一个适用于ClassIsland的新闻获取推送程序

Chronabys

注意：本程序基于Deep Seek v4 Flash构建，可能存在尚未解决的问题
本程序从中新网（"https://www.chinanews.com.cn/rss/scroll-news.xml","https://www.chinanews.com.cn/rss/edu.xml"，可自行调整，详见源码）获取当日新闻，在格式化后推送至 http://localhost:36000/news （可自行调整，详见源码）
使用方法：
安装ClassIsland的PutIsland插件；
修改源码中所要获取的新闻地址（详见 https://www.chinanews.cn/rss ）；
使用适配C++11版本的g++编译器将本程序构建为exe可执行程序并放置在合理位置；
在ClassIsland的组件设置中添加 网络文本 组件并将 Token 设置为 news ；
添加ClassIsland自动化以实现自动获取新闻文本；
建议设置对应组件自动隐藏以减少不必要的影响.
以下为对应源码：
// 编译命令 g++ -std=c++11 PathToThis.cpp -o Program.exe -lwinhttp

#include <windows.h>
#include <winhttp.h>
#include <string>
#include <vector>
#include <sstream>
#include <iostream>
#include <fstream> 

#pragma comment(lib, "winhttp.lib")

// 新闻条目结构体
struct NewsItem {
    std::string title;
    std::string desc;
};

// ---------- 1. HTTP GET 获取 RSS ----------
std::string httpGet(const std::string& url) {
    std::string result;
    std::wstring wurl(url.begin(), url.end());
    
    URL_COMPONENTS urlComp = { 0 };
    urlComp.dwStructSize = sizeof(urlComp);
    
    wchar_t hostName[512] = { 0 };  
    wchar_t urlPath[2048] = { 0 };
    urlComp.lpszHostName = hostName;
    urlComp.dwHostNameLength = 512;
    urlComp.lpszUrlPath = urlPath;
    urlComp.dwUrlPathLength = 2048;
    
    if (!WinHttpCrackUrl(wurl.c_str(), 0, 0, &urlComp))
        return result;

    HINTERNET hSession = WinHttpOpen(L"RSS Reader/1.0", WINHTTP_ACCESS_TYPE_DEFAULT_PROXY, NULL, NULL, 0);
    if (!hSession) return result;
    
    HINTERNET hConnect = WinHttpConnect(hSession, hostName, urlComp.nPort, 0);
    if (!hConnect) { WinHttpCloseHandle(hSession); return result; }
    
    DWORD flags = (urlComp.nScheme == INTERNET_SCHEME_HTTPS) ? WINHTTP_FLAG_SECURE : 0;
    HINTERNET hRequest = WinHttpOpenRequest(hConnect, L"GET", urlPath, NULL, NULL, NULL, flags);
    if (!hRequest) { WinHttpCloseHandle(hConnect); WinHttpCloseHandle(hSession); return result; }

    BOOL bResults = WinHttpSendRequest(hRequest, WINHTTP_NO_ADDITIONAL_HEADERS, 0, WINHTTP_NO_REQUEST_DATA, 0, 0, 0);
    if (bResults) bResults = WinHttpReceiveResponse(hRequest, NULL);
    
    if (bResults) {
        DWORD bytesRead = 0;
        char buffer[4096];
        while (WinHttpReadData(hRequest, buffer, sizeof(buffer), &bytesRead) && bytesRead > 0) {
            result.append(buffer, bytesRead);
        }
    }
    
    WinHttpCloseHandle(hRequest);
    WinHttpCloseHandle(hConnect);
    WinHttpCloseHandle(hSession);
    return result;
}

// ---------- 2. 简易 XML 解析 ----------
std::string extractTagContent(const std::string& xml, const std::string& tag) {
    std::string open = "<" + tag + ">";
    std::string close = "</" + tag + ">";
    size_t start = xml.find(open);
    if (start == std::string::npos) return "";
    start += open.length();
    size_t end = xml.find(close, start);
    if (end == std::string::npos) return "";
    
    std::string content = xml.substr(start, end - start);
    
    size_t cstart = content.find("<![CDATA[");
    if (cstart != std::string::npos) {
        content = content.substr(cstart + 9);
        size_t cend = content.find("]]>");
        if (cend != std::string::npos) content = content.substr(0, cend);
    }
    return content;
}

std::vector<NewsItem> parseRSS(const std::string& xml) {
    std::vector<NewsItem> items;
    size_t pos = 0;
    while (true) {
        size_t itemStart = xml.find("<item>", pos);
        if (itemStart == std::string::npos) break;
        size_t itemEnd = xml.find("</item>", itemStart);
        if (itemEnd == std::string::npos) break;
        
        std::string itemXml = xml.substr(itemStart, itemEnd - itemStart + 7);
        std::string title = extractTagContent(itemXml, "title");
        std::string desc = extractTagContent(itemXml, "description");
        
        if (!title.empty()) {
            items.push_back({title, desc});
        }
        pos = itemEnd + 7;
    }
    return items;
}

// 判断是否为标准空白字符
bool isSpaceChar(unsigned char c) {
    return c == ' ' || c == '\t' || c == '\r' || c == '\n' || c == '\v' || c == '\f';
}

// 强力清洗函数：剥离首尾的控制字符、半角空格以及【全角空格】
void trimString(std::string& str) {
    // 循环剥离前端
    while (!str.empty()) {
        unsigned char c1 = str.front();
        // 1. 检查普通空白
        if (isSpaceChar(c1)) {
            str.erase(str.begin());
            continue;
        }
        // 2. 检查全角空格 (全角空格在 UTF-8 为 E3 80 80，在 GBK 为 A1 A1)
        if (str.size() >= 3 && (unsigned char)str[0] == 0xE3 && (unsigned char)str[1] == 0x80 && (unsigned char)str[2] == 0x80) {
            str.erase(0, 3);
            continue;
        }
        if (str.size() >= 2 && (unsigned char)str[0] == 0xA1 && (unsigned char)str[1] == 0xA1) {
            str.erase(0, 2);
            continue;
        }
        break; // 既不是半角也不是全角空白，停止
    }

    // 循环剥离后端
    while (!str.empty()) {
        unsigned char cLast = str.back();
        if (isSpaceChar(cLast)) {
            str.pop_back();
            continue;
        }
        size_t len = str.size();
        if (len >= 3 && (unsigned char)str[len-3] == 0xE3 && (unsigned char)str[len-2] == 0x80 && (unsigned char)str[len-1] == 0x80) {
            str.pop_back(); str.pop_back(); str.pop_back();
            continue;
        }
        if (len >= 2 && (unsigned char)str[len-2] == 0xA1 && (unsigned char)str[len-1] == 0xA1) {
            str.pop_back(); str.pop_back();
            continue;
        }
        break;
    }
}

// ---------- 3. 格式化（严格清洗，完美贴合） ----------
std::string formatNews(const std::vector<NewsItem>& items) {
    std::string result;
    bool isFirst = true;

    for (const auto& item : items) {
        std::string clean_title = item.title;
        std::string clean_desc = item.desc;

        // 将新闻内部所有的换行符和制表符强制替换为空格，保证单行稳定
        for (char& c : clean_title) { if (isSpaceChar((unsigned char)c)) c = ' '; }
        for (char& c : clean_desc) { if (isSpaceChar((unsigned char)c)) c = ' '; }

        // 调用强力清洗，把新闻内容和标题自带的全角/半角空格全部拔除
        trimString(clean_title);
        trimString(clean_desc);

        // 如果描述内容和标题一模一样（很多新闻源会有这种情况），为了避免重复，直接清空描述
        if (clean_desc == "【" + clean_title + "】" || clean_desc == clean_title) {
            clean_desc = "";
        }

        // 只有在两条不同的新闻之间才插入 8 个空格作为分隔
        if (!isFirst) {
            result += "        "; 
        }
        isFirst = false;

        // 紧凑拼接：【标题】后面直接连正文，中间没有任何缝隙
        result += "【" + clean_title + "】";
        if (!clean_desc.empty()) {
            result += clean_desc;
        }
    }

    return result;
}

// ---------- 4. HTTP POST 推送 ----------
bool httpPost(const std::string& url, const std::string& data) {
    std::wstring wurl(url.begin(), url.end());
    URL_COMPONENTS urlComp = { 0 };
    urlComp.dwStructSize = sizeof(urlComp);
    
    wchar_t hostName[512] = { 0 };
    wchar_t urlPath[2048] = { 0 };
    urlComp.lpszHostName = hostName;
    urlComp.dwHostNameLength = 512;
    urlComp.lpszUrlPath = urlPath;
    urlComp.dwUrlPathLength = 2048;
    
    if (!WinHttpCrackUrl(wurl.c_str(), 0, 0, &urlComp))
        return false;

    HINTERNET hSession = WinHttpOpen(L"RSS Pusher/1.0", WINHTTP_ACCESS_TYPE_DEFAULT_PROXY, NULL, NULL, 0);
    if (!hSession) return false;
    
    HINTERNET hConnect = WinHttpConnect(hSession, hostName, urlComp.nPort, 0);
    if (!hConnect) { WinHttpCloseHandle(hSession); return false; }
    
    DWORD flags = (urlComp.nScheme == INTERNET_SCHEME_HTTPS) ? WINHTTP_FLAG_SECURE : 0;
    HINTERNET hRequest = WinHttpOpenRequest(hConnect, L"POST", urlPath, NULL, NULL, NULL, flags);
    if (!hRequest) { WinHttpCloseHandle(hConnect); WinHttpCloseHandle(hSession); return false; }

    std::wstring headers = L"Content-Type: text/plain\r\n";
    BOOL bResults = WinHttpSendRequest(hRequest, headers.c_str(), (DWORD)headers.length(), (LPVOID)data.c_str(), (DWORD)data.size(), (DWORD)data.size(), 0);
    
    if (bResults) bResults = WinHttpReceiveResponse(hRequest, NULL);
    
    DWORD statusCode = 0;
    if (bResults) {
        DWORD len = sizeof(statusCode);
        WinHttpQueryHeaders(hRequest, WINHTTP_QUERY_STATUS_CODE | WINHTTP_QUERY_FLAG_NUMBER, NULL, &statusCode, &len, NULL);
    }
    
    WinHttpCloseHandle(hRequest);
    WinHttpCloseHandle(hConnect);
    WinHttpCloseHandle(hSession);
    return (statusCode >= 200 && statusCode < 300);
}

// ---------- 主函数 ----------
int main() {
    const char* urls[] = {
        "https://www.chinanews.com.cn/rss/scroll-news.xml",
        "https://www.chinanews.com.cn/rss/edu.xml"
    };
    std::vector<NewsItem> allNews;

    for (const char* url : urls) {
        std::string xml = httpGet(url);
        if (xml.empty()) {
            std::cerr << "Failed to fetch: " << url << std::endl;
            continue;
        }
        auto items = parseRSS(xml);
        allNews.insert(allNews.end(), items.begin(), items.end());
    }

    std::string payload = formatNews(allNews);
    if (payload.empty()) {
        std::cerr << "No news to push." << std::endl;
        return 1;
    }

    // 调试输出：保存到本地
    // std::ofstream logFile("debug_payload.txt", std::ios::binary);
    // if (logFile.is_open()) {
    //     logFile << payload;
    //     logFile.close();
    // }

    // 推送给本地接收端
    bool ok = httpPost("http://localhost:36000/news", payload);
    if (ok)
        std::cout << "Pushed " << allNews.size() << " news items." << std::endl;
    else
        std::cerr << "POST failed." << std::endl;

    return ok ? 0 : 1;
}
HZYANG

借走了