objectionengine2 1.0.0__tar.gz

objectionengine2-1.0.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2021 Luis Mayo Valbuena
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

objectionengine2-1.0.0/PKG-INFO

@@ -0,0 +1,101 @@
+Metadata-Version: 2.4
+Name: objectionengine2
+Version: 1.0.0
+Summary: Independent continuation of Objection Engine for rendering Ace Attorney scenes
+Author: Luis Mayo Valbuena, Malcolm Anderson
+Author-email: Luis Mayo Valbuena <luismayovalbuena@outlook.es>, Malcolm Anderson <malcolminyo@gmail.com>
+License-Expression: MIT
+License-File: LICENSE
+Requires-Dist: pillow>=10.4,<13
+Requires-Dist: pydub>=0.25.1,<0.26
+Requires-Dist: audioop-lts>=0.2.2,<0.3 ; python_full_version >= '3.13'
+Requires-Dist: ffmpeg-python>=0.2,<0.3
+Requires-Dist: spacy>=3.8.16,<3.9
+Requires-Dist: opencv-python-headless>=4.14,<5
+Requires-Dist: fonttools>=4.60,<5
+Requires-Dist: numpy>=2.2,<3
+Requires-Dist: rich>=14,<16
+Requires-Dist: requests>=2.32,<3
+Requires-Dist: platformdirs>=4.4,<5
+Requires-Dist: toml>=0.10.2,<0.11
+Requires-Dist: transformers>=4.55,<6 ; extra == 'sentiment'
+Requires-Dist: torch>=2.8,<3 ; extra == 'sentiment'
+Requires-Dist: sentencepiece>=0.2,<0.3 ; extra == 'sentiment'
+Requires-Dist: protobuf>=5,<7 ; extra == 'sentiment'
+Requires-Python: >=3.10, <3.15
+Project-URL: Original, https://github.com/LuisMayo/objection_engine
+Provides-Extra: sentiment
+Description-Content-Type: text/markdown
+
+# objectionengine2
+
+objectionengine2 converts comment chains and scripted dialogue into Ace Attorney style videos. It is an independent continuation of [Objection Engine](https://github.com/LuisMayo/objection_engine), originally created by Luis Mayo Valbuena. The original copyright and MIT license are preserved in `LICENSE`.
+
+The distribution, repository, and Python import are all named `objectionengine2`. Update existing `objection_engine` imports when migrating scripts.
+
+## Requirements
+
+- Python 3.10 through 3.14
+- FFmpeg and `ffprobe` on `PATH`
+- The Objection Engine asset pack. The renderer downloads it on first use into
+  the user cache directory selected by `platformdirs` for `objectionengine2`.
+  An existing complete `./assets/` pack is copied there on first use.
+
+## Install
+
+From a clone, sync the locked environment with `uv`:
+
+```bash
+uv sync --python 3.14 --locked
+```
+
+The package supports Python 3.10–3.14; `.python-version` selects 3.14 by default. To enable Hugging Face sentiment, add `--extra sentiment` to the run command, for example `uv run --locked --extra sentiment python examples/example_easy.py`. This downloads a model on first use into the same user cache under `models/`. The base installation does not need a model. Find the cache location with `uv run python -c 'from objectionengine2.cache import cache_path; print(cache_path())'`.
+
+## Render comments
+
+```python
+from objectionengine2.beans.comment import Comment
+from objectionengine2.renderer import render_comment_list
+
+comments = [
+    Comment(user_name="Phoenix", text_content="The defense is ready."),
+    Comment(user_name="Edgeworth", text_content="The prosecution is ready."),
+]
+render_comment_list(comments, output_filename="outputs/output.mp4")
+```
+
+Run `uv run python examples/example_easy.py` from the repository root for a complete example. Additional scene, XML, and casting examples are in `examples/`; reference documentation is in `docs/`. Pass an `output_filename` in an existing directory. Scripts with literal `assets/...` paths still need a local asset directory; generated comment dialogue uses the user cache.
+
+## Sentiment settings
+
+By default, dialogue uses neutral sentiment and needs no machine learning dependency. Explicit `Comment(score=...)` values still select positive or negative sprites. Set `oe_sentiment_model=hf` after installing the `sentiment` extra to use the Hugging Face model. Setting `oe_bypass_sentiment` to a nonempty value forces neutral sentiment. The former Polyglot mode is no longer supported; an unsupported `oe_sentiment_model` value raises a clear error.
+
+Set `OE_DIRECT_H264_ENCODING=true` only if your OpenCV build provides an H.264 encoder. The default path encodes the intermediate video with OpenCV and creates the final H.264/AAC MP4 with FFmpeg.
+
+## Container (Podman)
+
+The Docker image uses Python 3.14 and includes FFmpeg. Build and run the included example from the repository root:
+
+```bash
+podman build -t objectionengine2 .
+podman volume create objectionengine2-cache
+podman run --rm \
+  --mount type=volume,src=objectionengine2-cache,dst=/root/.cache/objectionengine2 \
+  --mount type=bind,src="$(pwd)/outputs",dst=/app/outputs \
+  objectionengine2
+```
+
+The result is `outputs/output.mp4`. Ensure `outputs/` exists before running;
+the image saves downloaded assets and models in the mounted cache volume.
+To run your own script, mount it at `/app/render.py` and add `render.py` after
+the image name. Have the script write into `outputs/`.
+
+## Tests
+
+`uv sync --locked` installs the default `dev` group, including pytest. Run `uv run --locked pytest -q` for offline tests. They generate their own small media inputs and require FFmpeg and `ffprobe` but do not download assets or models.
+
+For a complete courtroom smoke test, run `uv run --locked pytest -q -m asset_integration`. The test calls `ensure_assets_are_available()` from `objectionengine2.utils` first, which populates the platform cache if needed. Keep that cache to avoid downloading the pack again.
+
+## Contributing
+
+Changes and fixes are welcome. Please run the offline suite before opening a pull request.

objectionengine2-1.0.0/README.md

@@ -0,0 +1,72 @@
+# objectionengine2
+
+objectionengine2 converts comment chains and scripted dialogue into Ace Attorney style videos. It is an independent continuation of [Objection Engine](https://github.com/LuisMayo/objection_engine), originally created by Luis Mayo Valbuena. The original copyright and MIT license are preserved in `LICENSE`.
+
+The distribution, repository, and Python import are all named `objectionengine2`. Update existing `objection_engine` imports when migrating scripts.
+
+## Requirements
+
+- Python 3.10 through 3.14
+- FFmpeg and `ffprobe` on `PATH`
+- The Objection Engine asset pack. The renderer downloads it on first use into
+  the user cache directory selected by `platformdirs` for `objectionengine2`.
+  An existing complete `./assets/` pack is copied there on first use.
+
+## Install
+
+From a clone, sync the locked environment with `uv`:
+
+```bash
+uv sync --python 3.14 --locked
+```
+
+The package supports Python 3.10–3.14; `.python-version` selects 3.14 by default. To enable Hugging Face sentiment, add `--extra sentiment` to the run command, for example `uv run --locked --extra sentiment python examples/example_easy.py`. This downloads a model on first use into the same user cache under `models/`. The base installation does not need a model. Find the cache location with `uv run python -c 'from objectionengine2.cache import cache_path; print(cache_path())'`.
+
+## Render comments
+
+```python
+from objectionengine2.beans.comment import Comment
+from objectionengine2.renderer import render_comment_list
+
+comments = [
+    Comment(user_name="Phoenix", text_content="The defense is ready."),
+    Comment(user_name="Edgeworth", text_content="The prosecution is ready."),
+]
+render_comment_list(comments, output_filename="outputs/output.mp4")
+```
+
+Run `uv run python examples/example_easy.py` from the repository root for a complete example. Additional scene, XML, and casting examples are in `examples/`; reference documentation is in `docs/`. Pass an `output_filename` in an existing directory. Scripts with literal `assets/...` paths still need a local asset directory; generated comment dialogue uses the user cache.
+
+## Sentiment settings
+
+By default, dialogue uses neutral sentiment and needs no machine learning dependency. Explicit `Comment(score=...)` values still select positive or negative sprites. Set `oe_sentiment_model=hf` after installing the `sentiment` extra to use the Hugging Face model. Setting `oe_bypass_sentiment` to a nonempty value forces neutral sentiment. The former Polyglot mode is no longer supported; an unsupported `oe_sentiment_model` value raises a clear error.
+
+Set `OE_DIRECT_H264_ENCODING=true` only if your OpenCV build provides an H.264 encoder. The default path encodes the intermediate video with OpenCV and creates the final H.264/AAC MP4 with FFmpeg.
+
+## Container (Podman)
+
+The Docker image uses Python 3.14 and includes FFmpeg. Build and run the included example from the repository root:
+
+```bash
+podman build -t objectionengine2 .
+podman volume create objectionengine2-cache
+podman run --rm \
+  --mount type=volume,src=objectionengine2-cache,dst=/root/.cache/objectionengine2 \
+  --mount type=bind,src="$(pwd)/outputs",dst=/app/outputs \
+  objectionengine2
+```
+
+The result is `outputs/output.mp4`. Ensure `outputs/` exists before running;
+the image saves downloaded assets and models in the mounted cache volume.
+To run your own script, mount it at `/app/render.py` and add `render.py` after
+the image name. Have the script write into `outputs/`.
+
+## Tests
+
+`uv sync --locked` installs the default `dev` group, including pytest. Run `uv run --locked pytest -q` for offline tests. They generate their own small media inputs and require FFmpeg and `ffprobe` but do not download assets or models.
+
+For a complete courtroom smoke test, run `uv run --locked pytest -q -m asset_integration`. The test calls `ensure_assets_are_available()` from `objectionengine2.utils` first, which populates the platform cache if needed. Keep that cache to avoid downloading the pack again.
+
+## Contributing
+
+Changes and fixes are welcome. Please run the offline suite before opening a pull request.