examples/deepzoom: switch from os.path to pathlib

Signed-off-by: Benjamin Gilbert <[email protected]>

Author: bgilbert

examples/deepzoom/deepzoom_multiserver.py

@@ -3,7 +3,7 @@
 # deepzoom_multiserver - Example web application for viewing multiple slides
 #
 # Copyright (c) 2010-2015 Carnegie Mellon University
-# Copyright (c) 2021-2023 Benjamin Gilbert
+# Copyright (c) 2021-2024 Benjamin Gilbert
 #
 # This library is free software; you can redistribute it and/or modify it
 # under the terms of version 2.1 of the GNU Lesser General Public License
@@ -27,6 +27,7 @@
 from collections.abc import Callable
 from io import BytesIO
 import os
+from pathlib import Path, PurePath
 from threading import Lock
 from typing import TYPE_CHECKING, Any, Literal
 import zlib
@@ -82,7 +83,7 @@
 
 
 class DeepZoomMultiServer(Flask):
-    basedir: str
+    basedir: Path
     cache: _SlideCache
 
 
@@ -94,7 +95,7 @@ class AnnotatedDeepZoomGenerator(DeepZoomGenerator):
 
 def create_app(
     config: dict[str, Any] | None = None,
-    config_file: str | None = None,
+    config_file: Path | None = None,
 ) -> Flask:
     # Create and configure app
     app = DeepZoomMultiServer(__name__)
@@ -116,7 +117,7 @@ def create_app(
         app.config.from_mapping(config)
 
     # Set up cache
-    app.basedir = os.path.abspath(app.config['SLIDE_DIR'])
+    app.basedir = Path(app.config['SLIDE_DIR']).resolve(strict=True)
     config_map = {
         'DEEPZOOM_TILE_SIZE': 'tile_size',
         'DEEPZOOM_OVERLAP': 'overlap',
@@ -131,16 +132,18 @@ def create_app(
     )
 
     # Helper functions
-    def get_slide(path: str) -> AnnotatedDeepZoomGenerator:
-        path = os.path.abspath(os.path.join(app.basedir, path))
-        if not path.startswith(app.basedir + os.path.sep):
-            # Directory traversal
+    def get_slide(user_path: PurePath) -> AnnotatedDeepZoomGenerator:
+        try:
+            path = (app.basedir / user_path).resolve(strict=True)
+        except OSError:
+            # Does not exist
             abort(404)
-        if not os.path.exists(path):
+        if path.parts[: len(app.basedir.parts)] != app.basedir.parts:
+            # Directory traversal
             abort(404)
         try:
             slide = app.cache.get(path)
-            slide.filename = os.path.basename(path)
+            slide.filename = path.name
             return slide
         except OpenSlideError:
             abort(404)
@@ -152,7 +155,7 @@ def index() -> str:
 
     @app.route('/<path:path>')
     def slide(path: str) -> str:
-        slide = get_slide(path)
+        slide = get_slide(PurePath(path))
         slide_url = url_for('dzi', path=path)
         return render_template(
             'slide-fullpage.html',
@@ -163,15 +166,15 @@ def slide(path: str) -> str:
 
     @app.route('/<path:path>.dzi')
     def dzi(path: str) -> Response:
-        slide = get_slide(path)
+        slide = get_slide(PurePath(path))
         format = app.config['DEEPZOOM_FORMAT']
         resp = make_response(slide.get_dzi(format))
         resp.mimetype = 'application/xml'
         return resp
 
     @app.route('/<path:path>_files/<int:level>/<int:col>_<int:row>.<format>')
     def tile(path: str, level: int, col: int, row: int, format: str) -> Response:
-        slide = get_slide(path)
+        slide = get_slide(PurePath(path))
         format = format.lower()
         if format != 'jpeg' and format != 'png':
             # Not supported by Deep Zoom
@@ -208,7 +211,7 @@ def __init__(
         self.dz_opts = dz_opts
         self.color_mode = color_mode
         self._lock = Lock()
-        self._cache: OrderedDict[str, AnnotatedDeepZoomGenerator] = OrderedDict()
+        self._cache: OrderedDict[Path, AnnotatedDeepZoomGenerator] = OrderedDict()
         # Share a single tile cache among all slide handles, if supported
         try:
             self._tile_cache: OpenSlideCache | None = OpenSlideCache(
@@ -217,7 +220,7 @@ def __init__(
         except OpenSlideVersionError:
             self._tile_cache = None
 
-    def get(self, path: str) -> AnnotatedDeepZoomGenerator:
+    def get(self, path: Path) -> AnnotatedDeepZoomGenerator:
         with self._lock:
             if path in self._cache:
                 # Move to end of LRU
@@ -286,13 +289,14 @@ def xfrm(img: Image.Image) -> None:
 
 
 class _Directory:
-    def __init__(self, basedir: str, relpath: str = ''):
-        self.name = os.path.basename(relpath)
+    _DEFAULT_RELPATH = PurePath('.')
+
+    def __init__(self, basedir: Path, relpath: PurePath = _DEFAULT_RELPATH):
+        self.name = relpath.name
         self.children: list[_Directory | _SlideFile] = []
-        for name in sorted(os.listdir(os.path.join(basedir, relpath))):
-            cur_relpath = os.path.join(relpath, name)
-            cur_path = os.path.join(basedir, cur_relpath)
-            if os.path.isdir(cur_path):
+        for cur_path in sorted((basedir / relpath).iterdir()):
+            cur_relpath = relpath / cur_path.name
+            if cur_path.is_dir():
                 cur_dir = _Directory(basedir, cur_relpath)
                 if cur_dir.children:
                     self.children.append(cur_dir)
@@ -301,9 +305,9 @@ def __init__(self, basedir: str, relpath: str = ''):
 
 
 class _SlideFile:
-    def __init__(self, relpath: str):
-        self.name = os.path.basename(relpath)
-        self.url_path = relpath
+    def __init__(self, relpath: PurePath):
+        self.name = relpath.name
+        self.url_path = relpath.as_posix()
 
 
 if __name__ == '__main__':
@@ -336,7 +340,7 @@ def __init__(self, relpath: str):
         ),
     )
     parser.add_argument(
-        '-c', '--config', metavar='FILE', dest='config', help='config file'
+        '-c', '--config', metavar='FILE', type=Path, dest='config', help='config file'
     )
     parser.add_argument(
         '-d',
@@ -396,6 +400,7 @@ def __init__(self, relpath: str):
     parser.add_argument(
         'SLIDE_DIR',
         metavar='SLIDE-DIRECTORY',
+        type=Path,
         nargs='?',
         help='slide directory',
     )

examples/deepzoom/deepzoom_server.py

@@ -26,6 +26,7 @@
 from collections.abc import Callable
 from io import BytesIO
 import os
+from pathlib import Path
 import re
 from typing import TYPE_CHECKING, Any, Literal, Mapping
 from unicodedata import normalize
@@ -93,7 +94,7 @@ class DeepZoomServer(Flask):
 
 def create_app(
     config: dict[str, Any] | None = None,
-    config_file: str | None = None,
+    config_file: Path | None = None,
 ) -> Flask:
     # Create and configure app
     app = DeepZoomServer(__name__)
@@ -113,9 +114,9 @@ def create_app(
         app.config.from_mapping(config)
 
     # Open slide
-    slidefile = app.config['DEEPZOOM_SLIDE']
-    if slidefile is None:
+    if app.config['DEEPZOOM_SLIDE'] is None:
         raise ValueError('No slide file specified')
+    slidefile = Path(app.config['DEEPZOOM_SLIDE'])
     config_map = {
         'DEEPZOOM_TILE_SIZE': 'tile_size',
         'DEEPZOOM_OVERLAP': 'overlap',
@@ -273,7 +274,7 @@ def xfrm(img: Image.Image) -> None:
         ),
     )
     parser.add_argument(
-        '-c', '--config', metavar='FILE', dest='config', help='config file'
+        '-c', '--config', metavar='FILE', type=Path, dest='config', help='config file'
     )
     parser.add_argument(
         '-d',
@@ -333,6 +334,7 @@ def xfrm(img: Image.Image) -> None:
     parser.add_argument(
         'DEEPZOOM_SLIDE',
         metavar='SLIDE',
+        type=Path,
         nargs='?',
         help='slide file',
     )

examples/deepzoom/deepzoom_tile.py

@@ -3,7 +3,7 @@
 # deepzoom_tile - Convert whole-slide images to Deep Zoom format
 #
 # Copyright (c) 2010-2015 Carnegie Mellon University
-# Copyright (c) 2022-2023 Benjamin Gilbert
+# Copyright (c) 2022-2024 Benjamin Gilbert
 #
 # This library is free software; you can redistribute it and/or modify it
 # under the terms of version 2.1 of the GNU Lesser General Public License
@@ -31,6 +31,7 @@
 from multiprocessing import JoinableQueue, Process
 import multiprocessing.queues
 import os
+from pathlib import Path
 import re
 import shutil
 import sys
@@ -87,7 +88,7 @@
         'ignore',
     ]
     TileQueue: TypeAlias = multiprocessing.queues.JoinableQueue[
-        tuple[str | None, int, tuple[int, int], str] | None
+        tuple[str | None, int, tuple[int, int], Path] | None
     ]
     Transform: TypeAlias = Callable[[Image.Image], None]
 
@@ -98,7 +99,7 @@ class TileWorker(Process):
     def __init__(
         self,
         queue: TileQueue,
-        slidepath: str,
+        slidepath: Path,
         tile_size: int,
         overlap: int,
         limit_bounds: bool,
@@ -196,7 +197,7 @@ class DeepZoomImageTiler:
     def __init__(
         self,
         dz: DeepZoomGenerator,
-        basename: str,
+        basename: Path,
         format: str,
         associated: str | None,
         queue: TileQueue,
@@ -214,16 +215,15 @@ def run(self) -> None:
 
     def _write_tiles(self) -> None:
         for level in range(self._dz.level_count):
-            tiledir = os.path.join("%s_files" % self._basename, str(level))
-            if not os.path.exists(tiledir):
-                os.makedirs(tiledir)
+            tiledir = self._basename.with_name(self._basename.name + '_files') / str(
+                level
+            )
+            tiledir.mkdir(parents=True, exist_ok=True)
             cols, rows = self._dz.level_tiles[level]
             for row in range(rows):
                 for col in range(cols):
-                    tilename = os.path.join(
-                        tiledir, '%d_%d.%s' % (col, row, self._format)
-                    )
-                    if not os.path.exists(tilename):
+                    tilename = tiledir / f'{col}_{row}.{self._format}'
+                    if not tilename.exists():
                         self._queue.put((self._associated, level, (col, row), tilename))
                     self._tile_done()
 
@@ -241,7 +241,7 @@ def _tile_done(self) -> None:
                 print(file=sys.stderr)
 
     def _write_dzi(self) -> None:
-        with open('%s.dzi' % self._basename, 'w') as fh:
+        with self._basename.with_name(self._basename.name + '.dzi').open('w') as fh:
             fh.write(self.get_dzi())
 
     def get_dzi(self) -> str:
@@ -253,8 +253,8 @@ class DeepZoomStaticTiler:
 
     def __init__(
         self,
-        slidepath: str,
-        basename: str,
+        slidepath: Path,
+        basename: Path,
         format: str,
         tile_size: int,
         overlap: int,
@@ -303,12 +303,12 @@ def _run_image(self, associated: str | None = None) -> None:
         if associated is None:
             image = self._slide
             if self._with_viewer:
-                basename = os.path.join(self._basename, VIEWER_SLIDE_NAME)
+                basename = self._basename / VIEWER_SLIDE_NAME
             else:
                 basename = self._basename
         else:
             image = ImageSlide(self._slide.associated_images[associated])
-            basename = os.path.join(self._basename, self._slugify(associated))
+            basename = self._basename / self._slugify(associated)
         dz = DeepZoomGenerator(
             image, self._tile_size, self._overlap, limit_bounds=self._limit_bounds
         )
@@ -335,9 +335,7 @@ def _write_html(self) -> None:
             # We're not running from a module (e.g. "python deepzoom_tile.py")
             # so PackageLoader('__main__') doesn't work in jinja2 3.x.
             # Load templates directly from the filesystem.
-            loader = jinja2.FileSystemLoader(
-                os.path.join(os.path.dirname(__file__), 'templates')
-            )
+            loader = jinja2.FileSystemLoader(Path(__file__).parent / 'templates')
         env = jinja2.Environment(loader=loader, autoescape=True)
         template = env.get_template('slide-multipane.html')
         associated_urls = {n: self._url_for(n) for n in self._slide.associated_images}
@@ -357,22 +355,20 @@ def _write_html(self) -> None:
             properties=self._slide.properties,
             dzi_data=json.dumps(self._dzi_data),
         )
-        with open(os.path.join(self._basename, 'index.html'), 'w') as fh:
+        with open(self._basename / 'index.html', 'w') as fh:
             fh.write(data)
 
     def _write_static(self) -> None:
-        basesrc = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'static')
-        basedst = os.path.join(self._basename, 'static')
+        basesrc = Path(__file__).absolute().parent / 'static'
+        basedst = self._basename / 'static'
         self._copydir(basesrc, basedst)
-        self._copydir(os.path.join(basesrc, 'images'), os.path.join(basedst, 'images'))
+        self._copydir(basesrc / 'images', basedst / 'images')
 
-    def _copydir(self, src: str, dest: str) -> None:
-        if not os.path.exists(dest):
-            os.makedirs(dest)
-        for name in os.listdir(src):
-            srcpath = os.path.join(src, name)
-            if os.path.isfile(srcpath):
-                shutil.copy(srcpath, os.path.join(dest, name))
+    def _copydir(self, src: Path, dest: Path) -> None:
+        dest.mkdir(parents=True, exist_ok=True)
+        for srcpath in src.iterdir():
+            if srcpath.is_file():
+                shutil.copy(srcpath, dest / srcpath.name)
 
     @classmethod
     def _slugify(cls, text: str) -> str:
@@ -444,6 +440,7 @@ def _shutdown(self) -> None:
         '-o',
         '--output',
         metavar='NAME',
+        type=Path,
         dest='basename',
         help='base name of output file',
     )
@@ -475,12 +472,13 @@ def _shutdown(self) -> None:
     parser.add_argument(
         'slidepath',
         metavar='SLIDE',
+        type=Path,
         help='slide file',
     )
 
     args = parser.parse_args()
     if args.basename is None:
-        args.basename = os.path.splitext(os.path.basename(args.slidepath))[0]
+        args.basename = Path(args.slidepath.stem)
 
     DeepZoomStaticTiler(
         args.slidepath,