Files
authoring-service/src/authoring_service/content.py
Zaine 87582d1e34
All checks were successful
Build Authoring Service / build (push) Successful in 8s
x
2026-05-13 23:32:39 +01:00

428 lines
15 KiB
Python
Executable File

"""Editable page, upload, and diagnostics operations."""
from __future__ import annotations
import os
import posixpath
import re
import struct
import sys
from datetime import datetime
from email.parser import BytesParser
from email.policy import default as email_default_policy
from pathlib import Path
from typing import Any
from .config import (
ALLOWED_UPLOAD_EXTENSIONS,
BLOGS_DIR,
EXCLUDED_CONTENT_DIR_NAMES,
GENERATED_CONTENT_NAMES,
HZONE_ASSETS_DIR,
IMAGE_ASSETS_DIR,
LIMA_DIR,
MONTH_NAMES,
POSTS_DIR,
ROOT,
)
from .models import ContentPage
from .utils import html_escape, normalise_tags, org_date, parse_org_datetime, slugify
def safe_relative_path(path: str) -> Path:
rel = Path(path)
if rel.is_absolute() or ".." in rel.parts:
raise ValueError("Path must stay inside this repository.")
full = (ROOT / rel).resolve()
if not full.is_relative_to(ROOT):
raise ValueError("Path must stay inside this repository.")
if full.name in GENERATED_CONTENT_NAMES or "sync-conflict" in full.name:
raise ValueError("Generated and sync-conflict files are not editable here.")
rel_parts = full.relative_to(ROOT).parts
if any(part in EXCLUDED_CONTENT_DIR_NAMES for part in rel_parts):
raise ValueError("This path is outside the editable content folders.")
if full.suffix == ".org":
return full
if full.suffix == ".md" and full.is_relative_to(LIMA_DIR):
return full
raise ValueError("Only .org content files and .md files under lima can be edited here.")
def safe_target_path(path: str, slug: str, page_type: str) -> Path:
candidate = path.strip()
if not candidate:
raise ValueError("Path is required.")
default_ext = ".md" if page_type == "lima" else ".org"
if candidate.endswith("/"):
candidate = f"{candidate}{slug}{default_ext}"
elif not Path(candidate).suffix:
candidate = f"{candidate}{default_ext}"
return safe_relative_path(candidate)
def markdown_title(content: str, fallback: str) -> str:
for line in content.splitlines():
match = re.match(r"^#{1,6}\s+(.+?)\s*$", line)
if match:
return match.group(1).strip()
return fallback.replace("-", " ").replace("_", " ").title()
def read_markdown_page(path: Path) -> ContentPage:
content = path.read_text(encoding="utf-8")
rel = path.relative_to(ROOT).as_posix()
title = markdown_title(content, path.stem)
return ContentPage(
path=rel,
page_type="lima",
title=title,
slug=path.stem,
tags=[],
content=content,
date="",
comments=True,
options="",
format="markdown",
)
def read_page(path: Path) -> ContentPage:
if path.suffix == ".md" and path.is_relative_to(LIMA_DIR):
return read_markdown_page(path)
text = path.read_text(encoding="utf-8")
meta: dict[str, str] = {}
body_lines: list[str] = []
in_header = True
for line in text.splitlines():
if in_header and line.startswith("#+"):
key, _, value = line[2:].partition(":")
meta[key.strip().upper()] = value.strip()
else:
in_header = False
body_lines.append(line)
rel = path.relative_to(ROOT).as_posix()
if path.is_relative_to(BLOGS_DIR):
page_type = "blog"
elif path.is_relative_to(POSTS_DIR):
page_type = "post"
else:
page_type = "page"
slug = meta.get("SLUG") or path.stem
tags = normalise_tags(meta.get("FILETAGS", ""))
return ContentPage(
path=rel,
page_type=page_type,
title=meta.get("TITLE", path.stem),
slug=slug,
tags=tags,
content="\n".join(body_lines).lstrip("\n"),
date=meta.get("DATE", org_date(datetime.fromtimestamp(path.stat().st_mtime))),
comments=meta.get("COMMENTS", "t").lower() == "t",
options=meta.get("OPTIONS", "num:nil"),
format="org",
wip=meta.get("WIP"),
)
def page_to_dict(page: ContentPage) -> dict[str, Any]:
return {
"path": page.path,
"pageType": page.page_type,
"title": page.title,
"slug": page.slug,
"tags": page.tags,
"content": page.content,
"date": page.date,
"comments": page.comments,
"options": page.options,
"format": page.format,
"wip": page.wip or "",
}
def server_diagnostics() -> dict[str, Any]:
pages = list_pages()
try:
cwd = Path.cwd().as_posix()
except OSError as exc:
cwd = f"<unavailable: {exc}>"
return {
"root": ROOT.as_posix(),
"cwd": cwd,
"executable": sys.executable,
"pid": os.getpid(),
"pageCount": len(pages),
"firstPage": pages[0]["path"] if pages else "",
}
def list_pages() -> list[dict[str, Any]]:
pages = []
org_paths = []
if ROOT.exists():
try:
for path in ROOT.rglob("*.org"):
try:
rel_parts = path.relative_to(ROOT).parts
except ValueError:
continue
if any(part in EXCLUDED_CONTENT_DIR_NAMES for part in rel_parts):
continue
org_paths.append(path)
except OSError:
org_paths = []
try:
md_paths = list(LIMA_DIR.rglob("*.md")) if LIMA_DIR.exists() else []
except OSError:
md_paths = []
for path in sorted(org_paths + md_paths):
if path.name in GENERATED_CONTENT_NAMES or "sync-conflict" in path.name:
continue
try:
page = read_page(path)
parsed = parse_org_datetime(page.date)
timestamp = parsed.timestamp() if parsed else path.stat().st_mtime
except (OSError, UnicodeDecodeError, ValueError):
continue
pages.append(
{
"path": page.path,
"pageType": page.page_type,
"title": page.title,
"slug": page.slug,
"tags": page.tags,
"date": page.date,
"format": page.format,
"timestamp": timestamp,
}
)
return sorted(pages, key=lambda item: item["timestamp"], reverse=True)
def target_path(data: dict[str, Any], existing_path: str | None) -> Path:
if existing_path:
return safe_relative_path(existing_path)
title = str(data.get("title") or "").strip()
slug = slugify(str(data.get("slug") or title))
page_type = str(data.get("pageType") or "blog")
explicit_path = str(data.get("targetPath") or "").strip()
if explicit_path:
return safe_target_path(explicit_path, slug, page_type)
if page_type == "blog":
dt = parse_org_datetime(str(data.get("date") or "")) or datetime.now()
folder = BLOGS_DIR / str(dt.year) / f"{dt.month:02d}-{MONTH_NAMES[dt.month - 1]}"
return folder / f"{slug}.org"
if page_type == "post":
raw_section = str(data.get("section") or "").strip()
section = slugify(raw_section) if raw_section else ""
folder = POSTS_DIR / section if section else POSTS_DIR
return folder / f"{slug}.org"
if page_type == "lima":
return LIMA_DIR / f"{slug}.md"
if page_type == "page":
return ROOT / f"{slug}.org"
raise ValueError("pageType must be blog, post, page, or lima.")
def render_markdown(data: dict[str, Any]) -> str:
content = str(data.get("content") or "").replace("\r\n", "\n").strip()
title = str(data.get("title") or "").strip()
if not title:
raise ValueError("Title is required.")
content = re.sub(
r'<a\b[^>]*>\s*<img\b[^>]*\bsrc="([^"]+)"[^>]*\balt="([^"]*)"[^>]*>\s*</a>',
lambda match: f"![{match.group(2)}]({match.group(1)})",
content,
flags=re.IGNORECASE,
)
if content:
if re.search(r"^#{1,6}[^\S\r\n]+.+?[^\S\r\n]*$", content, flags=re.MULTILINE):
content = re.sub(
r"^#{1,6}[^\S\r\n]+.+?[^\S\r\n]*$",
f"# {title}",
content,
count=1,
flags=re.MULTILINE,
)
else:
content = f"# {title}\n\n{content}"
return content + "\n"
return f"# {title}\n"
def render_org(data: dict[str, Any], previous: ContentPage | None) -> str:
title = str(data.get("title") or "").strip()
if not title:
raise ValueError("Title is required.")
slug = slugify(str(data.get("slug") or title))
tags = normalise_tags(data.get("tags", []))
content = str(data.get("content") or "").replace("\r\n", "\n").strip()
date = str(data.get("date") or "").strip()
if not parse_org_datetime(date):
date = previous.date if previous else org_date(datetime.now())
options = str(data.get("options") or (previous.options if previous else "num:nil")).strip()
comments = bool(data.get("comments", True))
lines = [
f"#+TITLE: {title}",
f"#+OPTIONS: {options}",
f"#+DATE: {date}",
f"#+filetags: {''.join(f':{tag}' for tag in tags)}:",
]
wip = str(data.get("wip") or (previous.wip if previous else "") or "").strip()
if wip:
lines.append(f"#+WIP: {wip}")
lines.extend(
[
f"#+COMMENTS: {'t' if comments else ''}",
f"#+SLUG: {slug}",
"",
content,
"",
]
)
return "\n".join(lines)
def save_page(data: dict[str, Any]) -> dict[str, Any]:
existing_path = data.get("path") or None
target = target_path(data, str(existing_path) if existing_path else None)
previous = read_page(target) if target.exists() else None
if not target.parent.exists():
target.parent.mkdir(parents=True)
if target.exists() and not existing_path:
raise ValueError(f"{target.relative_to(ROOT)} already exists.")
if target.suffix == ".md":
target.write_text(render_markdown(data), encoding="utf-8")
else:
target.write_text(render_org(data, previous), encoding="utf-8")
return page_to_dict(read_page(target))
def image_dimensions(payload: bytes, ext: str) -> tuple[int, int] | None:
if ext == ".png" and payload.startswith(b"\x89PNG\r\n\x1a\n") and len(payload) >= 24:
width, height = struct.unpack(">II", payload[16:24])
return width, height
if ext == ".gif" and payload[:6] in {b"GIF87a", b"GIF89a"} and len(payload) >= 10:
width, height = struct.unpack("<HH", payload[6:10])
return width, height
if ext in {".jpg", ".jpeg"} and payload.startswith(b"\xff\xd8"):
i = 2
while i + 9 < len(payload):
if payload[i] != 0xFF:
i += 1
continue
marker = payload[i + 1]
i += 2
if marker in {0xD8, 0xD9}:
continue
if i + 2 > len(payload):
break
size = int.from_bytes(payload[i:i + 2], "big")
if size < 2:
break
if marker in {0xC0, 0xC1, 0xC2, 0xC3, 0xC5, 0xC6, 0xC7, 0xC9, 0xCA, 0xCB, 0xCD, 0xCE, 0xCF}:
if i + 7 <= len(payload):
height = int.from_bytes(payload[i + 3:i + 5], "big")
width = int.from_bytes(payload[i + 5:i + 7], "big")
return width, height
break
i += size
return None
def relative_asset_path(page_path: str, asset_path: str) -> str:
page = safe_relative_path(page_path) if page_path else ROOT / "index.org"
page_rel = page.relative_to(ROOT).as_posix()
page_output_dir = posixpath.dirname(page_rel)
return posixpath.relpath(asset_path, page_output_dir or ".")
def gallery_image_html(filename: str, asset_path: str, page_path: str, payload: bytes, ext: str) -> str:
absolute_url = f"https://zainezq.com/{asset_path}"
relative_url = relative_asset_path(page_path, asset_path)
dims = image_dimensions(payload, ext)
width, height = dims if dims else (1920, 1080)
alt = html_escape(filename)
return (
f'<a href="{absolute_url}" data-img="{absolute_url}" data-alt="{alt}" '
f'data-width="{width}" data-height="{height}">'
f'<img src="{relative_url}" alt="{alt}" style="cursor: zoom-in;"></a>'
)
def attachment_image_dir(page_path: str) -> Path:
page = safe_relative_path(page_path) if page_path else ROOT / "index.org"
if page.suffix == ".md" and page.is_relative_to(LIMA_DIR):
return HZONE_ASSETS_DIR
if page.is_relative_to(POSTS_DIR):
rel = page.relative_to(POSTS_DIR)
section = rel.parts[0] if len(rel.parts) > 1 else "posts"
return IMAGE_ASSETS_DIR / slugify(section)
if page.is_relative_to(BLOGS_DIR):
return IMAGE_ASSETS_DIR / "blogs"
rel = page.relative_to(ROOT)
if len(rel.parts) > 1:
return IMAGE_ASSETS_DIR / slugify(rel.parts[0])
return IMAGE_ASSETS_DIR / "pages"
def save_upload(filename: str, payload: bytes, page_path: str = "") -> dict[str, str]:
original = Path(filename or "attachment").name
ext = Path(original).suffix.lower()
if ext not in ALLOWED_UPLOAD_EXTENSIONS:
raise ValueError("Only common image files can be uploaded.")
now = datetime.now()
target_dir = attachment_image_dir(page_path)
target_dir.mkdir(parents=True, exist_ok=True)
stem = slugify(Path(original).stem)
prefix = "" if re.match(r"^\d{4}-\d{2}-\d{2}-", stem) else f"{now.strftime('%Y-%m-%d')}-"
target = target_dir / f"{prefix}{stem}{ext}"
counter = 2
while target.exists():
target = target_dir / f"{prefix}{stem}-{counter}{ext}"
counter += 1
target.write_bytes(payload)
rel = target.relative_to(ROOT).as_posix()
absolute_url = f"https://zainezq.com/{rel}"
relative_url = relative_asset_path(page_path, rel)
is_markdown = page_path.endswith(".md")
insert_text = f"![{target.name}]({absolute_url})" if is_markdown else f"[[{relative_url}]]"
return {
"url": absolute_url,
"relativeUrl": relative_url,
"path": rel,
"markdown": insert_text,
"insertText": insert_text,
"filename": target.name,
}
def parse_upload_form(content_type: str, body: bytes) -> tuple[str, bytes, str]:
if not content_type.lower().startswith("multipart/form-data"):
raise ValueError("Uploads must use multipart/form-data.")
message = BytesParser(policy=email_default_policy).parsebytes(
f"Content-Type: {content_type}\r\nMIME-Version: 1.0\r\n\r\n".encode("utf-8") + body
)
if not message.is_multipart():
raise ValueError("Upload form is not multipart.")
filename = ""
payload = b""
page_path = ""
for part in message.iter_parts():
name = part.get_param("name", header="content-disposition")
if name == "attachment":
filename = part.get_filename() or ""
payload = part.get_payload(decode=True) or b""
elif name == "pagePath":
raw_value = part.get_payload(decode=True) or b""
page_path = raw_value.decode(part.get_content_charset() or "utf-8", errors="replace")
if not filename:
raise ValueError("No attachment was uploaded.")
if not payload:
raise ValueError("Attachment is empty.")
return filename, payload, page_path