Speed up the initial build when the image cache is cold

This takes the cold build from ~50s to ~7s.

.gitignore

@@ -2,5 +2,6 @@
 .jekyll-metadata
 .sass-cache
 _site
+*.pyc
 
 .DS_Store

src/_plugins/pillow/get_all_image_info.py

@@ -0,0 +1,44 @@
+#!/usr/bin/env python3
+"""
+Get the information about every image it can find.
+"""
+
+import json
+import pathlib
+
+from get_image_info import get_info
+
+
+def get_file_paths_under(root=".", *, suffix=""):
+    """
+    Generates the absolute paths to every matching file under ``root``.
+    """
+    root = pathlib.Path(root)
+
+    if root.exists() and not root.is_dir():
+        raise ValueError(f"Cannot find files under file: {root!r}")
+
+    if not root.is_dir():
+        raise FileNotFoundError(root)
+
+    for dirpath, _, filenames in root.walk():
+        for f in filenames:
+            if f == ".DS_Store":
+                continue
+
+            if f.endswith((".svg", ".mov")):
+                continue
+
+            p = dirpath / f
+
+            if p.is_file() and f.lower().endswith(suffix):
+                yield p
+
+
+if __name__ == "__main__":
+    result = {}
+
+    for p in get_file_paths_under("src/_images"):
+        result[str(p)] = get_info(p)
+
+    print(json.dumps(result))

src/_plugins/pillow/get_image_info.py

@@ -1,32 +1,32 @@
 #!/usr/bin/env python3
 """
-Get the information about one or more images.
+Get the information about an images.
 
-This takes one argument, which is a JSON-formatted list of image paths.
-It prints a JSON object whose keys are paths and values are image info.
+This takes a path to an image, prints the output as JSON.
 
 Example:
 
-    $ python3 convert_image.py '["cat.jpg", "dog.png"]'
-    {
-        "cat.jpg": {"width": 400, "height": 300, "format": "JPEG"},
-        "dog.png": {"width": 1600, "height": 900, "format": "PNG"}
-    }
+    $ python3 convert_image.py cat.jpg
+    {"width": 400, "height": 300, "format": "JPEG"}
 
 """
 
 import json
+import os
 import sys
 
 from PIL import Image
 
 
 def get_info(path):
     im = Image.open(path)
-    return {"width": im.width, "height": im.height, "format": im.format}
+    return {
+        "width": im.width,
+        "height": im.height,
+        "format": im.format,
+        "mtime": int(os.path.getmtime(path)),
+    }
 
 
 if __name__ == "__main__":
-    result = {path: get_info(path) for path in json.loads(sys.argv[1])}
-
-    print(json.dumps(result))
+    print(json.dumps(get_info(sys.argv[1])))

src/_plugins/pillow/get_image_info.rb

@@ -1,19 +1,48 @@
+# Use Python to get some basic info about an image, including the
+# dimensions and format.
+
 require 'json'
 require 'open3'
 
-def get_image_info(image_paths)
-  stdout, status = Open3.capture2('python3', 'src/_plugins/pillow/get_image_info.py', JSON.generate(image_paths))
-  raise "Unable to get info for images #{image_paths}" unless status.success?
-
-  JSON.parse(stdout)
-end
-
 def get_single_image_info(path)
   cache = Jekyll::Cache.new('ImageInfo')
-  mtime = File.mtime(path)
+
+  # How this works:
+  #
+  #   - When we have an empty cache, we get the info for every image
+  #     we can find in `_src/images` and warm the cache.
+  #
+  #   - If the image isn't in the cache (e.g. it was just added) or the
+  #     cache entry is invalid (the image has changed), we get the info
+  #     for just that image.
+  #
+  # We want to minimise the calls to Python, because each one involves
+  # a minor performance penalty we'd rather avoid.
+  #
+  # Jekyll will clear the cache every time `_config.yml` changes, or you
+  # change your options (e.g. adding the `--profile` flag), so we warm
+  # the cache to speed up that initial build.
+
+  unless cache.key?('-1')
+    puts 'Image info cache is empty, rebuilding...'
+
+    stdout, status = Open3.capture2('python3', 'src/_plugins/pillow/get_all_image_info.py')
+
+    if status.success?
+      JSON.parse(stdout).each do |path, info|
+        cache["#{path}--#{info['mtime']}"] = info
+      end
+
+      cache['-1'] = -1
+    end
+  end
+
+  mtime = File.mtime(path).to_i
 
   cache.getset("#{path}--#{mtime}") do
-    info = get_image_info([path])
-    info[path]
+    stdout, status = Open3.capture2('python3', 'src/_plugins/pillow/get_image_info.py', path)
+    raise "Unable to get info for images #{image_paths}" unless status.success?
+
+    JSON.parse(stdout)
   end
 end

src/_tests/test_pillow.rb

@@ -4,14 +4,8 @@
 require 'test/unit'
 
 require_relative '../_plugins/pillow/convert_image'
-require_relative '../_plugins/pillow/get_image_info'
 
 class TestPillow < Test::Unit::TestCase
-  def test_get_image_info
-    output = get_image_info(['src/_tests/images/gradient.png'])
-    assert_equal(output, { 'src/_tests/images/gradient.png' => { 'width' => 250, 'height' => 250, 'format' => 'PNG' } })
-  end
-
   def test_convert_image
     Dir.mktmpdir do |d|
       convert_image({