any-file to PDF (Phase 1): converters/ dispatcher, images/text/md/csv/json/xml/svg/pdf-passthrough, 112 tests green

converters/html_converter.py

@@ -0,0 +1,67 @@
+"""HTML / HTM -> PDF.
+
+A standalone .html upload is rendered as-is: the document's own <style>/
+<link> rules stay authoritative. app.BASE_CSS is injected at the TOP of
+<head> as a fallback layer (equal-specificity rules lose to the document's
+own later declarations, same layering as the EPUB path). Fragments without
+a <head> ("<p>...</p>") render bare.
+
+Declared charset is honored (windows-1252/latin-1 <meta charset> or XML
+declaration) via the pinned regexes in app — no mojibake. The document
+<title> becomes the export name and PDF /Title when present, else the
+file stem.
+"""
+from __future__ import annotations
+
+import re
+from pathlib import Path
+
+from . import common
+
+_TITLE_RE = re.compile(r"<title[^>]*>(.*?)</title>", re.IGNORECASE | re.DOTALL)
+
+
+def _extract_title(html_text: str) -> str | None:
+    m = _TITLE_RE.search(html_text)
+    if not m:
+        return None
+    title = re.sub(r"<[^>]+>", "", m.group(1))
+    title = " ".join(title.split())
+    return title or None
+
+
+def _inject_base_css(html_text: str) -> str:
+    """Insert BASE_CSS as a fallback <style> at the top of <head>.
+
+    Placed before the document's own styles so, at equal specificity,
+    the document wins (WeasyPrint resolves source order within one
+    stylesheet context like a browser). Fragments without <head> are
+    returned unchanged (a lone <style> outside <head> is legal HTML5
+    and still applies).
+    """
+    import app  # lazy: BASE_CSS lives with the EPUB pipeline
+
+    style = f"<style>\n{app.BASE_CSS}\n</style>"
+    m = re.search(r"<head[^>]*>", html_text, re.IGNORECASE)
+    if m:
+        return html_text[: m.end()] + style + html_text[m.end():]
+    return style + html_text
+
+
+def convert(path: Path, job_id: str, progress_cb=None, output_dir=None,
+            source_name=None) -> str:
+    """Render a standalone HTML/HTM file to a typeset PDF."""
+    raw = path.read_bytes()
+    html_text = common.decode_declared(raw)
+    if not html_text.strip():
+        raise ValueError("empty file: nothing to convert")
+    title = _extract_title(html_text)
+
+    def _build():
+        return _inject_base_css(html_text), path.name
+
+    return common.convert_single(
+        path, job_id, build_html=_build, title=title, css="",
+        progress_cb=progress_cb, output_dir=output_dir,
+        name_src=source_name or path.name,
+    )