Miscs improvements (Interro02)
This commit is contained in:
@@ -7,11 +7,17 @@ from pathlib import Path
|
||||
import pandas as pd
|
||||
from PIL import Image, ImageDraw, ImageFont
|
||||
|
||||
from copienator.configuration import FINAL_SCORE_FONT_PATH, FINAL_SCORE_ODS_PATH, FINAL_SCORE_OUTPUT_DIR
|
||||
from copienator.configuration import (
|
||||
FINAL_SCORE_FONT_PATH,
|
||||
FINAL_SCORE_HISTOGRAM_PATH,
|
||||
FINAL_SCORE_ODS_PATH,
|
||||
FINAL_SCORE_OUTPUT_DIR,
|
||||
)
|
||||
|
||||
# Configuration constants
|
||||
ODS_PATH = Path(FINAL_SCORE_ODS_PATH).expanduser()
|
||||
OUTPUT_DIR = Path(FINAL_SCORE_OUTPUT_DIR).expanduser()
|
||||
HISTOGRAM_PATH = Path(FINAL_SCORE_HISTOGRAM_PATH).expanduser()
|
||||
|
||||
|
||||
def score_font(size):
|
||||
@@ -33,6 +39,35 @@ def get_rounded_score(score):
|
||||
except (ValueError, TypeError):
|
||||
return None
|
||||
|
||||
|
||||
def copy_return_artifacts(source_dir: Path, destination_dir: Path) -> None:
|
||||
"""Copy the metadata and optional individual answers for one student."""
|
||||
for filename in ("score.json", "info.json"):
|
||||
source = source_dir / filename
|
||||
if source.is_file():
|
||||
shutil.copy2(source, destination_dir / filename)
|
||||
else:
|
||||
print(f"Warning: Missing '{source}'.")
|
||||
|
||||
answers_source = source_dir / "answers"
|
||||
answers_destination = destination_dir / "answers"
|
||||
if answers_destination.is_symlink():
|
||||
answers_destination.unlink()
|
||||
elif answers_destination.is_dir():
|
||||
shutil.rmtree(answers_destination)
|
||||
if answers_source.is_dir():
|
||||
shutil.copytree(answers_source, answers_destination)
|
||||
|
||||
|
||||
def copy_histogram(output_dir: Path) -> None:
|
||||
"""Copy the score histogram beside the per-student output folders."""
|
||||
if not HISTOGRAM_PATH.is_file():
|
||||
print(f"Warning: Missing histogram '{HISTOGRAM_PATH}'.")
|
||||
return
|
||||
destination = output_dir / "histogramme.pdf"
|
||||
shutil.copy2(HISTOGRAM_PATH, destination)
|
||||
print(f"Copied histogram: {destination}")
|
||||
|
||||
def process_images(base_dir, output_dir):
|
||||
# 1. Load Data
|
||||
try:
|
||||
@@ -56,60 +91,64 @@ def process_images(base_dir, output_dir):
|
||||
print(f"Error: Directory '{search_path}' not found.")
|
||||
sys.exit(1)
|
||||
|
||||
for img_path in sorted(search_path.glob("*/*.jpg")):
|
||||
student_name = img_path.stem # Filename without extension
|
||||
for student_source in sorted(path for path in search_path.iterdir() if path.is_dir()):
|
||||
image_paths = sorted(student_source.glob("*.jpg"))
|
||||
pdf_paths = sorted(student_source.glob("*.pdf"))
|
||||
media_paths = image_paths or pdf_paths
|
||||
if not media_paths:
|
||||
print(f"Error: No JPG or PDF found in '{student_source}'.")
|
||||
continue
|
||||
|
||||
student_name = media_paths[0].stem
|
||||
student_output = output_dir / student_name
|
||||
student_output.mkdir(parents=True, exist_ok=True)
|
||||
# Remove files produced by the former flat output layout when migrating
|
||||
# an existing export directory.
|
||||
for suffix in (".jpg", ".pdf"):
|
||||
legacy_output = output_dir / f"{student_name}{suffix}"
|
||||
if legacy_output.is_file() or legacy_output.is_symlink():
|
||||
legacy_output.unlink()
|
||||
copy_return_artifacts(student_source, student_output)
|
||||
|
||||
# 4. Find Score
|
||||
if student_name not in score_db:
|
||||
print(f"Error: Student '{student_name}' not found in ODS file.")
|
||||
continue
|
||||
else:
|
||||
raw_score = score_db[student_name]
|
||||
score = get_rounded_score(raw_score)
|
||||
|
||||
raw_score = score_db[student_name]
|
||||
score = get_rounded_score(raw_score)
|
||||
if score is None:
|
||||
print(f"Error: Invalid score '{raw_score}' for '{student_name}'.")
|
||||
else:
|
||||
# 5. Process Images
|
||||
for img_path in image_paths:
|
||||
try:
|
||||
with Image.open(img_path) as img:
|
||||
img = img.convert("RGB")
|
||||
draw = ImageDraw.Draw(img)
|
||||
width, _height = img.size
|
||||
|
||||
if score is None:
|
||||
print(f"Error: Invalid score '{raw_score}' for '{student_name}'.")
|
||||
continue
|
||||
font_size = int(width * 0.08)
|
||||
font = score_font(font_size)
|
||||
text = str(score)
|
||||
|
||||
# 5. Process Image
|
||||
try:
|
||||
with Image.open(img_path) as img:
|
||||
img = img.convert("RGB")
|
||||
draw = ImageDraw.Draw(img)
|
||||
width, height = img.size
|
||||
bbox = draw.textbbox((0, 0), text, font=font)
|
||||
text_w = bbox[2] - bbox[0]
|
||||
|
||||
# Dynamic font size (15% of image height)
|
||||
font_size = int(width * 0.08)
|
||||
# 30px padding, top right.
|
||||
x = width - text_w - 30
|
||||
y = 30
|
||||
draw.text((x, y), text, fill=(255, 0, 0), font=font)
|
||||
|
||||
font = score_font(font_size)
|
||||
img.save(student_output / img_path.name)
|
||||
print(f"Processed: {student_name} -> {score}")
|
||||
except Exception as e:
|
||||
print(f"Error processing image for '{student_name}': {e}")
|
||||
|
||||
text = str(score)
|
||||
for pdf_path in pdf_paths:
|
||||
shutil.copy2(pdf_path, student_output / pdf_path.name)
|
||||
|
||||
# Calculate text size and position (Top Right)
|
||||
bbox = draw.textbbox((0, 0), text, font=font)
|
||||
text_w = bbox[2] - bbox[0]
|
||||
text_h = bbox[3] - bbox[1]
|
||||
|
||||
# 30px padding
|
||||
x = width - text_w - 30
|
||||
y = 30
|
||||
|
||||
# Draw Text (Red)
|
||||
draw.text((x, y), text, fill=(255, 0, 0), font=font)
|
||||
|
||||
# Save
|
||||
save_path = output_dir / f"{student_name}.jpg"
|
||||
img.save(save_path)
|
||||
print(f"Processed: {student_name} -> {score}")
|
||||
|
||||
except Exception as e:
|
||||
print(f"Error processing image for '{student_name}': {e}")
|
||||
|
||||
for pdf_path in sorted(search_path.glob("*/*.pdf")):
|
||||
student_name = pdf_path.stem # Filename without extension
|
||||
save_path = output_dir / f"{student_name}.pdf"
|
||||
|
||||
shutil.copy(str(pdf_path), str(save_path))
|
||||
copy_histogram(output_dir)
|
||||
|
||||
|
||||
def main(argv=None):
|
||||
|
||||
@@ -85,6 +85,8 @@ def flush_thread_log(tid=None):
|
||||
# --- Lock for thread-safe file writing ---
|
||||
io_lock = threading.Lock()
|
||||
pro_lock = threading.Lock()
|
||||
group_index_lock = threading.Lock()
|
||||
reserved_group_indices: dict[str, int] = {}
|
||||
pro_count = 0
|
||||
flash_count = 0
|
||||
pro_quota_exhausted = False
|
||||
@@ -155,6 +157,7 @@ def configure_runtime(
|
||||
completed_tasks = []
|
||||
results = {label: [] for _file, label in tasks}
|
||||
thread_logs.clear()
|
||||
reserved_group_indices.clear()
|
||||
pro_count = 0
|
||||
flash_count = 0
|
||||
pro_quota_exhausted = False
|
||||
@@ -294,6 +297,16 @@ def get_next_group_idx(label):
|
||||
if not existing: return 0
|
||||
return max([int(f.stem.split("_")[1]) for f in existing])
|
||||
|
||||
|
||||
def reserve_next_group_idx(label: str) -> int:
|
||||
"""Reserve a unique zero-based group index for this correction run."""
|
||||
with group_index_lock:
|
||||
if label not in reserved_group_indices:
|
||||
reserved_group_indices[label] = get_next_group_idx(label)
|
||||
idx = reserved_group_indices[label]
|
||||
reserved_group_indices[label] = idx + 1
|
||||
return idx
|
||||
|
||||
def handle_label_errors(pid, label, res, pdf_path):
|
||||
"""Handles Gemini labeling errors, moves/copies files, and returns new tasks."""
|
||||
new_tasks = []
|
||||
@@ -334,7 +347,7 @@ def handle_label_errors(pid, label, res, pdf_path):
|
||||
if pdf_path != old_pdf_path:
|
||||
shutil.move(str(pdf_path), str(old_pdf_path))
|
||||
|
||||
idx = get_next_group_idx(new_label)
|
||||
idx = reserve_next_group_idx(new_label)
|
||||
height = grouping.get_pdf_height(str(new_pdf_path))
|
||||
grouping.create_jpg(new_label, idx, [(pid, str(new_pdf_path), height)], GROUPS_DIR)
|
||||
tprint(f"\t\tMaking {new_label} group {idx+1}")
|
||||
@@ -366,7 +379,7 @@ def handle_label_errors(pid, label, res, pdf_path):
|
||||
if not base_add_pdf_path.exists() and not add_pdf_path.exists():
|
||||
shutil.copy(str(pdf_path), str(add_pdf_path))
|
||||
tprint(f"\t\tCopying Copie{pid} : {label} -> {add_label}")
|
||||
idx = get_next_group_idx(add_label)
|
||||
idx = reserve_next_group_idx(add_label)
|
||||
tprint(f"\t\tMaking {add_label} group {idx+1}")
|
||||
height = grouping.get_pdf_height(str(add_pdf_path))
|
||||
grouping.create_jpg(add_label, idx, [(pid, str(add_pdf_path), height)], GROUPS_DIR)
|
||||
@@ -584,7 +597,7 @@ def resolve_delayed_moves():
|
||||
if pdf_path != old_pdf_path:
|
||||
shutil.move(str(pdf_path), str(old_pdf_path))
|
||||
|
||||
idx = get_next_group_idx(target_label)
|
||||
idx = reserve_next_group_idx(target_label)
|
||||
height = grouping.get_pdf_height(str(new_pdf_path))
|
||||
grouping.create_jpg(target_label, idx, [(pid, str(new_pdf_path), height)], GROUPS_DIR)
|
||||
new_tasks.append((str(GROUPS_DIR / target_label / f"Group_{idx+1}.jpg"), target_label, False))
|
||||
@@ -602,7 +615,7 @@ def resolve_delayed_moves():
|
||||
resolved_any = True
|
||||
|
||||
shutil.copy(str(pdf_path), str(add_pdf_path))
|
||||
idx = get_next_group_idx(target_label)
|
||||
idx = reserve_next_group_idx(target_label)
|
||||
height = grouping.get_pdf_height(str(add_pdf_path))
|
||||
grouping.create_jpg(target_label, idx, [(pid, str(add_pdf_path), height)], GROUPS_DIR)
|
||||
new_tasks.append((str(GROUPS_DIR / target_label / f"Group_{idx+1}.jpg"), target_label, False))
|
||||
@@ -688,7 +701,7 @@ def run_configured(args: argparse.Namespace) -> ExitCode:
|
||||
# pdf_path = copie_dir / f"{label}_old.pdf"
|
||||
|
||||
if pdf_path.exists():
|
||||
idx = get_next_group_idx(label)
|
||||
idx = reserve_next_group_idx(label)
|
||||
height = grouping.get_pdf_height(str(pdf_path))
|
||||
grouping.create_jpg(label, idx, [(pid, str(pdf_path), height)], GROUPS_DIR)
|
||||
new_group_path = str(GROUPS_DIR / label / f"Group_{idx+1}.jpg")
|
||||
|
||||
@@ -29,9 +29,20 @@ def build_parser() -> argparse.ArgumentParser:
|
||||
choices=ANNOTATION_CHOICES,
|
||||
help="Annotation directory to use",
|
||||
)
|
||||
parser.add_argument(
|
||||
"--update",
|
||||
action="store_true",
|
||||
help=(
|
||||
"Update only the individual images in existing A Rendre/answers "
|
||||
"directories, matching folders by their trailing copy ID"
|
||||
),
|
||||
)
|
||||
return parser
|
||||
|
||||
|
||||
RETURN_COPY_ID = re.compile(r"\((\d+)\)$")
|
||||
|
||||
|
||||
def _read_expected_names(workspace: EvaluationWorkspace) -> set[str]:
|
||||
names_path = workspace.names_file()
|
||||
if not names_path.exists():
|
||||
@@ -47,6 +58,70 @@ def _read_expected_names(workspace: EvaluationWorkspace) -> set[str]:
|
||||
}
|
||||
|
||||
|
||||
def _annotation_source(
|
||||
workspace: EvaluationWorkspace,
|
||||
annotation_dir_name: str,
|
||||
copy_id: str,
|
||||
) -> Path | None:
|
||||
selected = workspace.root / annotation_dir_name / f"Copie{copy_id}"
|
||||
fallback = workspace.annotation_dir("simple") / f"Copie{copy_id}"
|
||||
for candidate in (selected, fallback):
|
||||
if (candidate / "score.json").is_file() and (
|
||||
(candidate / "Concat.jpg").is_file()
|
||||
or (candidate / "info.json").is_file()
|
||||
):
|
||||
return candidate
|
||||
return None
|
||||
|
||||
|
||||
def update_named_return_answers(
|
||||
workspace: EvaluationWorkspace,
|
||||
annotation_dir_name: str,
|
||||
) -> ExitCode:
|
||||
"""Refresh only answers/ in existing returns, preserving manual names."""
|
||||
workspace.require_directories(annotation_dir_name, "A Rendre")
|
||||
had_errors = False
|
||||
found = False
|
||||
for destination in sorted(workspace.return_dir.iterdir()):
|
||||
if not destination.is_dir():
|
||||
continue
|
||||
match = RETURN_COPY_ID.search(destination.name)
|
||||
if match is None:
|
||||
print(
|
||||
f"Warning: cannot identify a copy ID in {destination.name!r}; skipped",
|
||||
file=sys.stderr,
|
||||
)
|
||||
had_errors = True
|
||||
continue
|
||||
found = True
|
||||
copy_id = match.group(1)
|
||||
source_folder = _annotation_source(
|
||||
workspace, annotation_dir_name, copy_id
|
||||
)
|
||||
if source_folder is None:
|
||||
print(
|
||||
f"Warning: no annotation source found for Copie{copy_id}; skipped",
|
||||
file=sys.stderr,
|
||||
)
|
||||
had_errors = True
|
||||
continue
|
||||
try:
|
||||
publish_answer_returns(
|
||||
workspace.root,
|
||||
source_folder,
|
||||
destination,
|
||||
answers_only=True,
|
||||
)
|
||||
print(f"Updated answers for {destination.name} from Copie{copy_id}")
|
||||
except (OSError, TypeError, ValueError) as exc:
|
||||
print(f"Error updating answers for {destination.name}: {exc}", file=sys.stderr)
|
||||
had_errors = True
|
||||
if not found:
|
||||
print("Warning: no identifiable student folders found in A Rendre", file=sys.stderr)
|
||||
had_errors = True
|
||||
return ExitCode.PARTIAL if had_errors else ExitCode.SUCCESS
|
||||
|
||||
|
||||
def prepare_named_returns(
|
||||
workspace: EvaluationWorkspace,
|
||||
annotation_dir_name: str,
|
||||
@@ -74,9 +149,6 @@ def prepare_named_returns(
|
||||
had_errors = True
|
||||
|
||||
assigned_names: set[str] = set()
|
||||
selected_annotations = workspace.root / annotation_dir_name
|
||||
fallback_annotations = workspace.annotation_dir("simple")
|
||||
|
||||
for name, copy_ids in copies_map.items():
|
||||
if name == "Unknown":
|
||||
print(
|
||||
@@ -92,16 +164,9 @@ def prepare_named_returns(
|
||||
|
||||
safe_name = safe_filename(name)
|
||||
for copy_id in copy_ids:
|
||||
selected = selected_annotations / f"Copie{copy_id}"
|
||||
fallback = fallback_annotations / f"Copie{copy_id}"
|
||||
source_folder = None
|
||||
for candidate in (selected, fallback):
|
||||
if (candidate / "score.json").is_file() and (
|
||||
(candidate / "Concat.jpg").is_file()
|
||||
or (candidate / "info.json").is_file()
|
||||
):
|
||||
source_folder = candidate
|
||||
break
|
||||
source_folder = _annotation_source(
|
||||
workspace, annotation_dir_name, copy_id
|
||||
)
|
||||
if source_folder is None:
|
||||
continue
|
||||
|
||||
@@ -160,7 +225,10 @@ def run(
|
||||
workspace: EvaluationWorkspace,
|
||||
*,
|
||||
annotation_dir: str,
|
||||
update: bool = False,
|
||||
) -> ExitCode:
|
||||
if update:
|
||||
return update_named_return_answers(workspace, annotation_dir)
|
||||
return prepare_named_returns(workspace, annotation_dir)
|
||||
|
||||
|
||||
@@ -172,6 +240,7 @@ def main(argv: Sequence[str] | None = None) -> int:
|
||||
lambda args: run(
|
||||
workspace_from_args(args, repository=Path.cwd()),
|
||||
annotation_dir=args.annotation_dir,
|
||||
update=args.update,
|
||||
),
|
||||
)
|
||||
|
||||
|
||||
@@ -19,6 +19,10 @@ from copienator import (
|
||||
|
||||
WORD_LIST_FILE = Path(__file__).resolve().parents[1] / "data" / "liste_francais.txt"
|
||||
ACCENT_PATTERN = re.compile(r"[éèêëàâäîïôöùûüçœÉÈÊËÀÂÄÎÏÔÖÙÛÜÇŒ]")
|
||||
MATH_PATTERN = re.compile(
|
||||
r"(\$\$.*?\$\$|\$.*?\$|\\\(.*?\\\)|\\\[.*?\\\])",
|
||||
re.DOTALL,
|
||||
)
|
||||
|
||||
|
||||
def build_parser() -> argparse.ArgumentParser:
|
||||
@@ -26,22 +30,37 @@ def build_parser() -> argparse.ArgumentParser:
|
||||
|
||||
|
||||
def escape_latex_underscores(text: str) -> str:
|
||||
r"""Escape underscores outside LaTeX math environments."""
|
||||
math_pattern = re.compile(
|
||||
r"(\$\$.*?\$\$|\$.*?\$|\\\(.*?\\\)|\\\[.*?\\\])",
|
||||
re.DOTALL,
|
||||
)
|
||||
r"""Escape underscores outside math without double-escaping existing ones."""
|
||||
|
||||
def escape_plain(value: str) -> str:
|
||||
# Collapse any existing escape run as well, making cleanup idempotent.
|
||||
return re.sub(r"\\*_", lambda _match: r"\_", value)
|
||||
|
||||
parts: list[str] = []
|
||||
last_end = 0
|
||||
for match in math_pattern.finditer(text):
|
||||
for match in MATH_PATTERN.finditer(text):
|
||||
start, end = match.span()
|
||||
parts.append(text[last_end:start].replace("_", r"\_"))
|
||||
parts.append(escape_plain(text[last_end:start]))
|
||||
parts.append(match.group(0))
|
||||
last_end = end
|
||||
parts.append(text[last_end:].replace("_", r"\_"))
|
||||
parts.append(escape_plain(text[last_end:]))
|
||||
return "".join(parts)
|
||||
|
||||
|
||||
def normalize_overescaped_latex_commands(text: str) -> str:
|
||||
r"""Collapse doubled command escapes inside LaTeX math environments.
|
||||
|
||||
Model responses occasionally contain ``\\mathbb`` after JSON decoding where
|
||||
LaTeX requires ``\mathbb``. A doubled backslash followed by whitespace is a
|
||||
legitimate row break (for example in ``cases``), so it must be preserved.
|
||||
"""
|
||||
|
||||
def normalize_math(match: re.Match[str]) -> str:
|
||||
return re.sub(r"\\\\(?=[A-Za-z{}])", r"\\", match.group(0))
|
||||
|
||||
return MATH_PATTERN.sub(normalize_math, text)
|
||||
|
||||
|
||||
def build_lookup_map(word_list_path: Path = WORD_LIST_FILE) -> dict[str, str]:
|
||||
words = word_list_path.read_text(encoding="utf-8").splitlines()
|
||||
lookup: dict[str, str] = {}
|
||||
@@ -68,8 +87,23 @@ def fix_hex_corruption_safe(text: str) -> str:
|
||||
)
|
||||
|
||||
|
||||
def some_other_replacements(text: str) -> str:
|
||||
return text.replace("\neq", "\\neq").replace("\not", "\\not")
|
||||
def repair_json_escape_corruption(text: str) -> str:
|
||||
r"""Restore observed LaTeX commands consumed as JSON control escapes."""
|
||||
replacements = (
|
||||
("\x0crac", r"\frac"),
|
||||
("\x0ceuille", r"\equiv"),
|
||||
("\theta", r"\theta"),
|
||||
("\times", r"\times"),
|
||||
("\textbackslash ", "\\"),
|
||||
("\negthinspace", r"\negthinspace"),
|
||||
("\neq", r"\neq"),
|
||||
("\not", r"\not"),
|
||||
("∈", r"\ensuremath{\in}"),
|
||||
("⊂", r"\ensuremath{\subset}"),
|
||||
)
|
||||
for broken, repaired in replacements:
|
||||
text = text.replace(broken, repaired)
|
||||
return text
|
||||
|
||||
|
||||
def clean_string(text: str, lookup: dict[str, str]) -> str:
|
||||
@@ -80,7 +114,9 @@ def clean_string(text: str, lookup: dict[str, str]) -> str:
|
||||
text = re.sub(r" \x00{1,2} ", " à ", text)
|
||||
if "\x00" in text:
|
||||
text = fast_fix(text, lookup).replace("\x00", "")
|
||||
return escape_latex_underscores(some_other_replacements(text))
|
||||
text = repair_json_escape_corruption(text)
|
||||
text = normalize_overescaped_latex_commands(text)
|
||||
return escape_latex_underscores(text)
|
||||
|
||||
|
||||
def clean_obj(value: Any, lookup: dict[str, str]) -> Any:
|
||||
|
||||
@@ -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)
|
||||
|
||||
|
||||
@@ -9,6 +9,7 @@ from typing import Any
|
||||
|
||||
from PIL import Image, ImageDraw
|
||||
|
||||
from copienator import configuration
|
||||
from copienator import (
|
||||
EvaluationWorkspace,
|
||||
ExitCode,
|
||||
@@ -29,9 +30,11 @@ from copienator.commands.reading_annotations import (
|
||||
has_significant_notes,
|
||||
)
|
||||
from copienator.filesystem import staged_files
|
||||
from copienator.return_answers import save_return_answer_options
|
||||
|
||||
LabelNotes = dict[str, dict[str, Any]]
|
||||
ScanResult = tuple[dict[str, list[dict[str, Any]]], dict[str, LabelNotes]]
|
||||
SCAN_WORKERS = 2
|
||||
|
||||
|
||||
def get_extra_pdfs_as_images(
|
||||
@@ -191,9 +194,11 @@ def apply_actions_and_regenerate_grouped(
|
||||
output_dir = workspace.root / annotation_dir / f"Copie{student_id}"
|
||||
labels_data = data.get(student_id, {})
|
||||
apply_checkbox_actions(labels_data, actions, logs.append)
|
||||
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", logs.append
|
||||
labels_data, score_path, logs.append
|
||||
)
|
||||
|
||||
selected_labels = selected_labels if selected_labels is not None else set()
|
||||
@@ -350,7 +355,8 @@ def apply_actions_and_regenerate_grouped(
|
||||
atomic_write_json(staging / "refaire_simple_layout.json", simple_layout)
|
||||
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, touched
|
||||
))
|
||||
@@ -364,6 +370,8 @@ def apply_actions_and_regenerate_grouped(
|
||||
[image for group in filtered_groups for image in group]
|
||||
)
|
||||
filtered_image.save(staging / "Concat_F.jpg")
|
||||
if preserve_score_file:
|
||||
logs.append(f" Preserved existing score.json in {output_dir}")
|
||||
logs.append(f" Saved regenerated files in {output_dir}")
|
||||
status = ExitCode.PARTIAL if incomplete else ExitCode.SUCCESS
|
||||
return status, "\n".join(logs)
|
||||
@@ -461,6 +469,9 @@ def run(
|
||||
refaire: bool = False,
|
||||
update_score: bool = False,
|
||||
annotation_dir: str = "BGnot",
|
||||
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", annotation_dir)
|
||||
@@ -470,6 +481,25 @@ def run(
|
||||
workspace.require_files("refaire.json")
|
||||
workspace.require_directories("BRnot")
|
||||
refaire_list, refaire_by_student = _read_refaire(workspace)
|
||||
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)
|
||||
@@ -494,7 +524,10 @@ def run(
|
||||
and path.is_dir()
|
||||
and not path.name.startswith("Copie")
|
||||
]
|
||||
with concurrent.futures.ThreadPoolExecutor(max_workers=6) as executor:
|
||||
# Each worker decodes a full-height returned group and its reference image.
|
||||
# Keep this stage deliberately narrow; answer regeneration below has its own
|
||||
# parallel executor and a much smaller per-task memory footprint.
|
||||
with concurrent.futures.ThreadPoolExecutor(max_workers=SCAN_WORKERS) as executor:
|
||||
futures = [
|
||||
executor.submit(_scan_annotation_directory, path, only_ids)
|
||||
for path in group_dirs
|
||||
@@ -599,7 +632,28 @@ 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
|
||||
|
||||
@@ -615,6 +669,9 @@ def main(argv: Sequence[str] | None = None) -> int:
|
||||
refaire=args.refaire,
|
||||
update_score=args.update_score,
|
||||
annotation_dir=args.annotation_dir,
|
||||
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)
|
||||
|
||||
@@ -39,6 +39,7 @@ RETURN_ANSWERS_ENABLED = False
|
||||
RETURN_ANSWERS_CONTEXT = False
|
||||
RETURN_ANSWERS_QUESTION = True
|
||||
RETURN_ANSWERS_SOLUTION = False
|
||||
FINAL_SCORE_HISTOGRAM_PATH = Path("histogramme.pdf")
|
||||
for _name in dir(_configuration):
|
||||
if not _name.startswith("_"):
|
||||
globals()[_name] = getattr(_configuration, _name)
|
||||
|
||||
@@ -8,9 +8,56 @@ from copienator import atomic_write_json, configuration, read_json, utils
|
||||
from copienator.filesystem import staged_directory
|
||||
from copienator.platform import safe_filename
|
||||
|
||||
RETURN_ANSWER_OPTIONS_FILE = Path(".copienator") / "return_answers.json"
|
||||
|
||||
def publish_answer_returns(root: Path, source: Path, destination: Path) -> None:
|
||||
"""Publish individual reviewed answers and per-question information."""
|
||||
|
||||
def configured_return_answer_options() -> dict[str, bool]:
|
||||
return {
|
||||
"context": bool(configuration.RETURN_ANSWERS_CONTEXT),
|
||||
"question": bool(configuration.RETURN_ANSWERS_QUESTION),
|
||||
"solution": bool(configuration.RETURN_ANSWERS_SOLUTION),
|
||||
}
|
||||
|
||||
|
||||
def save_return_answer_options(
|
||||
root: Path,
|
||||
*,
|
||||
context: bool,
|
||||
question: bool,
|
||||
solution: bool,
|
||||
) -> None:
|
||||
path = Path(root) / RETURN_ANSWER_OPTIONS_FILE
|
||||
path.parent.mkdir(parents=True, exist_ok=True)
|
||||
atomic_write_json(
|
||||
path,
|
||||
{"context": context, "question": question, "solution": solution},
|
||||
)
|
||||
|
||||
|
||||
def load_return_answer_options(root: Path) -> dict[str, bool]:
|
||||
options = configured_return_answer_options()
|
||||
path = Path(root) / RETURN_ANSWER_OPTIONS_FILE
|
||||
if not path.is_file():
|
||||
return options
|
||||
loaded = read_json(path)
|
||||
if not isinstance(loaded, dict):
|
||||
raise ValueError(f"Expected a return-answer options object in {path}")
|
||||
for name in options:
|
||||
if name in loaded:
|
||||
if type(loaded[name]) is not bool:
|
||||
raise ValueError(f"Expected a boolean for {name!r} in {path}")
|
||||
options[name] = loaded[name]
|
||||
return options
|
||||
|
||||
|
||||
def publish_answer_returns(
|
||||
root: Path,
|
||||
source: Path,
|
||||
destination: Path,
|
||||
*,
|
||||
answers_only: bool = False,
|
||||
) -> None:
|
||||
"""Publish reviewed answers, optionally without touching return metadata."""
|
||||
scores = read_json(source / "score.json")
|
||||
if not isinstance(scores, dict):
|
||||
raise ValueError(f"Expected a score object in {source}")
|
||||
@@ -36,6 +83,7 @@ def publish_answer_returns(root: Path, source: Path, destination: Path) -> None:
|
||||
if answers_dir.is_symlink():
|
||||
raise ValueError(f"Expected a real answer directory: {answers_dir}")
|
||||
if configuration.RETURN_ANSWERS_ENABLED:
|
||||
options = load_return_answer_options(root)
|
||||
labels = [label for label, entry in info.items() if entry["present"] and entry["not_empty"]]
|
||||
|
||||
# Import the rendering backend only when individual images are requested.
|
||||
@@ -44,13 +92,13 @@ def publish_answer_returns(root: Path, source: Path, destination: Path) -> None:
|
||||
|
||||
all_labels = utils.read_all_labels(root)
|
||||
with staged_directory(answers_dir) as staging:
|
||||
for index, label in enumerate(sorted(labels, key=utils.natural_key), 1):
|
||||
for label in sorted(labels, key=utils.natural_key):
|
||||
paths = []
|
||||
if configuration.RETURN_ANSWERS_CONTEXT:
|
||||
if options["context"]:
|
||||
paths.extend(utils.pdf_images_of_contexts(root, label, all_labels))
|
||||
if configuration.RETURN_ANSWERS_QUESTION:
|
||||
if options["question"]:
|
||||
paths.append(utils.pdf_image_of_enonce(root, label))
|
||||
if configuration.RETURN_ANSWERS_SOLUTION:
|
||||
if options["solution"]:
|
||||
paths.append(utils.pdf_image_of_solution(root, label))
|
||||
images = []
|
||||
for path in paths:
|
||||
@@ -62,11 +110,16 @@ def publish_answer_returns(root: Path, source: Path, destination: Path) -> None:
|
||||
with Image.open(source / f"{label}.jpg") as answer:
|
||||
images.append(answer.convert("RGB"))
|
||||
image = concatenate(images)
|
||||
# Numbering avoids collisions between sanitized label filenames.
|
||||
image.save(staging / f"{index:03d} - {safe_filename(label)}.jpg")
|
||||
output = staging / f"{safe_filename(label)}.jpg"
|
||||
if output.exists():
|
||||
raise ValueError(
|
||||
f"Answer labels produce the same filename in {answers_dir}: {label}"
|
||||
)
|
||||
image.save(output)
|
||||
elif answers_dir.exists():
|
||||
# Replace the managed directory with an empty one to remove stale exports.
|
||||
with staged_directory(answers_dir):
|
||||
pass
|
||||
atomic_write_json(destination / "info.json", info)
|
||||
(destination / "touched.json").unlink(missing_ok=True)
|
||||
if not answers_only:
|
||||
atomic_write_json(destination / "info.json", info)
|
||||
(destination / "touched.json").unlink(missing_ok=True)
|
||||
|
||||
+21
-6
@@ -160,23 +160,38 @@ def compile_to_pdf(text, output_pdf_path):
|
||||
# env['TEXINPUTS'] = f".:{current_dir}:"
|
||||
|
||||
try:
|
||||
subprocess.run(
|
||||
result = subprocess.run(
|
||||
['pdflatex', '-interaction=nonstopmode', tex_filename],
|
||||
cwd=temp_dir,
|
||||
stdout=subprocess.DEVNULL,
|
||||
stderr=subprocess.DEVNULL,
|
||||
stdout=subprocess.PIPE,
|
||||
stderr=subprocess.STDOUT,
|
||||
text=True,
|
||||
check=False
|
||||
)
|
||||
if "minted" in text:
|
||||
subprocess.run(
|
||||
result = subprocess.run(
|
||||
['pdflatex', '-interaction=nonstopmode', tex_filename],
|
||||
cwd=temp_dir,
|
||||
stdout=subprocess.DEVNULL,
|
||||
stderr=subprocess.DEVNULL,
|
||||
stdout=subprocess.PIPE,
|
||||
stderr=subprocess.STDOUT,
|
||||
text=True,
|
||||
check=False)
|
||||
|
||||
if result.returncode != 0:
|
||||
error_lines = [
|
||||
line.strip() for line in result.stdout.splitlines()
|
||||
if line.lstrip().startswith("!")
|
||||
]
|
||||
detail = f": {error_lines[0]}" if error_lines else ""
|
||||
print(
|
||||
f"Warning: LaTeX compilation failed for {output_pdf_path} "
|
||||
f"(exit code {result.returncode}){detail}"
|
||||
)
|
||||
|
||||
generated_pdf = os.path.join(temp_dir, pdf_filename)
|
||||
if os.path.exists(generated_pdf):
|
||||
shutil.move(generated_pdf, output_pdf_path)
|
||||
else:
|
||||
print(f"Warning: LaTeX compilation produced no PDF for {output_pdf_path}")
|
||||
except Exception as e:
|
||||
print(f"Compilation error for {output_pdf_path}: {e}")
|
||||
|
||||
Reference in New Issue
Block a user