Add script to reduce example html

Author: osolmaz

examples/notion/frontend/reduce_html.py

@@ -0,0 +1,90 @@
+import argparse
+import re
+from pathlib import Path
+
+from bs4 import BeautifulSoup, NavigableString, Tag
+
+# ---------- Argument Parsing ----------
+DATA_URI_RE = re.compile(r"^\s*data:", re.I)
+
+
+# ---------- Helper functions ----------
+def redact_binary_attrs(tag: Tag) -> None:
+    """Replace base64‐encoded `src` / `srcset` with a placeholder."""
+    if tag.name == "img" and tag.has_attr("src") and DATA_URI_RE.match(tag["src"]):
+        tag["src"] = "redacted"
+    if (
+        tag.name == "source"
+        and tag.has_attr("srcset")
+        and DATA_URI_RE.match(tag["srcset"])
+    ):
+        tag["srcset"] = "redacted"
+
+
+def is_leaf(tag: Tag) -> bool:
+    """Returns True if `tag` has no child *elements* (text is allowed)."""
+    return not any(isinstance(c, Tag) for c in tag.contents)
+
+
+def clean_node(tag: Tag) -> None:
+    """Recursively clean a BeautifulSoup element in-place."""
+    if isinstance(tag, NavigableString):
+        return
+
+    # 1. strip style attributes
+    tag.attrs.pop("style", None)
+
+    # 2. redact binary embeds
+    redact_binary_attrs(tag)
+
+    # 3. special-case SVG: keep the wrapper to preserve structure
+    if tag.name == "svg":
+        tag.clear()
+        tag.append("<!-- redacted -->")
+
+    # 4. recurse
+    for child in list(tag.children):
+        clean_node(child)
+
+
+def prune_top_level(soup: BeautifulSoup) -> None:
+    """Remove high-volume, low-value nodes entirely."""
+    for n in soup.find_all(["script", "style", "noscript"]):
+        n.decompose()
+    for link in soup.find_all("link", rel=lambda x: x and "stylesheet" in x):
+        link.decompose()
+
+
+def clean_html(raw_html: str) -> str:
+    soup = BeautifulSoup(raw_html, "lxml")
+    prune_top_level(soup)
+    clean_node(soup)
+    # Pretty formatting keeps the tree readable yet compact
+    return soup.prettify(formatter="minimal")
+
+
+# ---------- Execute cleaning ----------
+def main():
+    parser = argparse.ArgumentParser(
+        description="Clean and reduce HTML files by removing unnecessary content."
+    )
+    parser.add_argument("input_path", help="Path to the input HTML file")
+    parser.add_argument("output_path", help="Path where the cleaned HTML will be saved")
+    args = parser.parse_args()
+    raw_html = Path(args.input_path).read_text(encoding="utf-8", errors="ignore")
+
+    # Update MAX_LEAF_CHARS based on command line argument
+
+    cleaned_html = clean_html(raw_html)
+
+    # Save to disk
+    Path(args.output_path).write_text(cleaned_html, encoding="utf-8")
+
+    # Show a short preview to the user
+    print(f"Saved cleaned HTML to {args.output_path}")
+    print(f"Original size: {len(raw_html):,} characters")
+    print(f"Cleaned size : {len(cleaned_html):,} characters")
+
+
+if __name__ == "__main__":
+    main()

python/pyproject.toml

@@ -21,6 +21,7 @@ convert_jsondoc = "jsondoc.bin.convert_jsondoc:main"
 dev = [
     "datamodel-code-generator>=0.25.9,<0.26",
     "ipdb>=0.13.13",
+    "lxml>=5.4.0",
     "pre-commit>=4.1.0",
     "pytest>=8.3.5",
     "python-dotenv>=1.0.1",