251 lines
9.5 KiB
Python
251 lines
9.5 KiB
Python
#!/usr/bin/env python3
|
|
"""
|
|
pipeline.py
|
|
───────────────────────────────────────────────────────────────
|
|
Full chapter translation pipeline for Dandadan_059_2022_Digital
|
|
|
|
Flow per page:
|
|
1. Run manga-translator.py → output.txt + bubbles.json
|
|
2. Run manga-renderer.py → translated image
|
|
3. Collect all translated images → .cbz
|
|
|
|
Folder structure produced:
|
|
Dandadan_059_2022_Digital_1r0n/
|
|
├── 00.jpg ← original (untouched)
|
|
├── ...
|
|
└── translated/
|
|
├── 00/
|
|
│ ├── output.txt ← raw translations
|
|
│ ├── bubbles.json ← bubble boxes
|
|
│ ├── debug_clusters.png ← cluster debug
|
|
│ └── 00_translated.jpg ← rendered output
|
|
├── 01/
|
|
│ └── ...
|
|
└── ...
|
|
Dandadan_059_translated.cbz ← final output
|
|
"""
|
|
|
|
import os
|
|
import sys
|
|
import shutil
|
|
import zipfile
|
|
import importlib.util
|
|
from pathlib import Path
|
|
|
|
|
|
# ─────────────────────────────────────────────
|
|
# CONFIG — edit these as needed
|
|
# ─────────────────────────────────────────────
|
|
CHAPTER_DIR = "/Users/guillemhernandezsola/Downloads/Dandadan_059_2022_Digital_1r0n"
|
|
OUTPUT_CBZ = "/Users/guillemhernandezsola/Downloads/Dandadan_059_translated.cbz"
|
|
SOURCE_LANG = "en"
|
|
TARGET_LANG = "ca"
|
|
|
|
# manga-translator.py settings
|
|
CONFIDENCE_THRESHOLD = 0.10
|
|
MIN_TEXT_LENGTH = 2
|
|
CLUSTER_EPS = "auto"
|
|
PROXIMITY_PX = 80
|
|
FILTER_SFX = True
|
|
QUALITY_THRESHOLD = 0.5
|
|
UPSCALE_FACTOR = 2.5
|
|
BBOX_PADDING = 5
|
|
DEBUG = True
|
|
|
|
# manga-renderer.py settings
|
|
FONT_PATH = "fonts/ComicRelief-Regular.ttf"
|
|
FONT_COLOR = (0, 0, 0)
|
|
|
|
|
|
# ─────────────────────────────────────────────
|
|
# DYNAMIC MODULE LOADER
|
|
# Loads manga-translator.py and manga-renderer.py
|
|
# by file path (handles hyphens in filenames)
|
|
# ─────────────────────────────────────────────
|
|
def load_module(name, filepath):
|
|
spec = importlib.util.spec_from_file_location(name, filepath)
|
|
module = importlib.util.module_from_spec(spec)
|
|
spec.loader.exec_module(module)
|
|
return module
|
|
|
|
|
|
# ─────────────────────────────────────────────
|
|
# HELPERS
|
|
# ─────────────────────────────────────────────
|
|
def sorted_pages(chapter_dir):
|
|
"""
|
|
Returns sorted list of image paths in the chapter folder.
|
|
Supports .jpg, .jpeg, .png, .webp
|
|
"""
|
|
exts = {".jpg", ".jpeg", ".png", ".webp"}
|
|
pages = [
|
|
p for p in Path(chapter_dir).iterdir()
|
|
if p.suffix.lower() in exts
|
|
]
|
|
return sorted(pages, key=lambda p: p.stem)
|
|
|
|
|
|
def make_page_workdir(chapter_dir, page_stem):
|
|
"""
|
|
Creates and returns:
|
|
<chapter_dir>/translated/<page_stem>/
|
|
"""
|
|
workdir = Path(chapter_dir) / "translated" / page_stem
|
|
workdir.mkdir(parents=True, exist_ok=True)
|
|
return workdir
|
|
|
|
|
|
def pack_cbz(translated_dir, output_cbz):
|
|
"""
|
|
Collects all *_translated.* images from translated/*/
|
|
sorted by page stem, packs into a .cbz file.
|
|
"""
|
|
images = sorted(
|
|
translated_dir.rglob("*_translated.*"),
|
|
key=lambda p: p.parent.name
|
|
)
|
|
|
|
if not images:
|
|
print("⚠️ No translated images found — CBZ not created.")
|
|
return
|
|
|
|
with zipfile.ZipFile(output_cbz, "w",
|
|
compression=zipfile.ZIP_STORED) as zf:
|
|
for img in images:
|
|
# Archive name keeps the page stem for ordering
|
|
arcname = img.name
|
|
zf.write(img, arcname)
|
|
print(f" 📄 Added: {arcname}")
|
|
|
|
print(f"\n✅ CBZ saved → {output_cbz} "
|
|
f"({len(images)} page(s))")
|
|
|
|
|
|
# ─────────────────────────────────────────────
|
|
# PER-PAGE PIPELINE
|
|
# ─────────────────────────────────────────────
|
|
def process_page(page_path, workdir,
|
|
translator_module, renderer_module):
|
|
"""
|
|
Runs translator + renderer for a single page.
|
|
All intermediate files land in workdir.
|
|
Returns path to the translated image, or None on failure.
|
|
"""
|
|
page_stem = page_path.stem
|
|
suffix = page_path.suffix
|
|
|
|
# Paths inside workdir
|
|
output_txt = str(workdir / "output.txt")
|
|
bubbles_json= str(workdir / "bubbles.json")
|
|
debug_png = str(workdir / "debug_clusters.png")
|
|
translated = str(workdir / f"{page_stem}_translated{suffix}")
|
|
|
|
print(f"\n{'─'*60}")
|
|
print(f" PAGE: {page_path.name}")
|
|
print(f"{'─'*60}")
|
|
|
|
# ── Step 1: Translate ─────────────────────────────────────────
|
|
print(f"\n[1/2] Translating...")
|
|
try:
|
|
# Temporarily redirect file outputs to workdir
|
|
orig_dir = os.getcwd()
|
|
os.chdir(workdir)
|
|
|
|
translator_module.translate_manga_text(
|
|
image_path = str(page_path.resolve()),
|
|
source_lang = SOURCE_LANG,
|
|
target_lang = TARGET_LANG,
|
|
confidence_threshold = CONFIDENCE_THRESHOLD,
|
|
export_to_file = "output.txt",
|
|
export_bubbles_to = "bubbles.json",
|
|
min_text_length = MIN_TEXT_LENGTH,
|
|
cluster_eps = CLUSTER_EPS,
|
|
proximity_px = PROXIMITY_PX,
|
|
filter_sound_effects = FILTER_SFX,
|
|
quality_threshold = QUALITY_THRESHOLD,
|
|
upscale_factor = UPSCALE_FACTOR,
|
|
bbox_padding = BBOX_PADDING,
|
|
debug = DEBUG,
|
|
)
|
|
except Exception as e:
|
|
print(f" ❌ Translation failed: {e}")
|
|
os.chdir(orig_dir)
|
|
return None
|
|
finally:
|
|
os.chdir(orig_dir)
|
|
|
|
# ── Step 2: Render ────────────────────────────────────────────
|
|
print(f"\n[2/2] Rendering...")
|
|
try:
|
|
renderer_module.render_translations(
|
|
input_image = str(page_path.resolve()),
|
|
output_image = translated,
|
|
translations_file = output_txt,
|
|
bubbles_file = bubbles_json,
|
|
font_path = str(Path(orig_dir) / FONT_PATH),
|
|
font_color = FONT_COLOR,
|
|
)
|
|
except Exception as e:
|
|
print(f" ❌ Rendering failed: {e}")
|
|
return None
|
|
|
|
print(f" ✅ Done → {translated}")
|
|
return translated
|
|
|
|
|
|
# ─────────────────────────────────────────────
|
|
# MAIN
|
|
# ─────────────────────────────────────────────
|
|
def main():
|
|
# ── Load modules ──────────────────────────────────────────────
|
|
print("Loading modules...")
|
|
try:
|
|
translator = load_module(
|
|
"manga_translator", "manga-translator.py")
|
|
renderer = load_module(
|
|
"manga_renderer", "manga-renderer.py")
|
|
except FileNotFoundError as e:
|
|
print(f"❌ Could not load module: {e}")
|
|
sys.exit(1)
|
|
|
|
# ── Discover pages ────────────────────────────────────────────
|
|
pages = sorted_pages(CHAPTER_DIR)
|
|
if not pages:
|
|
print(f"❌ No images found in: {CHAPTER_DIR}")
|
|
sys.exit(1)
|
|
|
|
print(f"\n📖 Chapter : {CHAPTER_DIR}")
|
|
print(f" Pages : {len(pages)}")
|
|
print(f" Source : {SOURCE_LANG} → Target: {TARGET_LANG}")
|
|
print(f" Output : {OUTPUT_CBZ}\n")
|
|
|
|
# ── Process each page ─────────────────────────────────────────
|
|
translated_dir = Path(CHAPTER_DIR) / "translated"
|
|
succeeded = []
|
|
failed = []
|
|
|
|
for page_path in pages:
|
|
workdir = make_page_workdir(CHAPTER_DIR, page_path.stem)
|
|
result = process_page(page_path, workdir,
|
|
translator, renderer)
|
|
if result:
|
|
succeeded.append(result)
|
|
else:
|
|
failed.append(page_path.name)
|
|
|
|
# ── Summary ───────────────────────────────────────────────────
|
|
print(f"\n{'═'*60}")
|
|
print(f" PIPELINE COMPLETE")
|
|
print(f" ✅ {len(succeeded)} page(s) succeeded")
|
|
if failed:
|
|
print(f" ❌ {len(failed)} page(s) failed: {failed}")
|
|
print(f"{'═'*60}\n")
|
|
|
|
# ── Pack CBZ ──────────────────────────────────────────────────
|
|
print("Packing CBZ...")
|
|
pack_cbz(translated_dir, OUTPUT_CBZ)
|
|
|
|
|
|
if __name__ == "__main__":
|
|
main()
|