searchwith

IDEA插件:search with bing、search with baidu

佐手、 提交于 2020-11-25 04:10:31
//转载请注明出处: https://www.cnblogs.com/nreg/p/11267169.html 当项目出现错误时,经常需要复制错误信息粘贴到浏览器查询,但是手动复制再粘贴太麻烦了, 因此IDEA官方给了右键菜单search with google, 感觉上方便了许多,但是goolgle在咱这个地方是查询不到的, 因此一位名叫Xu Wang的网友开发了一款叫search with baidu的插件,在IDEA的插件库可以查询到, 我用了一下他的最新版本1.3,在复制一大串错误信息时,使用search with baidu不生效,IDEA没有反应,应该是有BUG, 盲猜他没有对要搜索的字符串转换为url请求编码格式,因为我开发插件时也遇到了这个问题,但是我解决了: 1 // 将搜索的字符串转为URL请求编码格式(盲猜他应该没有写这一行) 2 String seclectStr = URLEncoder.encode(selectedText.trim(), "UTF-8" ); 3 // 使用bing引擎搜索 4 String url = "https://cn.bing.com/search?q=" + seclectStr; 我开发了两款插件,一款叫search with bing,一款叫search with baidu, 已经上传到IDEA插件库,目前正在审核,

如何延迟.keyup()处理函数,直到用户停止键入?

☆樱花仙子☆ 提交于 2020-08-15 17:56:45
问题: I've got a search field. 我有一个搜索字段。 Right now it searches for every keyup. 现在,它会搜索每个键。 So if someone types “Windows”, it will make a search with AJAX for every keyup: “W”, “Wi”, “Win”, “Wind”, “Windo”, “Window”, “Windows”. 因此,如果有人键入“ Windows”,它将使用AJAX搜索每个键入的内容:“ W”,“ Wi”,“ Win”,“ Wind”,“ Windo”,“ Window”,“ Windows”。 I want to have a delay, so it only searches when the user stops typing for 200 ms. 我希望有一个延迟,因此它仅在用户停止键入200毫秒时才搜索。 There is no option for this in the keyup function, and I have tried setTimeout , but it didn't work. 在 keyup 函数中没有用于此的选项,我已经尝试过 setTimeout ,但是它没有用。 How can I do that

CondenseNet:可学习分组卷积,原作对DenseNet的轻量化改造 | CVPR 2018

房东的猫 提交于 2020-08-14 13:33:39
CondenseNet特点在于可学习分组卷积的提出,结合训练过程进行剪枝,不仅能准确地剪枝,还能继续训练,使网络权重更平滑,是个很不错的工作   来源:晓飞的算法工程笔记 公众号 论文:Neural Architecture Search with Reinforcement Learning 论文地址: https://arxiv.org/abs/1711.09224 论文代码: https://github.com/ShichenLiu/CondenseNet Introduction   DenseNet基于特征复用,能够达到很好的性能,但是论文认为其内在连接存在很多冗余,早期的特征不需要复用到较后的层。为此,论文基于可学习分组卷积提出CondenseNet,能够在训练阶段自动稀疏网络结构,选择最优的输入输出连接模式,并在最后将其转换成常规的分组卷积分组卷积结构。 CondenseNets   分组卷积能够有效地降低网络参数,对于稠密的网络结构而言,可以将$3\times 3$卷积变为$3\times 3$分组卷积。然而,若将$1\times 1$卷积变为$1\times 1$分组卷积,则会造成性能的大幅下降,主要由于$1\times 1$卷积的输入一般有其内在的联系,并且输入有较大的多样性,不能这样硬性地人为分组。随机打乱能够一定程度地缓解性能的降低

SharePoint 2010 vs SharePoint 2013

泄露秘密 提交于 2020-05-04 03:21:02
* From a document collaboration perspective, the structures of both versions are the same – so if you create a metadata architecture for documents in 2010 it should be fully upgradable to 2013. * The most significant upgrades in document management are in the user experience – including drag and drop to upload documents and the ability to edit managed metadata in a datasheet view. *The primary differences are in the social experiences, especially with discussion boards. The 2013 discussion board (with Community features enabled) creates and engaging “Facebook-like” activity stream, which is