#!/usr/bin/env python3 """Optional build-time relocation of generated static output, never HTTP rewriting. HTML URL attributes and CSS URLs/URL selectors only. JavaScript, JSON, authored text and normal build output are untouched. This is not a JavaScript compiler: sites constructing absolute asset URLs in JS need their own base-path option. Published previews are symlinks to complete generations; old generations are retained so open readers and a failed next build cannot lose their output. """ import argparse from html.parser import HTMLParser import html as html_lib from pathlib import Path import os import re import shutil import tempfile def check_prefix(prefix): if not re.fullmatch(r"/(?:[A-Za-z0-9_-]+/)*[A-Za-z0-9_-]+/?", prefix): raise ValueError("prefix must be an absolute path of named segments, e.g. /sites/cafe/") return prefix.rstrip("/") def url(value, prefix): if value.startswith("/") and not value.startswith("//"): return prefix + value return value def string_end(text, start): quote = text[start] i = start + 1 while i < len(text): if text[i] == "\\": i += 2 elif text[i] == quote: return i + 1 else: i += 1 raise ValueError("unterminated CSS string") def css(text, prefix): parts, i = [], 0 while i < len(text): start = i if text.startswith("/*", i): end = text.find("*/", i + 2) i = len(text) if end < 0 else end + 2 parts.append(text[start:i]) elif text[i:i+4].lower() == "url(" and (i == 0 or not (text[i-1].isalnum() or text[i-1] in "_-")): i += 4 while i < len(text) and text[i].isspace(): i += 1 if i < len(text) and text[i] in "\"'": end = string_end(text, i) parts.extend((text[start:i+1], url(text[i+1:end-1], prefix), text[end-1:end])) i = end else: end = text.find(")", i) if end < 0: raise ValueError("unterminated CSS url()") parts.extend((text[start:i], url(text[i:end], prefix))) i = end elif text[i] in "\"'": i = string_end(text, i) # URL attribute selectors must continue matching relocated links. context = text[:start] is_url = re.search(r"(?:@import\s*|\[\s*(?:href|src|action)\s*=\s*)$", context, re.I) value = text[start+1:i-1] parts.append(text[start] + (url(value, prefix) if is_url else value) + text[i-1]) else: parts.append(text[i]) i += 1 return "".join(parts) ATTR = re.compile(r"([\w:-]+)(\s*=\s*)(\"[^\"]*\"|'[^']*'|[^\s>]+)") URL_ATTRS = {"href", "src", "poster", "action", "formaction", "cite", "data", "xlink:href"} def srcset(text, prefix): # Data URLs may contain commas; do not blindly split every comma. parts, i = [], 0 while i < len(text): start = i while i < len(text) and (text[i].isspace() or text[i] == ','): i += 1 parts.append(text[start:i]) start = i while i < len(text) and not text[i].isspace(): i += 1 value = text[start:i] parts.append(url(value, prefix)) if value.endswith(','): continue start = i depth = 0 while i < len(text): ch = text[i] if ch == '(': depth += 1 elif ch == ')': depth -= 1 i += 1 if ch == ',' and depth == 0: break parts.append(text[start:i]) return ''.join(parts) class Document(HTMLParser): def __init__(self, prefix): super().__init__(convert_charrefs=False) self.prefix, self.parts, self.raw = prefix, [], None def tag(self, tag, closed=False): def attr(m): name, eq, raw = m.groups() quote = raw[0] if raw[0] in "\"'" else "" value = raw[1:-1] if quote else raw original = value decoded = html_lib.unescape(value) if name.lower() in URL_ATTRS: value = url(decoded, self.prefix) elif name.lower() == "style": value = css(decoded, self.prefix) elif name.lower() == "srcset": value = srcset(decoded, self.prefix) else: return m[0] value = html_lib.escape(value, quote=True) if value != decoded else original return name + eq + quote + value + quote self.parts.append(ATTR.sub(attr, self.get_starttag_text())) if not closed and tag in ("script", "style"): self.raw = tag def handle_starttag(self, tag, attrs): self.tag(tag) def handle_startendtag(self, tag, attrs): self.tag(tag, True) def handle_endtag(self, tag): self.parts.append(f"") if tag == self.raw: self.raw = None def handle_data(self, data): self.parts.append(css(data, self.prefix) if self.raw == "style" else data) def handle_entityref(self, name): self.parts.append(f"&{name};") def handle_charref(self, name): self.parts.append(f"&#{name};") def handle_comment(self, data): self.parts.append(f"") def handle_decl(self, decl): self.parts.append(f"") def handle_pi(self, data): self.parts.append(f"") def html(text, prefix): parser = Document(prefix) parser.feed(text) parser.close() return "".join(parser.parts) def publish(source, output, prefix): prefix = check_prefix(prefix) source, output = Path(source).resolve(), Path(output).absolute() if not (source / "index.html").is_file(): raise ValueError("source must contain an already-built index.html") if source == output.resolve() or source in output.resolve().parents: raise ValueError("preview output must be outside the normal build output") if output.exists() and not output.is_symlink(): raise ValueError("refusing to replace a non-symlink output; choose a new preview path") output.parent.mkdir(parents=True, exist_ok=True) stage = Path(tempfile.mkdtemp(prefix=f".{output.name}-build-", dir=output.parent)) link = stage.with_name(stage.name + ".link") try: for path in source.rglob("*"): if path.is_symlink(): raise ValueError(f"symlink in build output: {path}") target = stage / path.relative_to(source) if path.is_dir(): target.mkdir(exist_ok=True) elif path.is_file(): target.parent.mkdir(parents=True, exist_ok=True) if path.suffix.lower() in (".html", ".css"): text = path.read_text(encoding="utf-8") transform = html if path.suffix.lower() == ".html" else css target.write_text(transform(text, prefix), encoding="utf-8") else: shutil.copy2(path, target) link.symlink_to(stage.name, target_is_directory=True) os.replace(link, output) except BaseException: link.unlink(missing_ok=True) shutil.rmtree(stage) raise print(f"preview: {output} at {prefix}/ (normal output unchanged)") def main(): parser = argparse.ArgumentParser(description=__doc__) parser.add_argument("--url-prefix", required=True) parser.add_argument("--output", required=True) parser.add_argument("source") args = parser.parse_args() publish(args.source, args.output, args.url_prefix) if __name__ == "__main__": main()