diff --git a/.DS_Store b/.DS_Store new file mode 100644 index 0000000..83fc51f Binary files /dev/null and b/.DS_Store differ diff --git a/dependency-reduced-pom.xml b/dependency-reduced-pom.xml new file mode 100644 index 0000000..3c9566f --- /dev/null +++ b/dependency-reduced-pom.xml @@ -0,0 +1,87 @@ + + + 4.0.0 + com.ets + ets-playwright + 1.0.0 + + + + maven-compiler-plugin + 3.14.0 + + 25 + + + + org.codehaus.mojo + exec-maven-plugin + 3.5.0 + + com.ets.scraper.EtsScraper + + + + maven-surefire-plugin + 3.5.2 + + + maven-shade-plugin + 3.6.0 + + + package + + shade + + + + + com.ets.scraper.EtsScraper + + + + + *:* + + META-INF/*.SF + META-INF/*.DSA + META-INF/*.RSA + + + + + + + + + + + + org.junit.jupiter + junit-jupiter + 5.12.1 + test + + + junit-jupiter-api + org.junit.jupiter + + + junit-jupiter-params + org.junit.jupiter + + + junit-jupiter-engine + org.junit.jupiter + + + + + + UTF-8 + 25 + 25 + 1.55.0 + + diff --git a/src/main/java/com/ets/scraper/EtsScraper.java b/src/main/java/com/ets/scraper/EtsScraper.java index 303005d..a8f4bca 100644 --- a/src/main/java/com/ets/scraper/EtsScraper.java +++ b/src/main/java/com/ets/scraper/EtsScraper.java @@ -34,7 +34,7 @@ public class EtsScraper { private static final String PROXY_HOST = "http://127.0.0.1:8081"; public static void main(String[] args) throws Exception { - // Parse CLI arguments + // Parse CLI arguments String proxyHost = null; String proxyUser = null; String proxyPass = null; @@ -61,22 +61,22 @@ public class EtsScraper { System.err.println("[-] Unknown option: " + args[i]); printHelp(); return; - } - } + } + } if (proxyHost == null || proxyUser == null || proxyPass == null || dateStr == null) { System.err.println("[-] Missing required arguments"); printHelp(); return; - } + } java.time.LocalDate targetDate; try { targetDate = java.time.LocalDate.parse(dateStr); - } catch (Exception e) { + } catch (Exception e) { System.err.println("[-] Invalid date format: " + dateStr + ", expected yyyy-MM-dd"); return; - } + } String dateStrFormatted = targetDate.format(java.time.format.DateTimeFormatter.ofPattern("yyyy-MM-dd")); String dateStrFileName = targetDate.format(DateTimeFormatter.ofPattern("yyyyMMdd")); @@ -89,53 +89,53 @@ public class EtsScraper { System.out.println("[+] File size: " + java.nio.file.Files.size(savedFile) + " bytes"); autoImportBill(savedFile, proxyHost, proxyUser, proxyPass); return; - } + } try { createDirectories(SCREENSHOT_DIR); - } catch (Exception e) { + } catch (Exception e) { System.err.println("Failed to create directories: " + e.getMessage()); - } + } try (Playwright playwright = Playwright.create()) { Browser browser = playwright.chromium().launch( - new BrowserType.LaunchOptions().setHeadless(true) - ); + new BrowserType.LaunchOptions().setHeadless(true) + ); BrowserContext context = browser.newContext( - new Browser.NewContextOptions() - .setIgnoreHTTPSErrors(true) - .setViewportSize(1920, 1080) - .setUserAgent("Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36") - ); + new Browser.NewContextOptions() + .setIgnoreHTTPSErrors(true) + .setViewportSize(1920, 1080) + .setUserAgent("Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/135.0.0.0 Safari/537.36") + ); Page page = context.newPage(); try { - // Navigate to frame.html first to establish session/cookies + // Navigate to frame.html first to establish session/cookies System.out.println("[*] Establishing session via " + FRAME_URL); page.navigate(FRAME_URL, new Page.NavigateOptions() - .setTimeout(30000) - .setWaitUntil(WaitUntilState.DOMCONTENTLOADED)); + .setTimeout(30000) + .setWaitUntil(WaitUntilState.DOMCONTENTLOADED)); sleep(3000); - // Navigate directly to the login page + // Navigate directly to the login page System.out.println("[*] Navigating to login page: " + LOGIN_URL); page.navigate(LOGIN_URL, new Page.NavigateOptions() - .setTimeout(30000) - .setWaitUntil(WaitUntilState.NETWORKIDLE)); + .setTimeout(30000) + .setWaitUntil(WaitUntilState.NETWORKIDLE)); sleep(2000); - // Close notification dialog FIRST (before filling credentials) + // Close notification dialog FIRST (before filling credentials) closeNotificationDialog(page); screenshot(page, "after_close_dialog"); - // Download captcha image + // Download captcha image downloadCaptcha(page); - // Close dialog again after page reload + // Close dialog again after page reload closeNotificationDialog(page); - // Recognize captcha and perform login + // Recognize captcha and perform login boolean loggedin = doLoginWithCaptcha(page); if (loggedin) { @@ -147,7 +147,7 @@ public class EtsScraper { System.out.println("[+] Page title: " + page.title()); System.out.println("[+] Page URL: " + page.url()); - // 点击三联单菜单 + // 点击三联单菜单 System.out.println("[*] Clicking 三联单 menu..."); page.locator("#module_2094F683-C542-4904-B33E-0D227C4DE199").first().click(); sleep(3000); @@ -155,113 +155,104 @@ public class EtsScraper { screenshot(page, "after_sanliandan"); System.out.println("[+] 三联单 page title: " + page.title()); - // 设置日期筛选 + // 设置日期筛选 System.out.println("[*] Setting date filter to: " + dateStrFormatted); - // 检查元素是否存在 + // 检查元素是否存在 boolean startDateExists = page.locator("#Search_ThreeBillList_startWdate").count() > 0; boolean endDateExists = page.locator("#Search_ThreeBillList_endWdate").count() > 0; boolean queryBtnExists = page.locator("#Search_ThreeBillList_Button").count() > 0; System.out.println("[*] Elements found - startDate: " + startDateExists + ", endDate: " + endDateExists + ", queryBtn: " + queryBtnExists); - // 直接设置日期值(WdatePicker 类型输入框) + // 直接设置日期值(WdatePicker 类型输入框) if (startDateExists) { System.out.println("[*] Setting start date to: " + dateStrFormatted); page.locator("#Search_ThreeBillList_startWdate").first().fill(dateStrFormatted); sleep(500); - } else { + } else { System.out.println("[!] Start date element not found"); - } + } - // 设置结束日期 + // 设置结束日期 if (endDateExists) { System.out.println("[*] Setting end date to: " + dateStrFormatted); page.locator("#Search_ThreeBillList_endWdate").first().fill(dateStrFormatted); sleep(500); - } else { + } else { System.out.println("[!] End date element not found"); - } + } - // 点击查询按钮,等待列表加载 + // 点击查询按钮,等待列表加载 if (queryBtnExists) { System.out.println("[*] Clicking query button..."); page.locator("#Search_ThreeBillList_Button").first().click(); - // 等待列表内容出现 + // 等待列表内容出现 try { page.waitForSelector("tbody tr", new Page.WaitForSelectorOptions() - .setTimeout(30000)); + .setTimeout(30000)); System.out.println("[+] Query completed, list loaded"); - } catch (Exception e) { + } catch (Exception e) { System.out.println("[!] Wait for list timeout, but query was submitted"); - } - } else { + } + } else { System.out.println("[!] Query button not found"); - } + } screenshot(page, "after_query"); - // 点击导出按钮 + // 点击导出按钮 if (page.locator("#Export_ThreeBillList_Button").count() > 0) { System.out.println("[*] Clicking export button..."); - // 设置下载目录 - // 点击主导出按钮打开对话框,再用 JS click 触发对话框内导出按钮 - Download dl = page.waitForDownload( - new Page.WaitForDownloadOptions().setTimeout(300000), - () -> { - page.locator("#Export_ThreeBillList_Button").first().click(); - sleep(2000); - System.out.println("[*] Triggering dialog export via JS..."); - page.evaluate("document.querySelectorAll('button').forEach(b => { if (b.textContent.trim() === '导出') b.click(); })"); - }); + // 设置下载目录 + // 点击主导出按钮打开对话框,再用 JS click 触发对话框内导出按钮 + Download dl = page.waitForDownload(new Page.WaitForDownloadOptions().setTimeout(300000), + () -> { + page.locator("#Export_ThreeBillList_Button").first().click(); + sleep(2000); + System.out.println("[*] Triggering dialog export via JS..."); + page.evaluate("document.querySelectorAll('button').forEach(b => { if (b.textContent.trim() === '导出') b.click(); })"); + }); System.out.println("[*] Waiting for download to complete..."); dl.saveAs(savedFile); System.out.println("[+] Download saved to: " + savedFile); if (java.nio.file.Files.size(savedFile) == 0) { System.out.println("[-] Downloaded file is empty"); - } else { + } else { System.out.println("[+] Download size: " + java.nio.file.Files.size(savedFile) + " bytes"); - // Auto-import to ets-proxy - autoImportBill(savedFile, proxyHost, proxyUser, proxyPass); + // Auto-import to ets-proxy + autoImportBill(savedFile, proxyHost, proxyUser, proxyPass); } } screenshot(page, "after_export"); System.out.println("[+] Query and export completed!"); - - String content = page.textContent("body"); - if (content != null) { - String preview = content.length() > 500 - ? content.substring(0, 500) + ".." - : content; - System.out.println("[+] Page content preview:\n" + preview); - } - } else { + } else { System.out.println("[-] Login failed. Check screenshots/ for debugging."); screenshot(page, "login_failed"); - } - } finally { + } + } finally { browser.close(); - } - } - } + } + } + } private static void printHelp() { System.out.println(""" - ETS 三联单爬虫 - 导出并导入三联单 Excel 数据 - - 用法: java -jar ets-playwright.jar [选项] - - 选项: - -s ets-proxy 服务器地址 - -u ets-proxy 用户名 - -p ets-proxy 密码 - -d 查询日期,格式 yyyy-MM-dd - -h 显示此帮助信息 - - 示例: - java -jar ets-playwright.jar -s https://api.ets.niko.red -u admin -p 123456 -d 2026-05-04 - """); - } + ETS 三联单爬虫 - 导出并导入三联单 Excel 数据 + + 用法: java -jar ets-playwright.jar [选项] + + 选项: + -s ets-proxy 服务器地址 + -u ets-proxy 用户名 + -p ets-proxy 密码 + -d 查询日期,格式 yyyy-MM-dd + -h 显示此帮助信息 + + 示例: + java -jar ets-playwright.jar -s https://api.ets.niko.red -u admin -p 123456 -d 2026-05-04 + """); + } public static boolean doLoginWithCaptcha(Page page) throws Exception { // Recognize captcha first @@ -486,7 +477,7 @@ public class EtsScraper { .build(); java.net.http.HttpRequest request = java.net.http.HttpRequest.newBuilder() .uri(uri) - .POST(java.net.http.HttpRequest.BodyPublishers.noBody()) + .POST(java.net.http.HttpRequest.BodyPublishers.noBody()) .header("Content-Type", "application/json") .build(); java.net.http.HttpResponse response = client.send(request, java.net.http.HttpResponse.BodyHandlers.ofString());