Upload hf_backend/ebook_convert.py with huggingface_hub

hf_backend/ebook_convert.py

@@ -0,0 +1,51 @@
+from __future__ import annotations
+
+import os
+import subprocess
+from pathlib import Path
+
+
+def normalize_epub_for_translation(config, source_path: Path, work_dir: Path) -> Path:
+    convert_dir = work_dir / "converted"
+    convert_dir.mkdir(parents=True, exist_ok=True)
+
+    mobi_path = convert_dir / f"{source_path.stem}.mobi"
+    normalized_epub = convert_dir / source_path.name
+    env = os.environ.copy()
+    env["PYTHONNOUSERSITE"] = "1"
+
+    epub_to_mobi = subprocess.run(
+        [config.ebook_convert_bin, str(source_path), str(mobi_path)],
+        check=False,
+        capture_output=True,
+        text=True,
+        encoding="utf-8",
+        errors="replace",
+        env=env,
+    )
+    if epub_to_mobi.returncode != 0 or not mobi_path.exists():
+        output = "\n".join(
+            part for part in [epub_to_mobi.stdout.strip(), epub_to_mobi.stderr.strip()] if part
+        )
+        raise RuntimeError(
+            f"ebook-convert EPUB->MOBI failed for {source_path.name}\n{output[-3000:]}"
+        )
+
+    mobi_to_epub = subprocess.run(
+        [config.ebook_convert_bin, str(mobi_path), str(normalized_epub)],
+        check=False,
+        capture_output=True,
+        text=True,
+        encoding="utf-8",
+        errors="replace",
+        env=env,
+    )
+    if mobi_to_epub.returncode != 0 or not normalized_epub.exists():
+        output = "\n".join(
+            part for part in [mobi_to_epub.stdout.strip(), mobi_to_epub.stderr.strip()] if part
+        )
+        raise RuntimeError(
+            f"ebook-convert MOBI->EPUB failed for {source_path.name}\n{output[-3000:]}"
+        )
+
+    return normalized_epub