flowocr 0.1.0__py3-none-any.whl

flowocr/typeset/fx/__init__.py

@@ -0,0 +1,11 @@
+"""阶段 4 的扩展效果（`flowocr.extensions` 的 `fx` 种类）放这里的是内置的；现在没有，底板、编号、逐字显出在 `core` 里。
+
+一个效果模块：
+
+* `NAME`：它在字幕稿 Effect 字段 `fo:…` 里的参数名，源行带了这个参数才调它；
+* `elements(line, value) -> list[core.Element]`（可选）：第 2 步，读排好版的 `core.Line`（只读），加元素；
+* `tags(line, value, element) -> None`（可选）：第 4 步，改每个元素的 tag（`element.tail` / `element.tail2` / `element.body`）；
+  在动的行按轨迹拆成几段时，每一段的元素定稿之后各调一次。
+
+两个挂点至少有一个。元素的位置按条目框左上角排，第 3 步和底板、正文一起按轨迹走。例子见 `examples/fx_minimal.py`。
+"""

flowocr-0.1.0.dist-info/METADATA

@@ -0,0 +1,109 @@
+Metadata-Version: 2.4
+Name: flowocr
+Version: 0.1.0
+Summary: 从视频里抽全部屏幕文字、带起止时间、按区域聚类，导出 SRT / ASS
+Author: caca2331
+License-Expression: GPL-3.0-or-later
+Project-URL: Homepage, https://github.com/caca2331/FlowOCR
+Project-URL: Documentation, https://github.com/caca2331/FlowOCR/blob/main/docs/manual/README.md
+Project-URL: Issues, https://github.com/caca2331/FlowOCR/issues
+Project-URL: Changelog, https://github.com/caca2331/FlowOCR/blob/main/CHANGELOG.md
+Keywords: ocr,subtitle,video,srt,ass,onnxruntime
+Classifier: Development Status :: 3 - Alpha
+Classifier: Environment :: Console
+Classifier: Environment :: GPU :: NVIDIA CUDA
+Classifier: Intended Audience :: End Users/Desktop
+Classifier: Natural Language :: Chinese (Simplified)
+Classifier: Operating System :: Microsoft :: Windows
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Multimedia :: Video
+Classifier: Topic :: Scientific/Engineering :: Image Recognition
+Requires-Python: <3.13,>=3.12
+Description-Content-Type: text/markdown
+License-File: LICENSE
+License-File: LICENSES/Apache-2.0.txt
+License-File: LICENSES/PP-OCRv6-NOTICE.md
+Requires-Dist: numpy<3,>=2.3
+Requires-Dist: opencv-contrib-python<4.11,>=4.10
+Requires-Dist: pillow>=10
+Requires-Dist: psutil>=6
+Requires-Dist: pyclipper>=1.3
+Requires-Dist: pyyaml>=6
+Requires-Dist: zstandard>=0.22
+Provides-Extra: nvidia
+Requires-Dist: onnxruntime-gpu[cuda,cudnn]==1.29.0; extra == "nvidia"
+Requires-Dist: onnx>=1.17; extra == "nvidia"
+Requires-Dist: nvidia-cuda-runtime==13.0.88; extra == "nvidia"
+Requires-Dist: nvidia-cudnn-cu13==9.13.0.50; extra == "nvidia"
+Requires-Dist: nvidia-cublas==13.0.2.14; extra == "nvidia"
+Requires-Dist: nvidia-cufft==12.0.0.61; extra == "nvidia"
+Requires-Dist: nvidia-curand==10.4.0.35; extra == "nvidia"
+Requires-Dist: nvidia-cuda-nvrtc==13.0.*; extra == "nvidia"
+Requires-Dist: nvidia-nvjitlink==13.0.88; extra == "nvidia"
+Provides-Extra: cpu
+Requires-Dist: onnxruntime==1.29.0; extra == "cpu"
+Requires-Dist: onnx>=1.17; extra == "cpu"
+Provides-Extra: learned
+Requires-Dist: scikit-learn>=1.5; extra == "learned"
+Dynamic: license-file
+
+# flowocr
+
+把视频里**出现过的所有文字**读出来——不只是底部字幕，还有对话框、名牌、画面上的标注——每条带起止时间（精确到帧），
+并按位置自动分成若干条"文字轨"（对白、字幕、界面、水印……）。产物是 SRT，或叠加在原画面上、每条字留在原位的 ASS 字幕。
+
+## 用来做什么
+
+- **画面文字翻译**：画面各处的文字都读出来、按位置分好轨；在字幕稿里把原文改成译文，叠加时每条译文留在它原来的位置上。
+- **字幕与原文对齐**：游戏、影视常常能找到原文剧本。把读到的字幕对上原文，拿到准确的文本和时间轴，用来做翻译字幕。
+  原神、崩坏：星穹铁道、绝区零有现成的提取工具，按它的说明自己打出文本包（见[相关项目](https://github.com/caca2331/FlowOCR#相关项目)）。
+
+## 能读什么、在哪跑
+
+- **文字**：识别模型是 PaddlePaddle 的 PP-OCRv6（官方称支持约 50 种语言）。实际验证过的是日文（几十小时实况录像和四款游戏的直播录像）与简体中文；
+  英文只做过零星试跑；竖排文字没测过。
+- **系统**：Windows 10 / 11。有 NVIDIA 显卡时走显卡；没有也能跑，用 CPU，慢一到两个数量级。
+- 另要 [ffmpeg](https://ffmpeg.org/)（完整构建）。
+
+## 快速开始
+
+要先装 [uv](https://docs.astral.sh/uv/) 和完整版 [ffmpeg](https://ffmpeg.org/)；从源码装、参数和产物说明都在 [使用手册](https://github.com/caca2331/FlowOCR/blob/main/docs/manual/README.md)。
+
+```powershell
+uv tool install "flowocr[nvidia]"                   # 没有 NVIDIA 显卡用 "flowocr[cpu]"
+uv tool update-shell                                # 第一次用 uv tool 时，然后重开终端
+```
+
+有原文剧本时，推荐**开启匹配器，使用默认配置，先做到阶段 3 的字幕稿**；导入 [Aegisub](https://aegisub.org/) 并加载视频，检查文字、时间和位置，改好保存后再跑阶段 4 渲染特效。
+下面以星穹铁道为例，先按 [game-text-data](https://github.com/caca2331/game-text-data) 的说明准备文本包，放进 `gametext\`：
+
+```powershell
+flowocr-ocr video.mp4 --out out\v\v.jsonl           # 阶段 1：读字
+flowocr-gamescript out\v\v.jsonl --out out\v\v-ref.json --game starrail
+flowocr-tracks out\v\v.jsonl --outdir out\v --tag v --matcher gametext --matcher-ctx game=starrail
+flowocr-match out\v\v-tracks.json --matcher gametext --ctx ref=out\v\v-ref.json --out out\v\v-matched.json
+flowocr-render out\v\v-matched.json --preset script # 阶段 3：只写 v-jp-main.script.ass，默认保留正文与名牌
+```
+
+在 Aegisub 中检查并保存 `out\v\v-jp-main.script.ass`，然后用 `default` 预设出特效字幕：
+
+```powershell
+flowocr-typeset out\v\v-jp-main.script.ass --preset default # 阶段 4：写 v-jp-main.ass
+```
+
+原神 / 绝区零把 `starrail` 换成 `genshin` / `zzz`；没有剧本时省略剧本提取与匹配，用 `flowocr-tracks` 不带匹配参数建轨，再对 `v-tracks.json` 运行阶段 3，字幕稿为 `v-main.script.ass`。
+
+## 相关项目
+
+- [game-text-data](https://github.com/caca2331/game-text-data)：从原神、崩坏：星穹铁道、绝区零的社区解包数据里提取对白文本（中日英对齐），
+  打成 flowocr 能直接读的文本包。它只公开提取代码，文本本身不分发——要用的话按它的说明自己打包。
+
+## 问题与参与
+
+- 用起来有问题、或者在别的显卡 / 机器上跑出了结果，欢迎开 [issue](https://github.com/caca2331/FlowOCR/issues)。
+- Pull request 请以 `main` 为目标。开发环境、代码结构和验证方法见 [开发指南](https://github.com/caca2331/FlowOCR/blob/main/docs/dev-guide/README.md)，全部文档见 [文档地图](https://github.com/caca2331/FlowOCR/blob/main/docs/README.md)。
+- 版本变化见 [CHANGELOG](https://github.com/caca2331/FlowOCR/blob/main/CHANGELOG.md)。
+
+## 许可
+
+GPL-3.0-or-later，全文见 [LICENSE](https://github.com/caca2331/FlowOCR/blob/main/LICENSE)。识别用的 PP-OCRv6 模型由 PaddlePaddle 以 Apache-2.0 发布，来源与我们的改动见 [LICENSES/](https://github.com/caca2331/FlowOCR/blob/main/LICENSES/PP-OCRv6-NOTICE.md)。

flowocr-0.1.0.dist-info/RECORD

@@ -0,0 +1,91 @@
+flowocr/__init__.py,sha256=f48OWGLkK90ugd-brJbVzhRF_amXYX7MKUGvrzJ2MKI,724
+flowocr/extensions.py,sha256=QuKZ_RYJy4ImIxM68wcvDpFavXLZcwJJwUfY5kNxji0,4525
+flowocr/models.py,sha256=6ZBIYDeuCpRnnTjNy2ciRXhwi954xw90W59kYIcXFFU,13220
+flowocr/paths.py,sha256=KEQu9Vu8O_jnVt6EQ75hqm3iOsSd5zw0OEqUIM1sLLk,8207
+flowocr/provenance.py,sha256=p0LOFhDKuXsN_BtINEtcKyVdniNqZUsrFkFAgyDSFio,6409
+flowocr/analyze/__init__.py,sha256=RKRJu7avF0Se-zr9ABTjZKdpt-8U6AkrgGTxXppWQPY,685
+flowocr/analyze/align.py,sha256=L7ftPboNWVy6s70Sr-liALotjRYk-aY4XUufZF8VqDM,7130
+flowocr/analyze/build_tracks.py,sha256=Tw3UrrvBElCcoqwAHsn6WsFesAvJwpND0DaJdvplGk8,175585
+flowocr/analyze/cluster_layers.py,sha256=3ScPZPSry0yVnZj_1fzou7-8yd_hgPbYyqMiCgjNZS0,8355
+flowocr/analyze/game_align.py,sha256=0WnyNjWtu_Kv3vpJ0Bw__zegVy3qWa9t59gqu7TE4hk,38328
+flowocr/analyze/gamescript.py,sha256=vThdRXKANJsW2Rayo7CnzegZwLz-gPLwSvIz-ZmVAHs,70852
+flowocr/analyze/gtdbundle.py,sha256=xJYb5lgXkuBfc6NXnYSeXd18gT899zUnb7zXvjA4yQY,15938
+flowocr/analyze/match.py,sha256=VkGHLjoyxdMkOz7pAJq7QgKo7y4zV_J1Lphqxu2JbFs,3409
+flowocr/analyze/merge_nameplate.py,sha256=zzwAzTE7yLcpvEq3ixjgTbq57x0qmjmQt1H0JetjG7o,10616
+flowocr/analyze/nameplate.py,sha256=rhu-0A3UhEDN2UlebPP01McJt2yzp0hPfeg5R02EYo0,8770
+flowocr/analyze/pair_features.py,sha256=V3iRzTqleGtNyJ3FMzm2c-V8MLhjIBN9LS-RUgwZKl0,9497
+flowocr/analyze/pair_model.py,sha256=edNpENP0dptE0xCYKk0BOOpNDiSUPTAZwCgPq6Ks1qs,6283
+flowocr/analyze/refine_boundaries.py,sha256=QBUDqUOGMEz31cR-ix7nSu8CAwZSoKq2ZywfTWrgBcw,26983
+flowocr/analyze/script_align.py,sha256=MehyccUNK4CBWyWKvgWs1iathIkGTN_Wu5gDFV7VUr4,35752
+flowocr/analyze/scriptmatch.py,sha256=0-ep_44ChqNlloCrH9RF4xsEHOJ6L0-6-zOcIrgfVhc,66663
+flowocr/analyze/slot_learned.py,sha256=jApWHIOFCQ8Jx0F-hOPXBZ9KNKjwqZfdXWZITFpKy68,16956
+flowocr/analyze/slot_lines.py,sha256=CL4v0Dq_P3-qlMx2CZ958UscplBz8wUryl3sLJSmd60,12460
+flowocr/analyze/slot_modes.py,sha256=k4ewdF8CZnhjiprG_dPnvBJSf7XONbo3hhw_dSmq47Q,7278
+flowocr/analyze/slot_pairs.py,sha256=SB8Bj1ZlybOzAZ89wydiZuWNlcwlnICMnU6HhHrljfk,29368
+flowocr/analyze/slot_veto.py,sha256=wwfr2_rx9oKJBVIAqNy2XrnkJqgKl33fN9Y65UeBZ1A,5076
+flowocr/analyze/uigate.py,sha256=BlHMs4KWByCm0HFrVvXVr5Iy-O5qgTmrovpkjZMAZFs,19702
+flowocr/analyze/matchers/__init__.py,sha256=hGA8eY5evGerDpc_ck62sHr0Rm4CeLn-nZDD8cWpBRc,339
+flowocr/analyze/matchers/gametext.py,sha256=jW2IZH_ZVxclodhvdyqWuauUjsdz7tMat5DqjZ_-ArA,4219
+flowocr/analyze/models/slot_pair.json,sha256=N0pmuQLDvsJi_vvzUhmVcdeUXiZagCdN6kWb0hbsRx8,2938
+flowocr/artifacts/__init__.py,sha256=WbhFG7HwdIQNkRF9pKK3Trd3G4P_55QNFSUeEWmxC64,320
+flowocr/artifacts/evalkit.py,sha256=ZWJQPuYeGyeJn1J--CCgTLVV7OTYNNgKFJiWM04BweU,5251
+flowocr/artifacts/matchedio.py,sha256=wdxEgYMSxSXZ3_zh4tkiJmBs2j2zexlj7qqXPDLzhgk,4483
+flowocr/artifacts/srtio.py,sha256=Hb0V5llnjn7v1tCkQ6_pH4AMraqWog70UgWbSfLur-Q,6766
+flowocr/artifacts/tracksio.py,sha256=6HExInREfnqYwrl9yO8D9DJyBp5ElWiss2IIei9YW1A,19571
+flowocr/extract/__init__.py,sha256=QaDUPc42_jEZ0AtPPUL92nrHSpoMYkojqnR90wweJzo,736
+flowocr/extract/childproc.py,sha256=446Vb-9yqsCHOYCmsvvUepddaEx1d-9pzJjcsnFacTI,5071
+flowocr/extract/decode_proc.py,sha256=W149NjXoan9FhmHrOjlhlf33Nro0-z6weJcF2zuWmTE,22498
+flowocr/extract/decode_shards.py,sha256=eb4VCdCgTQtAbnyRNv48ZbMEFjhQc2KOQEeWxyDk8PQ,30803
+flowocr/extract/detpost.py,sha256=kVW47kliUtIdx13TWEFXW0ZtWbuq_IQF_pEoa0KMuJk,11871
+flowocr/extract/edge_proc.py,sha256=yzXMagPGyLW36_3R-FId2WULpA9d1UNAfPJvpFCKovs,15740
+flowocr/extract/edge_refine.py,sha256=uwKb5IxtRaBsqr5GiQxclFBsYA8rKt089b-By1auIWA,37057
+flowocr/extract/fast_det.py,sha256=VlUWr7l1VdW_y0HWuNS5N3FCRU_YASk7WvPMv9HYV-E,11819
+flowocr/extract/ffcheck.py,sha256=Js_1FeVEse3pUmXa61rUhPXpaldH9CSECofxIbtGFLc,4771
+flowocr/extract/framegrid.py,sha256=pa1LFKi0QKXIN8Shemy6wSZ7DeFYADM_A3b49spuJ4k,12930
+flowocr/extract/framesource.py,sha256=lI_ZefIAeMX5GZ7Fwn6ejLEEvUEb_7jBNLndYz1n8i0,73546
+flowocr/extract/ocr_args.py,sha256=LhB0q9up6CbR_epUAbfJ_AXbBfKYMbkUy08XEfzDgPc,70143
+flowocr/extract/ocr_complete.py,sha256=FMqSzRjQ1LBtHpx8WXGNjNzMDNtmBfhQ8fLGYYi06pc,14486
+flowocr/extract/ocr_parallel.py,sha256=TPGSk1l1qdq6le3HGnjhny-JW1wppx2abulqK3D9eJE,12326
+flowocr/extract/ort_server.py,sha256=wyO0mNUiCTSKPKoygJ38WCiW6MyENV26Ux_CrA4jmDY,38945
+flowocr/extract/ortclient.py,sha256=7Suc2VO8zFalkRUT1D2bOrb8ygGxhMpS27FW1MAX5KA,20461
+flowocr/extract/ptsclock.py,sha256=EAA0sE1aP6q54QVJa0zMZHXDej2IQ-_YsDokFYxLiYI,7015
+flowocr/extract/recdecode.py,sha256=CwyQgLLhXTyULQ9dOdiXJzNYSMjWAf3kOa35gJgFv3A,4162
+flowocr/extract/recort.py,sha256=bCWlmrmb7r6Vi62jZLmI1DUH34BBxY2JxIGK0B4WPjk,9515
+flowocr/extract/recpack.py,sha256=tNmYUCGU-txQJs_SvzzyVpT6Flgbcs5itywnoDj-h0E,5155
+flowocr/extract/recpool.py,sha256=umaSGguVDBhlDyzAkiWP-qwIJBageB5FwhsBSyfViRk,12235
+flowocr/extract/recprep.py,sha256=2F6hyFLTGkiEr9iW0ZlRqbJCcIw-tGqQvK87-ZNN6xw,4203
+flowocr/extract/refine_video.py,sha256=7WAG_zblD3nImhruSfhuj1bJRBTEp6qhysWiDFT4wWY,15274
+flowocr/extract/regions.py,sha256=xwPcrJOxGyLtLML5Ygd_ajK2sPxemptp3KbrDZVdft0,19726
+flowocr/extract/reuse_v2.py,sha256=3e8IH6R7mwN2UsxGjQilSunjo9wcmAS53g2iOa6M998,39729
+flowocr/extract/run_groups.py,sha256=DS8qhnfO25CGrYpkFeQWszRmFMtCKGb_AsDQUZes84E,14263
+flowocr/extract/run_ocr2.py,sha256=4qhDn-S6fZLLF6k1vhV3u3-0IQYcVZ68X6x4hrw-kZ8,99403
+flowocr/extract/supervisor.py,sha256=ysWssXJ6qiUnr6N74GY1_iJUqtv3wNro5e9XoXlECP8,14722
+flowocr/extract/timeline.py,sha256=wu3aKP4asz6mEmLjMUNOTcpB7MTTMesAqDVUYIEvWyc,3559
+flowocr/extract/typewriter_fuse.py,sha256=BLvffQFT2hEDoKky-tpUyxjQyvXl3GMUCFBF-7G6CIU,21447
+flowocr/output/__init__.py,sha256=z_o5eIKOKcmsHLcCACjdhko1tu_ZMaeKs8S6nJ_xTNc,260
+flowocr/output/export.py,sha256=jSU2nr78SqecuU8ibm9XdI2c3tWqs9hqjm3u62bCVIQ,11410
+flowocr/output/layout.py,sha256=-ZygdiZuhI3H5i1t7P6A0WO8sILy-Z45sbOkmFxDxZ8,11565
+flowocr/output/render.py,sha256=VEZSVmNKjKgC4qOhTJ7JZca-OC7fmER-Jqf5qGSNb-w,4249
+flowocr/output/run_srt.py,sha256=MsyvobmSa3Lx1lWcWVc1ivMUnAKkt5TjADNS0La3Gok,5422
+flowocr/output/script.py,sha256=PjyYKRh3DlKD0sssR3n3EBXuPXdj8MI2GBDKei8t4r8,32831
+flowocr/output/presets/__init__.py,sha256=OpUnVz9lC7Bqo0RuLHX8xeKuN-oq7K9ok1K79Top0VI,554
+flowocr/output/presets/_overlay.py,sha256=STfWp-VDxhYcVlk7WleKQAKw_vEqeXjp1YWDcoAszvk,2464
+flowocr/output/presets/default.py,sha256=fVgLXif1EUKkwx5DPHRp4Ti1nFpOEJt2U9_L9DQlqSU,1131
+flowocr/output/presets/default_all.py,sha256=XsJt1QmIrc9Phsphx8kAV00Vj5cTo1lpnshP4Ey54sQ,1059
+flowocr/output/presets/dev.py,sha256=929xEFWf-ksoR_1iGAwAaQNHGzhYNXoNy1LgkthSzT4,1258
+flowocr/output/presets/matched_srt.py,sha256=YBRuT53yT3LVIOJkBQUiEN5I8W-TGKN6MGTcE7Mq5Rk,1518
+flowocr/output/presets/script.py,sha256=L3lEebYV_m1IxreKuhvlC9apxsNQ413VMGlr86eUWC8,887
+flowocr/output/presets/srt_main.py,sha256=LsSSgPdNvm6MuW3R0_o29R6wEVD-j7_rxe6JJHwnsRo,1496
+flowocr/typeset/__init__.py,sha256=j9-IU-MlM-iZRMcLcRmx_vbIMha5dybww6FYfmRZxl0,706
+flowocr/typeset/__main__.py,sha256=sF72uIKK0IHpP9yVjIM9cfJfRK4qktErvrky8MbXg1E,2002
+flowocr/typeset/assfile.py,sha256=A6d7Kf4XyIUpeoBligqyq6AF3SlAxW2flEGuzBDkRMA,7747
+flowocr/typeset/core.py,sha256=tUILsAmpUSkanHK4wdqU0eVXzb3e2VPs-EIODUH6X_g,43118
+flowocr/typeset/fx/__init__.py,sha256=tijDwWseNnsChI-lJpOO07HUqBLzIvhgM-RWY7H_vV0,813
+flowocr-0.1.0.dist-info/licenses/LICENSE,sha256=jOtLnuWt7d5Hsx6XXB2QxzrSe2sWWh3NgMfFRetluQM,35147
+flowocr-0.1.0.dist-info/licenses/LICENSES/Apache-2.0.txt,sha256=OEDFwMYcKUJk0t13uHd75t3ZASHvTg5kq80i7epYHW4,11376
+flowocr-0.1.0.dist-info/licenses/LICENSES/PP-OCRv6-NOTICE.md,sha256=OuR6CjbgXkEQT_WXnN4Q760xTBDa-LP104Tg3AC8wJ8,1421
+flowocr-0.1.0.dist-info/METADATA,sha256=0mdu5Xvx_yjvvQgIMuujW7lSC0RqEPpEBni8vJQreHU,6570
+flowocr-0.1.0.dist-info/WHEEL,sha256=YVMoNqKzERt-wjUZwJ33xBGAwnFl-4cqbYkTtWa4itE,91
+flowocr-0.1.0.dist-info/entry_points.txt,sha256=6zYakPBGYRVLvxCau1ZXkQBJbtermYXfIxFHHLN3OGk,419
+flowocr-0.1.0.dist-info/top_level.txt,sha256=0gpG_kfItuR_W6q9upEGO6spFGn5sDmX9E_1ESUCjfA,8
+flowocr-0.1.0.dist-info/RECORD,,

flowocr-0.1.0.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (84.0.0)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

flowocr-0.1.0.dist-info/entry_points.txt

@@ -0,0 +1,10 @@
+[console_scripts]
+flowocr-export = flowocr.output.export:main
+flowocr-gamescript = flowocr.analyze.gamescript:main
+flowocr-match = flowocr.analyze.match:main
+flowocr-models = flowocr.models:main
+flowocr-ocr = flowocr.extract.supervisor:main
+flowocr-paths = flowocr.paths:main
+flowocr-render = flowocr.output.render:main
+flowocr-tracks = flowocr.analyze.build_tracks:main
+flowocr-typeset = flowocr.typeset.__main__:main