rp 0.1.1136__tar.gz → 0.1.1138__tar.gz

{rp-0.1.1136 → rp-0.1.1138}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 1.0
 Name: rp
-Version: 0.1.1136
+Version: 0.1.1138
 Summary: Ryan's Python
 Home-page: https://github.com/RyannDaGreat/Quick-Python
 Author: Ryan Burgert

{rp-0.1.1136 → rp-0.1.1138}/rp/r.py

@@ -29032,6 +29032,31 @@ def save_video_avi(frames,path:str=None,framerate:int=30):
 def _get_default_video_path(extension='mp4'):
     return get_unique_copy_path('video.'+extension)
 
+_named_video_bitrates = {
+    "low"    : 100000,
+    "medium" : 1000000,
+    "high"   : 10000000,
+    "max"    : 10000000000,
+}
+_named_video_qualities = {
+    "low"    : 10,
+    "medium" : 50,
+    "high"   : 95,
+    "max"    : 100,
+}
+def _as_video_bitrate(video_bitrate:str) -> int:
+    """ As a bitrate """
+    if video_bitrate in _named_video_bitrates:
+        video_bitrate = _named_video_bitrates[video_bitrate]
+    return video_bitrate
+def _as_video_quality(video_quality:str) -> int:
+    """ As a percent """
+    if video_quality in _named_video_qualities:
+        video_quality = _named_video_qualities[video_quality]
+    if video_quality > 100:
+        video_quality = 100
+    return video_quality
+
 class VideoWriterMP4:
     #Todo: If this ever gets fucky, try https://github.com/imageio/imageio-ffmpeg - it looks pretty good!
 
@@ -29047,8 +29072,7 @@ class VideoWriterMP4:
 
         rp.pip_import('ffmpeg', 'ffmpeg-python')
 
-        if isinstance(video_bitrate,str) and video_bitrate in 'low medium high max':
-            video_bitrate = {'low':100000,'medium':1000000,'high':10000000,'max':10000000000}[video_bitrate]
+        video_bitrate = _as_video_bitrate(video_bitrate)
 
         assert path.endswith('.mp4')
         assert isinstance(video_bitrate,int)
@@ -29131,23 +29155,136 @@ class VideoWriterMP4:
         self.process.wait()
         self.finished=True
         
-def save_video_mp4(frames, path=None, framerate=60, *, video_bitrate='high', height=None, width=None, show_progress=True):
+
+def _cv_save_video_mp4(
+    video,
+    path,
+    *,
+    framerate,
+    show_progress
+):
+    pip_import("cv2")
+    import cv2
+
+    video = as_rgb_images(video)
+    video = as_byte_images(video)
+    video = as_numpy_array(video)
+
+    if path is None:
+        path = get_unique_copy_path('video.mp4')
+
+    if not has_file_extension(path):
+        path = with_file_extension(path, 'mp4')
+
+    fourcc = cv2.VideoWriter_fourcc(*"mp4v")
+    _, h, w, c = video.shape
+    video_writer = cv2.VideoWriter(path, fourcc, fps=framerate, frameSize=(w, h))
+
+    video_iter = video
+    if show_progress:
+        video_iter = eta(video_iter, title='rp.r.'+get_current_function_name())
+
+    for frame in video_iter:
+        frame = cv_rgb_bgr_swap(frame)
+        video_writer.write(img)
+
+    return path
+
+def _cv_save_video_mp4(
+    video,
+    path,
+    *,
+    framerate,
+    show_progress,
+    quality
+):
+    """ 
+    Uses quality instead of bitrate because of how opencv works: https://docs.opencv.org/3.4/d4/d15/group__videoio__flags__base.html#gga41c5cfa7859ae542b71b1d33bbd4d2b4ace3320e146f4f95f3d58b32b0e1237b1 
+    NOTE: This quality parameter...doesn't seem to do anything. So it will be hidden in save_video_mp4's docstring for now.
+    """
+    pip_import("cv2")
+    import cv2
+
+    video = as_rgb_images(video)
+    video = as_byte_images(video)
+    video = as_numpy_array(video)
+
+    quality = _as_video_quality(quality)
+
+    if path is None:
+        path = get_unique_copy_path('video.mp4')
+
+    if not has_file_extension(path):
+        path = with_file_extension(path, 'mp4')
+
+    fourcc = cv2.VideoWriter_fourcc(*"mp4v")
+    _, h, w, c = video.shape
+    video_writer = cv2.VideoWriter(path, fourcc, framerate, (w, h))
+
+    # Set quality if provided (in percent)
+    if quality is not None:
+        video_writer.set(cv2.VIDEOWRITER_PROP_QUALITY, quality)
+
+    video_iter = video
+    if show_progress:
+        video_iter = eta(video_iter, title='rp.r.'+get_current_function_name())
+
+    for frame in video_iter:
+        frame = cv_rgb_bgr_swap(frame)
+        video_writer.write(frame)  # Fixed typo: 'img' -> 'frame'
+
+    video_writer.release()  # Ensure proper file closure
+
+    return path
+
+_save_video_mp4_default_backend = 'ffmpeg'
+
+def set_save_video_mp4_default_backend(backend):
+    assert backend in ('ffmpeg', 'opencv'), backend
+    global _save_video_mp4_default_backend
+    _save_video_mp4_default_backend = backend
+
+def save_video_mp4(frames, path=None, framerate=60, *, video_bitrate='high', height=None, width=None, show_progress=True, backend=None):
     """
     frames: a list of images as defined by rp.is_image(). Saves an .mp4 file at the path
         - frames can also contain strings, if those strings are image file paths
         - frames can also be a glob or a folder of images, and if so they will be sorted by number
     Note that frames can also be a generator, as opposed to a numpy array.
     This can let you save larger videos that would otherwise make your computer run out of memory.
+
+    frames: the video
+    path: the path to save the video to. Defaults to 'video.mp4'
+    video_bitrate: controls the quality of the output. If your backend is opencv, this parameter has no effect!
+    height, width: If frames have various sizes, and are given as a generator, use this to set the height and width or else it will use the first frame's height and width
+    show_progress: Whether to show the saving progress
+    backend: Defaults to 'ffmpeg'. Can also be 'opencv' if you can't install 'ffmpeg' for some reason
+
+    If you can't install ffmpeg, please set the backend to 'opencv'
+    If you need this to be the default, call rp.r.set_save_video_mp4_default_backend('opencv') instead of 'ffmpeg', the default
     
     EXAMPLE BITRATES (used for the Sunkist soda example):
-     100000    : ( 345KB) is decent, and very compressed. It starts out a bit mushy though
-     1000000   : ( 3.3MB) I believe this is close to ffmpeg's default rate. It looks okay, but it does look a tiny bit mushy
-     10000000  : (32.7MB)
-     100000000 : (93.0MB)
-     1000000000: (93.0MB) It seems to be the maximum size
+     10^5: 100000    : ( 345KB) is decent, and very compressed. It starts out a bit mushy though
+     10^6: 1000000   : ( 3.3MB) I believe this is close to ffmpeg's default rate. It looks okay, but it does look a tiny bit mushy
+     10^7: 10000000  : (32.7MB)
+     10^8: 100000000 : (93.0MB)
+     10^9: 1000000000: (93.0MB) It seems to be the maximum size
     """
+
+    if backend is None:
+        backend = _save_video_mp4_default_backend
+
+    assert backend in ('ffmpeg', 'cv2'), backend
     assert not isinstance(frames, str), 'The first argument should be the sequence of video frames, not the path!'
-    
+
+    if backend=='cv2':
+        return _cv_save_video_mp4(
+            frames,
+            path,
+            framerate=framerate,
+            show_progress=show_progress,
+            quality=video_bitrate, #Included for future updates of CV2...but right now it's inert. This parameter has no effect on the saved videos.
+        )
+
     height = None if height is None else height
     width  = None if width  is None else width 
     
@@ -33957,10 +34094,10 @@ __import__("sys").path.append(__import__("os").getcwd())
 #__import__('rp').r._add_pterm_prefix_shortcut("fu","!!fileutil")
 
 ## Custom pterm commands
-# _add_pterm_command_shortcuts('''
+#__import__('rp').r._add_pterm_command_shortcuts('''
 #     CLC $r._pterm_cd("~/CleanCode")
 #     RZG $os.system(f"cd {$get_path_parent($get_module_path(rp)} ; lazygit")
-# ''')
+#''')
 
 #Added protected folders to CDH and CDC.
 #When you add a directory to this list, if any file inside it doesn't exist but the prefix also doesn't exist, it will be shown as blue when running CDH and it won't be deleted during CDH CLEAN.

{rp-0.1.1136 → rp-0.1.1138}/rp/rp_ptpython/completer.py

@@ -320,6 +320,12 @@ class PythonCompleter(Completer):
             return path_before_cursor
 
 
+        def scandir_path_before_cursor():
+            try:
+                return os.scandir(get_path_before_cursor())
+            except Exception:
+                #Please ignore for  now
+                pass
                 
         if starts_with_any(before_line,'CDH ') and not ('\n' in before) and not after:#not after and not '\n' in before and re.fullmatch(before_line):
             import os
@@ -334,12 +340,12 @@ class PythonCompleter(Completer):
             import rp
             import os
             from rp import is_a_directory
-            yield from yield_from_candidates(*pathsmod(x for x in os.scandir(get_path_before_cursor()) if x.is_dir()))
+            yield from yield_from_candidates(*pathsmod(x for x in scandir_path_before_cursor() if x.is_dir()))
             return 
         if starts_with_any(before_line,'TAKE ','MKDIR ') and not ('\n' in before) and not after:#not after and not '\n' in before and re.fullmatch(before_line):
             import os
             from rp import is_a_directory
-            yield from yield_from_candidates(*pathsmod(os.scandir(get_path_before_cursor()), priority= lambda x:not x.is_dir()))
+            yield from yield_from_candidates(*pathsmod(scandir_path_before_cursor(), priority= lambda x:not x.is_dir()))
             return 
         if starts_with_any(before_line,'PYM ','APYM ') and not ('\n' in before) and not after:#not after and not '\n' in before and re.fullmatch(before_line):
             def get_module_names():
@@ -371,7 +377,7 @@ class PythonCompleter(Completer):
             return 
         if starts_with_any(before_line,'RN ','RM ','VIM ','OPEN ','FD ','WANS ','MV ','PIP ') and not ('\n' in before) and not after:#not after and not '\n' in before and re.fullmatch(before_line):
             import os
-            yield from yield_from_candidates(*pathsmod(os.scandir(get_path_before_cursor())))
+            yield from yield_from_candidates(*pathsmod(scandir_path_before_cursor()))
             return 
         if re.fullmatch(r'(CAT |NCAT |CCAT |OPEN |ACAT |TAB |RUN |PY |APY |PYM |APYM |LSS |LSR |FCOPY ).*',before) and not ('\n' in before) and not after:#not after and not '\n' in before and re.fullmatch(before_line):
             import os
@@ -384,7 +390,7 @@ class PythonCompleter(Completer):
 
             yield from yield_from_candidates(
                 *pathsmod(
-                    os.scandir(get_path_before_cursor()),
+                    scandir_path_before_cursor(),
                     priority=lambda x: (
                         x.is_dir() #  + (2 * x.path[len("./") :].startswith(".")) * (not after_cmd.startswith('.'))  #Also commented out because it interferes with enter-completions from pterm LSS
                     ),

rp-0.1.1138/rp/version.py

@@ -0,0 +1 @@
+1138

{rp-0.1.1136 → rp-0.1.1138}/rp.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 1.0
 Name: rp
-Version: 0.1.1136
+Version: 0.1.1138
 Summary: Ryan's Python
 Home-page: https://github.com/RyannDaGreat/Quick-Python
 Author: Ryan Burgert

rp-0.1.1138/version.py

@@ -0,0 +1 @@
+1138

rp-0.1.1136/rp/version.py

@@ -1 +0,0 @@
-1136

rp-0.1.1136/version.py

@@ -1 +0,0 @@
-1136