Compare commits

...

5 Commits

Author SHA1 Message Date
Matt Miller
ad697bf3cc
Merge 058692a79b into 6880614319 2026-07-07 18:36:58 -07:00
comfyanonymous
6880614319
Update AGENTS.md (#14819) 2026-07-07 18:36:13 -07:00
Barish Ozbay
51bf508a0b
feat: Implement basic text overlay node (CORE-137) (#14610)
Some checks are pending
Detect Unreviewed Merge / detect (push) Waiting to run
Python Linting / Run Ruff (push) Waiting to run
Python Linting / Run Pylint (push) Waiting to run
Full Comfy CI Workflow Runs / test-stable (12.1, , linux, 3.10, [self-hosted Linux], stable) (push) Waiting to run
Full Comfy CI Workflow Runs / test-stable (12.1, , linux, 3.11, [self-hosted Linux], stable) (push) Waiting to run
Full Comfy CI Workflow Runs / test-stable (12.1, , linux, 3.12, [self-hosted Linux], stable) (push) Waiting to run
Full Comfy CI Workflow Runs / test-unix-nightly (12.1, , linux, 3.11, [self-hosted Linux], nightly) (push) Waiting to run
Execution Tests / test (macos-latest) (push) Waiting to run
Execution Tests / test (ubuntu-latest) (push) Waiting to run
Execution Tests / test (windows-latest) (push) Waiting to run
Test server launches without errors / test (push) Waiting to run
Unit Tests / test (macos-latest) (push) Waiting to run
Unit Tests / test (ubuntu-latest) (push) Waiting to run
Unit Tests / test (windows-2022) (push) Waiting to run
2026-07-07 21:26:52 +08:00
Matt Miller
058692a79b fix(assets): treat explicit empty hash filter as exact-match miss
Some checks failed
Python Linting / Run Ruff (push) Has been cancelled
Python Linting / Run Pylint (push) Has been cancelled
Address review findings on the listAssets contract PR:

- Empty `?hash=` no longer collapses to "no filter" (which returned a full
  unfiltered page); it now stays "" and is matched exactly, yielding an
  empty page — consistent with the documented "malformed -> empty page"
  contract. Omitting the param entirely still disables the filter.
- Guard `list_references_page` on `asset_hash is not None` (page + count)
  so the present-empty vs omitted distinction is preserved.
- Add tests for hash normalization (uppercase/whitespace) and for the
  explicit-empty -> empty-page behavior.
- Clarify the `include_public` comment: core reads are owner-scoped with no
  separate public pool, so the flag is inert here; cloud enforces it in its
  own service layer.
2026-07-01 00:43:15 -07:00
Matt Miller
58df1a3564 feat: implement remaining listAssets contract fields (display_name, hash filter, include_public)
Bring GET /api/assets into param-for-param parity with the projected
openapi.yaml listAssets contract for the three remaining fields:

- Add display_name to the Asset response schema (nullable, mirrors name)
  and populate it from ref.name in _build_asset_response, covering list,
  get, create, update, and upload responses uniformly.
- Add the hash query param to ListAssetsQuery (named hash per the spec,
  not asset_hash) with before-validation strip/lower normalization, and
  thread it through list_assets_page -> list_references_page, filtering
  both the page and count statements on Asset.hash for consistency.
- Accept include_public (bool, default true) for contract parity; it is
  inert in core (no public asset pool) and intentionally not passed to
  the service layer.

Cursor pagination and size optionality are untouched.

Add integration tests covering display_name mirroring, exact hash match,
unknown-hash empty page, and include_public acceptance.
2026-06-30 02:22:06 -07:00
9 changed files with 317 additions and 0 deletions

View File

@ -127,6 +127,8 @@
- Do not add unnecessary `try`/`except` blocks. Use them for optional dependency,
platform, or backend capability detection only when the program has a useful
fallback. Prefer specific exception types when changing new code.
- If a library version is pinned in `requirements.txt`, do not add code to
ComfyUI to handle older versions of that library.
- Remove any workarounds for PyTorch versions that ComfyUI no longer officially
supports. Deprecated workarounds include catching an exception and rerunning
the same op with the input cast to float. If a workaround does not have a

View File

@ -165,6 +165,8 @@ def _build_asset_response(result: schemas.AssetDetailResult | schemas.UploadResu
return schemas_out.Asset(
id=result.ref.id,
name=result.ref.name,
# Mirror name for backwards compatibility; ref.name is always set.
display_name=result.ref.name,
hash=asset_content_hash,
asset_hash=asset_content_hash,
size=int(result.asset.size_bytes) if result.asset else None,
@ -219,6 +221,7 @@ async def list_assets_route(request: web.Request) -> web.Response:
exclude_tags=q.exclude_tags,
name_contains=q.name_contains,
metadata_filter=q.metadata_filter,
asset_hash=q.hash,
limit=q.limit,
offset=q.offset,
sort=sort,

View File

@ -54,6 +54,18 @@ class ListAssetsQuery(BaseModel):
exclude_tags: list[str] = Field(default_factory=list)
name_contains: str | None = None
# Filter to assets whose content hash matches exactly. Param name is `hash`
# per the projected openapi.yaml listAssets contract (the response-body field
# is `asset_hash`; the query param is `hash`).
hash: str | None = None
# Declared for cloud/core contract parity. In core, reads are owner-scoped
# (owner_id == "") and there is no separate shared/public pool for this flag
# to include or exclude, so it is inert here and intentionally not threaded
# into the query. Accepted (not rejected) so the FE needs no isCloud branch;
# cloud enforces the flag in its own service layer.
include_public: bool = True
# Accept either a JSON string (query param) or a dict
metadata_filter: dict[str, Any] | None = None
@ -86,6 +98,19 @@ class ListAssetsQuery(BaseModel):
return out
return v
@field_validator("hash", mode="before")
@classmethod
def _normalize_hash(cls, v):
# Normalize for an exact match against stored hashes (which are
# lowercase `blake3:<hex>`). Liberal in what we accept — no pattern
# enforcement; a non-matching value simply yields an empty page.
# An explicitly-supplied-but-empty value (`?hash=`) stays `""` so it
# is treated as an exact-match miss (empty page), not silently dropped
# to "no filter" — omit the param entirely to disable the filter.
if isinstance(v, str):
return v.strip().lower()
return v
@field_validator("metadata_filter", mode="before")
@classmethod
def _parse_metadata_json(cls, v):

View File

@ -10,6 +10,8 @@ class Asset(BaseModel):
id: str
name: str
# Mirrors `name` for backwards compatibility.
display_name: str | None = None
hash: str | None = None
asset_hash: str | None = None
size: int | None = None

View File

@ -261,6 +261,7 @@ def list_references_page(
limit: int = 100,
offset: int = 0,
name_contains: str | None = None,
asset_hash: str | None = None,
include_tags: Sequence[str] | None = None,
exclude_tags: Sequence[str] | None = None,
metadata_filter: dict | None = None,
@ -293,6 +294,11 @@ def list_references_page(
escaped, esc = escape_sql_like_string(name_contains)
base = base.where(AssetReference.name.ilike(f"%{escaped}%", escape=esc))
# `is not None` (not truthiness): an explicit empty hash is an exact-match
# miss (empty page), while an omitted hash (None) disables the filter.
if asset_hash is not None:
base = base.where(Asset.hash == asset_hash)
base = apply_tag_filters(base, include_tags, exclude_tags)
base = apply_metadata_filter(base, metadata_filter)
@ -345,6 +351,8 @@ def list_references_page(
count_stmt = count_stmt.where(
AssetReference.name.ilike(f"%{escaped}%", escape=esc)
)
if asset_hash is not None:
count_stmt = count_stmt.where(Asset.hash == asset_hash)
count_stmt = apply_tag_filters(count_stmt, include_tags, exclude_tags)
count_stmt = apply_metadata_filter(count_stmt, metadata_filter)

View File

@ -274,6 +274,7 @@ def list_assets_page(
exclude_tags: Sequence[str] | None = None,
name_contains: str | None = None,
metadata_filter: dict | None = None,
asset_hash: str | None = None,
limit: int = 20,
offset: int = 0,
sort: str = "created_at",
@ -319,6 +320,7 @@ def list_assets_page(
exclude_tags=exclude_tags,
name_contains=name_contains,
metadata_filter=metadata_filter,
asset_hash=asset_hash,
limit=fetch_limit,
offset=offset,
sort=sort,

View File

@ -0,0 +1,150 @@
import numpy as np
import torch
from PIL import Image as PILImage, ImageColor, ImageDraw, ImageFont
from typing_extensions import override
from comfy_api.latest import ComfyExtension, IO
class TextOverlay(IO.ComfyNode):
@classmethod
def define_schema(cls):
return IO.Schema(
node_id="TextOverlay",
display_name="Draw Text Overlay",
category="text",
description="Draw text overlay on an image or batch of images.",
search_aliases=["text", "label", "caption", "subtitle", "watermark", "title", "addlabel", "overlay"],
inputs=[
IO.Image.Input("images"),
IO.String.Input("text", multiline=True, default=""),
IO.Float.Input("font_size", default=5.0, min=0.5, max=50.0, step=0.5, tooltip="Font size as a percentage of the image height."),
IO.Color.Input("color", default="#ffffff", tooltip="Color of the text."),
IO.Combo.Input("position", options=["top", "bottom"], default="top"),
IO.Combo.Input("align", options=["left", "center", "right"], default="left"),
IO.Boolean.Input("outline", default=True, tooltip="Draw a black outline around the text."),
],
outputs=[IO.Image.Output(display_name="images")],
)
@classmethod
def execute(cls, images, text, font_size, color, position, align, outline) -> IO.NodeOutput:
if text.strip() == "":
return IO.NodeOutput(images)
text = text.replace("\\n", "\n").replace("\\t", "\t")
text_rgba = cls.parse_color_to_rgba(color)
outline_rgba = (0, 0, 0, 255) if outline else (0, 0, 0, 0)
# Render the overlay once and composite it across all frames in the batch
height = images.shape[1]
width = images.shape[2]
overlay_rgb, overlay_alpha = cls.render_overlay_text(width, height, text, position, align, font_size, text_rgba, outline_rgba)
overlay_rgb = overlay_rgb.to(device=images.device, dtype=images.dtype)
overlay_alpha = overlay_alpha.to(device=images.device, dtype=images.dtype)
result = images * (1.0 - overlay_alpha) + overlay_rgb * overlay_alpha
return IO.NodeOutput(result)
@staticmethod
def parse_color_to_rgba(color_string):
parsed = ImageColor.getrgb(color_string)
if len(parsed) == 3:
return (*parsed, 255)
return parsed
@classmethod
def render_overlay_text(cls, width, height, text, position, align, font_size, text_rgba, outline_rgba):
line_spacing = 1.2
margin_percent = 1.0
min_font_percent = 2.0
min_font_pixels = 10
outline_thickness_factor = 0.04
# Draw onto a transparent layer so the result can be alpha-composited over any frame.
layer = PILImage.new("RGBA", (width, height), (0, 0, 0, 0))
draw = ImageDraw.Draw(layer)
margin = int(round(margin_percent / 100.0 * min(width, height)))
max_width = max(1, width - 2 * margin)
max_height = max(1, height - 2 * margin)
# Font scales with resolution, then shrinks to fit the height.
size = max(1, int(round(font_size / 100.0 * height)))
floor = min(size, max(min_font_pixels, int(round(min_font_percent / 100.0 * height))))
while True:
font = ImageFont.load_default(size=size)
stroke = max(1, int(round(size * outline_thickness_factor))) if outline_rgba[3] > 0 else 0
block = "\n".join(cls.wrap_text(text, font, max_width))
# convert line spacing to pixel spacing
single = draw.textbbox((0, 0), "Ay", font=font, stroke_width=stroke)
double = draw.multiline_textbbox((0, 0), "Ay\nAy", font=font, spacing=0, stroke_width=stroke)
natural_advance = (double[3] - double[1]) - (single[3] - single[1])
pixel_spacing = int(round(size * line_spacing - natural_advance))
box = draw.multiline_textbbox((0, 0), block, font=font, spacing=pixel_spacing, stroke_width=stroke)
block_height = box[3] - box[1]
if block_height <= max_height or size <= floor:
break
size = max(floor, int(size * 0.9))
anchor_h, x = {"left": ("l", margin), "center": ("m", width / 2), "right": ("r", width - margin)}[align]
# Offset y so the rendered text sits flush against the margin
if position == "bottom":
y = height - margin - box[3]
else:
y = margin - box[1]
draw.multiline_text((x, y), block, font=font, fill=text_rgba, anchor=anchor_h + "a",
align=align, spacing=pixel_spacing, stroke_width=stroke, stroke_fill=outline_rgba)
overlay = np.array(layer).astype(np.float32) / 255.0
overlay_rgb = torch.from_numpy(overlay[:, :, :3])
overlay_alpha = torch.from_numpy(overlay[:, :, 3:4])
return overlay_rgb, overlay_alpha
@staticmethod
def wrap_text(text, font, max_width):
lines = []
for raw_line in text.split("\n"):
words = raw_line.split()
if not words:
lines.append("")
continue
current = ""
# Break the line into words and split words that are too long
for word in words:
while font.getlength(word) > max_width and len(word) > 1:
cut = 1
while cut < len(word) and font.getlength(word[:cut + 1]) <= max_width:
cut += 1
if current:
lines.append(current)
current = ""
lines.append(word[:cut])
word = word[cut:]
candidate = word if not current else current + " " + word
if not current or font.getlength(candidate) <= max_width:
current = candidate
else:
lines.append(current)
current = word
if current:
lines.append(current)
return lines
class TextOverlayExtension(ComfyExtension):
@override
async def get_node_list(self) -> list[type[IO.ComfyNode]]:
return [TextOverlay]
async def comfy_entrypoint() -> TextOverlayExtension:
return TextOverlayExtension()

View File

@ -2478,6 +2478,7 @@ async def init_builtin_extra_nodes():
"nodes_glsl.py",
"nodes_lora_debug.py",
"nodes_textgen.py",
"nodes_text_overlay.py",
"nodes_color.py",
"nodes_toolkit.py",
"nodes_replacements.py",

View File

@ -306,6 +306,130 @@ def test_list_assets_invalid_query_rejected(http: requests.Session, api_base: st
assert body["error"]["code"] == error_code
def test_list_assets_display_name_mirrors_name(http, api_base, asset_factory, make_asset_bytes):
"""`display_name` is emitted and mirrors `name` for every populated asset."""
scope = f"lf-dispname-{uuid.uuid4().hex[:6]}"
tags = ["models", "checkpoints", "unit-tests", scope]
asset_factory("dn_a.safetensors", tags, {}, make_asset_bytes("dn_a", 700))
asset_factory("dn_b.safetensors", tags, {}, make_asset_bytes("dn_b", 700))
r = http.get(
api_base + "/api/assets",
params={"include_tags": f"unit-tests,{scope}", "limit": "50"},
timeout=120,
)
body = r.json()
assert r.status_code == 200, body
assert body["assets"], "expected at least one asset"
for asset in body["assets"]:
assert "display_name" in asset, "populated asset must emit display_name"
assert asset["display_name"] == asset["name"]
def test_list_assets_hash_filter_exact_match(http, api_base, asset_factory, make_asset_bytes):
"""`hash` filters to assets whose content hash matches exactly."""
scope = f"lf-hash-{uuid.uuid4().hex[:6]}"
tags = ["models", "checkpoints", "unit-tests", scope]
a = asset_factory("hf_a.safetensors", tags, {}, make_asset_bytes("hf_a", 1024))
b = asset_factory("hf_b.safetensors", tags, {}, make_asset_bytes("hf_b", 2048))
target = a["hash"]
assert target and a["hash"] != b["hash"], "fixtures must have distinct content hashes"
r = http.get(
api_base + "/api/assets",
params={"hash": target, "limit": "50"},
timeout=120,
)
body = r.json()
assert r.status_code == 200, body
names = [x["name"] for x in body["assets"]]
assert names == [a["name"]]
assert body["total"] == 1
def test_list_assets_hash_filter_no_match(http, api_base, asset_factory, make_asset_bytes):
"""A well-formed but unknown hash returns an empty page (200)."""
scope = f"lf-hash-none-{uuid.uuid4().hex[:6]}"
tags = ["models", "checkpoints", "unit-tests", scope]
asset_factory("hn_a.safetensors", tags, {}, make_asset_bytes("hn_a", 800))
unknown = "blake3:" + ("0" * 64)
r = http.get(
api_base + "/api/assets",
params={"hash": unknown, "limit": "50"},
timeout=120,
)
body = r.json()
assert r.status_code == 200, body
assert body["assets"] == []
assert body["total"] == 0
def test_list_assets_hash_filter_normalizes_case_and_whitespace(
http, api_base, asset_factory, make_asset_bytes
):
"""`hash` is trimmed and lowercased before matching, so an upper-cased,
space-padded value still matches the stored lowercase hash."""
scope = f"lf-hashnorm-{uuid.uuid4().hex[:6]}"
tags = ["models", "checkpoints", "unit-tests", scope]
a = asset_factory("hnorm_a.safetensors", tags, {}, make_asset_bytes("hnorm_a", 1024))
target = a["hash"]
assert target == target.lower(), "stored hash is expected to be lowercase"
messy = f" {target.upper()} "
r = http.get(
api_base + "/api/assets",
params={"hash": messy, "limit": "50"},
timeout=120,
)
body = r.json()
assert r.status_code == 200, body
names = [x["name"] for x in body["assets"]]
assert names == [a["name"]]
assert body["total"] == 1
def test_list_assets_hash_filter_empty_returns_empty_page(
http, api_base, asset_factory, make_asset_bytes
):
"""An explicitly-supplied but empty `hash` (`?hash=`) is an exact-match miss
and returns an empty page, rather than silently disabling the filter."""
scope = f"lf-hashempty-{uuid.uuid4().hex[:6]}"
tags = ["models", "checkpoints", "unit-tests", scope]
asset_factory("he_a.safetensors", tags, {}, make_asset_bytes("he_a", 800))
r = http.get(
api_base + "/api/assets",
params={"hash": "", "limit": "50"},
timeout=120,
)
body = r.json()
assert r.status_code == 200, body
assert body["assets"] == []
assert body["total"] == 0
def test_list_assets_include_public_accepted(http, api_base, asset_factory, make_asset_bytes):
"""`include_public` is accepted for contract parity; core results are always
the caller's own assets regardless of its value (the param is inert)."""
scope = f"lf-incpub-{uuid.uuid4().hex[:6]}"
tags = ["models", "checkpoints", "unit-tests", scope]
a = asset_factory("ip_a.safetensors", tags, {}, make_asset_bytes("ip_a", 900))
for value in ("false", "true"):
r = http.get(
api_base + "/api/assets",
params={"include_tags": f"unit-tests,{scope}", "include_public": value, "limit": "50"},
timeout=120,
)
body = r.json()
assert r.status_code == 200, body
names = [x["name"] for x in body["assets"]]
assert a["name"] in names, f"caller's own asset must be returned (include_public={value})"
def test_list_assets_name_contains_literal_underscore(
http,
api_base,