack.addFirst(headerNode.childNode(i)); } } else // 对于叶子节点和已经访问过的非叶子节点则入队列 { queue.addLast(stack.removeFirst()); } } while (!queue.isEmpty()) { Node headerNode = queue.removeFirst(); if (headerNode instanceof Element) { Element tag = (Element) headerNode; CountInfo countInfo = new CountInfo(); for (Node childNode : headerNode.childNodes()) { CountInfo childCountInfo = nodeInfoMap.get(childNode); countInfo.textCount += childCountInfo.textCount; countInfo.linkTextCount += childCountInfo.linkTextCount; countInfo.tagCount += childCountInfo.tagCount; countInfo.linkTagCount += childCountInfo.linkTagCount; countInfo.leafList.addAll(childCountInfo.leafList); countInfo.densitySum += childCountInfo.density; countInfo.pCount += childCountInfo.pCount; } countInfo.tagCount++; String tagName = tag.tagName(); if (tagName.equals("a")) { countInfo.linkTextCount = countInfo.textCount; countInfo.linkTagCount++; } else if (tagName.equals("p")) { countInfo.pCount++; } int pureLen = countInfo.textCount - countInfo.linkTextCount; int len = countInfo.tagCount - countInfo.linkTagCount; if (pureLen == 0 || len == 0) { countInfo.density = 0; } else { countInfo.density = (pureLen + 0.0) / len; } infoMap.put(tag, countInfo); nodeInfoMap.put(headerNode, countInfo); } else { nodeInfoMap.put(headerNode, computeLeafNodeInfo(headerNode)); } } return nodeInfoMap.get(node); } else { return computeLeafNodeInfo(node); } }` — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#116?email_source=notifications&email_token=AAZZQYKEXPGVWJ45RXR6FL3QWCGGBA5CNFSM4JPPZSCKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEFNZMXQ#issuecomment-559650398>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAZZQYINCF74MUFYOUDUFODQWCGGBANCNFSM4JPPZSCA> .