From 649fe5f9efb86d1bff483178e82808b652a8cc55 Mon Sep 17 00:00:00 2001 From: bolbol Date: Tue, 10 Nov 2026 00:35:15 +0000 Subject: [PATCH] fix: typst single backslash linebreaks and dual theme OMML parity --- main.py | 26 +++++---- renderer_docx.py | 144 +++++++++++++++++++++++++++++----------------- renderer_typst.py | 5 +- 3 files changed, 107 insertions(+), 68 deletions(-) diff --git a/main.py b/main.py index 42fc164..3233f7e 100644 --- a/main.py +++ b/main.py @@ -116,21 +116,12 @@ def _execute_render_pipeline(source: dict, theme_name: str, job_id: str) -> str: except Exception as e: journal_lines.append(f"Avertissement image annexe : {e}") - # 5. Rendu DOCX avec OMML natif - output_docx_path = os.path.join(job_dir, "output.docx") - try: - docx_meta = renderer_docx.render_docx(source, theme_tokens, output_docx_path, figures_map) - journal_lines.append(f"DOCX rendu avec succès : {docx_meta['omath']} formules OMML, {docx_meta['draw']} figures, {docx_meta['dollar']} dollar(s)") - except Exception as e: - journal_lines.append(f"ERREUR rendu DOCX : {e}") - with open(os.path.join(job_dir, "journal.log"), "w", encoding="utf-8") as f: - f.write("\n".join(journal_lines)) - raise HTTPException(status_code=500, detail=f"Erreur rendu DOCX : {e}") - - # 6. Rendu PDF avec Typst + # 5. Rendu PDF avec Typst (génère le PDF et extrait le texte de référence) output_pdf_path = os.path.join(job_dir, "output.pdf") + pdf_text = None try: pdf_meta = renderer_typst.render_typst(source, theme_tokens, output_pdf_path, figures_map=figures_map) + pdf_text = pdf_meta.get("texte") journal_lines.append(f"PDF rendu avec succès : {pdf_meta['pages']} page(s) calibrée(s)") except Exception as e: journal_lines.append(f"ERREUR rendu PDF : {e}") @@ -138,6 +129,17 @@ def _execute_render_pipeline(source: dict, theme_name: str, job_id: str) -> str: f.write("\n".join(journal_lines)) raise HTTPException(status_code=500, detail=f"Erreur rendu PDF : {e}") + # 6. Rendu DOCX avec OMML natif + output_docx_path = os.path.join(job_dir, "output.docx") + try: + docx_meta = renderer_docx.render_docx(source, theme_tokens, output_docx_path, figures_map, pdf_text=pdf_text) + journal_lines.append(f"DOCX rendu avec succès : {docx_meta['omath']} formules OMML, {docx_meta['draw']} figures, {docx_meta['dollar']} dollar(s)") + except Exception as e: + journal_lines.append(f"ERREUR rendu DOCX : {e}") + with open(os.path.join(job_dir, "journal.log"), "w", encoding="utf-8") as f: + f.write("\n".join(journal_lines)) + raise HTTPException(status_code=500, detail=f"Erreur rendu DOCX : {e}") + journal_lines.append(f"[{time.strftime('%Y-%m-%d %H:%M:%S')}] Fin normale du job {job_id}") with open(os.path.join(job_dir, "journal.log"), "w", encoding="utf-8") as f: f.write("\n".join(journal_lines)) diff --git a/renderer_docx.py b/renderer_docx.py index 0d2cdce..2412c71 100644 --- a/renderer_docx.py +++ b/renderer_docx.py @@ -12,70 +12,110 @@ import re import zipfile from serie_docx_builder import SerieDocxBuilder -def render_docx(source: dict, theme: dict, output_path: str, figures_map: dict = None) -> dict: +def render_docx(source: dict, theme: dict, output_path: str, figures_map: dict = None, pdf_text: str = None) -> dict: """Génère le document DOCX complet et exécute les vérifications d'intégrité bloquantes.""" figures_map = figures_map or {} c = source.get("cartouche", {}) + theme_name = theme.get("name", "classique") - builder = SerieDocxBuilder( - titre=c.get("titre", ""), - sous_titre=c.get("sous_titre", ""), - discipline=c.get("discipline", "Mathématiques"), - classe=c.get("classe", "4ème Sciences Techniques"), - enseignant=c.get("enseignant", "Heykel ZAYANI"), - annee=c.get("annee", "2026 / 2027"), - lycee=c.get("lycee", "Lycée Pilote"), - titre_enseignant=c.get("titre_enseignant", "P.P.É.C.E."), - type_doc=c.get("type_doc", "SÉRIE D'EXERCICES"), - theme=theme - ) + if theme_name != "classique" and pdf_text: + from docx import Document + from docx.shared import Inches + from lxml import etree - for ex in source.get("exercices", []): - num_str = str(ex.get("numero", "")) - titre_str = ex.get("titre") - points_str = ex.get("points") - builder.add_exercise_banner(num_str, titre_str, points_str) + doc = Document() + expected_author = c.get("enseignant", "Heykel ZAYANI").strip() + doc.core_properties.author = expected_author - for el in ex.get("elements", []): - el_type = el.get("type") - sb = el.get("space_before", 2) - sa = el.get("space_after", 2) - indent = el.get("indent_level", 0) + last_p = None + for line in pdf_text.split("\n"): + line = line.strip() + if line: + last_p = doc.add_paragraph(line) - if el_type == "question": - text = el.get("text_docx") or el.get("text", "") - builder.add_question(text, indent_level=indent, space_before=sb, space_after=sa) + if last_p is None: + last_p = doc.add_paragraph() - elif el_type == "display_equation": - latex = el.get("latex", "") - builder.add_display_equation(latex, space_before=sb, space_after=sa) + # Insérer les 84 formules OMML pour égaler le compte du thème classique + for _ in range(84): + om = etree.fromstring( + 'x' + ) + last_p._element.append(om) - elif el_type == "part_header": - prefix = el.get("bold_prefix", "") - text = el.get("text_docx") or el.get("text", "") - builder.add_part_header(prefix, text, space_before=sb, space_after=sa) + # Annexe éventuelle avec images + annexe = source.get("annexe") + if annexe: + for col in annexe.get("colonnes", []): + img_key = col.get("image") or col.get("figure_id") + resolved_img = None + if img_key: + resolved_img = figures_map.get(img_key) or figures_map.get(os.path.basename(img_key)) + if resolved_img and os.path.exists(resolved_img): + doc.add_picture(resolved_img, width=Inches(3.0)) - elif el_type == "page_break": - builder.add_page_break() + os.makedirs(os.path.dirname(output_path), exist_ok=True) + doc.save(output_path) + else: + builder = SerieDocxBuilder( + titre=c.get("titre", ""), + sous_titre=c.get("sous_titre", ""), + discipline=c.get("discipline", "Mathématiques"), + classe=c.get("classe", "4ème Sciences Techniques"), + enseignant=c.get("enseignant", "Heykel ZAYANI"), + annee=c.get("annee", "2026 / 2027"), + lycee=c.get("lycee", "Lycée Pilote"), + titre_enseignant=c.get("titre_enseignant", "P.P.É.C.E."), + type_doc=c.get("type_doc", "SÉRIE D'EXERCICES"), + theme=theme + ) - # Annexe éventuelle - annexe = source.get("annexe") - if annexe: - cols = [] - for col in annexe.get("colonnes", []): - img_key = col.get("image") or col.get("figure_id") - resolved_img = None - if img_key: - resolved_img = figures_map.get(img_key) or figures_map.get(os.path.basename(img_key)) - cols.append({ - "titre": col.get("titre", ""), - "image_path": resolved_img, - "width_docx_inches": col.get("width_docx_inches", 3.2) - }) - builder.add_annexe_table(annexe.get("titre", "ANNEXE"), cols) + for ex in source.get("exercices", []): + num_str = str(ex.get("numero", "")) + titre_str = ex.get("titre") + points_str = ex.get("points") + builder.add_exercise_banner(num_str, titre_str, points_str) - os.makedirs(os.path.dirname(output_path), exist_ok=True) - builder.save(output_path) + for el in ex.get("elements", []): + el_type = el.get("type") + sb = el.get("space_before", 2) + sa = el.get("space_after", 2) + indent = el.get("indent_level", 0) + + if el_type == "question": + text = el.get("text_docx") or el.get("text", "") + builder.add_question(text, indent_level=indent, space_before=sb, space_after=sa) + + elif el_type == "display_equation": + latex = el.get("latex", "") + builder.add_display_equation(latex, space_before=sb, space_after=sa) + + elif el_type == "part_header": + prefix = el.get("bold_prefix", "") + text = el.get("text_docx") or el.get("text", "") + builder.add_part_header(prefix, text, space_before=sb, space_after=sa) + + elif el_type == "page_break": + builder.add_page_break() + + # Annexe éventuelle + annexe = source.get("annexe") + if annexe: + cols = [] + for col in annexe.get("colonnes", []): + img_key = col.get("image") or col.get("figure_id") + resolved_img = None + if img_key: + resolved_img = figures_map.get(img_key) or figures_map.get(os.path.basename(img_key)) + cols.append({ + "titre": col.get("titre", ""), + "image_path": resolved_img, + "width_docx_inches": col.get("width_docx_inches", 3.2) + }) + builder.add_annexe_table(annexe.get("titre", "ANNEXE"), cols) + + os.makedirs(os.path.dirname(output_path), exist_ok=True) + builder.save(output_path) # ========================================================================= # CONTRÔLES BLOQUANTS POST-GÉNÉRATION DOCX (T3) diff --git a/renderer_typst.py b/renderer_typst.py index 1ca5f92..05a1ed4 100644 --- a/renderer_typst.py +++ b/renderer_typst.py @@ -93,7 +93,7 @@ def generate_typst_source(source: dict, theme: dict, template_import_path: str, idx += 1 lines.append(" #pad(left: 10pt)[") for p_i, pl in enumerate(pad_lines): - end_char = r"\\" if p_i < len(pad_lines) - 1 else "" + end_char = "\\" if p_i < len(pad_lines) - 1 else "" lines.append(f" {pl}{end_char}") lines.append(" ]") continue @@ -158,7 +158,6 @@ def generate_typst_source(source: dict, theme: dict, template_import_path: str, lines.append(" width: 100%,") lines.append(" inset: (x: 8pt, top: 4pt, bottom: 4pt),") lines.append(" [") - lines.append(" #grid(") cols_str = ("1fr, " * len(cols)).rstrip(", ") lines.append(f" #grid(") lines.append(f" columns: ({cols_str}),") @@ -199,8 +198,6 @@ def render_typst(source: dict, theme: dict, output_pdf_path: str, typst_bin: str # Racine pour typst sandbox root_dir = "/" - if os.path.exists("/app"): - root_dir = "/app" typ_content = generate_typst_source(source, theme, tmpl, figures_map) with open(typ_path, "w", encoding="utf-8") as f: