dlibrary/dlibrary/dlibrary.py

#!/usr/bin/env python3

import argparse
import asyncio
import importlib_resources as resources
from pathlib import Path
from os import getenv
from os.path import relpath, splitext
import re
import shutil
import sqlite3
import textwrap
from urllib.parse import urlparse
import zipfile

from dlsite_async import DlsiteAPI
import fitz
from jinja2 import Environment, PackageLoader, select_autoescape
import requests

NUMBER_REGEX = re.compile('[0-9]+')

DLSITE_ID_REGEX = re.compile('^[BR]J[0-9]+$')

IMAGE_FILE_EXTENSIONS = ['.png', '.jpg', '.jpeg', '.gif', '.tiff']

IGNOREABLE_FILES = ['Thumbs.db', '__MACOSX', '.DS_Store']
IGNOREABLE_EXTENSIONS = ['.txt', '.html', '.htm']

def open_zipfile_with_encoding(path):
    try:
        return zipfile.ZipFile(path, metadata_encoding="utf-8")
    except UnicodeDecodeError:
        pass

    try:
        return zipfile.ZipFile(path, metadata_encoding="shift-jis")
    except UnicodeDecodeError:
        pass

    return zipfile.ZipFile(path, metadata_encoding="shift-jisx0213")

def extract(args):
    for zip_path in args.zipfiles:
        work_id = zip_path.stem
        work_extract_path = args.destdir / 'extract' / work_id
        work_extract_path.mkdir(parents=True)

        with open_zipfile_with_encoding(zip_path) as z:
            z.extractall(path=work_extract_path)

        if args.remove:
            zip_path.unlink()

async def fetch_async(args):
    con = sqlite3.connect(args.destdir / 'meta.db')
    cur = con.cursor()

    cur.execute("CREATE TABLE IF NOT EXISTS works(id TEXT PRIMARY KEY, title TEXT, circle TEXT, date TEXT, description TEXT, series TEXT, virtual INT)")
    cur.execute("CREATE TABLE IF NOT EXISTS authors(author TEXT, work TEXT, FOREIGN KEY(work) REFERENCES works(id))")
    cur.execute("CREATE TABLE IF NOT EXISTS tags(tag TEXT, work TEXT, FOREIGN KEY(work) REFERENCES works(id))")

    thumbnails_dir = args.destdir / 'site' / 'thumbnails'
    thumbnails_dir.mkdir(parents=True, exist_ok=True)

    async with DlsiteAPI(locale=args.locale) as api:
        for work_path in (args.destdir / 'extract').iterdir():
            work_id = work_path.name

            res = cur.execute("SELECT id FROM works WHERE id = ?", (work_id,))
            if res.fetchone() is not None:
                continue

            if not DLSITE_ID_REGEX.fullmatch(work_id):
                print(f"{work_id} doesn't seem to be from DLSite, not attempting to fetch metadata")
                continue

            print(f'Fetching metadata for {work_id}')
            metadata = await api.get_work(work_id)

            cur.execute(
                "INSERT INTO works(id, title, circle, date, description, series) VALUES(:id, :title, :circle, :date, :description, :series)",
                {
                    "id": work_id,
                    "title": metadata.work_name,
                    "circle": metadata.circle,
                    "date": metadata.regist_date.date().isoformat(),
                    "description": metadata.description,
                    "series": metadata.series,
                },
            )
            cur.executemany(
                "INSERT INTO authors VALUES(:author, :work)",
                [{ "author": author, "work": work_id } for author in (metadata.author or [])],
            )
            cur.executemany(
                "INSERT INTO tags VALUES(:tag, :work)",
                [{ "tag": tag, "work": work_id } for tag in (metadata.genre or [])],
            )

            thumbnail_url = metadata.work_image
            if thumbnail_url.startswith('//'):
                thumbnail_url = 'https:' + thumbnail_url
            ext = url_file_ext(thumbnail_url)
            dest_file = thumbnails_dir / (work_id + ext)
            print(f'Downloading thumbnail for {work_id} from {thumbnail_url}')
            with open(dest_file, 'wb') as fd:
                with requests.get(thumbnail_url, stream=True) as r:
                    for chunk in r.iter_content(chunk_size=16384):
                        fd.write(chunk)

            con.commit()

        con.close()

def url_file_ext(url):
    return splitext(urlparse(url).path)[1]

def fetch(args):
    asyncio.run(fetch_async(args))


def image_xrefs(pdf):
    images_by_page = [page.get_images() for page in pdf]
    if all(len(images) == 1 for images in images_by_page):
        return [images[0][0] for images in images_by_page]

    print("Checking PDF images the quick way failed, trying the slow way")
    xrefs = []
    for (idx, page) in enumerate(pdf):
        print(f'\x1b[2K\r{idx}/{pdf.page_count} pages processed...', end='')
        images = page.get_image_info(xrefs=True)
        if len(images) != 1 or images[0]['xref'] == 0:
            print('\nFailed')
            return None
        xrefs.append(images[0]['xref'])

    print('\nSuccess')
    return xrefs

def link_pdf(src, dest, start_index=0):
    with fitz.open(src) as pdf:
        xrefs = image_xrefs(pdf)
        if xrefs is None:
            print(f'Support for weirder PDFs not yet implemented, skipping {src}')
            return

        dest.mkdir(parents=True, exist_ok=True)
        for (idx, xref) in enumerate(xrefs, start=start_index):
            image = pdf.extract_image(xref)
            file_path = dest / f'{idx:04d}.{image["ext"]}'
            with open(file_path, 'wb') as f:
                f.write(image["image"])

def complete_prefix_number_ordering(entries):
    matches = reversed(list(NUMBER_REGEX.finditer(entries[0].name)))
    for m in matches:
        pos = m.start()
        prefix = entries[0].name[:pos]
        if all(e.name.startswith(prefix) for e in entries):
            entries_with_indices = []
            indices = set()
            for e in entries:
                n = NUMBER_REGEX.match(e.name[pos:])
                if n is None:
                    return None
                i = int(n.group())
                if i in indices:
                    return None
                indices.add(i)
                entries_with_indices.append((e, i))
            entries_with_indices.sort(key=lambda ei: ei[1])
            return [e for (e, i) in entries_with_indices]
    return None

def link_ordered_files(ordering, dest, start_index=0):
    dest.mkdir(parents=True, exist_ok=True)

    for (idx, src_path) in enumerate(ordering, start=start_index):
        ext = src_path.suffix.lower()
        link_path = dest / f'{idx:04d}{ext}'
        link_path.symlink_to(relpath(src_path, dest))

def ls_ignore(directory):
    return [
        path for path in directory.iterdir()
        if path.name not in IGNOREABLE_FILES and path.suffix.lower() not in IGNOREABLE_EXTENSIONS
    ]

def collate(args):
    con = sqlite3.connect(args.destdir / 'meta.db')
    cur = con.cursor()

    extraction_dir = args.destdir / 'extract'
    hint_map = {hint.relative_to(extraction_dir).parents[-2].name: hint for hint in args.hints}

    for work_path in extraction_dir.iterdir():
        work_id = work_path.name

        collation_dir = args.destdir / 'site' / 'images' / work_id
        if collation_dir.exists():
            continue

        virtual = cur.execute("SELECT virtual FROM works WHERE id = ?", (work_id,)).fetchone()
        if virtual == (1,):
            continue

        if work_id in hint_map:
            hint = hint_map[work_id]
            entries = [hint] if hint.is_file() else ls_ignore(hint)
        else:
            search_dir = work_path
            while True:
                entries = ls_ignore(search_dir)
                if len(entries) == 1 and entries[0].is_dir():
                    search_dir = entries[0]
                else:
                    break

        if len(entries) == 1 and entries[0].suffix.lower() == '.pdf':
            print(f'Extracting images from {entries[0]} for {work_id}')
            link_pdf(entries[0], collation_dir)
            continue

        if len(entries) == 0:
            print(f'{work_id} contains no files? skipping')
            continue

        if all(entry.is_file() and entry.suffix.lower() in IMAGE_FILE_EXTENSIONS for entry in entries):
            ordering = complete_prefix_number_ordering(entries)
            if ordering:
                print(f'Symlinking image files for {work_id}')
                link_ordered_files(ordering, collation_dir)
                continue

        print(f'Unable to deduce file structure for {work_id}, skipping')

    con.close()

def self_and_parents(path):
    return [path] + list(path.parents)

def manual_collate(args):
    work_id = self_and_parents(args.paths[0].relative_to(args.destdir / 'extract'))[-2].name

    collation_dir = args.destdir / 'site' / 'images' / work_id
    if collation_dir.exists() and len(list(collation_dir.iterdir())) > 0:
        print(f'Collation directory already exists!')
        return

    nonexistent = [path for path in args.paths if not path.exists()]
    if len(nonexistent) > 0:
        print(f'Nonexistent paths: {nonexistent}')
        return

    collation_dir.mkdir(parents=True, exist_ok=True)

    index = 0
    for path in args.paths:
        if path.is_dir():
            entries = [p for p in path.iterdir() if p.suffix.lower() in IMAGE_FILE_EXTENSIONS]
            ordering = complete_prefix_number_ordering(entries)
            if ordering is None:
                ordering = entries
                ordering.sort()
            link_ordered_files(ordering, collation_dir, start_index=index)
            index += len(ordering)
        elif path.suffix.lower() in IMAGE_FILE_EXTENSIONS:
            link_ordered_files([path], collation_dir, start_index=index)
            index += 1
        elif path.suffix.lower() == ".pdf":
            link_pdf(path, collation_dir, start_index=index)
            with fitz.open(path) as pdf:
                index += pdf.page_count
        else:
            print(f'Unknown file type {path}, stopping')
            return

def metadata(args):
    con = sqlite3.connect(args.destdir / 'meta.db')
    cur = con.cursor()

    if args.virtual is not None:
        cur.execute("UPDATE works SET virtual = ? WHERE id = ?", (1 if args.virtual else 0, args.work_id))
        con.commit()

    res = cur.execute(
        "SELECT title, circle, date, description, series, virtual FROM works WHERE id = ?",
        (args.work_id,),
    ).fetchone()

    if res is None:
        print(f'Work id {args.work_id} not found!')
        return

    (title, circle, date, description, series, virtual) = res
    print(f'Work ID:     {args.work_id}')
    print(f'Title:       {title}')
    print(f'Circle:      {circle}')
    print(f'Pub date:    {date}')
    print(f'Description: {description}')
    print(f'Series:      {series}')
    print(f'Virtual:     {"Yes" if virtual == 1 else "No"}')

    con.close()

def copy_recursive(src, dest):
    dest.mkdir(parents=True, exist_ok=True)
    for item in src.iterdir():
        if item.is_dir() and not item.is_symlink():
            copy_recursive(item, dest / item.name)
        else:
            shutil.copyfile(item, dest / item.name)

def generate(args):
    jenv = Environment(
        loader=PackageLoader("dlibrary"),
        autoescape=select_autoescape()
    )
    viewer_template = jenv.get_template("viewer.html")
    list_template = jenv.get_template("list.html")
    categorization_template = jenv.get_template("categorization.html")
    work_template = jenv.get_template("work.html")
    index_template = jenv.get_template("index.html")

    con = sqlite3.connect(args.destdir / 'meta.db')
    cur = con.cursor()

    collated_work_ids = {p.name for p in (args.destdir / 'site' / 'images').iterdir()}

    actual_series = {series for (series,) in cur.execute('SELECT series FROM works GROUP BY series HAVING count(series) > 1')}

    works = []
    for (work_id, title, circle, date, description, series) in cur.execute('SELECT id, title, circle, date, description, series FROM works ORDER BY date DESC').fetchall():
        if work_id not in collated_work_ids:
            continue
        authors = [author for (author,) in cur.execute('SELECT author FROM authors WHERE work = ?', (work_id,))]
        tags = [tag for (tag,) in cur.execute('SELECT tag FROM tags WHERE work = ?', (work_id,))]
        work = {
            'id': work_id,
            'title': title,
            'circle': circle,
            'date': date,
            'description': description,
            'series': series,
            'authors': authors,
            'tags': tags,
        }
        works.append(work)

        images = [path.name for path in (args.destdir / 'site' / 'images' / work_id).iterdir()]
        images.sort()

        work_dir = args.destdir / 'site' / 'works' / work_id
        viewer_dir = work_dir / 'view'
        viewer_dir.mkdir(parents=True, exist_ok=True)
        with open(work_dir / 'index.html', 'w') as f:
            f.write(work_template.render(depth=2, work=work, title=title, images=images))
        with open(viewer_dir / 'index.html', 'w') as f:
            f.write(viewer_template.render(depth=3, work=work, title=title, images=images))

    def make_categorization(categorization, query, work_filter, work_style_cards=False):
        categorization_dir = args.destdir / 'site' / categorization

        cats = [cat for (cat,) in cur.execute(query)]
        cat_samples = {}
        for cat in cats:
            cat_works = list(filter(work_filter(cat), works))
            cat_samples[cat] = cat_works[0] if len(cat_works) > 0 else None

            safeish_cat = cat.replace('/', ' ')
            cat_dir = categorization_dir / safeish_cat
            cat_dir.mkdir(parents=True, exist_ok=True)
            with open(cat_dir / 'index.html', 'w') as f:
                f.write(list_template.render(
                    depth=2,
                    works=cat_works,
                    title=cat,
                    categorization=categorization,
                ))

        categorization_dir.mkdir(parents=True, exist_ok=True)
        with open(categorization_dir / 'index.html', 'w') as f:
            f.write(categorization_template.render(
                depth=1,
                categorization=categorization,
                categories=cats,
                samples=cat_samples,
                work_style_cards=work_style_cards,
            ))

    make_categorization(
        'authors',
        'SELECT DISTINCT author FROM authors ORDER BY author',
        lambda author: lambda work: author in work['authors'],
    )
    make_categorization(
        'tags',
        'SELECT DISTINCT tag FROM tags ORDER BY tag',
        lambda tag: lambda work: tag in work['tags'],
    )
    make_categorization(
        'circles',
        'SELECT DISTINCT circle FROM works WHERE circle NOT NULL ORDER BY circle',
        lambda circle: lambda work: work['circle'] == circle,
    )
    make_categorization(
        'series',
        'SELECT DISTINCT series FROM works WHERE series NOT NULL ORDER BY series',
        lambda series: lambda work: work['series'] == series,
        work_style_cards=True,
    )

    with resources.as_file(resources.files("dlibrary")) as r:
        copy_recursive(r / 'static', args.destdir / 'site' / 'static')

    with open(args.destdir / 'site' / 'index.html', 'w') as f:
        f.write(index_template.render(depth=0, works=works))

    con.close()


argparser = argparse.ArgumentParser(
    prog='dlibrary',
    formatter_class=argparse.RawDescriptionHelpFormatter,
    description=textwrap.dedent("""\
    Organize DRM-free works purchased from DLSite into a library
    that can be viewed in a web browser.

    Intended workflow:
    - `extract` a collection of zipfiles downloaded from DLSite
      into DLibrary's data directory, giving each work its own
      subfolder.
    - `fetch` metadata and thumbnail images for extracted works
      from DLSite.
    - `collate` and/or `manual-collate` extracted works,
      producing a single sequence of image files (or symlinks
      into the extracted data, when possible) for each work.
    - Manually adjust works' `metadata` when necessary.
    - `generate` a static website providing a catalog and viewer
      for all collated works.
    """),
)

argparser.add_argument(
    '-d', '--destdir',
    type=Path,
    default=Path(getenv('DLIBRARY_DIR', './dlibrary')),
    help='directory to store dlibrary content and metadata to (default: $DLIBRARY_DIR or ./dlibrary)',
)
subparsers = argparser.add_subparsers(title="subcommands", required=True)

parser_extract = subparsers.add_parser('extract', help='extract zipfiles')
parser_extract.add_argument(
    '-r', '--remove',
    action='store_true',
    help='remove original zipfiles after extraction',
)
parser_extract.add_argument(
    'zipfiles',
    metavar='FILE',
    type=Path,
    nargs='+',
    help='zipfiles to extract',
)
parser_extract.set_defaults(func=extract)

parser_fetch = subparsers.add_parser('fetch', help='fetch metadata and thumbnails')
parser_fetch.add_argument(
    '-l', '--locale',
    type=str,
    default=getenv('DLIBRARY_LOCALE', 'en_US'),
    help=('locale to use when requesting metadata (e.g. "ja_JP", "en_US"). '
          'May still fall back to Japanese if metadata in other languages is unavailable. '
          '(default: $DLIBRARY_LOCALE or en_US)'),
)
parser_fetch.set_defaults(func=fetch)

parser_collate = subparsers.add_parser(
    'collate',
    help='collate each work into a sequence of image files',
    formatter_class=argparse.RawDescriptionHelpFormatter,
    description=textwrap.dedent("""\
    For each extracted work that has not already been collated,
    DLibrary will attempt to intuit its structure as follows:

    - Enter the work's directory. If the directory contains
      nothing except a single subdirectory (ignoring a few types
      of files that are definitely not relevant), traverse
      downwards repeatedly.
    - If the current directory contains nothing except a single
      PDF (again, ignoring irrelevant files), attempt to extract
      a series of images from the PDF. This process expects that
      each page of the PDF consists of a single embedded image,
      which will be extracted at full resolution. Support for
      more complex PDFs is not yet implemented.
    - If the current directory contains nothing except image
      files, and the image files are named in a way that clearly
      indicates a complete numerical order (each filename
      consists of a shared prefix followed by a distinct
      number), symlink files in the inferred order.
    - Otherwise, skip processing this work for now.

    DLibrary can be given "collation hints" which provide
    alternative starting points for this search process. A hint
    is a path under $DLIBRARY_DIR/extract/[work id]/
    indicating a different directory or PDF file to begin the
    search process for that work, rather than starting at the
    top level of the extracted data. There can be at most one
    hint per work; for more complicated scenarios where a work
    includes multiple folders that need to be collated together,
    or where filenames do not clearly indicate an ordering, use
    `manual-collate` instead.
    """),
)
parser_collate.add_argument(
    'hints',
    metavar='PATH',
    type=Path,
    nargs='*',
    help='paths within extraction folders as collation hints'
)
parser_collate.set_defaults(func=collate)

parser_manual_collate = subparsers.add_parser(
    'manual-collate',
    help='collate a single work manually',
    formatter_class=argparse.RawDescriptionHelpFormatter,
    description=textwrap.dedent("""\
    All provided paths must be under $DLIBRARY_DIR/extract/[work id]/
    for the work being manually collated. `manual-collate` can
    only handle one work at a time. Paths are used as follows:

    - If a path is a directory, all *image files* immediately
      inside that directory will be appended to the sequence. If
      files are named in a way which indicates a clear ordering,
      that ordering will be used. Otherwise, filenames will be
      sorted lexicographically. Non-image files and
      subdirectories will be ignored.
    - If a path is an image file, that image file will be
      appended to the sequence.
    - If a path is a PDF file, page images will be extracted
      from that PDF and appended to the sequence.
"""),
)
parser_manual_collate.add_argument(
    'paths',
    metavar='PATH',
    type=Path,
    nargs='+',
    help='paths within a single work to be collated in sequence',
)
parser_manual_collate.set_defaults(func=manual_collate)

parser_metadata = subparsers.add_parser('metadata', help='view or modify metadata for a work')
parser_metadata.add_argument('work_id')
parser_metadata.add_argument(
    '--virtual',
    action=argparse.BooleanOptionalAction,
    help='set work as virtual',
)
parser_metadata.set_defaults(func=metadata)

parser_generate = subparsers.add_parser(
    'generate',
    help='generate HTML/CSS/JS for library site',
    formatter_class=argparse.RawDescriptionHelpFormatter,
    description=textwrap.dedent("""\
    The static site will be generated under $DLIBRARY_DIR/site/
    and can be served by pointing an HTTP server at that
    directory. Note that some files inside the static site
    hierarchy will be symlinks into $DLIBRARY_DIR/extract/
    outside the site hierarchy, so make sure your HTTP server
    will allow those symlinks to be read.
    """),
)
parser_generate.set_defaults(func=generate)

def main():
    args = argparser.parse_args()
    args.func(args)

if __name__ == "__main__":
    main()