Merge 320fcfa29b into b08debceca

Signed-off-by: bigcat88 <bigcat88@icloud.com>

.coderabbit.yaml

@@ -4,12 +4,12 @@ early_access: false
 tone_instructions: "Only comment on issues introduced by this PR's changes. Do not flag pre-existing problems in moved, re-indented, or reformatted code."
 
 reviews:
-  profile: "chill"
-  request_changes_workflow: false
+  profile: "assertive"
+  request_changes_workflow: true
   high_level_summary: false
   poem: false
   review_status: false
-  review_details: false
+  review_details: true
   commit_status: true
   collapse_walkthrough: true
   changed_files_summary: false
@@ -39,6 +39,14 @@ reviews:
     - path: "**"
       instructions: |
         IMPORTANT: Only comment on issues directly introduced by this PR's code changes.
+        Treat AGENTS.md as mandatory repository policy, not optional style guidance.
+        Flag PR changes that violate AGENTS.md even when the code is otherwise functional.
+        In particular, enforce architecture boundaries, dtype/device/memory rules,
+        interface contracts, import style, no unnecessary try/except blocks, no inline
+        imports, no outbound internet paths in core ComfyUI, and narrow scoped fixes.
+        Prefer direct findings over suggestions when a rule is violated. Only ignore
+        AGENTS.md when it clearly conflicts with a newer explicit maintainer instruction
+        in the PR.
         Do NOT flag pre-existing issues in code that was merely moved, re-indented,
         de-indented, or reformatted without logic changes. If code appears in the diff
         only due to whitespace or structural reformatting (e.g., removing a `with:` block),
@@ -123,5 +131,10 @@ chat:
 
 knowledge_base:
   opt_out: false
+  code_guidelines:
+    enabled: true
+    filePatterns:
+      - files: "AGENTS.md"
+        applyTo: "**"
   learnings:
     scope: "auto"

comfy/sd1_clip.py

@@ -543,18 +543,24 @@ class SDTokenizer:
     def _try_get_embedding(self, embedding_name:str):
         '''
         Takes a potential embedding name and tries to retrieve it.
-        Returns a Tuple consisting of the embedding and any leftover string, embedding can be None.
+        Returns a Tuple consisting of the embedding, the cleaned embedding name, and any leftover string, embedding can be None.
         '''
         split_embed = embedding_name.split()
         embedding_name = split_embed[0]
         leftover = ' '.join(split_embed[1:])
+
+        match = re.search(r'[<\[]', embedding_name)
+        if match is not None:
+            leftover = embedding_name[match.start():] + (" " + leftover if leftover else "")
+            embedding_name = embedding_name[:match.start()]
+
         embed = load_embed(embedding_name, self.embedding_directory, self.embedding_size, self.embedding_key)
         if embed is None:
             stripped = embedding_name.strip(',')
             if len(stripped) < len(embedding_name):
                 embed = load_embed(stripped, self.embedding_directory, self.embedding_size, self.embedding_key)
-                return (embed, "{} {}".format(embedding_name[len(stripped):], leftover))
-        return (embed, leftover)
+                return (embed, embedding_name, "{} {}".format(embedding_name[len(stripped):], leftover))
+        return (embed, embedding_name, leftover)
 
     def pad_tokens(self, tokens, amount):
         if self.pad_left:
@@ -585,7 +591,7 @@ class SDTokenizer:
         tokens = []
         for weighted_segment, weight in parsed_weights:
             to_tokenize = unescape_important(weighted_segment)
-            split = re.split(' {0}|\n{0}'.format(self.embedding_identifier), to_tokenize)
+            split = re.split(r'(?<=\s){}'.format(re.escape(self.embedding_identifier)), to_tokenize)
             to_tokenize = [split[0]]
             for i in range(1, len(split)):
                 to_tokenize.append("{}{}".format(self.embedding_identifier, split[i]))
@@ -595,7 +601,7 @@ class SDTokenizer:
                 # if we find an embedding, deal with the embedding
                 if word.startswith(self.embedding_identifier) and self.embedding_directory is not None:
                     embedding_name = word[len(self.embedding_identifier):].strip('\n')
-                    embed, leftover = self._try_get_embedding(embedding_name)
+                    embed, embedding_name, leftover = self._try_get_embedding(embedding_name)
                     if embed is None:
                         logging.warning(f"warning, embedding:{embedding_name} does not exist, ignoring")
                     else:

comfy/text_encoders/llama.py

@@ -937,22 +937,41 @@ class BaseGenerate:
             return torch.argmax(logits, dim=-1, keepdim=True)
 
         # Sampling mode
-        if repetition_penalty != 1.0:
-            for i in range(logits.shape[0]):
-                for token_id in set(token_history):
-                    logits[i, token_id] *= repetition_penalty if logits[i, token_id] < 0 else 1/repetition_penalty
-
-        if presence_penalty is not None and presence_penalty != 0.0:
-            for i in range(logits.shape[0]):
-                for token_id in set(token_history):
-                    logits[i, token_id] -= presence_penalty
+        if len(token_history) > 0 and (repetition_penalty != 1.0 or (presence_penalty is not None and presence_penalty != 0.0)):
+            token_ids = torch.tensor(list(set(token_history)), device=logits.device)
+            token_logits = logits[:, token_ids]
+            if repetition_penalty != 1.0:
+                token_logits = torch.where(token_logits < 0, token_logits * repetition_penalty, token_logits / repetition_penalty)
+            if presence_penalty is not None and presence_penalty != 0.0:
+                token_logits = token_logits - presence_penalty
+            logits[:, token_ids] = token_logits
 
         if temperature != 1.0:
             logits = logits / temperature
 
         if top_k > 0:
-            indices_to_remove = logits < torch.topk(logits, top_k)[0][..., -1, None]
-            logits[indices_to_remove] = torch.finfo(logits.dtype).min
+            top_k = min(top_k, logits.shape[-1])
+            logits, top_indices = torch.topk(logits, top_k)
+
+            if min_p > 0.0:
+                probs_before_filter = torch.nn.functional.softmax(logits, dim=-1)
+                top_probs, _ = probs_before_filter.max(dim=-1, keepdim=True)
+                min_threshold = min_p * top_probs
+                indices_to_remove = probs_before_filter < min_threshold
+                logits[indices_to_remove] = torch.finfo(logits.dtype).min
+
+            if top_p < 1.0:
+                sorted_logits, sorted_indices = torch.sort(logits, descending=True)
+                cumulative_probs = torch.cumsum(torch.nn.functional.softmax(sorted_logits, dim=-1), dim=-1)
+                sorted_indices_to_remove = cumulative_probs > top_p
+                sorted_indices_to_remove[..., 0] = False
+                indices_to_remove = torch.zeros_like(logits, dtype=torch.bool)
+                indices_to_remove.scatter_(1, sorted_indices, sorted_indices_to_remove)
+                logits[indices_to_remove] = torch.finfo(logits.dtype).min
+
+            probs = torch.nn.functional.softmax(logits, dim=-1)
+            next_token = torch.multinomial(probs, num_samples=1, generator=generator)
+            return top_indices.gather(1, next_token)
 
         if min_p > 0.0:
             probs_before_filter = torch.nn.functional.softmax(logits, dim=-1)

comfy_api_nodes/nodes_bria.py

@@ -251,7 +251,12 @@ class BriaRemoveVideoBackground(IO.ComfyNode):
             node_id="BriaRemoveVideoBackground",
             display_name="Bria Remove Video Background",
             category="partner/video/Bria",
-            description="Remove the background from a video using Bria. ",
+            description="Remove the background from a video using Bria. Pick a color to composite "
+            "the foreground over a solid background (returns an MP4), or pick 'Transparent' to get "
+            "a WebM/VP9 video with a real alpha channel. For per-frame compositing (images + mask "
+            "sockets) or saving the transparent result to disk with alpha preserved, use 'Bria "
+            "Remove Video Background (Transparent)' instead -- Save Video currently writes to .mp4 "
+            "by default, which drops the VP9 alpha plane.",
             inputs=[
                 IO.Video.Input("video"),
                 IO.Combo.Input(
@@ -267,8 +272,13 @@ class BriaRemoveVideoBackground(IO.ComfyNode):
                         "Cyan",
                         "Magenta",
                         "Orange",
+                        "Transparent",
                     ],
-                    tooltip="Background color for the output video.",
+                    default="Black",
+                    tooltip="Background color for the output video. Colors composite the "
+                    "foreground over a solid background and return an MP4. 'Transparent' returns "
+                    "a WebM/VP9 video with an alpha channel; for per-frame compositing or saving "
+                    "the alpha to disk, use 'Bria Remove Video Background (Transparent)' instead.",
                 ),
                 IO.Int.Input(
                     "seed",
@@ -301,13 +311,18 @@ class BriaRemoveVideoBackground(IO.ComfyNode):
         seed: int,
     ) -> IO.NodeOutput:
         validate_video_duration(video, max_duration=60.0)
+        # Bria returns 422 if background_color="Transparent" is paired with mp4_h264
+        # (no alpha plane), so request webm_vp9 instead. VP9 carries the alpha in a side
+        # layer; the bytes pass through ComfyUI's VIDEO type as a file reference but
+        # SaveVideo's default .mp4 extension currently drops it on mux -- see PR notes.
+        output_container_and_codec = "webm_vp9" if background_color == "Transparent" else "mp4_h264"
         response = await sync_op(
             cls,
             ApiEndpoint(path="/proxy/bria/v2/video/edit/remove_background", method="POST"),
             data=BriaRemoveVideoBackgroundRequest(
                 video=await upload_video_to_comfyapi(cls, video),
                 background_color=background_color,
-                output_container_and_codec="mp4_h264",
+                output_container_and_codec=output_container_and_codec,
                 seed=seed,
             ),
             response_model=BriaStatusResponse,
@@ -507,8 +522,10 @@ class BriaTransparentVideoBackground(IO.ComfyNode):
             display_name="Bria Remove Video Background (Transparent)",
             category="partner/video/Bria",
             description="Remove the background from a video using Bria and return the cut-out frames "
-            "plus an alpha mask. Connect both to a compositing node, or feed them to Save WEBM to "
-            "write a transparent video.",
+            "plus an alpha mask, ready for per-frame compositing or feeding into Save WEBM to "
+            "persist a transparent video to disk. This is currently the only way to save the "
+            "alpha channel: Save Video on the consolidated 'Bria Remove Video Background' node's "
+            "Transparent output writes a .mp4 by default and drops the VP9 alpha plane on mux.",
             inputs=[
                 IO.Video.Input("video"),
                 IO.Int.Input(

comfy_api_nodes/util/request_logger.py

@@ -9,6 +9,7 @@ from typing import Any
 import folder_paths
 
 logger = logging.getLogger(__name__)
+_SENSITIVE_HEADERS = {"authorization", "x-api-key"}
 
 
 def get_log_directory():
@@ -73,6 +74,10 @@ def _format_data_for_logging(data: Any) -> str:
     return str(data)
 
 
+def _redact_headers(headers: dict) -> dict:
+    return {k: ("***" if k.lower() in _SENSITIVE_HEADERS else v) for k, v in headers.items()}
+
+
 def log_request_response(
     operation_id: str,
     request_method: str,
@@ -101,7 +106,7 @@ def log_request_response(
         log_content.append(f"Method: {request_method}")
         log_content.append(f"URL: {request_url}")
         if request_headers:
-            log_content.append(f"Headers:\n{_format_data_for_logging(request_headers)}")
+            log_content.append(f"Headers:\n{_format_data_for_logging(_redact_headers(request_headers))}")
         if request_params:
             log_content.append(f"Params:\n{_format_data_for_logging(request_params)}")
         if request_data is not None:

comfy_extras/nodes_color.py

@@ -16,23 +16,30 @@ class ColorToRGBInt(io.ComfyNode):
             ],
             outputs=[
                 io.Int.Output(display_name="rgb_int"),
-                io.Color.Output(display_name="hex")
+                io.Color.Output(display_name="hex"),
+                io.Float.Output(display_name="alpha"),
             ],
         )
 
     @classmethod
     def execute(cls, color: str) -> io.NodeOutput:
-        # expect format #RRGGBB
-        if len(color) != 7 or color[0] != "#":
-            raise ValueError("Color must be in format #RRGGBB")
+        # expect format #RRGGBB or #RRGGBBAA
+        if len(color) not in (7, 9) or color[0] != "#":
+            raise ValueError("Color must be in format #RRGGBB or #RRGGBBAA")
         try:
             int(color[1:], 16)
         except ValueError:
-            raise ValueError("Color must be in format #RRGGBB") from None
+            raise ValueError("Color must be in format #RRGGBB or #RRGGBBAA") from None
+
+        alpha = 1.0
+        if len(color) == 9:
+            alpha = int(color[7:9], 16) / 255.0
+            color = color[:7]
+
         r, g, b = hex_to_rgb(color)
 
         rgb_int = r * 256 * 256 + g * 256 + b
-        return io.NodeOutput(rgb_int, color)
+        return io.NodeOutput(rgb_int, color, alpha)
 
 
 class ColorExtension(ComfyExtension):

requirements.txt

@@ -1,6 +1,6 @@
 comfyui-frontend-package==1.45.20
 comfyui-workflow-templates==0.11.2
-comfyui-embedded-docs==0.5.6
+comfyui-embedded-docs==0.5.7
 torch
 torchsde
 torchvision

tests-unit/comfy_api_test/video_types_test.py

@@ -5,10 +5,13 @@ import os
 import av
 import io
 from fractions import Fraction
+import numpy as np
+
 from comfy_api.input_impl.video_types import VideoFromFile, VideoFromComponents
-from comfy_api.util.video_types import VideoComponents
+from comfy_api.util.video_types import VideoComponents, VideoContainer, VideoCodec
 from comfy_api.input.basic_types import AudioInput
 from av.error import InvalidDataError
+from av.codec import CodecContext
 
 EPSILON = 0.0001
 
@@ -237,3 +240,70 @@ def test_duration_consistency(video_components):
     manual_duration = float(components.images.shape[0] / components.frame_rate)
 
     assert duration == pytest.approx(manual_duration)
+
+
+def _make_webm_vp9_with_alpha(width=32, height=32, frames=5, fps=10) -> bytes:
+    buf = io.BytesIO()
+    with av.open(buf, mode="w", format="webm") as container:
+        stream = container.add_stream("libvpx-vp9", rate=fps)
+        stream.width = width
+        stream.height = height
+        stream.pix_fmt = "yuva420p"
+        for i in range(frames):
+            rgba = np.zeros((height, width, 4), dtype=np.uint8)
+            rgba[..., 0] = 200
+            rgba[..., 1] = 50 + i * 30
+            rgba[..., 2] = 80
+            rgba[..., 3] = (i + 1) * 50
+            frame = av.VideoFrame.from_ndarray(rgba, format="rgba")
+            frame = frame.reformat(format="yuva420p")
+            for packet in stream.encode(frame):
+                container.mux(packet)
+        for packet in stream.encode(None):
+            container.mux(packet)
+    return buf.getvalue()
+
+
+def _per_frame_alpha_means_libvpx(source) -> list:
+    means = []
+    with av.open(source, mode="r") as container:
+        stream = container.streams.video[0]
+        decoder = CodecContext.create("libvpx-vp9", "r")
+        for packet in container.demux(stream):
+            for frame in decoder.decode(packet):
+                rgba = frame.to_ndarray(format="rgba")
+                means.append(float(rgba[..., 3].mean()))
+        try:
+            for frame in decoder.decode(None):
+                rgba = frame.to_ndarray(format="rgba")
+                means.append(float(rgba[..., 3].mean()))
+        except EOFError:
+            pass
+    return means
+
+
+def test_video_from_file_save_to_bytesio_stream_copies_vp9_alpha():
+    """VideoFromFile.save_to(BytesIO, format=AUTO, codec=AUTO) stream-copies the source
+    container/codec without re-encoding, so a WebM/VP9 file with an alpha plane
+    round-trips unchanged. Note: this covers ONLY the in-memory BytesIO path. Saving
+    to a .mp4 file (SaveVideo's default extension) muxes VP9 into MP4 and drops the
+    alpha plane -- intentionally not covered here; see PR notes."""
+    raw = _make_webm_vp9_with_alpha()
+    original_means = _per_frame_alpha_means_libvpx(io.BytesIO(raw))
+    assert len(original_means) > 0, "test fixture failed to produce frames"
+    assert original_means == sorted(original_means), (
+        "test fixture alpha did not encode as monotonically increasing"
+    )
+
+    video = VideoFromFile(io.BytesIO(raw))
+    out_buf = io.BytesIO()
+    video.save_to(out_buf, format=VideoContainer.AUTO, codec=VideoCodec.AUTO)
+
+    out_buf.seek(0)
+    saved_means = _per_frame_alpha_means_libvpx(out_buf)
+
+    assert len(saved_means) == len(original_means)
+    for original, saved in zip(original_means, saved_means):
+        assert original == pytest.approx(saved, abs=EPSILON), (
+            f"alpha lost across stream-copy round-trip: {original} -> {saved}"
+        )