sebastien/Copies
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Miscs improvements (Interro02)

Browse Source
This commit is contained in:
sebastien
2026-09-15 14:18:22 +02:00
parent 0a86403ca6
commit 9b22a8a137
15 changed files with 893 additions and 118 deletions
Download Patch File Download Diff File Expand all files Collapse all files
copienator/commands/reading_annotations.py
+83 -15
View File
@@ -10,7 +10,7 @@ from pdf2image import convert_from_path
from PIL import Image, ImageChops, ImageDraw, ImageFilter
from copienator.commands import annotating
from copienator import utils
from copienator import configuration, utils
from copienator import (
EvaluationWorkspace,
ExitCode,
@@ -24,6 +24,7 @@ from copienator.annotation_actions import apply_checkbox_actions, apply_score_ov
from copienator.answer_info import build_answer_info
from copienator.annotation_data import AnnotationData, load_annotation_data
from copienator.filesystem import staged_files
from copienator.return_answers import save_return_answer_options
Image.MAX_IMAGE_PIXELS = None
@@ -69,10 +70,11 @@ def detect_checks_and_notes(
print(f" Resizing annotated PDF from {user_image.size} to {reference.size}")
user_image = user_image.resize(reference.size, Image.Resampling.LANCZOS)
difference = np.abs(
np.array(reference).astype(int) - np.array(user_image).astype(int)
).astype(np.uint8)
difference_gray = np.mean(difference, axis=2)
# Keep the full-size difference in uint8. Converting both tall group images
# to the platform ``int`` dtype used several gigabytes per scan worker.
difference = np.asarray(
ImageChops.difference(reference, user_image), dtype=np.uint8
)
keep_mask = Image.new("L", reference.size, 255)
mask_draw = ImageDraw.Draw(keep_mask)
actions: list[dict[str, Any]] = []
@@ -83,10 +85,14 @@ def detect_checks_and_notes(
x1, y1, x2, y2 = map(int, raw_box["global_box"])
x1, y1 = max(0, x1), max(0, y1)
x2, y2 = min(reference.width, x2), min(reference.height, y2)
region = difference_gray[y1 + 5 : y2 - 5, x1 + 5 : x2 - 5]
region = difference[y1 + 5 : y2 - 5, x1 + 5 : x2 - 5]
if region.size == 0:
continue
density = np.sum(region > 30) / region.size
# Preserve the previous mean-across-RGB threshold, but allocate its
# temporary float array only for the small checkbox region.
density = np.count_nonzero(np.mean(region, axis=2) > 30) / (
region.shape[0] * region.shape[1]
)
if density > 0.05:
actions.append(raw_box)
mask_draw.rectangle([x1 - 15, y1 - 15, x2 + 15, y2 + 15], fill=0)
@@ -95,13 +101,17 @@ def detect_checks_and_notes(
if raw_box.get("type") == "score" and raw_box.get("value") == 0.0:
mask_draw.rectangle([0, y1 - 10, reference.width, y2 + 10], fill=0)
del difference
reference_blur = reference.filter(ImageFilter.GaussianBlur(2))
user_blur = user_image.filter(ImageFilter.GaussianBlur(2))
diff_image = ImageChops.difference(reference_blur, user_blur).convert("L")
alpha = np.where(np.array(diff_image) > 50, 255, 0).astype(np.uint8)
final_alpha = np.minimum(alpha, np.array(keep_mask))
del reference_blur, user_blur
alpha = np.asarray(diff_image, dtype=np.uint8).copy()
np.greater(alpha, 50, out=alpha)
alpha *= np.uint8(255)
np.minimum(alpha, np.asarray(keep_mask, dtype=np.uint8), out=alpha)
notes = user_image.convert("RGBA")
notes.putalpha(Image.fromarray(final_alpha))
notes.putalpha(Image.fromarray(alpha))
return actions, notes
@@ -150,8 +160,10 @@ def apply_actions_and_regenerate(
labels_data = data[student_id]
apply_checkbox_actions(labels_data, actions, print)
score_path = output_dir / "score.json"
preserve_score_file = update_score and score_path.is_file()
if update_score:
apply_score_overrides(labels_data, output_dir / "score.json", print)
apply_score_overrides(labels_data, score_path, print)
scores = dict.fromkeys(all_labels, "")
answer_labels: list[str] = []
@@ -220,7 +232,8 @@ def apply_actions_and_regenerate(
"touched.json", "answer_labels.json")) as staging:
for label, image in dirty_images.items():
image.save(staging / f"{label}.jpg")
atomic_write_json(staging / "score.json", scores)
if not preserve_score_file:
atomic_write_json(staging / "score.json", scores)
atomic_write_json(staging / "info.json", build_answer_info(
scores, labels_data, answer_labels
))
@@ -229,13 +242,41 @@ def apply_actions_and_regenerate(
if filtered_image is not None:
filtered_image.save(staging / "Concat_F.jpg")
if preserve_score_file:
print(f" Preserved existing score.json in {output_dir}")
print(f" Saved regenerated files in {output_dir}")
return ExitCode.PARTIAL if incomplete else ExitCode.SUCCESS
def run(workspace: EvaluationWorkspace, *, update_score: bool = False) -> ExitCode:
def run(
workspace: EvaluationWorkspace,
*,
update_score: bool = False,
return_answers_context: bool | None = None,
return_answers_question: bool | None = None,
return_answers_solution: bool | None = None,
) -> ExitCode:
workspace.require_files("labels", "correction.json")
workspace.require_directories("Copies", "Par label", "Bnot")
if configuration.RETURN_ANSWERS_ENABLED:
save_return_answer_options(
workspace.root,
context=(
configuration.RETURN_ANSWERS_CONTEXT
if return_answers_context is None
else return_answers_context
),
question=(
configuration.RETURN_ANSWERS_QUESTION
if return_answers_question is None
else return_answers_question
),
solution=(
configuration.RETURN_ANSWERS_SOLUTION
if return_answers_solution is None
else return_answers_solution
),
)
all_labels = utils.read_all_labels(workspace.root)
loaded = load_annotation_data(workspace)
for warning in loaded.warnings:
@@ -276,7 +317,28 @@ def build_parser() -> argparse.ArgumentParser:
parser.add_argument(
"--update-score",
action="store_true",
help="Override generated scores with values from existing score.json files",
help=(
"Regenerate images with current statement/solution PDFs while "
"preserving and applying existing score.json values"
),
)
parser.add_argument(
"--return-answers-context",
action=argparse.BooleanOptionalAction,
default=configuration.RETURN_ANSWERS_CONTEXT,
help="Include applicable context pages in individual answer exports",
)
parser.add_argument(
"--return-answers-question",
action=argparse.BooleanOptionalAction,
default=configuration.RETURN_ANSWERS_QUESTION,
help="Include the current question PDF in individual answer exports",
)
parser.add_argument(
"--return-answers-solution",
action=argparse.BooleanOptionalAction,
default=configuration.RETURN_ANSWERS_SOLUTION,
help="Include the current solution PDF in individual answer exports",
)
return parser
@@ -285,7 +347,13 @@ def main(argv: Sequence[str] | None = None) -> int:
parser = build_parser()
def handle(args: argparse.Namespace) -> ExitCode:
return run(workspace_from_args(args), update_score=args.update_score)
return run(
workspace_from_args(args),
update_score=args.update_score,
return_answers_context=args.return_answers_context,
return_answers_question=args.return_answers_question,
return_answers_solution=args.return_answers_solution,
)
return execute(parser, argv, handle)