Compare commits
| Author | SHA1 | Date | |
|---|---|---|---|
 hz4th_coder
|
add2cea545 | ||
|
hz4th_coder
|
44852f7be7 | ||
|
hz4th_coder
|
e8541a8545 | ||
|
hz4th_coder
|
e70a3a877f |
@@ -23,6 +23,8 @@
|
||||
|
||||
### 1. 前端管理界面
|
||||
- **总体统计区**：任务总数 / 运行中 / 累计运行次数 / 成功失败页面 / 图片数 / 磁盘占用
|
||||
- **🔍 基本搜索**：顶部搜索框，按网址 / 标题 / 任务名即时查找已爬内容，可直接预览文件、查看元数据、跳转任务详情
|
||||
- **🗑️ 回收站**：删除的任务统一进回收站（可恢复）；只有手动在回收站中才能彻底删除或清空，防止误删
|
||||
- **日间/夜间双主题**：右上角按钮一键切换，自动记忆选择
|
||||
- 任务卡片总览：状态、进度、统计、下次调度时间一目了然
|
||||
- 一键操作：开始 / 暂停 / 恢复 / 终止 / 编辑 / 删除
|
||||
@@ -31,7 +33,10 @@
|
||||
- 运行中的任务参数支持**热更新**（修改后从下一页起生效）
|
||||
|
||||
### 2. 批量爬取模式
|
||||
一次粘贴多个网址（每行一个，`#` 注释），可配置：
|
||||
一次粘贴多个网址（每行一个，`#` 注释），或点击「📂 导入网址文件」上传 .txt 文件批量导入：
|
||||
- **处理方式可选**：追加（保留已有，默认）或覆盖（清空已有）
|
||||
- 自动识别 UTF-8/GBK 编码、自动去重、自动清理行内注释
|
||||
可配置：
|
||||
- 项目名称、输出目录（默认 `out/<任务ID>`，可填绝对路径）
|
||||
- 爬取间隔（随机秒数区间，防封 IP）、单页超时
|
||||
- 失败重试次数 / 重试间隔
|
||||
@@ -76,6 +81,11 @@
|
||||
| POST | `/api/tasks/<id>/pause` | 暂停 |
|
||||
| POST | `/api/tasks/<id>/resume` | 恢复 |
|
||||
| POST | `/api/tasks/<id>/stop` | 终止 |
|
||||
| GET | `/api/trash` | 回收站列表 |
|
||||
| GET | `/api/search?q=` | 基本搜索（网址/标题/任务名子串匹配，轻量） |
|
||||
| POST | `/api/trash/<id>/restore` | 从回收站恢复任务 |
|
||||
| DELETE | `/api/trash/<id>` | 彻底删除单个（默认输出目录一并清理） |
|
||||
| DELETE | `/api/trash` | 清空回收站 |
|
||||
| POST | `/api/probe` | 试爬取（表单规则预览链接清单） |
|
||||
| POST | `/api/tasks/<id>/probe` | 对已保存的自动任务试爬取 |
|
||||
| GET | `/api/runs/<rid>` | 运行详情（结果+日志） |
|
||||
|
||||
@@ -143,7 +143,7 @@ def api_status():
|
||||
|
||||
@app.route("/api/tasks", methods=["GET"])
|
||||
def api_tasks():
|
||||
tasks = store.load_tasks()
|
||||
tasks = [t for t in store.load_tasks() if not t.get("deleted_at")]
|
||||
for t in tasks:
|
||||
runs = store.get_runs(t["id"])
|
||||
t["latest_run"] = runs[-1] if runs else None
|
||||
@@ -154,6 +154,19 @@ def api_tasks():
|
||||
return jsonify(tasks)
|
||||
|
||||
|
||||
def _clean_urls(lines):
|
||||
"""清洗网址行: 去空白, 去行内 # 注释 (URL 本身不含空格, 安全), 自动补 https 前缀"""
|
||||
out = []
|
||||
for u in lines or []:
|
||||
u = str(u).split(" #")[0].strip()
|
||||
if not u:
|
||||
continue
|
||||
if not u.startswith("http"):
|
||||
u = "https://" + u
|
||||
out.append(u)
|
||||
return out
|
||||
|
||||
|
||||
@app.route("/api/tasks", methods=["POST"])
|
||||
def api_create_task():
|
||||
body = request.get_json(force=True) or {}
|
||||
@@ -171,7 +184,7 @@ def api_create_task():
|
||||
"created_at": now_str(),
|
||||
"updated_at": now_str(),
|
||||
"config": {**DEFAULT_CONFIG, **(body.get("config") or {})},
|
||||
"urls": [u.strip() for u in (body.get("urls") or []) if u.strip()],
|
||||
"urls": _clean_urls(body.get("urls") or []),
|
||||
}
|
||||
|
||||
if mode == "auto":
|
||||
@@ -221,6 +234,8 @@ def api_update_task(tid):
|
||||
task = store.get_task(tid)
|
||||
if not task:
|
||||
return jsonify({"error": "任务不存在"}), 404
|
||||
if task.get("deleted_at"):
|
||||
return jsonify({"error": "任务在回收站中，请先恢复"}), 400
|
||||
body = request.get_json(force=True) or {}
|
||||
with JOBS_LOCK:
|
||||
job = JOBS.get(tid)
|
||||
@@ -228,7 +243,7 @@ def api_update_task(tid):
|
||||
if "name" in body and str(body["name"]).strip():
|
||||
task["name"] = str(body["name"]).strip()
|
||||
if "urls" in body:
|
||||
task["urls"] = [u.strip() for u in body["urls"] if u.strip()]
|
||||
task["urls"] = _clean_urls(body["urls"])
|
||||
if "config" in body:
|
||||
merged = {**task.get("config", {}), **body["config"]}
|
||||
task["config"] = merged
|
||||
@@ -250,11 +265,71 @@ def api_update_task(tid):
|
||||
|
||||
@app.route("/api/tasks/<tid>", methods=["DELETE"])
|
||||
def api_delete_task(tid):
|
||||
"""删除任务 -> 移入回收站 (可恢复)"""
|
||||
task = store.get_task(tid)
|
||||
if not task:
|
||||
return jsonify({"error": "任务不存在"}), 404
|
||||
if task.get("deleted_at"):
|
||||
return jsonify({"error": "任务已在回收站中"}), 400
|
||||
_stop_job(tid)
|
||||
with JOBS_LOCK:
|
||||
JOBS.pop(tid, None)
|
||||
store.delete_task(tid)
|
||||
return jsonify({"ok": True})
|
||||
store.soft_delete_task(tid, now_str())
|
||||
return jsonify({"ok": True, "msg": "已移入回收站"})
|
||||
|
||||
|
||||
# ---------------- API: 回收站 ----------------
|
||||
|
||||
def _purge_out_dir(out_dir):
|
||||
"""删除任务输出目录; 仅当目录位于项目 out/ 下才删 (自定义目录保留), 返回是否删除"""
|
||||
try:
|
||||
d = os.path.realpath(out_dir)
|
||||
base = os.path.realpath(os.path.join(HERE, "out"))
|
||||
if d.startswith(base + os.sep) and os.path.isdir(d):
|
||||
import shutil
|
||||
shutil.rmtree(d, ignore_errors=True)
|
||||
return True
|
||||
except Exception:
|
||||
pass
|
||||
return False
|
||||
|
||||
|
||||
@app.route("/api/trash", methods=["GET"])
|
||||
def api_trash_list():
|
||||
items = store.list_trash()
|
||||
for t in items:
|
||||
t["runs_count"] = len(store.get_runs(t["id"]))
|
||||
t["out_dir"] = resolve_out_dir(t)
|
||||
return jsonify(items)
|
||||
|
||||
|
||||
@app.route("/api/trash/<tid>/restore", methods=["POST"])
|
||||
def api_trash_restore(tid):
|
||||
task = store.get_task(tid)
|
||||
if not task or not task.get("deleted_at"):
|
||||
return jsonify({"error": "任务不在回收站中"}), 404
|
||||
store.restore_task(tid)
|
||||
return jsonify({"ok": True, "msg": "已恢复"})
|
||||
|
||||
|
||||
@app.route("/api/trash/<tid>", methods=["DELETE"])
|
||||
def api_trash_purge(tid):
|
||||
task = store.get_task(tid)
|
||||
if not task or not task.get("deleted_at"):
|
||||
return jsonify({"error": "任务不在回收站中"}), 404
|
||||
out_dir = resolve_out_dir(task)
|
||||
removed = _purge_out_dir(out_dir)
|
||||
store.purge_task(tid)
|
||||
return jsonify({"ok": True, "purged": True, "files_removed": removed, "out_dir": out_dir})
|
||||
|
||||
|
||||
@app.route("/api/trash", methods=["DELETE"])
|
||||
def api_trash_clear():
|
||||
items = store.list_trash()
|
||||
dirs = [resolve_out_dir(t) for t in items]
|
||||
store.purge_trash()
|
||||
removed = sum(1 for d in dirs if _purge_out_dir(d))
|
||||
return jsonify({"ok": True, "purged": len(items), "dirs_removed": removed})
|
||||
|
||||
|
||||
# ---------------- API: 运行控制 ----------------
|
||||
@@ -301,7 +376,8 @@ def api_resume(tid):
|
||||
|
||||
@app.route("/api/stats")
|
||||
def api_stats():
|
||||
tasks = store.load_tasks()
|
||||
tasks = [t for t in store.load_tasks() if not t.get("deleted_at")]
|
||||
trash_count = sum(1 for t in store.load_tasks() if t.get("deleted_at"))
|
||||
total_runs = ok = fail = imgs = 0
|
||||
for t in tasks:
|
||||
for r in store.get_runs(t["id"]):
|
||||
@@ -334,6 +410,7 @@ def api_stats():
|
||||
"fail": fail,
|
||||
"images": imgs,
|
||||
"disk_mb": round(size / 1048576, 1),
|
||||
"trash": trash_count,
|
||||
})
|
||||
|
||||
|
||||
@@ -379,6 +456,69 @@ def api_task_probe(tid):
|
||||
return jsonify(result)
|
||||
|
||||
|
||||
# ---------------- API: 搜索 ----------------
|
||||
|
||||
@app.route("/api/search")
|
||||
def api_search():
|
||||
"""基本搜索: 对任务名/网址/标题做子串匹配 (仅遍历本地内存数据, 轻量)"""
|
||||
q = (request.args.get("q") or "").strip()
|
||||
if not q:
|
||||
return jsonify({"q": q, "count": 0, "results": []})
|
||||
ql = q.lower()
|
||||
results = []
|
||||
seen = set()
|
||||
for task in store.load_tasks():
|
||||
if task.get("deleted_at"):
|
||||
continue # 回收站任务不参与搜索
|
||||
auto = task.get("auto") or {}
|
||||
task_hit = (
|
||||
ql in (task.get("name") or "").lower()
|
||||
or any(ql in u.lower() for u in task.get("urls", []))
|
||||
or ql in ((auto.get("seed_url") or "").lower())
|
||||
)
|
||||
for run in store.get_runs(task["id"]):
|
||||
for r in run.get("results", []):
|
||||
url = r.get("url") or ""
|
||||
title = r.get("title") or ""
|
||||
if ql not in url.lower() and ql not in title.lower():
|
||||
continue
|
||||
key = (task["id"], run.get("id"), url)
|
||||
if key in seen:
|
||||
continue
|
||||
seen.add(key)
|
||||
results.append({
|
||||
"type": "page",
|
||||
"task_id": task["id"],
|
||||
"task_name": task.get("name", ""),
|
||||
"mode": task.get("mode", ""),
|
||||
"run_id": run.get("id", ""),
|
||||
"run_status": run.get("status", ""),
|
||||
"url": url,
|
||||
"title": title,
|
||||
"status": r.get("status", ""),
|
||||
"crawl_time": r.get("crawl_time", ""),
|
||||
"html_file": r.get("html_file", ""),
|
||||
"txt_file": r.get("txt_file", ""),
|
||||
"meta_file": r.get("meta_file", ""),
|
||||
"images": len(r.get("images", []) or []),
|
||||
})
|
||||
if len(results) >= 100:
|
||||
break
|
||||
if len(results) >= 100:
|
||||
break
|
||||
if task_hit and len(results) < 100:
|
||||
results.append({
|
||||
"type": "task",
|
||||
"task_id": task["id"],
|
||||
"task_name": task.get("name", ""),
|
||||
"mode": task.get("mode", ""),
|
||||
"urls_count": len(task.get("urls", [])),
|
||||
"created_at": task.get("created_at", ""),
|
||||
"seed_url": auto.get("seed_url", ""),
|
||||
})
|
||||
return jsonify({"q": q, "count": len(results), "results": results})
|
||||
|
||||
|
||||
# ---------------- API: 运行记录与文件 ----------------
|
||||
|
||||
@app.route("/api/runs/<rid>")
|
||||
|
||||
+197
-4
@@ -63,6 +63,7 @@ async function loadStats() {
|
||||
$("stImgs").textContent = s.images;
|
||||
$("stDisk").textContent = s.disk_mb >= 1024
|
||||
? (s.disk_mb / 1024).toFixed(1) + " GB" : s.disk_mb + " MB";
|
||||
$("trashCount").textContent = s.trash || 0;
|
||||
} catch (e) { /* 统计失败忽略 */ }
|
||||
}
|
||||
|
||||
@@ -152,10 +153,197 @@ async function actTask(tid, act) {
|
||||
|
||||
async function delTask(tid) {
|
||||
const t = state.tasks.find((x) => x.id === tid);
|
||||
if (!confirm(`确定删除任务「${t ? t.name : tid}」？\n运行记录与任务配置将被删除（已爬取的文件保留在磁盘上）。`)) return;
|
||||
if (!confirm(`确定删除任务「${t ? t.name : tid}」？\n任务将移入回收站，可在回收站中恢复或彻底删除。`)) return;
|
||||
try {
|
||||
await api(`/api/tasks/${tid}`, { method: "DELETE" });
|
||||
toast("已删除");
|
||||
const r = await api(`/api/tasks/${tid}`, { method: "DELETE" });
|
||||
toast(r.msg || "已删除");
|
||||
loadTasks();
|
||||
} catch (e) { toast(e.message, true); }
|
||||
}
|
||||
|
||||
/* ---------------- 网址文件导入 ---------------- */
|
||||
const MAX_IMPORT_SIZE = 2 * 1024 * 1024; // 2MB 上限
|
||||
|
||||
async function readFileSmart(file) {
|
||||
/* 自动识别 UTF-8 / GBK 编码 */
|
||||
const buf = await file.arrayBuffer();
|
||||
let text = new TextDecoder("utf-8").decode(buf);
|
||||
if (text.includes("\uFFFD")) {
|
||||
try { text = new TextDecoder("gbk").decode(buf); } catch (e) { /* 保留 utf-8 结果 */ }
|
||||
}
|
||||
return text;
|
||||
}
|
||||
|
||||
function importUrlsText(text, mode) {
|
||||
const ta = $("taskForm").elements["urls"];
|
||||
const clean = (s) => s.split(" #")[0].trim(); // 去掉行内注释 (URL 不含空格, 安全)
|
||||
const existing = new Set(
|
||||
mode === "overwrite" ? [] : ta.value.split(/\r?\n/).map(clean).filter(Boolean)
|
||||
);
|
||||
const fresh = text.split(/\r?\n/).map(clean).filter(Boolean);
|
||||
let added = 0;
|
||||
for (const line of fresh) {
|
||||
if (!existing.has(line)) { existing.add(line); added++; }
|
||||
}
|
||||
ta.value = Array.from(existing).join("\n");
|
||||
return { total: fresh.length, added };
|
||||
}
|
||||
|
||||
$("btnImportUrls").onclick = () => $("urlFileInput").click();
|
||||
$("urlFileInput").onchange = async (e) => {
|
||||
const file = e.target.files && e.target.files[0];
|
||||
e.target.value = ""; // 允许重复选择同一文件
|
||||
if (!file) return;
|
||||
if (file.size > MAX_IMPORT_SIZE) {
|
||||
toast("文件过大（上限 2MB），请拆分后导入", true);
|
||||
return;
|
||||
}
|
||||
try {
|
||||
const text = await readFileSmart(file);
|
||||
const mode = $("importMode").value;
|
||||
const r = importUrlsText(text, mode);
|
||||
formDirty = true;
|
||||
if (mode === "overwrite") {
|
||||
toast(`已导入 ${file.name}：共 ${r.total} 行（覆盖原列表，新增 ${r.added} 条）`);
|
||||
} else {
|
||||
toast(`已导入 ${file.name}：共 ${r.total} 行，新增 ${r.added} 条网址（追加）`);
|
||||
}
|
||||
} catch (err) {
|
||||
toast("文件读取失败: " + err.message, true);
|
||||
}
|
||||
};
|
||||
|
||||
/* ---------------- 搜索 ---------------- */
|
||||
async function doSearch() {
|
||||
const q = $("searchInput").value.trim();
|
||||
if (!q) { toast("请输入搜索关键词", true); return; }
|
||||
try {
|
||||
const d = await api(`/api/search?q=${encodeURIComponent(q)}`);
|
||||
renderSearch(d);
|
||||
showModal("searchModal");
|
||||
} catch (e) { toast(e.message, true); }
|
||||
}
|
||||
|
||||
function renderSearch(d) {
|
||||
const body = $("searchBody");
|
||||
if (!d.results.length) {
|
||||
body.innerHTML = `<div class="empty" style="padding:40px 0"><div class="empty-icon">🔍</div><p>没有找到与「${esc(d.q)}」相关的内容</p></div>`;
|
||||
return;
|
||||
}
|
||||
const items = d.results.map((r) => {
|
||||
const badge = `<span class="badge ${esc(r.mode)}">${MODE_LABEL[r.mode] || esc(r.mode)}</span>`;
|
||||
if (r.type === "task") {
|
||||
return `
|
||||
<div class="search-item">
|
||||
<div class="row">
|
||||
${badge}
|
||||
<span class="title">📋 ${esc(r.task_name)}</span>
|
||||
<span class="badge">任务</span>
|
||||
</div>
|
||||
<div class="url">${r.seed_url ? "起始: " + esc(r.seed_url) : "网址 " + (r.urls_count || 0) + " 个"} · 创建于 ${esc(r.created_at || "")}</div>
|
||||
<div class="row">
|
||||
<button class="btn sm" onclick="searchOpenDetail('${r.task_id}')">📋 查看任务</button>
|
||||
</div>
|
||||
</div>`;
|
||||
}
|
||||
const stCls = r.status === "OK" ? "t-ok" : "t-fail";
|
||||
const html = r.html_file ? `<span class="file-link" onclick="searchPreview('${r.task_id}','${esc(r.html_file)}','HTML: ${esc(r.url)}')">HTML</span>` : "";
|
||||
const txt = r.txt_file ? `<span class="file-link" onclick="searchPreview('${r.task_id}','${esc(r.txt_file)}','文本: ${esc(r.url)}')">TXT</span>` : "";
|
||||
const meta = r.meta_file ? `<span class="file-link" onclick="searchMeta('${r.task_id}','${esc(r.meta_file)}')">📋 元数据</span>` : "";
|
||||
return `
|
||||
<div class="search-item">
|
||||
<div class="row">
|
||||
${badge}
|
||||
<span class="title" title="${esc(r.title)}">${esc(r.title) || "（无标题）"}</span>
|
||||
<span class="badge ${stCls}">${r.status}</span>
|
||||
<span class="badge">${esc(r.run_status)}</span>
|
||||
<span class="card-line">${esc(r.crawl_time || "")}</span>
|
||||
</div>
|
||||
<div class="url">🌐 ${esc(r.url)}${r.images ? ` · 🖼️ ${r.images}` : ""}</div>
|
||||
<div class="row">
|
||||
${html}${txt}${meta}
|
||||
<button class="btn sm" onclick="searchOpenDetail('${r.task_id}')">📋 任务详情</button>
|
||||
</div>
|
||||
</div>`;
|
||||
}).join("");
|
||||
body.innerHTML = `<div class="probe-summary"><span>关键词: <b>${esc(d.q)}</b></span><span>共 <b>${d.count}</b> 条结果</span></div>${items}`;
|
||||
}
|
||||
|
||||
/* 搜索结果的跳转动作: 先关搜索弹窗再打开目标 */
|
||||
function searchOpenDetail(tid) {
|
||||
hideModal("searchModal");
|
||||
openDetail(tid);
|
||||
}
|
||||
function searchPreview(tid, path, title) {
|
||||
hideModal("searchModal");
|
||||
previewFile(tid, path, title);
|
||||
}
|
||||
function searchMeta(tid, metaFile) {
|
||||
hideModal("searchModal");
|
||||
showMeta(tid, metaFile);
|
||||
}
|
||||
|
||||
/* ---------------- 回收站 ---------------- */
|
||||
async function openTrash() {
|
||||
try {
|
||||
const items = await api("/api/trash");
|
||||
state.trashItems = items;
|
||||
renderTrash(items);
|
||||
showModal("trashModal");
|
||||
} catch (e) { toast(e.message, true); }
|
||||
}
|
||||
|
||||
function renderTrash(items) {
|
||||
const body = $("trashBody");
|
||||
if (!items.length) {
|
||||
body.innerHTML = `<div class="empty" style="padding:40px 0"><div class="empty-icon">🗑️</div><p>回收站是空的</p></div>`;
|
||||
return;
|
||||
}
|
||||
body.innerHTML = items.map((t) => `
|
||||
<div class="trash-item">
|
||||
<div class="card-head" style="min-width:0">
|
||||
<span class="card-name" title="${esc(t.name)}">${esc(t.name)}</span>
|
||||
<span class="badge ${esc(t.mode)}">${MODE_LABEL[t.mode] || esc(t.mode)}</span>
|
||||
</div>
|
||||
<div class="card-line">删除时间: <b>${esc(t.deleted_at)}</b> · 运行次数: <b>${t.runs_count || 0}</b></div>
|
||||
<div class="card-line">输出目录: <b>${esc(t.out_dir || "")}</b></div>
|
||||
<div class="card-actions">
|
||||
<button class="btn sm primary" onclick="restoreTrash('${t.id}')">↩️ 恢复</button>
|
||||
<button class="btn sm danger" onclick="purgeTrashTask('${t.id}')">🗑️ 彻底删除</button>
|
||||
</div>
|
||||
</div>`).join("")
|
||||
+ `<div class="trash-foot">
|
||||
<span class="card-line">共 ${items.length} 项 · 彻底删除将移除任务配置与运行记录（默认输出目录一并清理）</span>
|
||||
<button class="btn danger" onclick="clearTrash()">🧹 清空回收站</button>
|
||||
</div>`;
|
||||
}
|
||||
|
||||
async function restoreTrash(tid) {
|
||||
try {
|
||||
const r = await api(`/api/trash/${tid}/restore`, { method: "POST" });
|
||||
toast(r.msg || "已恢复");
|
||||
openTrash();
|
||||
loadTasks();
|
||||
} catch (e) { toast(e.message, true); }
|
||||
}
|
||||
|
||||
async function purgeTrashTask(tid) {
|
||||
const t = state.trashItems && state.trashItems.find((x) => x.id === tid);
|
||||
if (!confirm(`彻底删除任务「${t ? t.name : tid}」？\n任务配置与所有运行记录将永久删除，不可恢复！`)) return;
|
||||
try {
|
||||
const r = await api(`/api/trash/${tid}`, { method: "DELETE" });
|
||||
toast(r.files_removed ? "已彻底删除（含输出文件）" : "已彻底删除（自定义输出目录已保留）");
|
||||
openTrash();
|
||||
loadTasks();
|
||||
} catch (e) { toast(e.message, true); }
|
||||
}
|
||||
|
||||
async function clearTrash() {
|
||||
if (!confirm(`确定清空回收站？\n回收站中所有任务将永久删除，不可恢复！`)) return;
|
||||
try {
|
||||
const r = await api("/api/trash", { method: "DELETE" });
|
||||
toast(`已清空 ${r.purged} 项`);
|
||||
openTrash();
|
||||
loadTasks();
|
||||
} catch (e) { toast(e.message, true); }
|
||||
}
|
||||
@@ -633,6 +821,9 @@ try {
|
||||
$("btnNew").onclick = openCreate;
|
||||
$("btnNew2").onclick = openCreate;
|
||||
$("btnRefresh").onclick = loadTasks;
|
||||
$("btnTrash").onclick = openTrash;
|
||||
$("btnSearch").onclick = doSearch;
|
||||
$("searchInput").addEventListener("keydown", (e) => { if (e.key === "Enter") doSearch(); });
|
||||
$("btnProbe").onclick = () => {
|
||||
const p = collectProbeFromForm();
|
||||
if (!p.seed_url) { toast("请先填写起始网址", true); return; }
|
||||
@@ -650,6 +841,8 @@ document.querySelectorAll("[data-close-detail]").forEach((b) => b.onclick = () =
|
||||
document.querySelectorAll("[data-close-preview]").forEach((b) => b.onclick = () => { $("previewFrame").src = "about:blank"; hideModal("previewModal"); });
|
||||
document.querySelectorAll("[data-close-probe]").forEach((b) => b.onclick = () => hideModal("probeModal"));
|
||||
document.querySelectorAll("[data-close-meta]").forEach((b) => b.onclick = () => hideModal("metaModal"));
|
||||
document.querySelectorAll("[data-close-trash]").forEach((b) => b.onclick = () => hideModal("trashModal"));
|
||||
document.querySelectorAll("[data-close-search]").forEach((b) => b.onclick = () => hideModal("searchModal"));
|
||||
|
||||
/* 表单改动监听 -> 脏标记 */
|
||||
$("taskForm").addEventListener("input", () => { formDirty = true; });
|
||||
@@ -676,7 +869,7 @@ document.addEventListener("keydown", (e) => {
|
||||
}
|
||||
stopLogPoll();
|
||||
$("previewFrame").src = "about:blank";
|
||||
["detailModal", "probeModal", "metaModal", "previewModal"].forEach((id) => $(id).classList.add("hidden"));
|
||||
["detailModal", "probeModal", "metaModal", "trashModal", "searchModal", "previewModal"].forEach((id) => $(id).classList.add("hidden"));
|
||||
}
|
||||
});
|
||||
|
||||
|
||||
+38
-1
@@ -10,8 +10,13 @@
|
||||
<header>
|
||||
<div class="logo">🕷️ 通用爬虫系统 <span id="version" class="version"></span></div>
|
||||
<div class="header-right">
|
||||
<div class="search-box">
|
||||
<input id="searchInput" placeholder="搜索网址 / 标题 / 任务名..." maxlength="100">
|
||||
<button id="btnSearch" class="btn sm primary" title="搜索">🔍</button>
|
||||
</div>
|
||||
<span class="status-pill" id="statusPill">运行中任务: 0</span>
|
||||
<button id="btnTheme" class="btn ghost" title="切换日间/夜间主题">☀️</button>
|
||||
<button id="btnTrash" class="btn ghost" title="回收站">🗑️ 回收站 <span id="trashCount" class="trash-count">0</span></button>
|
||||
<button id="btnRefresh" class="btn ghost">⟳ 刷新</button>
|
||||
<button id="btnNew" class="btn primary">＋ 新建任务</button>
|
||||
</div>
|
||||
@@ -68,7 +73,17 @@
|
||||
<div class="field"><label>通知邮箱</label><input name="notify_email" value="wlq@tphai.com"></div>
|
||||
</div>
|
||||
|
||||
<div class="field" id="urlsField"><label>网址列表（每行一个，# 开头为注释）</label>
|
||||
<div class="field" id="urlsField">
|
||||
<label>网址列表（每行一个，# 开头为注释）</label>
|
||||
<div class="row2" style="align-items:center">
|
||||
<button type="button" class="btn sm" id="btnImportUrls">📂 导入网址文件(.txt)</button>
|
||||
<select id="importMode" title="对已有网址的处理方式">
|
||||
<option value="append" selected>追加（保留已有）</option>
|
||||
<option value="overwrite">覆盖（清空已有）</option>
|
||||
</select>
|
||||
<span class="card-line">支持 UTF-8 / GBK，自动去重</span>
|
||||
<input type="file" id="urlFileInput" accept=".txt,.csv,.urls,text/plain" hidden>
|
||||
</div>
|
||||
<textarea name="urls" rows="5" placeholder="https://www.example.com/ https://www.example.com/page2"></textarea>
|
||||
</div>
|
||||
|
||||
@@ -143,6 +158,28 @@
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<!-- 搜索结果弹窗 -->
|
||||
<div id="searchModal" class="modal-overlay hidden">
|
||||
<div class="modal wide">
|
||||
<div class="modal-head">
|
||||
<span>🔍 搜索结果</span>
|
||||
<button class="btn ghost sm" data-close-search>✕</button>
|
||||
</div>
|
||||
<div id="searchBody" class="detail-wrap"></div>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<!-- 回收站弹窗 -->
|
||||
<div id="trashModal" class="modal-overlay hidden">
|
||||
<div class="modal wide">
|
||||
<div class="modal-head">
|
||||
<span>🗑️ 回收站（删除的任务统一在这里，可恢复或彻底删除）</span>
|
||||
<button class="btn ghost sm" data-close-trash>✕</button>
|
||||
</div>
|
||||
<div id="trashBody" class="detail-wrap"></div>
|
||||
</div>
|
||||
</div>
|
||||
|
||||
<!-- 试爬取结果弹窗 -->
|
||||
<div id="probeModal" class="modal-overlay hidden">
|
||||
<div class="modal wide">
|
||||
|
||||
@@ -215,6 +215,30 @@ td.title-cell { max-width: 220px; overflow: hidden; text-overflow: ellipsis; }
|
||||
/* ---------- preview ---------- */
|
||||
#previewFrame { flex: 1; border: none; background: var(--preview-bg); border-radius: 0 0 14px 14px; }
|
||||
|
||||
/* ---------------- 回收站 ---------------- */
|
||||
.search-box { display: flex; gap: 6px; align-items: center; }
|
||||
.search-box input { width: 250px; }
|
||||
.search-item {
|
||||
border: 1px solid var(--border); border-radius: 10px; padding: 10px 12px;
|
||||
display: flex; flex-direction: column; gap: 5px; background: var(--panel2);
|
||||
}
|
||||
.search-item .row { display: flex; align-items: center; gap: 8px; flex-wrap: wrap; }
|
||||
.search-item .title { font-weight: 600; max-width: 420px; overflow: hidden; text-overflow: ellipsis; white-space: nowrap; }
|
||||
.search-item .url { color: var(--muted); word-break: break-all; font-size: 12px; }
|
||||
.trash-count {
|
||||
display: inline-block; min-width: 18px; padding: 0 5px; border-radius: 9px;
|
||||
background: var(--red); color: #fff; font-size: 11px; line-height: 18px;
|
||||
text-align: center; margin-left: 2px;
|
||||
}
|
||||
.trash-item {
|
||||
border: 1px solid var(--border); border-radius: 10px; padding: 12px 14px;
|
||||
display: flex; flex-direction: column; gap: 6px; background: var(--panel2);
|
||||
}
|
||||
.trash-foot {
|
||||
display: flex; justify-content: space-between; align-items: center;
|
||||
gap: 10px; padding-top: 8px; border-top: 1px solid var(--border); margin-top: 4px;
|
||||
}
|
||||
|
||||
/* ---------- 试爬取结果 / 元数据 ---------- */
|
||||
.probe-summary { display: flex; flex-wrap: wrap; gap: 16px; font-size: 13px; }
|
||||
.probe-list {
|
||||
|
||||
@@ -67,6 +67,7 @@ def upsert_task(task):
|
||||
|
||||
|
||||
def delete_task(task_id):
|
||||
"""彻底删除任务 (任务 + 运行记录)"""
|
||||
with _lock:
|
||||
tasks = _load(TASKS_FILE, [])
|
||||
tasks = [t for t in tasks if t["id"] != task_id]
|
||||
@@ -76,6 +77,69 @@ def delete_task(task_id):
|
||||
_save(RUNS_FILE, runs)
|
||||
|
||||
|
||||
# ---------------- 回收站 ----------------
|
||||
|
||||
def soft_delete_task(task_id, ts):
|
||||
"""删除任务 -> 移入回收站 (软删除, 记录保留可恢复)"""
|
||||
with _lock:
|
||||
tasks = _load(TASKS_FILE, [])
|
||||
for t in tasks:
|
||||
if t["id"] == task_id:
|
||||
t["deleted_at"] = ts
|
||||
break
|
||||
_save(TASKS_FILE, tasks)
|
||||
|
||||
|
||||
def restore_task(task_id):
|
||||
"""从回收站恢复任务"""
|
||||
with _lock:
|
||||
tasks = _load(TASKS_FILE, [])
|
||||
for t in tasks:
|
||||
if t["id"] == task_id:
|
||||
t.pop("deleted_at", None)
|
||||
break
|
||||
_save(TASKS_FILE, tasks)
|
||||
|
||||
|
||||
def list_trash():
|
||||
"""回收站任务列表 (按删除时间倒序)"""
|
||||
with _lock:
|
||||
tasks = _load(TASKS_FILE, [])
|
||||
return sorted(
|
||||
[t for t in tasks if t.get("deleted_at")],
|
||||
key=lambda x: x.get("deleted_at", ""), reverse=True,
|
||||
)
|
||||
|
||||
|
||||
def purge_task(task_id):
|
||||
"""从回收站彻底删除单个任务 (任务 + 运行记录)"""
|
||||
with _lock:
|
||||
tasks = _load(TASKS_FILE, [])
|
||||
tasks = [t for t in tasks if t["id"] != task_id]
|
||||
_save(TASKS_FILE, tasks)
|
||||
runs = _load(RUNS_FILE, {})
|
||||
runs.pop(task_id, None)
|
||||
_save(RUNS_FILE, runs)
|
||||
|
||||
|
||||
def purge_trash():
|
||||
"""清空回收站, 返回被清空的任务 id 列表"""
|
||||
with _lock:
|
||||
tasks = _load(TASKS_FILE, [])
|
||||
kept, purged = [], []
|
||||
for t in tasks:
|
||||
if t.get("deleted_at"):
|
||||
purged.append(t["id"])
|
||||
else:
|
||||
kept.append(t)
|
||||
_save(TASKS_FILE, kept)
|
||||
runs = _load(RUNS_FILE, {})
|
||||
for tid in purged:
|
||||
runs.pop(tid, None)
|
||||
_save(RUNS_FILE, runs)
|
||||
return purged
|
||||
|
||||
|
||||
# ---------------- 运行记录 ----------------
|
||||
|
||||
def load_runs_map():
|
||||
|
||||
Reference in New Issue
Block a user