Rename numeric-prefixed pipeline scripts to clean module names

The 00_/01_/02_/03_/05_ prefixes read like a course assignment and can never be
imported as modules (an identifier can't start with a digit). Renamed:

  00_resolve_creators.py   -> resolve_creators.py
  01_fetch_videos.py       -> fetch_videos.py
  02_classify_and_stats.py -> classify_and_stats.py
  03_subtitle.py           -> subtitle.py
  05_comments.py           -> comments.py

These are standalone steps run as files (not imported, and not in viralens.py's
subprocess pipeline, which dispatches fetch_multi / compare_form / etc.). All
references were string mentions in docs/comments — README.md and the headers/
cross-references in resolve_creators.py, fetch_videos.py, classify_and_stats.py,
subtitle.py, comments.py, creators.py, fetch_bilibili.py — all updated. Verified:
no numeric-prefix tokens remain anywhere, compileall passes, `viralens --help` works.

Author: HarryXin0919

README.md

@@ -185,7 +185,7 @@ python scripts/compare_meme.py       # (opt-in, slow) cross-creator comment-enga
 python scripts/fetch_covers.py       # (opt-in, slow) cover-image metrics
 ```
 
-Don't know a Bilibili creator's UID? Run `python scripts/00_resolve_creators.py` — it searches by
+Don't know a Bilibili creator's UID? Run `python scripts/resolve_creators.py` — it searches by
 name and prints the top candidates by follower count so you pick the real account, not an impersonator.
 
 ---
@@ -216,7 +216,7 @@ creators.py ──▶ fetch_multi.py ──▶ data/<alias>_videos.json        (
 - **Signal scanner** (`scan_signals`): turns each video into a universal feature vector, then auto-tests
   every dimension (title patterns, length buckets, daypart, cover metrics…) for high/low-play separation,
   ranks by effect size, and reports which levers are *universal* vs *creator-specific*.
-- **L2 — text** (`03_subtitle`, `05_comments`): subtitles + hot comments → `jieba` keyword analysis.
+- **L2 — text** (`subtitle`, `comments`): subtitles + hot comments → `jieba` keyword analysis.
 - **Cross-creator / cross-zone gate**: a pattern earns a ✅ only if it survives the *same test* on
   multiple independent creators **and** more than one zone.
 

scripts/02_classify_and_stats.py scripts/classify_and_stats.pyscripts/02_classify_and_stats.py renamed to scripts/classify_and_stats.py

@@ -1,9 +1,9 @@
 """
-viralens · 02_classify_and_stats.py
+viralens · classify_and_stats.py
 读 bidao_videos.json,自动给视频打类型标签,按多个维度算对比统计。
 纯本地,零成本,零网络。
 
-跑: python 02_classify_and_stats.py
+跑: python classify_and_stats.py
 
 输出:
   - data/classified.json  (每个视频加了 type / 派生指标字段)

scripts/05_comments.py scripts/comments.pyscripts/05_comments.py renamed to scripts/comments.py

@@ -1,13 +1,13 @@
 """
-viralens · 05_comments.py
+viralens · comments.py
 抓 pilot 10 个视频(2025正经科普 高5+低5)的热评,做三类成分分析:
   共同 = 跨视频高频词(创作者标签性反馈,如"涨知识""哈哈哈")
   典型 = 单视频 TF-IDF top 词(这条视频独有的记忆点)
   差异 = 高播放组 vs 低播放组 词频差(观众"买账/不买账"时各说什么)
 
 依赖: bilibili-api-python jieba
 SESSDATA: 从 config_local.py 读(不进 git)
-跑: python 05_comments.py
+跑: python comments.py
 输出: data/comments_raw.json + data/comment_components.json + 终端三类成分
 """
 import asyncio

scripts/creators.py

@@ -13,13 +13,13 @@
             ⚠️ 跨平台别混用 zone:B站和 YouTube 受众体量不同,给 YouTube 用各自的 zone 名
             (例:B站知识区写 "知识",YouTube 科普写 "STEM-YT"),否则基准就是在跨平台硬比。
   —— Bilibili 专用 ——
-  uid       B站用户 UID(数字)。留 None 由 00_resolve_creators.py 按 name 搜索补全
+  uid       B站用户 UID(数字)。留 None 由 resolve_creators.py 按 name 搜索补全
   —— YouTube 专用 ——
   channel   "@handle"(推荐,稳定可核对) / "UCxxxx频道ID" / 频道名(兜底搜索)
   min_duration_sec  可选:过滤短视频(同时发 Shorts 的频道建议 180,排除 Shorts)
 """
 # 下面是示例 —— 换成你想分析的创作者。
-#   B站 UID:用 00_resolve_creators.py 按名字搜;YouTube @handle:在频道主页地址栏能看到。
+#   B站 UID:用 resolve_creators.py 按名字搜;YouTube @handle:在频道主页地址栏能看到。
 CREATORS = [
     # —— Bilibili(uid 留 None 会自动按 name 搜索补全)——
     {"name": "示例UP主A", "platform": "bilibili", "uid": None, "alias": "demo_b1", "zone": "知识"},

scripts/fetch_bilibili.py

@@ -33,7 +33,7 @@ async def fetch_creator(c, sessdata, num=40, tries=4):
         raise RuntimeError("没读到 SESSDATA(config_local.py)")
     uid = c.get("uid")
     if not uid:
-        raise RuntimeError("缺 UID,先跑 00_resolve_creators.py")
+        raise RuntimeError("缺 UID,先跑 resolve_creators.py")
     cred = Credential(sessdata=sessdata)
     u = user.User(uid=uid, credential=cred)
     raw = None

scripts/01_fetch_videos.py scripts/fetch_videos.pyscripts/01_fetch_videos.py renamed to scripts/fetch_videos.py

@@ -1,5 +1,5 @@
 """
-viralens · 01_fetch_videos.py
+viralens · fetch_videos.py
 抓一个 B 站 UP 主最近 N 个视频的元数据,输出 JSON 给后续分析用。
 (这是早期单人脚本;多创作者请用 fetch_multi.py。)
 
@@ -11,7 +11,7 @@
     2. F12 → Application/存储 → Cookies → https://www.bilibili.com → 找 SESSDATA
        复制 Value(看起来像 "abc123%2Cxxxx%2Cyyy...")
     3. 把 Value 粘贴到下方 SESSDATA 变量(引号内)
-    4. 跑: python 01_fetch_videos.py
+    4. 跑: python fetch_videos.py
 
 输出: ../data/<alias>_videos.json
 """
@@ -24,7 +24,7 @@
 from bilibili_api import user, Credential
 
 # ============ 配置 ============
-UID = 0                     # 换成你要抓的 UP 主 UID(00_resolve_creators.py 可按名字查)
+UID = 0                     # 换成你要抓的 UP 主 UID(resolve_creators.py 可按名字查)
 NUM_VIDEOS = 30             # 抓最近多少个
 try:
     from config_local import SESSDATA   # ← SESSDATA 统一放 config_local.py(已 gitignore,不进 git)

scripts/00_resolve_creators.py scripts/resolve_creators.pyscripts/00_resolve_creators.py renamed to scripts/resolve_creators.py

@@ -1,10 +1,10 @@
 """
-viralens · 00_resolve_creators.py
+viralens · resolve_creators.py
 按创作者名字搜 B站,补全 UID(选粉丝最多的账号,防同名假号)。
 跑完核对候选 → 我帮你把正确 UID 写回 creators.py。
 
 依赖: bilibili-api-python  /  SESSDATA 从 config_local.py 读
-跑: python 00_resolve_creators.py
+跑: python resolve_creators.py
 """
 import asyncio
 import sys

scripts/03_subtitle.py scripts/subtitle.pyscripts/03_subtitle.py renamed to scripts/subtitle.py

@@ -1,12 +1,12 @@
 """
-viralens · 03_subtitle.py
+viralens · subtitle.py
 抓 pilot 视频(2025正经科普 高5+低5)的 B站字幕,算"实验 vs 思辨"密度、语速、开场钩子。
 验证假设:高播放组 是不是 思辨密度更高、实验奇观更少。
 
 依赖: bilibili-api-python aiohttp (已装)
 SESSDATA: 从 01 脚本复制粘贴到下方
 
-跑: python 03_subtitle.py
+跑: python subtitle.py
 输出: data/subtitle_features.json + 终端高低组对比
 """
 import asyncio