stop automatic collation if the sequence of numbers doesn't seem to be indices in an ordering

dlibrary/dlibrary.py

@@ -24,10 +24,13 @@ DLSITE_ID_REGEX = re.compile('^[BR]J[0-9]+$')
 FANZA_ID_REGEX = re.compile('^d_[0-9]+$')
 FAKKU_ID_REGEX = re.compile('.*_FAKKU$')
 
+TEXTLESS_REGEX = re.compile('(台詞|セリフ)(な|無)し|notext|textless')
+ALT_VERSIONS = ['褐色', '日焼け']
+
 IMAGE_FILE_EXTENSIONS = ['.png', '.jpg', '.jpeg', '.gif', '.tiff']
 
 IGNOREABLE_FILES = ['Thumbs.db', '__MACOSX', '.DS_Store']
-IGNOREABLE_EXTENSIONS = ['.txt', '.html', '.htm']
+IGNOREABLE_EXTENSIONS = ['.txt', '.html', '.htm', '.psd']
 
 def open_zipfile_with_encoding(path):
     try:
@@ -111,7 +114,6 @@ async def fetch_async(args):
                 }
                 authors = dlsite_metadata.author or []
                 tags = dlsite_metadata.genre or []
-                thumbnail_local = False
                 thumbnail_url = dlsite_metadata.work_image
                 if thumbnail_url.startswith('//'):
                     thumbnail_url = 'https:' + thumbnail_url
@@ -120,17 +122,11 @@ async def fetch_async(args):
                 authors = db_row.pop('authors')
                 tags = db_row.pop('tags')
                 if FANZA_ID_REGEX.fullmatch(work_id):
-                    thumbnail_local = False
                     thumbnail_url = f'https://doujin-assets.dmm.co.jp/digital/comic/{work_id}/{work_id}pl.jpg'
                 elif FAKKU_ID_REGEX.fullmatch(work_id):
-                    thumbnail_local = True
-                    thumbnail_path = min(work_path.iterdir())
-                    if not thumbnail_path.is_file():
-                        print(f'Fakku thumbnail path {thumbnail_path} is not a file! Skipping {work_id}')
-                        continue
+                    thumbnail_url = None
                 else:
-                    thumbnail_local = False
-                    thumbnail_url = input('Thumbnail image URL: ')
+                    thumbnail_url = input('Thumbnail image URL [default: first page]: ')
 
             cur.execute(
                 "INSERT INTO works(id, title, circle, date, description, series) VALUES(:id, :title, :circle, :date, :description, :series)",
@@ -145,11 +141,7 @@ async def fetch_async(args):
                 [{ "tag": tag, "work": work_id } for tag in tags],
             )
 
-            if thumbnail_local:
-                ext = thumbnail_path.suffix
-                dest_path = thumbnails_dir / (work_id + ext)
-                dest_path.symlink_to(relpath(thumbnail_path, thumbnails_dir))
-            else:
+            if 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}')
@@ -202,24 +194,65 @@ def link_pdf(src, dest, start_index=0):
                 f.write(image["image"])
 
 def complete_prefix_number_ordering(entries):
+    if len(entries) == 1:
+        return entries
+
+    entries_by_version = {}
+    for entry in entries:
+        version = next(ver for ver in (ALT_VERSIONS + ['']) if ver in entry.name)
+        entries_by_version.setdefault(version, []).append(entry)
+
+    numberings_by_version = {ver: prefix_numbering(entries_by_version[ver]) for ver in entries_by_version}
+
+    unified_indices = set()
+    for numbering in numberings_by_version.values():
+        if numbering is None:
+            return None
+        unified_indices |= set(numbering.keys())
+    unified_indices = list(unified_indices)
+    unified_indices.sort()
+
+    if len(unified_indices) > 1 and min(unified_indices[i] - unified_indices[i-1] for i in range(1, len(unified_indices))) > 2:
+        return None
+
+    versions = list(numberings_by_version.keys())
+    versions.sort()
+
+    version_lengths = {ver: len(numberings_by_version[ver]) for ver in numberings_by_version}
+    inner_versions = []
+    outer_versions = [versions[0]]
+    for ver in versions[1:]:
+        if version_lengths[ver] >= version_lengths[versions[0]] - 2:
+            outer_versions.append(ver)
+        else:
+            inner_versions.append(ver)
+
+    result = []
+    for out_ver in outer_versions:
+        for i in unified_indices:
+            for ver in ([out_ver] + (inner_versions if out_ver == versions[0] else [])):
+                entries_i_ver = numberings_by_version[ver].get(i, [])
+                if len(entries_i_ver) <= 1:
+                    result += entries_i_ver
+                else:
+                    return None
+    return result
+
+def prefix_numbering(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()
+            entries_by_index = {}
             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]
+                entries_by_index.setdefault(i, []).append(e)
+            return entries_by_index
+
     return None
 
 def link_ordered_files(ordering, dest, start_index=0):
@@ -277,6 +310,19 @@ def collate(args):
 
         if all(entry.is_file() and entry.suffix.lower() in IMAGE_FILE_EXTENSIONS for entry in entries):
             ordering = complete_prefix_number_ordering(entries)
+            if not ordering:
+                with_text = []
+                textless = []
+                for entry in entries:
+                    if TEXTLESS_REGEX.search(entry.name):
+                        textless.append(entry)
+                    else:
+                        with_text.append(entry)
+                if with_text and textless:
+                    with_text_ordering = complete_prefix_number_ordering(with_text)
+                    textless_ordering = complete_prefix_number_ordering(textless)
+                    if with_text_ordering and textless_ordering:
+                        ordering = with_text_ordering + textless_ordering
             if ordering:
                 print(f'Symlinking image files for {work_id}')
                 link_ordered_files(ordering, collation_dir)
@@ -375,7 +421,9 @@ def generate(args):
     con = sqlite3.connect(args.destdir / 'meta.db')
     cur = con.cursor()
 
-    collated_work_ids = {p.name for p in (args.destdir / 'site' / 'images').iterdir()}
+    site_dir = args.destdir / 'site'
+
+    collated_work_ids = {p.name for p in (site_dir / 'images').iterdir()}
 
     actual_series = {series for (series,) in cur.execute('SELECT series FROM works GROUP BY series HAVING count(series) > 1')}
 
@@ -385,9 +433,16 @@ def generate(args):
             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,))]
-        thumbnail_filename = next(
-            f for f in (args.destdir / 'site' / 'thumbnails').iterdir() if f.stem == work_id
-        ).name
+
+        images = [path.name for path in (site_dir / 'images' / work_id).iterdir()]
+        images.sort()
+
+        try:
+            thumbnail_path = relpath(next(
+                f for f in (site_dir / 'thumbnails').iterdir() if f.stem == work_id
+            ), site_dir)
+        except StopIteration:
+            thumbnail_path = f'images/{work_id}/{images[0]}'
         work = {
             'id': work_id,
             'title': title,
@@ -397,14 +452,11 @@ def generate(args):
             'series': series,
             'authors': authors,
             'tags': tags,
-            'thumbnail_filename': thumbnail_filename,
+            'thumbnail_path': thumbnail_path,
         }
         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
+        work_dir = site_dir / '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:
@@ -413,7 +465,7 @@ def generate(args):
             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
+        categorization_dir = site_dir / categorization
 
         cats = [cat for (cat,) in cur.execute(query)]
         cat_samples = {}
@@ -465,9 +517,9 @@ def generate(args):
     )
 
     with resources.as_file(resources.files("dlibrary")) as r:
-        copy_recursive(r / 'static', args.destdir / 'site' / 'static')
+        copy_recursive(r / 'static', site_dir / 'static')
 
-    with open(args.destdir / 'site' / 'index.html', 'w') as f:
+    with open(site_dir / 'index.html', 'w') as f:
         f.write(index_template.render(depth=0, works=works))
 
     con.close()

dlibrary/static/index.js

@@ -48,7 +48,7 @@ document.addEventListener('DOMContentLoaded', () => {
             card.appendChild(link);
 
             const thumb = document.createElement('img');
-            thumb.src = `${ROOT}/thumbnails/${work.thumbnail_filename}`;
+            thumb.src = `${ROOT}/${work.thumbnail_path}`;
             link.appendChild(thumb);
 
             const creators = document.createElement('div');

dlibrary/templates/categorization.html

@@ -12,7 +12,7 @@
         {{ cat }}
       </div>
       {% if samples[cat] %}
-      <img src="{{ root() }}/thumbnails/{{ samples[cat]['thumbnail_filename'] }}">
+      <img src="{{ root() }}/{{ samples[cat]['thumbnail_path'] }}">
       {% endif %}
     </a>
   </div>

dlibrary/templates/list.html

@@ -7,7 +7,7 @@
   {% for work in works %}
   <div class="card">
     <a href="{{ root() }}/works/{{ work['id'] }}/{{ index() }}">
-      <img src="{{ root() }}/thumbnails/{{ work['thumbnail_filename'] }}">
+      <img src="{{ root() }}/{{ work['thumbnail_path'] }}">
       <div class="card-creators">
         [{% if work['circle'] %}{{ work['circle'] }}{% endif %}{% if work['circle'] and work['authors'] %} ({% endif %}{{ ', '.join(work['authors']) }}{% if work['circle'] and work['authors'] %}){% endif %}]
       </div>

dlibrary/templates/work.html

@@ -5,7 +5,7 @@
 <div class="work-container">
   <div class="work-preview">
     <a href="view/{{ index() }}">
-      <img src="{{ root() }}/thumbnails/{{ work['thumbnail_filename'] }}">
+      <img src="{{ root() }}/{{ work['thumbnail_path'] }}">
       <div class="start-link">Start Reading</div>
     </a>
   </div>