请看我的问题，相信我这很容易解决

Question

我试图实现异步并在spawn子进程中等待。但它不起作用。请看这个

预期产量

 *************
http://www.stevecostellolaw.com/
 *************
http://www.stevecostellolaw.com/personal-injury.html
http://www.stevecostellolaw.com/personal-injury.html
 *************
http://www.stevecostellolaw.com/#
http://www.stevecostellolaw.com/#
 *************
http://www.stevecostellolaw.com/home.html
http://www.stevecostellolaw.com/home.html
 *************
http://www.stevecostellolaw.com/about-us.html
http://www.stevecostellolaw.com/about-us.html
 *************
http://www.stevecostellolaw.com/
http://www.stevecostellolaw.com/

 *************

因为每次找到await生成子对象时，它都会返回python脚本并打印 *************它，然后打印URL。忽略此处相同url的2次打印

我得到的输出

C:\Users\ASUS\Desktop\searchermc>node app.js
server running on port 3000

DevTools listening on ws://127.0.0.1:52966/devtools/browser/933c20c7-e295-4d84-a4b8-eeb5888ecbbf
[3020:120:0402/105304.190:ERROR:device_event_log_impl.cc(214)] [10:53:04.188] USB: usb_device_handle_win.cc:1056 Failed to read descriptor from node connection: A device attached to the system is not functioning. (0x1F)
[3020:120:0402/105304.190:ERROR:device_event_log_impl.cc(214)] [10:53:04.189] USB: usb_device_handle_win.cc:1056 Failed to read descriptor from node connection: A device attached to the system is not functioning. (0x1F)

 *************
http://www.stevecostellolaw.com/
http://www.stevecostellolaw.com/personal-injury.html
http://www.stevecostellolaw.com/personal-injury.html
http://www.stevecostellolaw.com/#
http://www.stevecostellolaw.com/#
http://www.stevecostellolaw.com/home.html
http://www.stevecostellolaw.com/home.html
http://www.stevecostellolaw.com/about-us.html
http://www.stevecostellolaw.com/about-us.html
http://www.stevecostellolaw.com/
http://www.stevecostellolaw.com/

 *************

请参阅下面的app.js代码

// form submit request
app.post('/formsubmit', function(req, res){

    csvData = req.files.csvfile.data.toString('utf8');
    filteredArray = cleanArray(csvData.split(/\r?\n/))
    csvData = get_array_string(filteredArray)
    csvData = csvData.trim()
    
    var keywords = req.body.keywords
    keywords = keywords.trim()

    // Send request to python script
    var spawn = require('child_process').spawn;
    var process = spawn('python', ["./webextraction.py", csvData, keywords, req.body.full_search])

    var outarr = []

    // process.stdout.on('data', (data) => {
    //   console.log(`stdout: ${data}`);
    // });

    process.stdout.on('data', async function(data){

      console.log("\n ************* ")
      console.log(data.toString().trim())
      await outarr.push(data.toString().trim())
      console.log("\n ************* ")

    });

});

Python函数，当if条件匹配时发送URL

# Function for searching keyword start
def search_keyword(href, search_key):
    extension_list = ['mp3', 'jpg', 'exe', 'jpeg', 'png', 'pdf', 'vcf']
    if(href.split('.')[-1] not in extension_list):
        try:    
            content = selenium_calling(href)
            soup = BeautifulSoup(content,'html.parser')
            search_string = re.sub("\s+"," ", soup.body.text)
            search_string = search_string.lower()
            res = [ele for ele in search_key if(ele.lower() in search_string)]
            outstr = getstring(res)
            outstr = outstr.lstrip(", ")
            if(len(res) > 0):
                print(href)
                found_results.append(href)
                href_key_dict[href] = outstr
                return 1
            else:
                notfound_results.append(href)
        except Exception as err:
            pass

我之所以想这么做，是因为python脚本需要更多的时间来执行，因此每次都会出现超时错误，所以我想在我的nodejs脚本中获得python脚本的中间输出。您可以在下图中看到我遇到的错误

Answer 1

我不确定我是否完全理解你想做什么，但我会试一试，因为你似乎已经问了很多次这个问题（这通常不是一个好主意）。我认为你的问题不够清晰——如果你能澄清你的最终目标是什么（即，你希望它如何运作？），这会有很大帮助

我想你在这里提到了两个不同的问题。首先，您希望在从脚本返回的每个单独的数据之前放置一行新的“******”。这是一个不可依赖的问题——请查看此问题的答案以了解更多详细信息：Order of process.stdout.on( 'data', ... ) and process.stderr.on( 'data', ... )。数据将以块的形式传递给stdout处理程序，而不是逐行传递，根据管道中当前的数据量，可以一次提供任意数量的数据

我最困惑的部分是您的措辞“在我的nodejs脚本中获得python脚本的中间输出”。不一定有任何“即时”数据——您不能依赖于进程的标准输出处理程序在任何特定时间输入的数据，它将以Python脚本本身及其运行的进程确定的速度向您提供数据。话虽如此，听起来你的主要问题是你的帖子超时了。你永远不会结束你的请求-这就是为什么你会得到一个超时。我将假设您希望在发送回响应之前等待第一个数据块（不管它包含多少行）。在这种情况下，您需要添加res.send，如下所示：

    // form submit request
app.post('/formsubmit', function(req, res){

    csvData = req.files.csvfile.data.toString('utf8');
    filteredArray = cleanArray(csvData.split(/\r?\n/))
    csvData = get_array_string(filteredArray)
    csvData = csvData.trim()
    
    var keywords = req.body.keywords
    keywords = keywords.trim()

    // Send request to python script
    var spawn = require('child_process').spawn;
    var process = spawn('python', ["./webextraction.py", csvData, keywords, req.body.full_search])

    var outarr = []

    // process.stdout.on('data', (data) => {
    //   console.log(`stdout: ${data}`);
    // });
    
    // Keep track of whether we've already ended the request
    let responseSent = false;

    process.stdout.on('data', async function(data){

        console.log("\n ************* ")
        console.log(data.toString().trim())
        outarr.push(data.toString().trim())
        console.log("\n ************* ")
        
        // If the request hasn't already been ended, send back the current output from the script
        // and end the request
        if (!responseSent) {
            responseSent = true;
            res.send(outarr);
        }
    });

});