dsh-learning-assistant 0.0.1 → 0.1.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (69) hide show
  1. package/README.md +41 -7
  2. package/assets/collect/README.md +55 -0
  3. package/assets/skills/course-inbox-sync.md +182 -0
  4. package/assets/skills/course-material-digest.md +173 -0
  5. package/assets/skills/one-to-one-tutor.md +320 -0
  6. package/assets/tools/pdf/pdf-render.ps1 +30 -0
  7. package/assets/tools/pptx-digest.py +99 -0
  8. package/assets/tools/pptx-map.py +98 -0
  9. package/assets/tools/pptx-media.py +101 -0
  10. package/assets/tools/qq-group-msg.py +258 -0
  11. package/assets/tools/qq-group-sync.py +265 -0
  12. package/docs//345/215/217/344/275/234/345/274/200/345/217/221.md +18 -10
  13. package/lib/client.js +110 -0
  14. package/lib/{index.d.mts → index.d.ts} +3 -1
  15. package/lib/index.js +392 -0
  16. package/package.json +8 -7
  17. package/src/client/settings.tsx +6 -4
  18. package/src/index.ts +36 -11
  19. package/src/practice/learn/api.mjs +393 -0
  20. package/src/practice/learn/cli.mjs +175 -0
  21. package/src/practice/learn/curriculum.mjs +292 -0
  22. package/src/practice/learn/handoff.mjs +73 -0
  23. package/src/practice/learn/jsonschema.mjs +171 -0
  24. package/src/practice/learn/mapping.mjs +90 -0
  25. package/src/practice/learn/progress.mjs +247 -0
  26. package/src/practice/learn/schema/curriculum.schema.json +88 -0
  27. package/src/practice/learn/schema/handoff.schema.json +56 -0
  28. package/src/practice/learn/schema/node-practice-map.schema.json +28 -0
  29. package/src/practice/learn/schema/progress.schema.json +66 -0
  30. package/src/practice/learn/schemas.mjs +53 -0
  31. package/src/practice/learn/store.mjs +247 -0
  32. package/src/practice/public/app.js +883 -0
  33. package/src/practice/public/index.html +19 -0
  34. package/src/practice/public/ink.js +349 -0
  35. package/src/practice/public/learn.css +20 -0
  36. package/src/practice/public/learn.js +329 -0
  37. package/src/practice/public/style.css +326 -0
  38. package/src/practice/public/vendor/katex/contrib/auto-render.min.js +1 -0
  39. package/src/practice/public/vendor/katex/fonts/KaTeX_AMS-Regular.woff2 +0 -0
  40. package/src/practice/public/vendor/katex/fonts/KaTeX_Caligraphic-Bold.woff2 +0 -0
  41. package/src/practice/public/vendor/katex/fonts/KaTeX_Caligraphic-Regular.woff2 +0 -0
  42. package/src/practice/public/vendor/katex/fonts/KaTeX_Fraktur-Bold.woff2 +0 -0
  43. package/src/practice/public/vendor/katex/fonts/KaTeX_Fraktur-Regular.woff2 +0 -0
  44. package/src/practice/public/vendor/katex/fonts/KaTeX_Main-Bold.woff2 +0 -0
  45. package/src/practice/public/vendor/katex/fonts/KaTeX_Main-BoldItalic.woff2 +0 -0
  46. package/src/practice/public/vendor/katex/fonts/KaTeX_Main-Italic.woff2 +0 -0
  47. package/src/practice/public/vendor/katex/fonts/KaTeX_Main-Regular.woff2 +0 -0
  48. package/src/practice/public/vendor/katex/fonts/KaTeX_Math-BoldItalic.woff2 +0 -0
  49. package/src/practice/public/vendor/katex/fonts/KaTeX_Math-Italic.woff2 +0 -0
  50. package/src/practice/public/vendor/katex/fonts/KaTeX_SansSerif-Bold.woff2 +0 -0
  51. package/src/practice/public/vendor/katex/fonts/KaTeX_SansSerif-Italic.woff2 +0 -0
  52. package/src/practice/public/vendor/katex/fonts/KaTeX_SansSerif-Regular.woff2 +0 -0
  53. package/src/practice/public/vendor/katex/fonts/KaTeX_Script-Regular.woff2 +0 -0
  54. package/src/practice/public/vendor/katex/fonts/KaTeX_Size1-Regular.woff2 +0 -0
  55. package/src/practice/public/vendor/katex/fonts/KaTeX_Size2-Regular.woff2 +0 -0
  56. package/src/practice/public/vendor/katex/fonts/KaTeX_Size3-Regular.woff2 +0 -0
  57. package/src/practice/public/vendor/katex/fonts/KaTeX_Size4-Regular.woff2 +0 -0
  58. package/src/practice/public/vendor/katex/fonts/KaTeX_Typewriter-Regular.woff2 +0 -0
  59. package/src/practice/public/vendor/katex/katex.min.css +1 -0
  60. package/src/practice/public/vendor/katex/katex.min.js +1 -0
  61. package/src/practice//346/234/215/345/212/241.mjs +443 -0
  62. package/src/practice//346/240/241/351/252/214/351/242/230/347/233/256.mjs +114 -0
  63. package/src/practice//347/255/211/345/276/205/346/217/220/344/272/244.mjs +100 -0
  64. package/src/practice-service.ts +99 -0
  65. package/src/practice-tool.ts +91 -0
  66. package/src/skills.ts +156 -0
  67. package/lib/client.d.mts +0 -10
  68. package/lib/client.mjs +0 -87
  69. package/lib/index.mjs +0 -29
@@ -0,0 +1,258 @@
1
+ #!/usr/bin/env python
2
+ # -*- coding: utf-8 -*-
3
+ r"""QQ 群消息/群公告采集 + DDL 抽取(NapCat OneBot 11 HTTP)—— 插件参数化版
4
+
5
+ 用途:把指定 QQ 课程群的近期消息、群公告、群作业卡片抓下来,抽取其中的截止时间,
6
+ 写成 <数据根>/qq群消息.json 与 <数据根>/ddl_qq.json。
7
+
8
+ 配置来源(优先级:命令行 > 环境变量 > --config JSON > 默认):
9
+ - 端点:--endpoint / NAPCAT_ENDPOINT(默认 http://127.0.0.1:3000)
10
+ - 令牌:--token / NAPCAT_TOKEN
11
+ - 群号表:--config JSON 的 "groups",或 NAPCAT_GROUPS("gid=科目=群名,...")
12
+ - 输出根:--out-dir / QQ_MSG_OUT(默认当前目录;插件部署时指向数据根)
13
+
14
+ --config JSON 形如:
15
+ {"endpoint": "http://127.0.0.1:3000", "token": "<OneBot令牌>",
16
+ "groups": {"<群号>": ["<科目名>", "<群备注名>"]},
17
+ "out_dir": "<数据根>"}
18
+
19
+ 用法:
20
+ python qq-group-msg.py # 抓取 + 抽 DDL
21
+ python qq-group-msg.py --count 500 # 每群抓多少条(默认 300)
22
+ """
23
+ import argparse, base64, datetime, io, json, os, re, sys, time, urllib.request
24
+
25
+ ENDPOINT = os.environ.get("NAPCAT_ENDPOINT", "http://127.0.0.1:3000")
26
+ TOKEN = os.environ.get("NAPCAT_TOKEN", "")
27
+ GROUPS = {}
28
+ OUT_DIR = os.environ.get("QQ_MSG_OUT", os.getcwd())
29
+ DDL_PAT = re.compile(r"截止|死线|ddl|提交|上交|交回|答辩|考试|考核|测试|测验|作业|有效期|前交|晚于|逾期|延期|补交|点名|挂科", re.I)
30
+ FULL_RE = re.compile(r"(?:(\d{4})\s*年)?\s*(\d{1,2})\s*月\s*(\d{1,2})\s*[日号]")
31
+ SLASH_RE = re.compile(r"(?<!\d)(\d{1,2})\s*[/]\s*(\d{1,2})(?!\d)")
32
+ TIME_RE = re.compile(r"(\d{1,2})\s*[::]\s*(\d{2})")
33
+ WEEKDAY = {"一": 0, "二": 1, "三": 2, "四": 3, "五": 4, "六": 5, "日": 6, "天": 6}
34
+
35
+
36
+ def apply_config(path):
37
+ """读 --config JSON,把部署事实并入全局配置。命令行参数仍可覆盖。"""
38
+ global ENDPOINT, TOKEN, GROUPS, OUT_DIR
39
+ with io.open(path, "r", encoding="utf-8") as fp:
40
+ d = json.load(fp)
41
+ if d.get("endpoint"):
42
+ ENDPOINT = str(d["endpoint"])
43
+ if d.get("token"):
44
+ TOKEN = str(d["token"])
45
+ if d.get("out_dir"):
46
+ OUT_DIR = str(d["out_dir"])
47
+ g = {}
48
+ for gid, pair in (d.get("groups") or {}).items():
49
+ if isinstance(pair, (list, tuple)) and len(pair) >= 2:
50
+ g[str(gid)] = (str(pair[0]), str(pair[1]))
51
+ else:
52
+ g[str(gid)] = (str(pair), str(pair))
53
+ GROUPS = g
54
+
55
+
56
+ def env_groups():
57
+ """NAPCAT_GROUPS 形如 "gid1=科目1=群名1,gid2=科目2=群名2"。"""
58
+ raw = os.environ.get("NAPCAT_GROUPS", "")
59
+ out = {}
60
+ for part in raw.split(","):
61
+ seg = [x for x in part.split("=") if x]
62
+ if len(seg) >= 3:
63
+ out[seg[0]] = (seg[1], seg[2])
64
+ elif len(seg) == 2:
65
+ out[seg[0]] = (seg[1], seg[1])
66
+ return out
67
+
68
+
69
+ def ob(action, params=None):
70
+ req = urllib.request.Request(ENDPOINT + "/" + action, data=json.dumps(params or {}, ensure_ascii=False).encode("utf-8"),
71
+ method="POST", headers={"Content-Type": "application/json",
72
+ "Authorization": "Bearer " + TOKEN})
73
+ with urllib.request.urlopen(req, timeout=90) as r:
74
+ return json.loads(r.read().decode("utf-8", "replace"))
75
+
76
+
77
+ def b64(s):
78
+ try:
79
+ return base64.b64decode(s).decode("utf-8", "replace")
80
+ except Exception:
81
+ return ""
82
+
83
+
84
+ def seg(s):
85
+ """返回 (文本, 种类)"""
86
+ ty = s.get("type")
87
+ d = s.get("data") or {}
88
+ if ty == "text":
89
+ return (d.get("text") or "").strip(), "text"
90
+ if ty == "file":
91
+ nm = d.get("file") or d.get("name") or ""
92
+ return "[文件] " + str(nm), "file"
93
+ if ty == "image":
94
+ return "[图片]", "image"
95
+ if ty == "json":
96
+ try:
97
+ pay = json.loads(d.get("data") or "{}")
98
+ except Exception:
99
+ return "", None
100
+ meta = pay.get("meta") or {}
101
+ app = str(pay.get("app") or "")
102
+ if "mannounce" in app or "mannounce" in meta:
103
+ man = meta.get("mannounce") or {}
104
+ return ("【群公告】%s %s" % (b64(man.get("title") or ""), b64(man.get("text") or ""))).strip(), "announce"
105
+ news = meta.get("news") or {}
106
+ if "homework" in str(pay.get("bizsrc") or "") or "作业" in str(news.get("tag") or ""):
107
+ return "【群作业】" + str(news.get("title") or pay.get("prompt") or ""), "homework"
108
+ return str(pay.get("prompt") or ""), "card"
109
+ return "", None
110
+
111
+
112
+ def parse_when(text, base_dt):
113
+ """返回 (due ISO 或 None, time 字符串或 None, 依据说明)"""
114
+ due = None
115
+ note = ""
116
+ m = FULL_RE.search(text)
117
+ if m:
118
+ y, mo, dd = m.group(1), int(m.group(2)), int(m.group(3))
119
+ y = int(y) if y else base_dt.year
120
+ try:
121
+ d = datetime.date(y, mo, dd)
122
+ if not m.group(1) and (d - base_dt).days > 200:
123
+ d = datetime.date(y + 1, mo, dd)
124
+ due, note = d, "月日"
125
+ except Exception:
126
+ pass
127
+ else:
128
+ for kw, off in (("今天", 0), ("明天", 1), ("后天", 2)):
129
+ if kw in text:
130
+ due, note = base_dt + datetime.timedelta(days=off), kw
131
+ break
132
+ if due is None:
133
+ m2 = re.search(r"(本周|这周|下周|周)([一二三四五六日天])", text)
134
+ if m2:
135
+ wd = WEEKDAY[m2.group(2)]
136
+ monday = base_dt - datetime.timedelta(days=base_dt.weekday())
137
+ if m2.group(1).startswith("下"):
138
+ monday += datetime.timedelta(days=7)
139
+ due, note = monday + datetime.timedelta(days=wd), m2.group(1) + m2.group(2)
140
+ if due is None:
141
+ m3 = SLASH_RE.search(text)
142
+ if m3:
143
+ try:
144
+ due = datetime.date(base_dt.year, int(m3.group(1)), int(m3.group(2)))
145
+ note = "斜杠月日"
146
+ except Exception:
147
+ pass
148
+ tm = None
149
+ mt = TIME_RE.search(text)
150
+ if mt:
151
+ tm = "%02d:%s" % (int(mt.group(1)), mt.group(2))
152
+ elif re.search(r"上午|早上", text):
153
+ tm = "上午"
154
+ elif re.search(r"下午", text):
155
+ tm = "下午"
156
+ elif re.search(r"晚上", text):
157
+ tm = "晚上"
158
+ return (due.isoformat() if due else None), tm, note
159
+
160
+
161
+ def main():
162
+ global ENDPOINT, TOKEN, GROUPS, OUT_DIR
163
+ ap = argparse.ArgumentParser()
164
+ ap.add_argument("--config", default="", help="配置 JSON(endpoint/token/groups/out_dir)")
165
+ ap.add_argument("--endpoint", default="", help="OneBot HTTP 端点")
166
+ ap.add_argument("--token", default="", help="OneBot Bearer 令牌")
167
+ ap.add_argument("--out-dir", default="", help="输出目录(插件部署指向数据根)")
168
+ ap.add_argument("--count", type=int, default=300)
169
+ a = ap.parse_args()
170
+ if a.config:
171
+ apply_config(a.config)
172
+ if a.endpoint:
173
+ ENDPOINT = a.endpoint
174
+ if a.token:
175
+ TOKEN = a.token
176
+ if os.environ.get("NAPCAT_ENDPOINT"):
177
+ ENDPOINT = os.environ["NAPCAT_ENDPOINT"]
178
+ if os.environ.get("NAPCAT_TOKEN"):
179
+ TOKEN = os.environ["NAPCAT_TOKEN"]
180
+ if not GROUPS:
181
+ GROUPS = env_groups()
182
+ if a.out_dir:
183
+ OUT_DIR = a.out_dir
184
+ if not GROUPS:
185
+ print("ERR: 未配置群号白名单(--config JSON 的 groups,或 NAPCAT_GROUPS)")
186
+ return 2
187
+ os.makedirs(OUT_DIR, exist_ok=True)
188
+ OUT_MSG = os.path.join(OUT_DIR, "qq群消息.json")
189
+ OUT_DDL = os.path.join(OUT_DIR, "ddl_qq.json")
190
+ allmsgs, ddl = [], []
191
+ for gid, (subject, gname) in GROUPS.items():
192
+ try:
193
+ r = ob("get_group_msg_history", {"group_id": gid, "message_seq": 0, "count": a.count})
194
+ except Exception as e:
195
+ print("ERR %s: %s" % (subject, e))
196
+ continue
197
+ msgs = ((r or {}).get("data") or {}).get("messages") or []
198
+ for m in msgs:
199
+ t = int(m.get("time") or 0)
200
+ base_dt = datetime.date.fromtimestamp(t) if t else datetime.date.today()
201
+ parts, kinds = [], []
202
+ for sgm in (m.get("message") or []):
203
+ if not isinstance(sgm, dict):
204
+ continue
205
+ txt, k = seg(sgm)
206
+ if txt:
207
+ parts.append(txt)
208
+ if k:
209
+ kinds.append(k)
210
+ text = " ".join(parts).strip()
211
+ if not text:
212
+ continue
213
+ who = (m.get("sender") or {}).get("card") or (m.get("sender") or {}).get("nickname") or ""
214
+ kind = ("announce" if "announce" in kinds else
215
+ "homework" if "homework" in kinds else
216
+ "file" if kinds == ["file"] else "text")
217
+ rec = dict(group=gid, subject=subject, ts=t,
218
+ time=time.strftime("%Y-%m-%d %H:%M", time.localtime(t)) if t else "",
219
+ who=who, kind=kind, text=text[:800])
220
+ allmsgs.append(rec)
221
+ hit = DDL_PAT.search(text)
222
+ if not hit:
223
+ continue
224
+ if kind == "file" and not DDL_PAT.search(text.replace("[文件]", "")):
225
+ pass
226
+ due, tm, note = parse_when(text, base_dt)
227
+ if kind == "homework":
228
+ # 群作业卡片标题里的日期是【布置日】,不是截止日;截止时间要点开群作业查看
229
+ due, tm, note = None, None, "布置日 " + (rec["time"] or "")
230
+ if due is None and tm is None and kind not in ("announce", "homework"):
231
+ continue
232
+ ddl.append(dict(course=subject, group=gname, source=("QQ群公告" if kind == "announce" else
233
+ ("QQ群作业" if kind == "homework" else "QQ群消息")),
234
+ msg_time=rec["time"], who=who, due=due, time=tm,
235
+ sentence=text[:300], date_note=note))
236
+ # 去重
237
+ seen, uniq = set(), []
238
+ for d in ddl:
239
+ k = (d["course"], d["due"], d["sentence"][:60])
240
+ if k in seen:
241
+ continue
242
+ seen.add(k)
243
+ uniq.append(d)
244
+ uniq.sort(key=lambda x: (x["due"] is None, x["due"] or ""))
245
+ with io.open(OUT_MSG, "w", encoding="utf-8") as f:
246
+ json.dump(dict(generated=time.strftime("%Y-%m-%d %H:%M:%S"), count=len(allmsgs), messages=allmsgs), f,
247
+ ensure_ascii=False, indent=1)
248
+ with io.open(OUT_DDL, "w", encoding="utf-8") as f:
249
+ json.dump(dict(generated=time.strftime("%Y-%m-%d %H:%M:%S"), count=len(uniq), items=uniq), f,
250
+ ensure_ascii=False, indent=1)
251
+ print("消息 %d 条;DDL 候选 %d 条" % (len(allmsgs), len(uniq)))
252
+ for d in uniq:
253
+ print(" [%s] %s %s | %s | %s" % (d["course"], d["due"] or "?", d["time"] or "", d["source"], d["sentence"][:110].replace("\n", " ")))
254
+ return 0
255
+
256
+
257
+ if __name__ == "__main__":
258
+ sys.exit(main())
@@ -0,0 +1,265 @@
1
+ #!/usr/bin/env python
2
+ # -*- coding: utf-8 -*-
3
+ """QQ 群文件同步器(NapCat OneBot 11 HTTP)—— 插件参数化版
4
+
5
+ 与工作区 tools/ 版的差别:所有部署事实(端点、令牌、群号表、下载根、台账位置)
6
+ 不再写死,从命令行 / 环境变量 / 配置文件读取,因此可以随插件开源分发。
7
+
8
+ 配置来源(优先级:命令行 > 环境变量 > --config JSON > 默认):
9
+ - 端点:--endpoint / NAPCAT_ENDPOINT(默认 http://127.0.0.1:3000)
10
+ - 令牌:--token / NAPCAT_TOKEN
11
+ - 群号表:--config JSON 的 "groups",或 NAPCAT_GROUPS(格式见 env_groups())
12
+ - 下载根:--dest / QQ_SYNC_DEST
13
+ - 台账:--state / QQ_SYNC_STATE
14
+
15
+ --config JSON 形如:
16
+ {"endpoint": "http://127.0.0.1:3000", "token": "<OneBot令牌>",
17
+ "groups": {"<群号>": ["<科目名>", "<群备注名>"]},
18
+ "dest_root": "<下载根目录>", "state_file": "<数据根>/qq群文件台账.json"}
19
+
20
+ 用法:
21
+ python qq-group-sync.py --list # 只列不下载
22
+ python qq-group-sync.py # 列出并下载新文件
23
+ python qq-group-sync.py --only <gid> # 只处理指定群
24
+ python qq-group-sync.py --max-mb 80 # 单个文件上限(默认 60MB)
25
+ """
26
+ import argparse, hashlib, io, json, os, sys, time, urllib.request
27
+
28
+ ENDPOINT = os.environ.get("NAPCAT_ENDPOINT", "http://127.0.0.1:3000")
29
+ TOKEN = os.environ.get("NAPCAT_TOKEN", "")
30
+ GROUPS = {}
31
+ DEST_ROOT = os.environ.get("QQ_SYNC_DEST", "")
32
+ STATE = os.environ.get("QQ_SYNC_STATE", "")
33
+ UA = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) NapCat/4.18.41"
34
+
35
+
36
+ def apply_config(path):
37
+ """读 --config JSON,把部署事实并入全局配置。命令行参数仍可覆盖。"""
38
+ global ENDPOINT, TOKEN, GROUPS, DEST_ROOT, STATE
39
+ with io.open(path, "r", encoding="utf-8") as fp:
40
+ d = json.load(fp)
41
+ if d.get("endpoint"):
42
+ ENDPOINT = str(d["endpoint"])
43
+ if d.get("token"):
44
+ TOKEN = str(d["token"])
45
+ if d.get("dest_root"):
46
+ DEST_ROOT = str(d["dest_root"])
47
+ if d.get("state_file"):
48
+ STATE = str(d["state_file"])
49
+ g = {}
50
+ for gid, pair in (d.get("groups") or {}).items():
51
+ if isinstance(pair, (list, tuple)) and len(pair) >= 2:
52
+ g[str(gid)] = (str(pair[0]), str(pair[1]))
53
+ else:
54
+ g[str(gid)] = (str(pair), str(pair))
55
+ GROUPS = g
56
+
57
+
58
+ def env_groups():
59
+ """NAPCAT_GROUPS 形如 "gid1=科目1=群名1,gid2=科目2=群名2"。"""
60
+ raw = os.environ.get("NAPCAT_GROUPS", "")
61
+ out = {}
62
+ for part in raw.split(","):
63
+ seg = [x for x in part.split("=") if x]
64
+ if len(seg) >= 3:
65
+ out[seg[0]] = (seg[1], seg[2])
66
+ elif len(seg) == 2:
67
+ out[seg[0]] = (seg[1], seg[1])
68
+ return out
69
+
70
+
71
+ def ob(action, params=None, retries=2):
72
+ body = json.dumps(params or {}, ensure_ascii=False).encode("utf-8")
73
+ req = urllib.request.Request(ENDPOINT + "/" + action, data=body, method="POST",
74
+ headers={"Content-Type": "application/json",
75
+ "Authorization": "Bearer " + TOKEN})
76
+ last = None
77
+ for _ in range(retries + 1):
78
+ try:
79
+ with urllib.request.urlopen(req, timeout=90) as r:
80
+ return json.loads(r.read().decode("utf-8", "replace"))
81
+ except Exception as e:
82
+ last = e
83
+ time.sleep(1.5)
84
+ return {"status": "failed", "message": str(last)}
85
+
86
+
87
+ def list_group(gid):
88
+ """返回 (files, folders):递归展开一层文件夹(QQ 群文件夹一般只有一层)。"""
89
+ out_files, out_folders = [], []
90
+ r = ob("get_group_root_files", {"group_id": gid, "file_count": 200})
91
+ d = (r or {}).get("data") or {}
92
+ root_files = d.get("files") or []
93
+ root_folders = d.get("folders") or []
94
+ out_folders.extend(root_folders)
95
+ out_files.extend(root_files)
96
+ for fo in root_folders:
97
+ fid = fo.get("folder_id")
98
+ if not fid:
99
+ continue
100
+ r2 = ob("get_group_files_by_folder", {"group_id": gid, "folder_id": fid})
101
+ d2 = (r2 or {}).get("data") or {}
102
+ for f in (d2.get("files") or []):
103
+ f["_folder"] = fo.get("folder_name")
104
+ out_files.append(f)
105
+ out_folders.extend(d2.get("folders") or [])
106
+ return out_files, out_folders
107
+
108
+
109
+ def md5_file(p):
110
+ h = hashlib.md5()
111
+ with open(p, "rb") as fp:
112
+ while True:
113
+ b = fp.read(1024 * 1024)
114
+ if not b:
115
+ break
116
+ h.update(b)
117
+ return h.hexdigest()
118
+
119
+
120
+ def load_state():
121
+ try:
122
+ with io.open(STATE, "r", encoding="utf-8") as fp:
123
+ return json.load(fp)
124
+ except Exception:
125
+ return {}
126
+
127
+
128
+ def save_state(st):
129
+ os.makedirs(os.path.dirname(os.path.abspath(STATE)), exist_ok=True)
130
+ with io.open(STATE, "w", encoding="utf-8") as fp:
131
+ json.dump(st, fp, ensure_ascii=False, indent=1)
132
+
133
+
134
+ def download(url, dest):
135
+ tmp = dest + ".part"
136
+ req = urllib.request.Request(url, headers={"User-Agent": UA})
137
+ with urllib.request.urlopen(req, timeout=600) as r, open(tmp, "wb") as fp:
138
+ while True:
139
+ b = r.read(256 * 1024)
140
+ if not b:
141
+ break
142
+ fp.write(b)
143
+ os.replace(tmp, dest)
144
+
145
+
146
+ def safe_name(n):
147
+ n = n.replace("/", "_").replace("\\", "_").replace(":", "_")
148
+ n = "".join(c for c in n if 0xD800 > ord(c) or ord(c) > 0xDFFF)
149
+ return n.strip() or "unnamed"
150
+
151
+
152
+ def main():
153
+ global ENDPOINT, TOKEN, GROUPS, DEST_ROOT, STATE
154
+ ap = argparse.ArgumentParser()
155
+ ap.add_argument("--config", default="", help="配置 JSON(endpoint/token/groups/dest_root/state_file)")
156
+ ap.add_argument("--endpoint", default="", help="OneBot HTTP 端点")
157
+ ap.add_argument("--token", default="", help="OneBot Bearer 令牌")
158
+ ap.add_argument("--list", action="store_true", help="只列出,不下载")
159
+ ap.add_argument("--only", default="", help="只处理这些 group_id,逗号分隔")
160
+ ap.add_argument("--max-mb", type=float, default=60.0, help="单文件大小上限 MB")
161
+ ap.add_argument("--dest", default="", help="下载根目录")
162
+ ap.add_argument("--state", default="", help="台账 JSON 路径")
163
+ a = ap.parse_args()
164
+ if a.config:
165
+ apply_config(a.config)
166
+ if a.endpoint:
167
+ ENDPOINT = a.endpoint
168
+ if a.token:
169
+ TOKEN = a.token
170
+ if os.environ.get("NAPCAT_ENDPOINT"):
171
+ ENDPOINT = os.environ["NAPCAT_ENDPOINT"]
172
+ if os.environ.get("NAPCAT_TOKEN"):
173
+ TOKEN = os.environ["NAPCAT_TOKEN"]
174
+ if not GROUPS:
175
+ GROUPS = env_groups()
176
+ if a.dest:
177
+ DEST_ROOT = a.dest
178
+ if os.environ.get("QQ_SYNC_DEST") and not a.dest and not DEST_ROOT:
179
+ DEST_ROOT = os.environ["QQ_SYNC_DEST"]
180
+ if a.state:
181
+ STATE = a.state
182
+ if not GROUPS:
183
+ print("ERR: 未配置群号白名单(--config JSON 的 groups,或 NAPCAT_GROUPS)")
184
+ return 2
185
+ if not DEST_ROOT:
186
+ print("ERR: 未配置下载根(--dest / QQ_SYNC_DEST / --config 的 dest_root)")
187
+ return 2
188
+ if not STATE:
189
+ print("WARN: 未配置台账路径(--state / QQ_SYNC_STATE),本次不落台账")
190
+ STATE = os.path.join(os.getcwd(), "qq群文件台账.json")
191
+ only = [x for x in a.only.split(",") if x]
192
+ st = load_state()
193
+ log = []
194
+ total_new = 0
195
+ total_bytes = 0
196
+ for gid, (subject, gname) in GROUPS.items():
197
+ if only and gid not in only:
198
+ continue
199
+ files, folders = list_group(gid)
200
+ log.append("=== %s / %s (%s) files=%d folders=%d" % (subject, gname, gid, len(files), len(folders)))
201
+ st.setdefault(gid, {})
202
+ st[gid]["group_name"] = gname
203
+ st[gid]["subject"] = subject
204
+ st[gid].setdefault("files", {})
205
+ dest_dir = os.path.join(DEST_ROOT, subject)
206
+ seen_names = {}
207
+ for f in files:
208
+ name = safe_name(str(f.get("file_name") or ""))
209
+ size = int(f.get("file_size") or 0)
210
+ busid = f.get("busid")
211
+ fid = f.get("file_id")
212
+ if f.get("_folder"):
213
+ name = safe_name(str(f.get("_folder"))) + "__" + name
214
+ if not name or not fid:
215
+ continue
216
+ # 同名去重:第二次出现加 (2)
217
+ key = name.lower()
218
+ seen_names[key] = seen_names.get(key, 0) + 1
219
+ if seen_names[key] > 1:
220
+ base, ext = os.path.splitext(name)
221
+ name = "%s (%d)%s" % (base, seen_names[key], ext)
222
+ os.makedirs(dest_dir, exist_ok=True)
223
+ target = os.path.join(dest_dir, name)
224
+ local_ok = os.path.exists(target) and os.path.getsize(target) == size
225
+ rec = st[gid]["files"].get(name) or {}
226
+ flag = "SKIP"
227
+ if local_ok:
228
+ if rec.get("status") != "ok":
229
+ rec = {"size": size, "status": "ok", "md5": md5_file(target), "note": "已存在"}
230
+ elif a.list:
231
+ flag = "LIST"
232
+ elif size > a.max_mb * 1024 * 1024:
233
+ flag = "BIG"
234
+ else:
235
+ flag = "GET"
236
+ if flag == "GET":
237
+ r = ob("get_group_file_url", {"group_id": gid, "file_id": fid, "busid": busid})
238
+ url = ((r or {}).get("data") or {}).get("url")
239
+ if not url:
240
+ log.append(" [URL-FAIL] %s -> %s" % (name, json.dumps(r, ensure_ascii=False)[:160]))
241
+ rec = {"size": size, "status": "url-fail", "busid": busid}
242
+ flag = "ERR"
243
+ else:
244
+ try:
245
+ download(url, target)
246
+ got = os.path.getsize(target)
247
+ rec = {"size": size, "got": got, "status": "ok" if got == size else "size-mismatch",
248
+ "md5": md5_file(target), "busid": busid,
249
+ "downloaded_at": time.strftime("%Y-%m-%d %H:%M:%S")}
250
+ total_new += 1
251
+ total_bytes += got
252
+ except Exception as e:
253
+ rec = {"size": size, "status": "dl-fail", "err": str(e)[:200], "busid": busid}
254
+ flag = "ERR"
255
+ st[gid]["files"][name] = rec
256
+ log.append(" [%-4s] %-58s %10d B" % (flag, name[:58], size))
257
+ log.append(" -- 本地目录: %s" % dest_dir)
258
+ save_state(st)
259
+ print("\n".join(log))
260
+ print("NEW FILES: %d NEW BYTES: %d" % (total_new, total_bytes))
261
+ return 0
262
+
263
+
264
+ if __name__ == "__main__":
265
+ sys.exit(main())
@@ -1,9 +1,18 @@
1
1
  # 一起开发:仓库、分支、怎么装到自己机器上试
2
2
 
3
+ ## 0. 两个远端
4
+
5
+ 内容一致,日常协作走 Gitee(国内快、链接邀请进人),GitHub 是开源门面 + npm 发布源:
6
+
7
+ - **Gitee**:<https://gitee.com/lian-ziquan/dsh-learning-assistant>
8
+ - **GitHub**:<https://github.com/shinkontokitoki/dsh-learning-assistant>
9
+
10
+ 维护者合并 Gitee 上的 PR 后 `git push origin` 同步回 GitHub;发布永远只走 npm。
11
+
3
12
  ## 1. 一次性准备(每个开发者各做一遍)
4
13
 
5
14
  ```powershell
6
- git clone https://github.com/shinkontokitoki/dsh-learning-assistant.git
15
+ git clone https://gitee.com/lian-ziquan/dsh-learning-assistant.git
7
16
  cd dsh-learning-assistant
8
17
  npm install
9
18
  npm run build # 产出 lib/index.mjs、lib/client.mjs
@@ -64,23 +73,22 @@ npm publish # 首次若用了 scope 需加 --access public
64
73
  | 维护者 | 合并 PR、升版本、发 npm、答复 issue |
65
74
  | 开发者 | 自己的分支 + PR + 自测;装在自己机器上验证,不碰别人的 profile |
66
75
 
67
- ## 7. 仓库现在是 public:两种参与方式
76
+ ## 7. 加入与邀请(Gitee 链接邀请,无需逐个加好友)
68
77
 
69
- 仓库地址 <https://github.com/shinkontokitoki/dsh-learning-assistant> 已公开,任何人可 clone、读代码、提 issue。
78
+ 维护者:Gitee 仓库页 → **管理** → **仓库成员管理** → **添加仓库成员** → **链接邀请**
79
+ → 选 Developer 角色 → 生成链接 → 发给队友(QQ/微信均可)。
80
+ 勾选「需要管理员审核」可再加一道确认。
70
81
 
71
- **方式一:fork + PR(不用等邀请,推荐给不常一起改的人)**
72
- fork → 在自己的 fork 上开分支改 → 向上游提 PR。维护者合并。
82
+ 队友:点链接 → 注册/登录 Gitee → 接受邀请 → 直接 clone 上面 Gitee 地址开工。
83
+ 免费版私有仓库成员上限 5 人;本仓库当前是开源状态,不受限。
73
84
 
74
- **方式二:直接被加进仓库(能推分支,适合长期一起改的朋友)**
75
- 需要维护者按用户名或邮箱逐个邀请(GitHub 个人仓库没有「邀请链接」):
85
+ GitHub 侧的正式 collaborator 邀请(等主账号申诉解除后):
76
86
 
77
87
  ```powershell
78
- # permission: pull(只读) / push(可推) / admin
88
+ # 逐个加,permission: pull(只读) / push(可推) / admin
79
89
  gh api -X PUT repos/shinkontokitoki/dsh-learning-assistant/collaborators/<GitHub用户名> -f permission=push
80
90
  ```
81
91
 
82
- 或网页 Settings → Collaborators → Invite(可填用户名或邮箱)。对方收邮件/通知点接受即可。
83
-
84
92
  > 注意:仓库若日后要重新转回 private,已 fork 的副本不会收回——转 public 前请确认代码里没有密钥、个人路径。
85
93
 
86
94
  ## 8. 推不上去时(本机实测过)