bitwarden-print/render.py

#!/usr/bin/env python3
"""Render a Bitwarden JSON export as a dense PDF.

Every leaf value in the export is printed: well-known fields first in a
readable order, then every remaining value alphabetically. The PDF starts
with a summary and an index that carries page numbers.
"""

from __future__ import annotations

import argparse
import datetime as dt
import html
import json
import re
import sys
from collections import Counter
from pathlib import Path

TYPE_NAMES = {1: "login", 2: "secure note", 3: "card", 4: "identity", 5: "ssh key"}

# Preferred display order; matched by exact path, prefix or list index.
PRIORITY = [
    "type",
    "name",
    "folderId",
    "organizationId",
    "collectionIds",
    "favorite",
    "reprompt",
    "login.username",
    "login.password",
    "login.totp",
    "login.passwordRevisionDate",
    "login.uris",
    "login.fido2Credentials",
    "card.cardholderName",
    "card.brand",
    "card.number",
    "card.expMonth",
    "card.expYear",
    "card.code",
    "identity.title",
    "identity.firstName",
    "identity.middleName",
    "identity.lastName",
    "identity.address1",
    "identity.address2",
    "identity.address3",
    "identity.city",
    "identity.state",
    "identity.postalCode",
    "identity.country",
    "identity.company",
    "identity.email",
    "identity.phone",
    "identity.ssn",
    "identity.username",
    "identity.passportNumber",
    "identity.licenseNumber",
    "secureNote",
    "sshKey.privateKey",
    "sshKey.publicKey",
    "sshKey.fingerprint",
    "sshKey.keyFingerprint",
    "fields",
    "notes",
    "passwordHistory",
    "attachments",
    "id",
    "creationDate",
    "revisionDate",
    "deletedDate",
]

SECRET_RE = re.compile(
    r"(?:^|\.)(password|totp|code|ssn|pin|cvv|privateKey|passportNumber|licenseNumber|number)$"
)


class Empty:
    """Marks an empty dict/list so it is visible in the output."""

    def __init__(self, text: str) -> None:
        self.text = text


def esc(value: object) -> str:
    return html.escape(str(value), quote=True)


def flatten(obj, prefix: str = "", out=None):
    """Flatten nested JSON into (dotted.path[index], leaf) pairs."""
    if out is None:
        out = []
    if isinstance(obj, dict):
        if not obj:
            out.append((prefix, Empty("{}")))
            return out
        for key, value in obj.items():
            flatten(value, f"{prefix}.{key}" if prefix else str(key), out)
    elif isinstance(obj, list):
        if not obj:
            out.append((prefix, Empty("[]")))
            return out
        for index, value in enumerate(obj):
            flatten(value, f"{prefix}[{index}]", out)
    else:
        out.append((prefix, obj))
    return out


def natural_key(path: str):
    parts = re.split(r"(\d+)", path)
    return tuple((1, int(p)) if p.isdigit() else (0, p) for p in parts if p)


def rank(path: str):
    priority = len(PRIORITY)
    for index, prefix in enumerate(PRIORITY):
        if path == prefix or path.startswith(prefix + ".") or path.startswith(prefix + "["):
            priority = index
            break
    return (priority, natural_key(path))


def fmt_value(value):
    """Return (kind, text) so CSS can style empties, numbers and text."""
    if isinstance(value, Empty):
        return ("empty", value.text)
    if isinstance(value, bool):
        return ("code", "true" if value else "false")
    if value is None:
        return ("empty", "null")
    if isinstance(value, (int, float)):
        return ("code", repr(value) if isinstance(value, float) else str(value))
    text = str(value)
    if text == "":
        return ("empty", '""')
    return ("text", text)


def is_secret(path: str) -> bool:
    return bool(SECRET_RE.search(re.sub(r"\[\d+\]", "", path)))


def mask_text(text: str) -> str:
    return f"{'*' * 10} ({len(text)} chars)"


def type_num(item) -> int | None:
    value = item.get("type")
    return value if isinstance(value, int) and not isinstance(value, bool) else None


def type_label(item) -> str:
    value = type_num(item)
    if value is None:
        return "unknown"
    return TYPE_NAMES.get(value, f"type {value}")


def badge_class(item) -> str:
    value = type_num(item)
    return f"t-{value}" if value in TYPE_NAMES else "t-other"


def folder_paths(folders) -> dict:
    by_id = {
        f["id"]: f
        for f in folders
        if isinstance(f, dict) and isinstance(f.get("id"), str)
    }
    paths = {}
    for fid, folder in by_id.items():
        parts = [folder.get("name") or fid]
        parent = folder.get("parentId")
        seen = {fid}
        while isinstance(parent, str) and parent in by_id and parent not in seen:
            seen.add(parent)
            parts.append(by_id[parent].get("name") or parent)
            parent = by_id[parent].get("parentId")
        paths[fid] = " / ".join(reversed(parts))
    return paths


def sort_items(items, mode: str, folders: dict):
    def folder_of(item):
        fid = item.get("folderId")
        return folders.get(fid, str(fid)) if fid else ""

    def name_of(item):
        return str(item.get("name") or "").casefold()

    if mode == "as-is":
        return list(items)
    if mode == "name":
        key = lambda it: (name_of(it), folder_of(it))
    elif mode == "type":
        key = lambda it: (f"{type_num(it) or 99:03d}", folder_of(it), name_of(it))
    else:  # folder
        key = lambda it: (folder_of(it), name_of(it))
    return sorted(items, key=key)


CSS = """
@page {
  size: A4;
  margin: 9mm 8mm 11mm 8mm;
  @bottom-left {
    content: string(title) " - " string(source);
    font-family: "DejaVu Sans Mono", monospace;
    font-size: 5.8pt;
    color: #8a8a8a;
  }
  @bottom-right {
    content: counter(page) " / " counter(pages);
    font-family: "DejaVu Sans Mono", monospace;
    font-size: 6pt;
    color: #8a8a8a;
  }
}
html { font-family: "DejaVu Sans", sans-serif; font-size: 7.1pt; color: #111; }
body { margin: 0; }
a { color: inherit; text-decoration: none; }
h1 {
  string-set: title content();
  font-size: 15pt;
  margin: 0 0 0.8mm;
  letter-spacing: -0.2pt;
}
h2 {
  font-size: 8.4pt;
  text-transform: uppercase;
  letter-spacing: 0.5pt;
  color: #555;
  margin: 4.5mm 0 1.2mm;
  border-bottom: 0.6pt solid #333;
  padding-bottom: 0.5mm;
}
.sub { color: #666; font-size: 6.6pt; margin-bottom: 1mm; }
.sub .src { string-set: source content(); font-family: "DejaVu Sans Mono", monospace; }
table { width: 100%; border-collapse: collapse; table-layout: fixed; }
.mono, th, td.k { font-family: "DejaVu Sans Mono", monospace; }
table.stats th, table.stats td { text-align: left; padding: 0.35mm 1.5mm 0.35mm 0; vertical-align: top; }
table.stats th { width: 32%; color: #666; font-weight: 400; font-size: 6.4pt; }
table.stats td { overflow-wrap: anywhere; }
table.split { width: 100%; }
table.split td.col { vertical-align: top; padding: 0; width: 50%; }
table.split td.col + td.col { padding-left: 5mm; }
table.grid th {
  text-align: left;
  font-size: 6.1pt;
  font-weight: 700;
  text-transform: uppercase;
  letter-spacing: 0.3pt;
  color: #666;
  border-bottom: 0.5pt solid #999;
  padding: 0.5mm 1mm;
}
table.grid td {
  font-family: "DejaVu Sans Mono", monospace;
  font-size: 6.5pt;
  padding: 0.35mm 1mm;
  border-bottom: 0.3pt dotted #dcdcdc;
  overflow-wrap: anywhere;
  vertical-align: top;
}
table.grid tbody tr:nth-child(even) { background: #f6f6f6; }
td.r, th.r { text-align: right; }
.dot {
  display: inline-block;
  width: 1.4mm;
  height: 1.4mm;
  border-radius: 0.7mm;
  margin-right: 1mm;
  background: #777;
}
.t-1 { background: #1f6feb; }
.t-2 { background: #6e7781; }
.t-3 { background: #8250df; }
.t-4 { background: #b35c00; }
.t-5 { background: #1a7f37; }
.t-other { background: #555; }
.item { margin: 0 0 2.4mm; }
.item h3 {
  font-size: 7.8pt;
  margin: 0 0 0.6mm;
  padding: 0.5mm 0 0.5mm;
  border-bottom: 0.5pt solid #888;
  break-after: avoid;
}
.item h3 .idx {
  font-family: "DejaVu Sans Mono", monospace;
  color: #999;
  margin-right: 1.4mm;
}
.badge {
  font-family: "DejaVu Sans Mono", monospace;
  font-size: 5.8pt;
  font-weight: 700;
  text-transform: uppercase;
  letter-spacing: 0.2pt;
  color: #fff;
  padding: 0.25mm 1mm;
  border-radius: 0.6mm;
  margin-right: 1.2mm;
}
.item h3 .iname { font-weight: 700; }
.item h3 .meta {
  color: #666;
  font-size: 6.4pt;
  font-weight: 400;
  margin-left: 1.6mm;
}
table.kv th {
  width: 24%;
  text-align: left;
  font-family: "DejaVu Sans Mono", monospace;
  font-size: 6.1pt;
  font-weight: 400;
  color: #7a7a7a;
  padding: 0.4mm 1.6mm 0.4mm 0;
  vertical-align: top;
  overflow-wrap: anywhere;
}
table.kv td {
  font-family: "DejaVu Sans Mono", monospace;
  font-size: 6.9pt;
  padding: 0.4mm 0;
  vertical-align: top;
  white-space: pre-wrap;
  overflow-wrap: anywhere;
  border-bottom: 0.25pt dotted #e4e4e4;
}
table.kv tr { break-inside: avoid; }
td.empty { color: #a0a0a0; }
td.code { color: #0a5c2b; }
td.mask { color: #8a5a00; }
.summary-note { color: #666; font-size: 6.3pt; margin-top: 1mm; }
"""

PAPER_OVERRIDE = {"a4": "A4", "letter": "letter"}


def build_html(*, data, items, source_name, title, mask, sort_mode, paper):
    folders = [f for f in (data.get("folders") or []) if isinstance(f, dict)]
    fpaths = folder_paths(folders)
    collections = [c for c in (data.get("collections") or []) if isinstance(c, dict)]
    coll_names = {
        c.get("id"): (c.get("name") or c.get("id")) for c in collections if c.get("id")
    }

    items = sort_items(items, sort_mode, fpaths)

    leaves = sum(len(flatten(item)) for item in items)
    type_counts = Counter(type_label(item) for item in items)
    notes_count = sum(1 for item in items if item.get("notes"))
    field_count = sum(
        len(item.get("fields") or [])
        for item in items
        if isinstance(item.get("fields"), list)
    )
    attach_count = sum(
        len(item.get("attachments") or [])
        for item in items
        if isinstance(item.get("attachments"), list)
    )
    history_count = sum(
        len(item.get("passwordHistory") or [])
        for item in items
        if isinstance(item.get("passwordHistory"), list)
    )
    folder_counts = Counter(
        fpaths.get(item.get("folderId"), str(item.get("folderId")))
        if item.get("folderId")
        else "(no folder)"
        for item in items
    )

    def date_range(key):
        values = sorted(
            str(item[key]) for item in items if isinstance(item.get(key), str) and item[key]
        )
        return f"{values[0]} .. {values[-1]}" if values else ""

    def row(label, value, klass=""):
        cls = f' class="{klass}"' if klass else ""
        return f"<tr><th>{esc(label)}</th><td{cls}>{esc(value)}</td></tr>\n"

    meta_rows = [
        row("source", source_name),
        row("generated", dt.datetime.now().astimezone().strftime("%Y-%m-%d %H:%M %Z")),
        row("items", len(items)),
        row("folders", len(folders)),
    ]
    if collections:
        meta_rows.append(row("collections", len(collections)))
    meta_rows.extend(
        [
            row("values printed", leaves),
            row("items with notes", notes_count),
            row("custom fields", field_count),
            row("attachments", attach_count),
            row("password history", history_count),
            row("created range", date_range("creationDate")),
            row("revised range", date_range("revisionDate")),
        ]
    )
    for key, value in data.items():
        if key in {"items", "folders", "collections"}:
            continue
        kind, text = fmt_value(value)
        meta_rows.append(row(f"export.{key}", text, "empty" if kind == "empty" else ""))

    type_rows = "".join(
        f'<tr><td><span class="dot {badge_class({"type": next((n for n, t in TYPE_NAMES.items() if t == label), None)})}"></span>{esc(label)}</td>'
        f'<td class="r">{count}</td></tr>\n'
        for label, count in sorted(type_counts.items(), key=lambda kv: (-kv[1], kv[0]))
    )
    folder_rows = "".join(
        f"<tr><td>{esc(name)}</td><td class=\"r\">{count}</td></tr>\n"
        for name, count in sorted(folder_counts.items(), key=lambda kv: kv[0].casefold())
    )
    collection_rows = ""
    if collections:
        collection_rows = "".join(
            f"<tr><td>{esc(c.get('name') or c.get('id') or '')}</td>"
            f"<td>{esc(c.get('organizationId') or '')}</td>"
            f"<td>{esc(c.get('externalId') or '')}</td>"
            f"<td>{esc(c.get('readOnly'))}</td></tr>\n"
            for c in collections
        )

    index_rows = []
    for index, item in enumerate(items, 1):
        fid = item.get("folderId")
        folder = fpaths.get(fid, str(fid)) if fid else ""
        login = item.get("login") if isinstance(item.get("login"), dict) else {}
        username = login.get("username") or ""
        index_rows.append(
            f'<tr><td class="r">{index}</td>'
            f'<td><span class="dot {badge_class(item)}"></span>{esc(type_label(item))}</td>'
            f'<td><a href="#item-{index}">{esc(item.get("name") or "(unnamed)")}</a></td>'
            f"<td>{esc(folder)}</td><td>{esc(username)}</td>"
            f'<td class="r"><a class="pageref" href="#item-{index}"></a></td></tr>\n'
        )

    item_html = []
    for index, item in enumerate(items, 1):
        header_rows = []
        fid = item.get("folderId")
        folder = fpaths.get(fid, str(fid)) if fid else ""
        if fid:
            header_rows.append(row("folder", folder))
        cids = item.get("collectionIds")
        if isinstance(cids, list) and cids:
            header_rows.append(
                row("collections", ", ".join(str(coll_names.get(cid, cid)) for cid in cids))
            )
        for path, value in sorted(flatten(item), key=lambda kv: rank(kv[0])):
            kind, text = fmt_value(value)
            if mask and kind == "text" and is_secret(path):
                kind, text = "mask", mask_text(text)
            header_rows.append(row(path, text, kind))

        meta_bits = []
        if folder:
            meta_bits.append(folder)
        if isinstance(cids, list) and cids:
            meta_bits.append(", ".join(str(coll_names.get(cid, cid)) for cid in cids))
        meta = (
            f'<span class="meta">{esc(" · ".join(meta_bits))}</span>' if meta_bits else ""
        )

        item_html.append(
            f'<section class="item">\n'
            f'<h3 id="item-{index}"><span class="idx">{index}</span>'
            f'<span class="badge {badge_class(item)}">{esc(type_label(item))}</span>'
            f'<span class="iname">{esc(item.get("name") or "(unnamed)")}</span>{meta}</h3>\n'
            f'<table class="kv">\n{"".join(header_rows)}</table>\n'
            f"</section>\n"
        )

    paper_css = PAPER_OVERRIDE.get(paper, "A4")

    return f"""<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8">
<title>{esc(title)}</title>
<style>{CSS}
@page {{ size: {paper_css}; }}
.pageref::after {{ content: target-counter(attr(href), page); }}
</style>
</head>
<body>
<h1>{esc(title)}</h1>
<p class="sub"><span class="src">{esc(source_name)}</span> &middot; {len(items)} items &middot; {leaves} values &middot; generated {dt.datetime.now().astimezone().strftime("%Y-%m-%d %H:%M %Z")}</p>

<h2>Summary</h2>
<table class="split"><tr>
<td class="col"><table class="stats">{"".join(meta_rows)}</table></td>
<td class="col">
<table class="grid"><thead><tr><th>type</th><th class="r">items</th></tr></thead><tbody>{type_rows}</tbody></table>
<table class="grid" style="margin-top:3mm"><thead><tr><th>folder</th><th class="r">items</th></tr></thead><tbody>{folder_rows}</tbody></table>
</td></tr></table>
{f'<h2>Collections</h2><table class="grid"><thead><tr><th>name</th><th>organizationId</th><th>externalId</th><th>readOnly</th></tr></thead><tbody>{collection_rows}</tbody></table>' if collections else ''}

<h2>Index</h2>
<table class="grid"><thead><tr><th class="r" style="width:6%">#</th><th style="width:14%">type</th><th>name</th><th style="width:20%">folder</th><th style="width:16%">username</th><th class="r" style="width:6%">page</th></tr></thead>
<tbody>{"".join(index_rows)}</tbody></table>

<h2>Items</h2>
{"".join(item_html)}
</body>
</html>
"""


def load_export(path: Path):
    with path.open("r", encoding="utf-8") as handle:
        data = json.load(handle)
    if not isinstance(data, dict):
        raise SystemExit("error: export root must be a JSON object")
    if data.get("encrypted") is True:
        raise SystemExit(
            "error: this is an encrypted export and cannot be read.\n"
            "       Re-export from Bitwarden with 'Export as .json' and no file password."
        )
    items = data.get("items") or []
    if not isinstance(items, list):
        raise SystemExit("error: 'items' is not a list; is this a Bitwarden export?")
    items = [item for item in items if isinstance(item, dict)]
    return data, items


def main(argv=None) -> int:
    parser = argparse.ArgumentParser(
        description="Render a Bitwarden JSON export as a dense PDF."
    )
    parser.add_argument("input", type=Path, help="Bitwarden .json export")
    parser.add_argument("-o", "--output", type=Path, help="output PDF (default: input.pdf)")
    parser.add_argument("--title", help="document title")
    parser.add_argument(
        "--sort",
        choices=["folder", "name", "type", "as-is"],
        default="folder",
        help="item order (default: folder)",
    )
    parser.add_argument(
        "--mask-secrets",
        action="store_true",
        help="mask passwords, TOTP, card numbers/codes, SSN and private keys",
    )
    parser.add_argument("--paper", choices=["a4", "letter"], default="a4")
    parser.add_argument("--html-out", type=Path, help="also write the HTML source")
    args = parser.parse_args(argv)

    if not args.input.is_file():
        raise SystemExit(f"error: no such file: {args.input}")

    data, items = load_export(args.input)
    title = args.title or "Bitwarden export"
    document = build_html(
        data=data,
        items=items,
        source_name=args.input.name,
        title=title,
        mask=args.mask_secrets,
        sort_mode=args.sort,
        paper=args.paper,
    )

    if args.html_out:
        args.html_out.parent.mkdir(parents=True, exist_ok=True)
        args.html_out.write_text(document, encoding="utf-8")

    output = args.output or args.input.with_suffix(".pdf")
    output.parent.mkdir(parents=True, exist_ok=True)

    from weasyprint import HTML

    HTML(string=document, base_url=str(args.input.parent)).write_pdf(str(output))
    print(f"wrote {output} ({len(items)} items)")
    return 0


if __name__ == "__main__":
    sys.exit(main())