aimd-html 0.9.2__py3-none-any.whl

aimd_html/__init__.py

@@ -0,0 +1,5 @@
+"""Defuddle-backed HTML extraction helpers."""
+
+from .defuddle import DefuddleResult, extract_html_with_defuddle
+
+__all__ = ["DefuddleResult", "extract_html_with_defuddle"]

aimd_html/defuddle.py

@@ -0,0 +1,58 @@
+"""Small Python wrapper around the Defuddle CLI.
+
+Defuddle is distributed as a Node/TypeScript package. This wrapper keeps the
+Python package boundary explicit and invokes ``npx defuddle`` when callers opt
+into HTML extraction.
+"""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+import json
+from pathlib import Path
+import subprocess
+
+
+@dataclass(slots=True, frozen=True)
+class DefuddleResult:
+    """Parsed HTML content and metadata returned by Defuddle."""
+
+    content: str
+    title: str | None = None
+    author: str | None = None
+    source: str | None = None
+
+
+def extract_html_with_defuddle(
+    source: str | Path,
+    *,
+    markdown: bool = True,
+    npx_command: str = "npx",
+) -> DefuddleResult:
+    """Extract readable content from a URL, HTML file, or HTML piped source.
+
+    Requires Node.js/npm at runtime. ``source`` is passed to
+    ``npx defuddle parse`` as a URL or local file path.
+    """
+    command = [npx_command, "defuddle", "parse", str(source), "--json"]
+    if markdown:
+        command.append("--markdown")
+
+    completed = subprocess.run(
+        command,
+        capture_output=True,
+        text=True,
+        check=False,
+    )
+    if completed.returncode != 0:
+        stderr = completed.stderr.strip() or completed.stdout.strip()
+        raise RuntimeError(f"defuddle failed: {stderr}")
+
+    payload = json.loads(completed.stdout)
+    content = payload.get("contentMarkdown") or payload.get("content") or ""
+    return DefuddleResult(
+        content=content,
+        title=payload.get("title"),
+        author=payload.get("author"),
+        source=payload.get("source") or payload.get("url"),
+    )

aimd_html-0.9.2.dist-info/METADATA

@@ -0,0 +1,7 @@
+Metadata-Version: 2.3
+Name: aimd-html
+Version: 0.9.2
+Summary: Defuddle-backed HTML extraction helpers for aimd workflows.
+Author: Shu Li
+Author-email: Shu Li <zetarylee@gmail.com>
+Requires-Python: >=3.10, <3.13

aimd_html-0.9.2.dist-info/RECORD

@@ -0,0 +1,5 @@
+aimd_html/__init__.py,sha256=CdQEPlBc7E_fAOmf5QWw5rEgJHxs-fdzYH0Ty7K78w0,173
+aimd_html/defuddle.py,sha256=x4DKvyqFWMUrQiemaqOID8-TpC58BlYVElsMTlhXlnY,1661
+aimd_html-0.9.2.dist-info/WHEEL,sha256=eh7sammvW2TypMMMGKgsM83HyA_3qQ5Lgg3ynoecH3M,79
+aimd_html-0.9.2.dist-info/METADATA,sha256=JGTGx-qSU97oHJb67ZwEogPPK2TwkSQ0zjv-zZ4HXSk,211
+aimd_html-0.9.2.dist-info/RECORD,,

aimd_html-0.9.2.dist-info/WHEEL

@@ -0,0 +1,4 @@
+Wheel-Version: 1.0
+Generator: uv 0.8.24
+Root-Is-Purelib: true
+Tag: py3-none-any