md2substack.py

#!/usr/bin/env python3
"""Push a vault newsletter issue to Substack as a DRAFT (never publishes).

    python md2substack.py <issue.md> [--dry-run]

- Parses the issue file produced by /newsletter (frontmatter + '# SUBJECT:' +
  'PREVIEW:' + markdown body).
- Refuses to run if unresolved {{PLACEHOLDER}} markers remain in the body.
- Brand -> publication/section mapping and cookies path come from
  the config file (template below). Cookies come from
  your browser's cookie export (see README).
- Creates a Substack DRAFT via python-substack and prints the editor URL.
  Publishing stays a human action in the Substack UI.

Config template (~/.config/md2substack/config.json):
{
  "cookies_path": "~/.config/md2substack/cookies.json",
  "publications": {
    "myblog":  {"url": "https://myblog.substack.com", "section": null},
    "mybrand": {"url": "https://myblog.substack.com", "section": "My Section"}
  }
}
"""
import json
import os
import re
import sys

CONFIG = os.path.expanduser(os.environ.get("MD2SUBSTACK_CONFIG", "~/.config/md2substack/config.json"))


def parse_issue(path):
    text = open(path, encoding="utf-8").read()
    m = re.match(r"^---\n(.*?)\n---\n", text, re.S)
    if not m:
        raise SystemExit("No frontmatter found.")
    fm_raw, body = m.group(1), text[m.end():]
    brand = re.search(r"^brand:\s*(\S+)", fm_raw, re.M)
    if not brand:
        raise SystemExit("No 'brand:' in frontmatter.")
    hero = re.search(r"^hero_image:\s*(\S+)", fm_raw, re.M)
    hero_path = None
    if hero:
        hero_path = os.path.expanduser(hero.group(1))
        if not os.path.isabs(hero_path):
            hero_path = os.path.join(os.path.dirname(os.path.abspath(path)), hero_path)
        if not os.path.exists(hero_path):
            raise SystemExit(f"hero_image not found: {hero_path}")
    subj = re.search(r"^#\s*SUBJECT:\s*(.+)$", body, re.M)
    prev = re.search(r"^PREVIEW:\s*(.+)$", body, re.M)
    if not subj or not prev:
        raise SystemExit("Missing '# SUBJECT:' or 'PREVIEW:' line.")
    content = body[prev.end():].strip()
    if re.search(r"\{\{[A-Z_]+\}\}", content):
        raise SystemExit("Unresolved {{PLACEHOLDER}} markers in body — fill them first (see 'before_send' in frontmatter).")
    return brand.group(1), subj.group(1).strip(), prev.group(1).strip(), content, hero_path


def main():
    args = [a for a in sys.argv[1:] if not a.startswith("--")]
    dry = "--dry-run" in sys.argv
    if len(args) != 1:
        raise SystemExit(__doc__)
    brand, title, subtitle, content, hero_path = parse_issue(args[0])
    print(f"brand={brand}\ntitle={title}\nsubtitle={subtitle}\nbody={len(content)} chars\nhero={hero_path or 'none'}")
    if dry:
        print("DRY-RUN OK (parsed, no network)")
        return

    if not os.path.exists(CONFIG):
        raise SystemExit(f"Missing config {CONFIG} (template in this script's docstring).")
    cfg = json.load(open(CONFIG))
    pub = cfg["publications"].get(brand) or {}
    if not pub.get("url"):
        raise SystemExit(f"No Substack publication configured for brand '{brand}' in {CONFIG}.")

    from substack import Api
    from substack.post import Post

    api = Api(
        cookies_path=os.path.expanduser(cfg["cookies_path"]),
        publication_url=pub["url"],
    )
    post = Post(title=title, subtitle=subtitle, user_id=api.get_user_id())
    if hero_path:
        up = api.get_image(hero_path)
        hero_url = up.get("url")
        if not hero_url:
            raise SystemExit(f"hero upload failed: {up}")
        content = f"![{title}]({hero_url})\n\n{content}"
        print(f"hero uploaded: {hero_url[:80]}")
    post.from_markdown(content)
    draft = api.post_draft(post.get_draft())
    draft_id = draft.get("id")
    if pub.get("section"):
        # Assign section post-creation: the lib's get_sections()/pre-set path is
        # unreliable; PUT /drafts/<id> {draft_section_id} works.
        r = api._session.get(f"{pub['url']}/api/v1/publication/sections")
        r.raise_for_status()
        matches = [s for s in r.json() if s.get("name") == pub["section"]]
        if not matches:
            print(f"WARNING: section '{pub['section']}' not found — draft left in main feed.")
        else:
            api._session.put(f"{pub['url']}/api/v1/drafts/{draft_id}", json={"draft_section_id": matches[0]["id"]}).raise_for_status()
            print(f"section set: {pub['section']}")
    print(f"DRAFT CREATED (not published): {pub['url']}/publish/post/{draft_id}")


if __name__ == "__main__":
    main()