interactive-pipe 0.7.4__tar.gz → 0.7.6__tar.gz

{interactive_pipe-0.7.4/src/interactive_pipe.egg-info → interactive_pipe-0.7.6}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: interactive_pipe
-Version: 0.7.4
+Version: 0.7.6
 Summary: Library to create flexible interactive image processing pipelines and automatically add a graphical user interface without knowing anything about GUI coding!
 Author-email: Balthazar Neveu <balthazarneveu@gmail.com>
 Project-URL: Homepage, https://github.com/balthazarneveu/interactive_pipe
@@ -100,10 +100,13 @@ pip install -e ".[full]"
 
 
 ----------------------
+## :scroll:  Terminology
+
+![interactive_pipe_concept](static/interact-pipe-concept.svg)
 
 ## :scroll:  Features
 
-**Version 0.7.4**
+**Version 0.7.6**
 - Modular multi-image processing filters
 - Declarative: Easily make graphical user interface without having to learn anything about pyQt or matplotlib
 - Support in jupyter notebooks
@@ -147,7 +150,7 @@ Shortcuts while using the GUI (QT & matplotlib backends)
 | :star: | *PyQt / PySide* |  *Matplotlib*   | *Jupyter notebooks including Google collab*   | *Gradio* | 
 |:-----: |:-----:|:------:|:----: |:----: |
 | Backend name | `qt`  | `mpl`  | `nb`| `gradio` |
-| Preview | ![qt backend](/doc/images/qt_backend.jpg)   | ![mpl backend](/doc/images/mpl_backend.jpg)    |  ![nb backend](/doc/images/notebook_backend.jpg)  | | 
+| Preview | ![qt backend](/doc/images/qt_backend.jpg)   | ![mpl backend](/doc/images/mpl_backend.jpg)    |  ![nb backend](/doc/images/notebook_backend.jpg)  | ![mpl backend](/doc/images/gradio_backend.jpg)| 
 | Plot curves | :heavy_check_mark: | :heavy_check_mark: |   :heavy_check_mark: | :heavy_check_mark:|
 | Auto refreshed layout | :heavy_check_mark: | :heavy_check_mark: |   :heavy_check_mark: | :heavy_minus_sign: |
 | Keyboard shortcuts / fullscreen| :heavy_check_mark: | :heavy_check_mark: |  :heavy_minus_sign: | :heavy_minus_sign: |
@@ -158,26 +161,29 @@ Shortcuts while using the GUI (QT & matplotlib backends)
 
 
 # Tutorials
+### [Main tutorial](https://huggingface.co/spaces/balthou/interactive-pipe-tutorial)
+![tuto](/doc/images/tutorial.png)
 
 ### Learn by examples
-#### [Basic image processing](/demo/multi_image.py)
+#### [Basic image processing (python code sample for PyQT GUI)](/demo/multi_image.py)
 | GUI | Pipeline | 
 |:--:|:--:| 
 |![](/doc/images/image_selector.png) | ![](/doc/images/image_selector_graph.png) |
 
 ![](/doc/images/demo_multi_image.gif)
 
-#### [Speech exploration notebook](https://colab.research.google.com/drive/1mUX2FW0qflWn-v3nIx90P_KvRxnXlBpz#scrollTo=qDTaIwvaJQ6R)
+#### [Speech exploration notebook (colab, signal processing)](https://colab.research.google.com/drive/1mUX2FW0qflWn-v3nIx90P_KvRxnXlBpz#scrollTo=qDTaIwvaJQ6R)
 ![Speech processing exploration in a notebook](/doc/images/jupyter_integration_speech.png)
 
 
 
 
+## :rocket: Ultra short code
+
 
-# Tutorials
 *Since ipywidgets in notebooks are supported, the tutorial is also available in a [google collab notebook](https://colab.research.google.com/drive/1PZn8P_5TABVCugT3IcLespvZG-gxnFbO?usp=sharing)*
 
-## :rocket: Ultra short code
+
 Let's define 3 image processing very basic filters `exposure`, `black_and_white` & `blend`.
 
 By design:
@@ -253,11 +259,11 @@ import numpy as np
 
 COLOR_DICT = {"red": [1., 0., 0.],  "green": [0., 1.,0.], "blue": [0., 0., 1.], "gray": [0.5, 0.5, 0.5]}
 @interactive()
-def generate_flat_colored_image(color_choice=["red", "green", "blue", "gray"], global_params={}):
+def generate_flat_colored_image(color_choice=["red", "green", "blue", "gray"], context={}):
     '''Generate a constant colorful image
     '''
     flat_array =  np.array(COLOR_DICT.get(color_choice)) * np.ones((64, 64, 3))
-    global_params["avg"] = np.average(flat_array)
+    context["avg"] = np.average(flat_array)
     return flat_array
 ```
 
@@ -266,15 +272,15 @@ def generate_flat_colored_image(color_choice=["red", "green", "blue", "gray"], g
 ----------
 
 :bulb: Can filters communicate together?
-Yes, using the special keyword argument `global_params={}`. 
-- Check carefully how we stored the image average of the flat image in  global_params. 
+Yes, using the special keyword argument `context={}`. 
+- Check carefully how we stored the image average of the flat image in  context. 
 - This value will be available to other filters.
 `special_image_slice` is going to use that value to set the half bottom image to dark in case the average is high.
 
 ```python
 @interactive()
-def special_image_slice(img, global_params={}):
-    if global_params["avg"] > 0.4:
+def special_image_slice(img, context={}):
+    if context["avg"] > 0.4:
         out_img[out_img.shape[0]//2:, ...] = 0.
     return out_img
 ```
@@ -292,7 +298,7 @@ Note that you can create a filter to switch between several images. In `["pagedo
 
 ```python
 @interactive()
-def black_top_image_slice(img, top_slice_black=(True, "special", "k"), global_params={}):
+def black_top_image_slice(img, top_slice_black=(True, "special", "k"), context={}):
     out_img = img.copy()
     if top_slice_black:
         out_img[:out_img.shape[0]//2, ...] = 0.
@@ -320,7 +326,47 @@ if __name__ == '__main__':
 - September 2024: Gradio backend
 
 
+### FAQ
+- :question: Is there a difference between `global_params` and `context` ?
+> No, `global_params`, `global_parameters`, `global_state`, `global_context`, `context`, `state` all mean the same thing and are all supported for legacy reasons. `context` is the preferred wording.
+- :question: Do I have to remove `KeyboardSlider` when using gradio or notebook backends?
+> No, don't worry, these will be mapped back to regular sliders!
+- :question: How do I play audio live?
+> :sound: Inside a processing block, write the audio file to disk 
+> and use `context["__set_audio"](audio_file)`
+- :question: Do I have to decorate my processing block using the `@interactive` 
+> If you use the `@` decoration style, your function won't be useable in a regular manner (wich may be problematic in a serious development environment)
+```python
+@interactive(angle=(0., [-360., 360.]))
+def processing_block(angle=0.):
+    ...
+```
 
+> An alternative is to decorate the processing block outside... in a file dedicated to interactivity for instance
+```python
+# core_filter.py
+def processing_block(angle=0.):
+    ...
+```
+
+```python
+# graphical.py
+from core_filter import processing_block
+
+def add_interactivity():
+    interactive(angle=(0., [-360., 360.]))(processing_block)
+```
+- :question: Can I call the pipeline in a command line/batch fashion?
+> Yes, headless mode is supported. :soon: documentation needed.
+
+
+- :question: Can I use inplace operations?
+> Better avoid these in general. To avoid making extra copies, computing hashes everywhere and avoid loosing precious computation time, there are no checks that inputs are not modified in place.
+```python
+# Don't do that!
+def bad_processing_block(inp):
+    inp+=1
+``` 
 # Roadmap and todos
 :bug: Want to contribute or interested in adding new features? Enter a new [Github issue](https://github.com/balthazarneveu/interactive_pipe/issues)
 

{interactive_pipe-0.7.4 → interactive_pipe-0.7.6}/pyproject.toml

@@ -5,7 +5,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "interactive_pipe"
-version = "0.7.4"
+version = "0.7.6"
 authors = [
   { name="Balthazar Neveu", email="balthazarneveu@gmail.com" },
 ]

{interactive_pipe-0.7.4 → interactive_pipe-0.7.6}/readme.md

@@ -56,10 +56,13 @@ pip install -e ".[full]"
 
 
 ----------------------
+## :scroll:  Terminology
+
+![interactive_pipe_concept](static/interact-pipe-concept.svg)
 
 ## :scroll:  Features
 
-**Version 0.7.4**
+**Version 0.7.6**
 - Modular multi-image processing filters
 - Declarative: Easily make graphical user interface without having to learn anything about pyQt or matplotlib
 - Support in jupyter notebooks
@@ -103,7 +106,7 @@ Shortcuts while using the GUI (QT & matplotlib backends)
 | :star: | *PyQt / PySide* |  *Matplotlib*   | *Jupyter notebooks including Google collab*   | *Gradio* | 
 |:-----: |:-----:|:------:|:----: |:----: |
 | Backend name | `qt`  | `mpl`  | `nb`| `gradio` |
-| Preview | ![qt backend](/doc/images/qt_backend.jpg)   | ![mpl backend](/doc/images/mpl_backend.jpg)    |  ![nb backend](/doc/images/notebook_backend.jpg)  | | 
+| Preview | ![qt backend](/doc/images/qt_backend.jpg)   | ![mpl backend](/doc/images/mpl_backend.jpg)    |  ![nb backend](/doc/images/notebook_backend.jpg)  | ![mpl backend](/doc/images/gradio_backend.jpg)| 
 | Plot curves | :heavy_check_mark: | :heavy_check_mark: |   :heavy_check_mark: | :heavy_check_mark:|
 | Auto refreshed layout | :heavy_check_mark: | :heavy_check_mark: |   :heavy_check_mark: | :heavy_minus_sign: |
 | Keyboard shortcuts / fullscreen| :heavy_check_mark: | :heavy_check_mark: |  :heavy_minus_sign: | :heavy_minus_sign: |
@@ -114,26 +117,29 @@ Shortcuts while using the GUI (QT & matplotlib backends)
 
 
 # Tutorials
+### [Main tutorial](https://huggingface.co/spaces/balthou/interactive-pipe-tutorial)
+![tuto](/doc/images/tutorial.png)
 
 ### Learn by examples
-#### [Basic image processing](/demo/multi_image.py)
+#### [Basic image processing (python code sample for PyQT GUI)](/demo/multi_image.py)
 | GUI | Pipeline | 
 |:--:|:--:| 
 |![](/doc/images/image_selector.png) | ![](/doc/images/image_selector_graph.png) |
 
 ![](/doc/images/demo_multi_image.gif)
 
-#### [Speech exploration notebook](https://colab.research.google.com/drive/1mUX2FW0qflWn-v3nIx90P_KvRxnXlBpz#scrollTo=qDTaIwvaJQ6R)
+#### [Speech exploration notebook (colab, signal processing)](https://colab.research.google.com/drive/1mUX2FW0qflWn-v3nIx90P_KvRxnXlBpz#scrollTo=qDTaIwvaJQ6R)
 ![Speech processing exploration in a notebook](/doc/images/jupyter_integration_speech.png)
 
 
 
 
+## :rocket: Ultra short code
+
 
-# Tutorials
 *Since ipywidgets in notebooks are supported, the tutorial is also available in a [google collab notebook](https://colab.research.google.com/drive/1PZn8P_5TABVCugT3IcLespvZG-gxnFbO?usp=sharing)*
 
-## :rocket: Ultra short code
+
 Let's define 3 image processing very basic filters `exposure`, `black_and_white` & `blend`.
 
 By design:
@@ -209,11 +215,11 @@ import numpy as np
 
 COLOR_DICT = {"red": [1., 0., 0.],  "green": [0., 1.,0.], "blue": [0., 0., 1.], "gray": [0.5, 0.5, 0.5]}
 @interactive()
-def generate_flat_colored_image(color_choice=["red", "green", "blue", "gray"], global_params={}):
+def generate_flat_colored_image(color_choice=["red", "green", "blue", "gray"], context={}):
     '''Generate a constant colorful image
     '''
     flat_array =  np.array(COLOR_DICT.get(color_choice)) * np.ones((64, 64, 3))
-    global_params["avg"] = np.average(flat_array)
+    context["avg"] = np.average(flat_array)
     return flat_array
 ```
 
@@ -222,15 +228,15 @@ def generate_flat_colored_image(color_choice=["red", "green", "blue", "gray"], g
 ----------
 
 :bulb: Can filters communicate together?
-Yes, using the special keyword argument `global_params={}`. 
-- Check carefully how we stored the image average of the flat image in  global_params. 
+Yes, using the special keyword argument `context={}`. 
+- Check carefully how we stored the image average of the flat image in  context. 
 - This value will be available to other filters.
 `special_image_slice` is going to use that value to set the half bottom image to dark in case the average is high.
 
 ```python
 @interactive()
-def special_image_slice(img, global_params={}):
-    if global_params["avg"] > 0.4:
+def special_image_slice(img, context={}):
+    if context["avg"] > 0.4:
         out_img[out_img.shape[0]//2:, ...] = 0.
     return out_img
 ```
@@ -248,7 +254,7 @@ Note that you can create a filter to switch between several images. In `["pagedo
 
 ```python
 @interactive()
-def black_top_image_slice(img, top_slice_black=(True, "special", "k"), global_params={}):
+def black_top_image_slice(img, top_slice_black=(True, "special", "k"), context={}):
     out_img = img.copy()
     if top_slice_black:
         out_img[:out_img.shape[0]//2, ...] = 0.
@@ -276,7 +282,47 @@ if __name__ == '__main__':
 - September 2024: Gradio backend
 
 
+### FAQ
+- :question: Is there a difference between `global_params` and `context` ?
+> No, `global_params`, `global_parameters`, `global_state`, `global_context`, `context`, `state` all mean the same thing and are all supported for legacy reasons. `context` is the preferred wording.
+- :question: Do I have to remove `KeyboardSlider` when using gradio or notebook backends?
+> No, don't worry, these will be mapped back to regular sliders!
+- :question: How do I play audio live?
+> :sound: Inside a processing block, write the audio file to disk 
+> and use `context["__set_audio"](audio_file)`
+- :question: Do I have to decorate my processing block using the `@interactive` 
+> If you use the `@` decoration style, your function won't be useable in a regular manner (wich may be problematic in a serious development environment)
+```python
+@interactive(angle=(0., [-360., 360.]))
+def processing_block(angle=0.):
+    ...
+```
 
+> An alternative is to decorate the processing block outside... in a file dedicated to interactivity for instance
+```python
+# core_filter.py
+def processing_block(angle=0.):
+    ...
+```
+
+```python
+# graphical.py
+from core_filter import processing_block
+
+def add_interactivity():
+    interactive(angle=(0., [-360., 360.]))(processing_block)
+```
+- :question: Can I call the pipeline in a command line/batch fashion?
+> Yes, headless mode is supported. :soon: documentation needed.
+
+
+- :question: Can I use inplace operations?
+> Better avoid these in general. To avoid making extra copies, computing hashes everywhere and avoid loosing precious computation time, there are no checks that inputs are not modified in place.
+```python
+# Don't do that!
+def bad_processing_block(inp):
+    inp+=1
+``` 
 # Roadmap and todos
 :bug: Want to contribute or interested in adding new features? Enter a new [Github issue](https://github.com/balthazarneveu/interactive_pipe/issues)
 

{interactive_pipe-0.7.4 → interactive_pipe-0.7.6}/src/interactive_pipe/core/filter.py

@@ -5,6 +5,8 @@ from typing import Callable, List, Optional, Union, Tuple, Any
 from interactive_pipe.core.cache import CachedResults
 from interactive_pipe.core.signature import analyze_apply_fn_signature
 
+EQUIVALENT_STATE_KEYS = ["global_params", "global_parameters", "global_state", "global_context", "context", "state"]
+
 
 class PureFilter:
     def __init__(self, apply_fn: Optional[Callable] = None, name: Optional[str] = None, default_params: dict = {}):
@@ -12,6 +14,7 @@ class PureFilter:
             self.__class__.__name__ if not apply_fn else apply_fn.__name__)
         if apply_fn is not None:
             self.apply = apply_fn
+        self._global_params = {}
         self.__initialize_default_values()  # initialize default values from .apply method
         self.values = deepcopy(default_params)
 
@@ -40,8 +43,9 @@ class PureFilter:
         assert not hasattr(self, "_values")
         self.check_apply_signature()
         self._values = self.__kwargs_names
-        if "global_params" in self.__kwargs_names.keys():
-            self._values.pop("global_params")
+        for global_key in EQUIVALENT_STATE_KEYS:
+            if global_key in self.__kwargs_names.keys():
+                self._values.pop(global_key)
 
     @property
     def values(self):
@@ -60,11 +64,15 @@ class PureFilter:
         for key, val in self.values.items():
             assert key in self.__kwargs_names.keys(
             ), f"{self.name} : {key} not in {self.__kwargs_names.keys()}"
-        if "global_params" in self.__kwargs_names.keys():
-            # special key to provide the context dictionary
-            out = self.apply(
-                *imgs, global_params=self.global_params, **self.values)
-        else:
+        global_key_found = False
+        for global_key in EQUIVALENT_STATE_KEYS:
+            if global_key in self.__kwargs_names.keys():
+                # special key to provide the context dictionary
+                out = self.apply(
+                    *imgs, **{**{global_key: self.global_params}, **self.values})
+                global_key_found = True
+                break
+        if not global_key_found:
             out = self.apply(*imgs, **self.values)
         return out
 

{interactive_pipe-0.7.4 → interactive_pipe-0.7.6}/src/interactive_pipe/core/pipeline.py

@@ -11,13 +11,40 @@ class PipelineCore:
     - optionally, some inputs to process
     """
 
-    def __init__(self, filters: List[FilterCore], name="pipeline", cache=False, inputs: Optional[list] = None, parameters: dict = {}, global_params={}, outputs:  Optional[list] = None, safe_input_buffer_deepcopy: bool = True):
+    def __init__(
+        self,
+        filters: List[FilterCore],
+        name="pipeline",
+        cache=False,
+        inputs: Optional[list] = None,
+        parameters: dict = {},
+        global_params: Optional[dict] = None,
+        global_parameters: Optional[dict] = None,  # alias for global_params
+        global_state: Optional[dict] = None,  # alias for global_params
+        global_context: Optional[dict] = None,  # alias for global_params
+        context: Optional[dict] = None,  # alias for global_params
+        state: Optional[dict] = None,  # alias for global_params
+        outputs:  Optional[list] = None,
+        safe_input_buffer_deepcopy: bool = True
+    ):
         if not all(isinstance(f, FilterCore) for f in filters):
             raise ValueError(
                 f"All elements in 'filters' must be instances of 'Filter'. {[type(f) for f in filters]}")
         self.filters = filters
         self.engine = PipelineEngine(
             cache, safe_input_buffer_deepcopy=safe_input_buffer_deepcopy)
+        if global_parameters is not None:
+            global_params = global_parameters
+        elif global_context is not None:
+            global_params = global_context
+        elif global_state is not None:
+            global_params = global_state
+        elif state is not None:
+            global_params = state
+        elif context is not None:
+            global_params = context
+        if global_params is None:
+            global_params = {}
         self.global_params = global_params
         for filter in self.filters:
             # link each filter to global params
@@ -79,7 +106,8 @@ class PipelineCore:
         """
         available_filters_names = [filt.name for filt in self.filters]
         for filter_name in new_parameters.keys():
-            assert filter_name in available_filters_names, f"filter {filter_name} does not exist {available_filters_names}"
+            assert filter_name in available_filters_names, f"filter {filter_name}" + \
+                f"does not exist {available_filters_names}"
             self.filters[available_filters_names.index(
                 filter_name)].values = new_parameters[filter_name]
 
@@ -99,7 +127,8 @@ class PipelineCore:
             elif isinstance(inputs, list) or isinstance(inputs, tuple):
                 # inputs is a list or a tuple
                 assert len(inputs) == len(
-                    self.inputs_routing), f"wrong amount of inputs\nprovided {len(inputs)} inputs vs expected:{len(self.inputs_routing)}"
+                    self.inputs_routing), f"wrong amount of inputs\nprovided {len(inputs)}" + \
+                    f"inputs vs expected:{len(self.inputs_routing)}"
                 self.__inputs = {}
                 for idx, input_name in enumerate(self.inputs_routing):
                     self.__inputs[input_name] = inputs[idx]

{interactive_pipe-0.7.4 → interactive_pipe-0.7.6}/src/interactive_pipe/graphical/gradio_gui.py

@@ -51,14 +51,20 @@ class InteractivePipeGradio(InteractivePipeGUI):
         out_list_gradio_containers = []
         for idx in range(len(out_list)):
             for idy in range(len(out_list[idx])):
+                if self.window.image_canvas[idx][idy] is None:
+                    out_list_gradio_containers.append(gr.HTML())
+                    continue
                 title = self.window.image_canvas[idx][idy].get("title", f"{idx} {idy}")
                 title = title.replace("_", " ")
                 if isinstance(out_list[idx][idy], np.ndarray):
                     out_list_gradio_containers.append(gr.Image(label=title))
+                    self.window.image_canvas[idx][idy]["type"] = "image"
                 elif MPL_SUPPORT and isinstance(out_list[idx][idy], Curve):
                     # if out_list[idx][idy].title is not None:
                     #     title = out_list[idx][idy].title
                     out_list_gradio_containers.append(gr.Plot(label=title))
+                    self.window.image_canvas[idx][idy]["type"] = "curve"
+                # @TODO: https://github.com/balthazarneveu/interactive_pipe/issues/50 support audio!
                 else:
                     raise NotImplementedError(
                         f"output type {type(out_list[idx][idy])} not supported")
@@ -94,7 +100,9 @@ class MainWindow(InteractivePipeWindow):
             for idx in range(len(out)):
                 if isinstance(out[idx], list):
                     for idy in range(len(out[idx])):
-                        if MPL_SUPPORT and isinstance(out[idx][idy], Curve):
+                        if out[idx][idy] is None:
+                            flat_out.append("")
+                        elif MPL_SUPPORT and isinstance(out[idx][idy], Curve):
                             curve = out[idx][idy]
                             fig, ax = plt.subplots()
                             Curve._plot_curve(curve.data, ax=ax)
@@ -104,10 +112,12 @@ class MainWindow(InteractivePipeWindow):
                             flat_out.append(self.convert_image(out[idx][idy]))
                         else:
                             raise NotImplementedError(
-                                f"output type {type(out[idx][idy])} not supported")
+                                f"output type {type(out[idx][idy])} not suported")
                 else:
                     logging.info(f"CONVERTING IMAGE  {idx} ")
                     flat_out.append(self.convert_image(out[idx]))
+            if len(flat_out) == 1:
+                return flat_out[0]
             if out is None:
                 logging.warning("No output to display")
                 return
@@ -152,9 +162,13 @@ class MainWindow(InteractivePipeWindow):
             with gr.Blocks() as io:
                 with gr.Row(variant="compact"):
                     gr.Markdown("### " + self.name.replace("_", " "))
-                with gr.Row():
-                    for elem in outputs:
-                        elem.render()
+                for idy in range(len(self.image_canvas)):
+                    with gr.Row():
+                        for idx in range(len(self.image_canvas[idy])):
+                            elem = outputs[idy * len(self.image_canvas[idy]) + idx]
+                            if elem is not None:
+                                elem.render()
+                                
                 if self.audio:
                     outputs.append(self.audio)
                     with gr.Row():

interactive_pipe-0.7.6/src/interactive_pipe/helper/pipeline_decorator.py

@@ -0,0 +1,57 @@
+from interactive_pipe.headless.pipeline import HeadlessPipeline
+from interactive_pipe.graphical.gui import InteractivePipeGUI
+
+import functools
+from typing import Callable, Union, Optional
+from interactive_pipe.helper.choose_backend import get_interactive_pipeline_class
+
+None
+
+
+def pipeline(pipeline_function: Callable, **kwargs) -> HeadlessPipeline:
+    return HeadlessPipeline.from_function(pipeline_function, **kwargs)
+
+
+def interactive_pipeline(
+    gui="auto",
+    safe_input_buffer_deepcopy=True,
+    cache=False,
+    output_canvas=None,
+    global_params: Optional[dict] = None,
+    global_parameters: Optional[dict] = None,  # alias for global_params
+    global_state: Optional[dict] = None,  # alias for global_params
+    global_context: Optional[dict] = None,  # alias for global_params
+    context: Optional[dict] = None,  # alias for global_params
+    state: Optional[dict] = None,  # alias for global_params
+    **kwargs_gui,
+) -> Union[HeadlessPipeline, InteractivePipeGUI]:
+    """@interactive_pipeline
+
+    Function decorator to add some controls @interactive_pipeline
+    """
+    def wrapper(pipeline_function):
+        headless_pipeline = HeadlessPipeline.from_function(
+            pipeline_function,
+            safe_input_buffer_deepcopy=safe_input_buffer_deepcopy,
+            cache=cache,
+            global_params=global_params,
+            global_parameters=global_parameters,  # alias for global_params
+            global_state=global_state,  # alias for global_params
+            global_context=global_context,  # alias for global_params
+            context=context,  # alias for global_params
+            state=state,  # alias for global_params
+        )
+        if output_canvas is not None:
+            headless_pipeline.outputs = output_canvas
+        if gui is None:
+            return headless_pipeline
+        else:
+            InteractivePipeGui = get_interactive_pipeline_class(gui)
+            gui_pipeline = InteractivePipeGui(
+                pipeline=headless_pipeline, name=pipeline_function.__name__, **kwargs_gui)
+
+        @ functools.wraps(pipeline_function)
+        def inner(*args, **kwargs):
+            return gui_pipeline.__call__(*args, **kwargs)
+        return inner
+    return wrapper

{interactive_pipe-0.7.4 → interactive_pipe-0.7.6}/src/interactive_pipe/thirdparty/images_openai_api.py

@@ -6,7 +6,12 @@ from typing import Union
 from interactive_pipe.data_objects.parameters import Parameters
 from interactive_pipe.data_objects.image import Image
 import os
-import openai
+OPENAI_SUPPORT = True
+try:
+    import openai
+except ImportError:
+    OPENAI_SUPPORT = False
+    logging.warning("openai not installed, please install it with pip install openai")
 
 
 class ImageFromPrompt(Image):
@@ -57,6 +62,7 @@ class ImageFromPrompt(Image):
 
     @staticmethod
     def login(api_key=None, organization=None):
+        assert OPENAI_SUPPORT, "openai not installed"
         if api_key is not None:
             openai.api_key = api_key
             openai.organization = organization
@@ -75,6 +81,7 @@ class ImageFromPrompt(Image):
 
     @staticmethod
     def __check_login():
+        assert OPENAI_SUPPORT, "openai not installed"
         assert openai.api_key is not None
         try:
             response = openai.Model.list()
@@ -95,6 +102,7 @@ class ImageFromPrompt(Image):
                 path = Path(path)
                 path = path.with_suffix(ImageFromPrompt.SUFFIX)
             return path
+        assert OPENAI_SUPPORT, "openai not installed"
         try:
             # You need the API key at this moment
             ImageFromPrompt.login(api_key=api_key)

{interactive_pipe-0.7.4 → interactive_pipe-0.7.6/src/interactive_pipe.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: interactive_pipe
-Version: 0.7.4
+Version: 0.7.6
 Summary: Library to create flexible interactive image processing pipelines and automatically add a graphical user interface without knowing anything about GUI coding!
 Author-email: Balthazar Neveu <balthazarneveu@gmail.com>
 Project-URL: Homepage, https://github.com/balthazarneveu/interactive_pipe
@@ -100,10 +100,13 @@ pip install -e ".[full]"
 
 
 ----------------------
+## :scroll:  Terminology
+
+![interactive_pipe_concept](static/interact-pipe-concept.svg)
 
 ## :scroll:  Features
 
-**Version 0.7.4**
+**Version 0.7.6**
 - Modular multi-image processing filters
 - Declarative: Easily make graphical user interface without having to learn anything about pyQt or matplotlib
 - Support in jupyter notebooks
@@ -147,7 +150,7 @@ Shortcuts while using the GUI (QT & matplotlib backends)
 | :star: | *PyQt / PySide* |  *Matplotlib*   | *Jupyter notebooks including Google collab*   | *Gradio* | 
 |:-----: |:-----:|:------:|:----: |:----: |
 | Backend name | `qt`  | `mpl`  | `nb`| `gradio` |
-| Preview | ![qt backend](/doc/images/qt_backend.jpg)   | ![mpl backend](/doc/images/mpl_backend.jpg)    |  ![nb backend](/doc/images/notebook_backend.jpg)  | | 
+| Preview | ![qt backend](/doc/images/qt_backend.jpg)   | ![mpl backend](/doc/images/mpl_backend.jpg)    |  ![nb backend](/doc/images/notebook_backend.jpg)  | ![mpl backend](/doc/images/gradio_backend.jpg)| 
 | Plot curves | :heavy_check_mark: | :heavy_check_mark: |   :heavy_check_mark: | :heavy_check_mark:|
 | Auto refreshed layout | :heavy_check_mark: | :heavy_check_mark: |   :heavy_check_mark: | :heavy_minus_sign: |
 | Keyboard shortcuts / fullscreen| :heavy_check_mark: | :heavy_check_mark: |  :heavy_minus_sign: | :heavy_minus_sign: |
@@ -158,26 +161,29 @@ Shortcuts while using the GUI (QT & matplotlib backends)
 
 
 # Tutorials
+### [Main tutorial](https://huggingface.co/spaces/balthou/interactive-pipe-tutorial)
+![tuto](/doc/images/tutorial.png)
 
 ### Learn by examples
-#### [Basic image processing](/demo/multi_image.py)
+#### [Basic image processing (python code sample for PyQT GUI)](/demo/multi_image.py)
 | GUI | Pipeline | 
 |:--:|:--:| 
 |![](/doc/images/image_selector.png) | ![](/doc/images/image_selector_graph.png) |
 
 ![](/doc/images/demo_multi_image.gif)
 
-#### [Speech exploration notebook](https://colab.research.google.com/drive/1mUX2FW0qflWn-v3nIx90P_KvRxnXlBpz#scrollTo=qDTaIwvaJQ6R)
+#### [Speech exploration notebook (colab, signal processing)](https://colab.research.google.com/drive/1mUX2FW0qflWn-v3nIx90P_KvRxnXlBpz#scrollTo=qDTaIwvaJQ6R)
 ![Speech processing exploration in a notebook](/doc/images/jupyter_integration_speech.png)
 
 
 
 
+## :rocket: Ultra short code
+
 
-# Tutorials
 *Since ipywidgets in notebooks are supported, the tutorial is also available in a [google collab notebook](https://colab.research.google.com/drive/1PZn8P_5TABVCugT3IcLespvZG-gxnFbO?usp=sharing)*
 
-## :rocket: Ultra short code
+
 Let's define 3 image processing very basic filters `exposure`, `black_and_white` & `blend`.
 
 By design:
@@ -253,11 +259,11 @@ import numpy as np
 
 COLOR_DICT = {"red": [1., 0., 0.],  "green": [0., 1.,0.], "blue": [0., 0., 1.], "gray": [0.5, 0.5, 0.5]}
 @interactive()
-def generate_flat_colored_image(color_choice=["red", "green", "blue", "gray"], global_params={}):
+def generate_flat_colored_image(color_choice=["red", "green", "blue", "gray"], context={}):
     '''Generate a constant colorful image
     '''
     flat_array =  np.array(COLOR_DICT.get(color_choice)) * np.ones((64, 64, 3))
-    global_params["avg"] = np.average(flat_array)
+    context["avg"] = np.average(flat_array)
     return flat_array
 ```
 
@@ -266,15 +272,15 @@ def generate_flat_colored_image(color_choice=["red", "green", "blue", "gray"], g
 ----------
 
 :bulb: Can filters communicate together?
-Yes, using the special keyword argument `global_params={}`. 
-- Check carefully how we stored the image average of the flat image in  global_params. 
+Yes, using the special keyword argument `context={}`. 
+- Check carefully how we stored the image average of the flat image in  context. 
 - This value will be available to other filters.
 `special_image_slice` is going to use that value to set the half bottom image to dark in case the average is high.
 
 ```python
 @interactive()
-def special_image_slice(img, global_params={}):
-    if global_params["avg"] > 0.4:
+def special_image_slice(img, context={}):
+    if context["avg"] > 0.4:
         out_img[out_img.shape[0]//2:, ...] = 0.
     return out_img
 ```
@@ -292,7 +298,7 @@ Note that you can create a filter to switch between several images. In `["pagedo
 
 ```python
 @interactive()
-def black_top_image_slice(img, top_slice_black=(True, "special", "k"), global_params={}):
+def black_top_image_slice(img, top_slice_black=(True, "special", "k"), context={}):
     out_img = img.copy()
     if top_slice_black:
         out_img[:out_img.shape[0]//2, ...] = 0.
@@ -320,7 +326,47 @@ if __name__ == '__main__':
 - September 2024: Gradio backend
 
 
+### FAQ
+- :question: Is there a difference between `global_params` and `context` ?
+> No, `global_params`, `global_parameters`, `global_state`, `global_context`, `context`, `state` all mean the same thing and are all supported for legacy reasons. `context` is the preferred wording.
+- :question: Do I have to remove `KeyboardSlider` when using gradio or notebook backends?
+> No, don't worry, these will be mapped back to regular sliders!
+- :question: How do I play audio live?
+> :sound: Inside a processing block, write the audio file to disk 
+> and use `context["__set_audio"](audio_file)`
+- :question: Do I have to decorate my processing block using the `@interactive` 
+> If you use the `@` decoration style, your function won't be useable in a regular manner (wich may be problematic in a serious development environment)
+```python
+@interactive(angle=(0., [-360., 360.]))
+def processing_block(angle=0.):
+    ...
+```
 
+> An alternative is to decorate the processing block outside... in a file dedicated to interactivity for instance
+```python
+# core_filter.py
+def processing_block(angle=0.):
+    ...
+```
+
+```python
+# graphical.py
+from core_filter import processing_block
+
+def add_interactivity():
+    interactive(angle=(0., [-360., 360.]))(processing_block)
+```
+- :question: Can I call the pipeline in a command line/batch fashion?
+> Yes, headless mode is supported. :soon: documentation needed.
+
+
+- :question: Can I use inplace operations?
+> Better avoid these in general. To avoid making extra copies, computing hashes everywhere and avoid loosing precious computation time, there are no checks that inputs are not modified in place.
+```python
+# Don't do that!
+def bad_processing_block(inp):
+    inp+=1
+``` 
 # Roadmap and todos
 :bug: Want to contribute or interested in adding new features? Enter a new [Github issue](https://github.com/balthazarneveu/interactive_pipe/issues)
 

{interactive_pipe-0.7.4 → interactive_pipe-0.7.6}/test/test_core.py

@@ -7,7 +7,7 @@ class MultiplyFilter(FilterCore):
 
 
 def test_pure_filter():
-    def mult(img, global_parameters={}, scalar=8, add=5):
+    def mult(img, context={}, scalar=8, add=5):
         return [[value * scalar + add for value in row] for row in img]
     filt = PureFilter(apply_fn=mult, name="multiplication",
                       default_params={"scalar": 2, "add": 0})

interactive_pipe-0.7.4/src/interactive_pipe/helper/pipeline_decorator.py

@@ -1,34 +0,0 @@
-from interactive_pipe.headless.pipeline import HeadlessPipeline
-from interactive_pipe.graphical.gui import InteractivePipeGUI
-
-import functools
-from typing import Callable, Union
-from interactive_pipe.helper.choose_backend import get_interactive_pipeline_class
-
-
-def pipeline(pipeline_function: Callable, **kwargs) -> HeadlessPipeline:
-    return HeadlessPipeline.from_function(pipeline_function, **kwargs)
-
-
-def interactive_pipeline(gui="auto", safe_input_buffer_deepcopy=True, cache=False, output_canvas=None, **kwargs_gui) -> Union[HeadlessPipeline, InteractivePipeGUI]:
-    """@interactive_pipeline
-
-    Function decorator to add some controls @interactive_pipeline
-    """
-    def wrapper(pipeline_function):
-        headless_pipeline = HeadlessPipeline.from_function(
-            pipeline_function, safe_input_buffer_deepcopy=safe_input_buffer_deepcopy, cache=cache)
-        if output_canvas is not None:
-            headless_pipeline.outputs = output_canvas
-        if gui is None:
-            return headless_pipeline
-        else:
-            InteractivePipeGui = get_interactive_pipeline_class(gui)
-            gui_pipeline = InteractivePipeGui(
-                pipeline=headless_pipeline, name=pipeline_function.__name__, **kwargs_gui)
-
-        @functools.wraps(pipeline_function)
-        def inner(*args, **kwargs):
-            return gui_pipeline.__call__(*args, **kwargs)
-        return inner
-    return wrapper