Escape some newlines in question extraction

copienator/commands/clean.py

@@ -109,7 +109,10 @@ def _return_artifacts(workspace: EvaluationWorkspace) -> list[Path]:
 def _is_statement_text_file(workspace: EvaluationWorkspace, path: Path) -> bool:
     relative = path.relative_to(workspace.root)
     if len(relative.parts) == 1:
-        return relative.name in STATEMENT_ROOT_FILES
+        return relative.name in STATEMENT_ROOT_FILES or path in {
+            *workspace.input_candidates("enonce.tex"),
+            *workspace.input_candidates("correction.tex"),
+        }
     top_level = relative.parts[0]
     if top_level in STATEMENT_TEXT_DIRECTORIES:
         return path.suffix.casefold() in {"", ".json", ".tex", ".txt"}

copienator/commands/copies_tools.py

@@ -24,7 +24,7 @@ def copy_pdfs(directory: Path) -> list[Path]:
         (
             path
             for path in directory.glob("*.pdf")
-            if path.name.casefold() not in {"enonce.pdf", "énoncé.pdf"}
+            if not EvaluationWorkspace(directory).is_statement_pdf(path)
         ),
         key=lambda path: path.name.casefold(),
     )
@@ -134,4 +134,3 @@ def main(argv: Sequence[str] | None = None) -> int:
 
 if __name__ == "__main__":
     raise SystemExit(main())
-

copienator/commands/enonce_info.py

@@ -143,7 +143,7 @@ def save_split_content(text, path, base_fname, problem):
 def process_directory(workspace: EvaluationWorkspace) -> ExitCode:
     directory = str(workspace.root)
     # Find the first .tex file in the directory
-    enonce = workspace.root / "enonce.tex"
+    enonce = workspace.statement_source_file
     tex_files = [str(enonce)] if enonce.is_file() else sorted(glob.glob(os.path.join(directory, "*.tex")))
     if not tex_files:
         print(f"No .tex file found in {directory}. Looking in /Staging/Interro/")

copienator/commands/gemini_for_enonce.py

@@ -8,7 +8,7 @@ from pathlib import Path
 
 from google import genai
 from google.genai import types
-from pydantic import BaseModel, Field
+from pydantic import BaseModel, Field, field_validator
 
 from copienator import configuration as config
 from copienator import utils
@@ -22,7 +22,7 @@ from copienator import (
     workspace_from_args,
 )
 from copienator.platform import validate_windows_labels
-from copienator.latex_json import repair_latex_json
+from copienator.latex_json import repair_latex_json, normalize_literal_newlines
 from copienator.filesystem import staged_directory
 from copienator.utils import compile_to_pdf
 
@@ -43,7 +43,15 @@ MODEL_ID = config.MODEL_LITE_ID
 api_key = config.API_KEY
 
 # --- Modèles pour la Requête 1 ---
-class QuestionOnlyItem(BaseModel):
+class StatementContent(BaseModel):
+    @field_validator("question_content", "solution_content", "context_content",
+                     "rubric_content", "content", check_fields=False)
+    @classmethod
+    def normalize_newlines(cls, value: str) -> str:
+        return normalize_literal_newlines(value)
+
+
+class QuestionOnlyItem(StatementContent):
     label: str = Field(description="Label unique de la question (par exemple '1.a' ou 'Exercice 1').")
     question_content: str = Field(description="Texte source de la question, extrait exactement du fichier d’énoncé, SANS le label lui-même.")
 
@@ -51,7 +59,7 @@ class ExamQuestions(BaseModel):
     questions: list[QuestionOnlyItem]
 
 # --- Modèles pour la Requête 2 ---
-class SolutionOnlyItem(BaseModel):
+class SolutionOnlyItem(StatementContent):
     label: str = Field(description="Label exact de la question fourni en entrée, à conserver sans traduction.")
     solution_content: str = Field(description="Texte source de la solution, extrait exactement du fichier de correction.")
 
@@ -59,7 +67,7 @@ class ExamSolutions(BaseModel):
     solutions: list[SolutionOnlyItem]
 
 # --- Modèles pour la Requête 3 ---
-class ExtractedContext(BaseModel):
+class ExtractedContext(StatementContent):
     target_question_label: str = Field(description="Label exact de la PREMIÈRE question située immédiatement APRÈS cette information dans l’énoncé.")
     last_question_label: str = Field(description="Label exact de la DERNIÈRE question qui utilise cette information.")
     context_content: str = Field(description="Texte source des définitions, notations ou hypothèses, extrait de l’énoncé.")
@@ -68,7 +76,7 @@ class ExamContext(BaseModel):
     contexts: list[ExtractedContext]
 
 # --- Modèles pour la Requête 4 (Barèmes) ---
-class RubricItem(BaseModel):
+class RubricItem(StatementContent):
     label: str = Field(description="Label exact de la question, à conserver sans traduction.")
     rubric_content: str = Field(description="Barème détaillé sur 4 points, en français, avec formules LaTeX. Échapper chaque antislash LaTeX conformément à la syntaxe JSON.")
 
@@ -110,6 +118,8 @@ Exemple d’échappement dans un barème (adapte les clés au schéma demandé)
 Ne confonds pas les commandes LaTeX avec les échappements JSON :
 \\frac, \\rightarrow, \\not, \\theta et \\begin sont des commandes LaTeX ;
 \n est réservé à un véritable saut de ligne du texte.
+N'écris jamais \\n pour un saut de ligne : après décodage JSON, le texte
+doit contenir un véritable saut de ligne, pas une commande LaTeX \n.
 """
 
 
@@ -124,12 +134,12 @@ def validate_rubric_content(content: str) -> str:
     return content
 
 # --- Modèle fusionné (pour le reste du script) ---
-class QuestionItem(BaseModel):
+class QuestionItem(StatementContent):
     label: str
     question_content: str
     solution_content: str
 
-class ContextItem(BaseModel):
+class ContextItem(StatementContent):
     target_question_label: str
     last_question_label: str
     content: str  # Juste une string encapsulée pour le différencier facilement
@@ -331,14 +341,18 @@ def process_exam(
     cache_c_file = cache_dir / "gemini_context.json"
 
     # 1. Resolve files
-    pdf_path = folder / "enonce.pdf"
+    pdf_path = workspace.statement_pdf_file
     enonce_path = find_file(folder, "enonce")
+    if enonce_path is None:
+        enonce_path = workspace.statement_source_file
     correction_path = find_file(folder, "correction")
+    if correction_path is None:
+        correction_path = workspace.correction_source_file
 
     missing = []
     if not pdf_path.is_file(): missing.append("enonce.pdf")
-    if not enonce_path: missing.append("enonce.org or enonce.tex")
-    if not correction_path: missing.append("correction.org or correction.tex")
+    if not enonce_path.is_file(): missing.append("enonce.tex")
+    if not correction_path.is_file(): missing.append("correction.tex")
 
     if missing:
         raise CliError(
@@ -911,7 +925,7 @@ def process_exam(
         with open(text_dir / safe_group_filename, "w", encoding="utf-8") as f_text:
             f_text.write("\n".join(text_content_lines))
 
-    print(f"Success! Processed {len(grouped_extraction.groups)} groups.")
+    print(f"Wrote {len(grouped_extraction.groups)} groups; checking PDF compilation...")
     # ==========================================
     # PDF COMPILATION (4 Threads)
     # ==========================================
@@ -922,7 +936,7 @@ def process_exam(
         try:
             content = tex_file.read_text(encoding="utf-8")
             pdf_path = tex_file.with_suffix(".pdf")
-            compile_to_pdf(content, pdf_path)
+            compile_to_pdf(content, pdf_path, strict=True)
         except Exception as e:  # noqa: BLE001 - compiler worker boundary
             return f"Error compiling {tex_file.name}: {e}"
         return None

copienator/commands/plotting.py

@@ -476,17 +476,11 @@ class ImageViewer:
     def on_open_interro(self, event):
         if self.is_viewing and self.current_json_path:
         # Check local directory first
-            local_accent = self.base_dir / "énoncé.pdf"
-            local_plain = self.base_dir / "enonce.pdf"
-
-            if local_accent.exists():
-                pdf_path = str(local_accent)
-            elif local_plain.exists():
-                pdf_path = local_plain
-            else:
+            pdf_path = EvaluationWorkspace(self.base_dir).statement_pdf_file
+            if not pdf_path.is_file():
                 messagebox.showerror(
                     "PDF not found",
-                    f"Neither {local_accent.name} nor {local_plain.name} exists in {self.base_dir}.",
+                    f"No statement PDF exists in {self.base_dir}.",
                 )
                 return
 

copienator/utils.py

@@ -142,7 +142,7 @@ import subprocess
 from .configuration import LATEX_AFTER, LATEX_BEFORE
 
 
-def compile_to_pdf(text, output_pdf_path):
+def compile_to_pdf(text, output_pdf_path, *, strict=False):
     """Wraps text in standalone header/footer templates and compiles it to PDF."""
     latex_template = f"{LATEX_BEFORE}{text}{LATEX_AFTER}"
 
@@ -166,6 +166,7 @@ def compile_to_pdf(text, output_pdf_path):
                 stdout=subprocess.PIPE,
                 stderr=subprocess.STDOUT,
                 text=True,
+                encoding="utf-8", errors="replace",
                 check=False
             )
             if "minted" in text:
@@ -175,6 +176,7 @@ def compile_to_pdf(text, output_pdf_path):
                     stdout=subprocess.PIPE,
                     stderr=subprocess.STDOUT,
                     text=True,
+                    encoding="utf-8", errors="replace",
                     check=False)
 
             generated_pdf = os.path.join(temp_dir, pdf_filename)
@@ -183,17 +185,21 @@ def compile_to_pdf(text, output_pdf_path):
                 shutil.move(generated_pdf, output_pdf_path)
 
             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 ""
+                lines = result.stdout.splitlines()
+                first = next((i for i, line in enumerate(lines) if line.lstrip().startswith("!")), None)
+                detail = "\n" + "\n".join(lines[first:first + 5]) if first is not None else ""
                 pdf_status = "a PDF was generated" if pdf_generated else "no PDF was generated"
                 print(
                     f"Warning: LaTeX compilation reported errors for {output_pdf_path} "
                     f"(exit code {result.returncode}; {pdf_status}){detail}"
                 )
+                if strict:
+                    raise RuntimeError(f"LaTeX failed for {output_pdf_path}{detail}")
             elif not pdf_generated:
                 print(f"Warning: LaTeX compilation produced no PDF for {output_pdf_path}")
+                if strict:
+                    raise RuntimeError(f"No PDF generated for {output_pdf_path}")
         except Exception as e:
+            if strict:
+                raise
             print(f"Compilation error for {output_pdf_path}: {e}")

copienator/workspace.py

@@ -63,12 +63,49 @@ class EvaluationWorkspace:
             ".copienator",
             ".copienator-gui.json",
         )
-        return path.is_dir() and any((path / marker).exists() for marker in markers)
+        if not path.is_dir():
+            return False
+        aliases = (f"{path.name}.pdf", f"{path.name}.tex", f"{path.name}c.tex")
+        return any((path / marker).exists() for marker in (*markers, *aliases))
 
     @property
     def name(self) -> str:
         return self.root.name
 
+    def input_candidates(self, canonical_name: str) -> tuple[Path, ...]:
+        """Return preferred paths for one of the three statement inputs."""
+        aliases = {
+            "enonce.pdf": ("enonce.pdf", "énoncé.pdf", f"{self.name}.pdf"),
+            "enonce.tex": ("enonce.tex", f"{self.name}.tex"),
+            "correction.tex": ("correction.tex", f"{self.name}c.tex"),
+        }
+        names = aliases.get(canonical_name, (canonical_name,))
+        return tuple(self.root / name for name in names)
+
+    def resolve(self, relative_path: str) -> Path:
+        """Resolve a logical workspace path, including statement aliases."""
+        candidates = self.input_candidates(relative_path)
+        return next((path for path in candidates if path.exists()), candidates[0])
+
+    def resolve_file(self, relative_path: str) -> Path:
+        candidates = self.input_candidates(relative_path)
+        return next((path for path in candidates if path.is_file()), candidates[0])
+
+    @property
+    def statement_pdf_file(self) -> Path:
+        return self.resolve_file("enonce.pdf")
+
+    @property
+    def statement_source_file(self) -> Path:
+        return self.resolve_file("enonce.tex")
+
+    @property
+    def correction_source_file(self) -> Path:
+        return self.resolve_file("correction.tex")
+
+    def is_statement_pdf(self, path: Path) -> bool:
+        return path in self.input_candidates("enonce.pdf")
+
     @property
     def metadata_dir(self) -> Path:
         return self.root / ".copienator"
@@ -220,7 +257,7 @@ class EvaluationWorkspace:
         missing = [
             relative_path
             for relative_path in relative_paths
-            if not (self.root / relative_path).exists()
+            if not self.resolve(relative_path).exists()
         ]
         if missing:
             raise WorkspaceValidationError(self.root, missing)
@@ -229,7 +266,7 @@ class EvaluationWorkspace:
         missing = [
             relative_path
             for relative_path in relative_paths
-            if not (self.root / relative_path).is_file()
+            if not self.resolve_file(relative_path).is_file()
         ]
         if missing:
             raise WorkspaceValidationError(self.root, missing)

copienator_gui/app.py

@@ -173,7 +173,7 @@ def copy_pdf_paths(evaluation: Path) -> list[Path]:
             (
                 path
                 for path in location.glob("*.pdf")
-                if path.name.casefold() not in {"enonce.pdf", "énoncé.pdf"}
+                if not EvaluationWorkspace(evaluation).is_statement_pdf(path)
             ),
             key=lambda path: path.name.casefold(),
         )
@@ -578,7 +578,8 @@ class CopienatorApp(tk.Tk):
             if self.state_store.step(step.id).get("status") == "running":
                 self.state_store.update_step(step.id, status="interrupted")
         self._populate_tree()
-        missing = [name for name in ("enonce.pdf", "enonce.tex", "correction.tex") if not (evaluation / name).exists()]
+        workspace = EvaluationWorkspace(evaluation)
+        missing = [name for name in ("enonce.pdf", "enonce.tex", "correction.tex") if not workspace.resolve(name).exists()]
         names_found = (evaluation / "names").exists() or (self.repository / "names").exists()
         details = []
         if missing:
@@ -678,7 +679,8 @@ class CopienatorApp(tk.Tk):
             if any(char in pattern for char in "*?["):
                 exists = next(evaluation.glob(pattern), None) is not None
             else:
-                path = EvaluationWorkspace(evaluation).annotation_dir("refaire") if pattern == "BRnot" else evaluation / pattern
+                workspace = EvaluationWorkspace(evaluation)
+                path = workspace.annotation_dir("refaire") if pattern == "BRnot" else workspace.resolve(pattern)
                 exists = path.exists()
             if not exists:
                 missing.append(pattern)

copienator_gui/workflow.py

@@ -126,7 +126,8 @@ def build_workflow(show_personal_steps: bool) -> list[StepDefinition]:
             "inputs",
             "Préparation",
             "Vérifier les fichiers d’entrée",
-            "Le dossier doit contenir enonce.pdf, enonce.tex et correction.tex. "
+            "Le dossier doit contenir enonce.pdf, enonce.tex et correction.tex, ou, "
+            "s’il s’appelle A, les fichiers A.pdf, A.tex et Ac.tex. "
             "Le fichier names doit être présent dans l’évaluation ou à la racine du projet.",
             (manual("check"),),
             requires=("enonce.pdf", "enonce.tex", "correction.tex"),

default_config.py

@@ -28,7 +28,8 @@ FINAL_SCORE_HISTOGRAM_PATH = Path("histogramme.pdf")
 FINAL_SCORE_FONT_PATH = None
 
 # Modèle pour des choses très légères
-MODEL_LITE_ID = "gemini-3.5-flash-lite"
+# MODEL_LITE_ID = "gemini-3.5-flash-lite"
+MODEL_LITE_ID = "gemini-3.8-flash"
 
 # Modèle pour identifier visuellement les labels
 # MODEL_FOR_LABEL_ID = "gemini-3.5-flash-lite" # 3.5 flash lite marche moyennement, il insiste pour mettre les labels dans l'ordre, sans considérer ce qu'il y a écrit

rename_to_copie.sh

@@ -8,13 +8,14 @@ fi
 
 # Go to the directory
 cd "$1" || exit
+folder_name="$(basename "$PWD")"
 
 count=1
 for file in *.pdf; do
     # Handle case where no pdfs exist
     [ -e "$file" ] || continue
 
-    if [ "$file" = "enonce.pdf" ]; then
+    if [ "$file" = "enonce.pdf" ] || [ "$file" = "énoncé.pdf" ] || [ "$file" = "$folder_name.pdf" ]; then
         echo "Skipping: $file"
         continue
     fi

rotate_all.sh

@@ -8,12 +8,13 @@ fi
 
 # Try to change into the directory, exit on failure
 cd "$1" || { echo "Error: Cannot access directory '$1'"; exit 1; }
+folder_name="$(basename "$PWD")"
 
 # Enable nullglob: if no pdfs exist, the loop won't run once with "*.pdf"
 shopt -s nullglob
 
 for file in *.pdf; do
-    if [ "$file" = "enonce.pdf" ]; then
+    if [ "$file" = "enonce.pdf" ] || [ "$file" = "énoncé.pdf" ] || [ "$file" = "$folder_name.pdf" ]; then
         echo "Skipping: $file"
         continue
     fi

tests/test_gui_core.py

@@ -108,6 +108,33 @@ class WorkspaceTests(unittest.TestCase):
             workspace = EvaluationWorkspace.discover(artifact)
             self.assertEqual(workspace.root, root)
 
+    def test_folder_named_statement_inputs_are_logical_canonical_files(self) -> None:
+        with tempfile.TemporaryDirectory() as directory:
+            root = Path(directory) / "A"
+            root.mkdir()
+            for name in ("A.pdf", "A.tex", "Ac.tex"):
+                (root / name).touch()
+
+            workspace = EvaluationWorkspace.discover(root)
+            workspace.require_files("enonce.pdf", "enonce.tex", "correction.tex")
+
+            self.assertEqual(workspace.statement_pdf_file, root / "A.pdf")
+            self.assertEqual(workspace.statement_source_file, root / "A.tex")
+            self.assertEqual(workspace.correction_source_file, root / "Ac.tex")
+
+    def test_canonical_statement_inputs_take_precedence_over_folder_aliases(self) -> None:
+        with tempfile.TemporaryDirectory() as directory:
+            root = Path(directory) / "A"
+            root.mkdir()
+            for name in ("enonce.pdf", "A.pdf", "enonce.tex", "A.tex",
+                         "correction.tex", "Ac.tex"):
+                (root / name).touch()
+
+            workspace = EvaluationWorkspace(root)
+            self.assertEqual(workspace.statement_pdf_file, root / "enonce.pdf")
+            self.assertEqual(workspace.statement_source_file, root / "enonce.tex")
+            self.assertEqual(workspace.correction_source_file, root / "correction.tex")
+
     def test_discover_and_require_report_clear_errors(self) -> None:
         with tempfile.TemporaryDirectory() as directory:
             root = Path(directory)
@@ -2574,8 +2601,9 @@ class WorkflowTests(unittest.TestCase):
 class CrossPlatformFileTests(unittest.TestCase):
     def test_rotate_all_skips_statement(self) -> None:
         with tempfile.TemporaryDirectory() as directory:
-            root = Path(directory)
-            for filename in ("Copie.pdf", "enonce.pdf"):
+            root = Path(directory) / "A"
+            root.mkdir()
+            for filename in ("Copie.pdf", "A.pdf"):
                 writer = PdfWriter()
                 writer.add_blank_page(width=100, height=200)
                 with (root / filename).open("wb") as output:
@@ -2583,22 +2611,23 @@ class CrossPlatformFileTests(unittest.TestCase):
 
             self.assertEqual(rotate_all(root), 1)
             self.assertEqual(PdfReader(root / "Copie.pdf").pages[0].rotation, 180)
-            self.assertEqual(PdfReader(root / "enonce.pdf").pages[0].rotation, 0)
+            self.assertEqual(PdfReader(root / "A.pdf").pages[0].rotation, 0)
 
     def test_rename_all_is_collision_safe(self) -> None:
         with tempfile.TemporaryDirectory() as directory:
-            root = Path(directory)
+            root = Path(directory) / "A"
+            root.mkdir()
             (root / "z.pdf").write_bytes(b"z")
             (root / "a.pdf").write_bytes(b"a")
             (root / "Copie01.pdf").write_bytes(b"existing")
-            (root / "enonce.pdf").write_bytes(b"statement")
+            (root / "A.pdf").write_bytes(b"statement")
 
             rename_all(root)
 
             self.assertEqual((root / "Copie01.pdf").read_bytes(), b"a")
             self.assertEqual((root / "Copie02.pdf").read_bytes(), b"existing")
             self.assertEqual((root / "Copie03.pdf").read_bytes(), b"z")
-            self.assertEqual((root / "enonce.pdf").read_bytes(), b"statement")
+            self.assertEqual((root / "A.pdf").read_bytes(), b"statement")
 
     def test_link_falls_back_to_copy(self) -> None:
         with tempfile.TemporaryDirectory() as directory:

tests/test_utils_compile.py

@@ -10,6 +10,16 @@ from copienator import utils
 
 
 class CompileToPdfTests(unittest.TestCase):
+    def test_strict_compilation_propagates_failure_with_context(self):
+        result = subprocess.CompletedProcess(args=[], returncode=1,
+            stdout="! Undefined control sequence.\nl.13 broken command\n")
+        with tempfile.TemporaryDirectory() as directory, \
+                patch.object(utils.subprocess, "run", return_value=result) as run:
+            with self.assertRaisesRegex(RuntimeError, "LaTeX failed") as error:
+                utils.compile_to_pdf("broken", Path(directory) / "out.pdf", strict=True)
+            self.assertIn("l.13 broken command", str(error.exception))
+            self.assertEqual(run.call_args.kwargs["errors"], "replace")
+
     def test_warns_when_latex_fails_even_if_partial_pdf_exists(self):
         with tempfile.TemporaryDirectory() as temporary:
             output = Path(temporary) / "result.pdf"