z230 - Joplin mirror skripty, store_cda_batch v1.2, vyrazeni Evernote exportu a en_backup.db z gitu

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
2026-06-11 19:45:04 +02:00
parent c4d353a343
commit d2e8a70bfe
11 changed files with 990 additions and 0 deletions
Binary file not shown.
@@ -0,0 +1,69 @@
# evernote_to_joplin_mirror_v1.0
**Verze:** 1.0 | **Datum:** 2026-06-11 | **Autor:** Claude (pro V. Buzalku)
Jednosměrné zrcadlení **Evernote → self-hosted Joplin Server**. Záložní „únikový
východ" pro případ, že by Evernote (Bending Spoons) přestal vyhovovat — kompletní
kopie poznámek na vlastním Toweru, odkud lze plynule přejít na Joplin.
## Architektura
```
Evernote cloud --(evernote-backup sync)--> en_backup.db --(tento skript)--> Joplin Server
```
- **Čtení**: přímo z `en_backup.db` přes knihovnu `evernote_backup.note_storage.SqliteStorage`
(žádný ENEX mezikrok — knihovna je vlastní kód nástroje, čtení je spolehlivé).
- **Zápis**: Joplin sync API `PUT /api/items/root:/<id>.md:/content`, autentizace
session tokenem z `POST /api/sessions` (header `X-API-AUTH`).
## Mapování
| Evernote | Joplin | poznámka |
|----------|--------|----------|
| notebook | folder (`type_ 2`) | |
| stack | nadřazený folder | notebooky ve stacku dostanou parent |
| note | note (`type_ 1`) | tělo = ENML → HTML (`markup_language: 2`) |
| en-media (příloha) | resource (`type_ 4`) + blob | odkaz v těle `:/id` |
| en-todo | ☐ / ☑ | |
**ID v Joplinu** = `md5("evernote-<typ>:<evernote-guid>")` → 32 hex. Deterministická,
takže opakované běhy **aktualizují** (upsert), neduplikují.
## Použití
```bash
# pilot na jednom notebooku
python evernote_to_joplin_mirror_v1.0.py --notebook "CL2-78989-011"
# test s limitem poznámek
python evernote_to_joplin_mirror_v1.0.py --notebook "Recepty" --limit 5
# plné zrcadlení
python evernote_to_joplin_mirror_v1.0.py --all
# náhled bez zápisu
python evernote_to_joplin_mirror_v1.0.py --all --dry-run
```
Po běhu spustit v Joplin klientovi synchronizaci, aby se položky stáhly ze serveru.
## Ověřeno (2026-06-11)
- Pilot `CL2-78989-011`: 4 poznámky, text + HTML formátování věrně přeneseno.
- Pilot `Recepty --limit 1`: 1 poznámka + 20 příloh; blob byte-identický (md5 shoda),
tělo správně odkazuje `:/resid`.
## Známá omezení / TODO v1.1
- **Upsert only** — položky smazané/přejmenované v Evernote se v Joplinu zatím NEMAŽOU.
Pro pravý mirror doplnit mazání (porovnat sadu GUID, smazat osiřelé `deleted_time`).
- Heslo je v hlavičce skriptu plaintextem — zvážit přesun do env/konfig.
- ENML→HTML je „best effort" (en-media, en-todo). Exotické prvky (en-crypt, tabulky
se zvláštním stylingem) neřešeny.
- Plánování zatím ručně na Z230; přesun na Tower (User Scripts cron) plánován.
## Předchůdce
evernote-backup `sync` musí proběhnout před tímto skriptem (naplní/aktualizuje
`en_backup.db`). Plná pipeline: `evernote-backup sync && python evernote_to_joplin_mirror_v1.0.py --all`.
+347
View File
@@ -0,0 +1,347 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
# =============================================================================
# Nazev: evernote_to_joplin_mirror_v1.0.py
# Verze: 1.0
# Datum: 2026-06-11
# Autor: Claude (pro Vladimira Buzalku)
# Popis: Jednosmerne zrcadleni Evernote -> self-hosted Joplin Server.
# Cte poznamky primo z lokalni databaze nastroje evernote-backup
# (en_backup.db, pres knihovnu evernote_backup) a zapisuje je do
# Joplin Serveru pres jeho sync API (PUT /api/items, X-API-AUTH).
#
# Evernote notebook -> Joplin folder (stack -> nadrazeny folder)
# Evernote note -> Joplin note (telo = ENML prevedeny na HTML)
# Evernote resource -> Joplin resource (priloha + binarni blob)
#
# ID v Joplinu jsou odvozena deterministicky z Evernote GUID
# (md5), takze opakovane behy poznamky AKTUALIZUJI, neduplikuji.
#
# POZOR (v1.0): upsert only. Poznamky/notebooky smazane v Evernote tento
# skript zatim v Joplinu NEMAZE (planovano do v1.1). Mirror je
# jednosmerny - do Joplinu rucne nepiste, prepise se.
#
# Pouziti:
# # pilotni beh na jednom notebooku:
# python evernote_to_joplin_mirror_v1.0.py --notebook "CL2-78989-011"
# # vic notebooku + limit poznamek (test):
# python evernote_to_joplin_mirror_v1.0.py --notebook "Recepty" --limit 5
# # plne zrcadleni vseho:
# python evernote_to_joplin_mirror_v1.0.py --all
# # jen vypsat co by se delalo, nic nezapisovat:
# python evernote_to_joplin_mirror_v1.0.py --all --dry-run
# =============================================================================
import argparse
import hashlib
import json
import re
import sys
import urllib.request
import urllib.error
from datetime import datetime, timezone
from pathlib import Path
# --- konfigurace ------------------------------------------------------------
DB_PATH = Path(__file__).with_name("en_backup.db")
JOPLIN_BASE = "https://joplin.buzalka.cz"
JOPLIN_EMAIL = "vladimir.buzalka@buzalka.cz"
JOPLIN_PASSWORD = "Vlado7309208104++" # heslo = Vlado + RC + ++ (shodne s Postgres)
# Joplin item typy
T_NOTE = 1
T_FOLDER = 2
T_RESOURCE = 4
IMG_MIMES = ("image/",)
# --- pomocne -----------------------------------------------------------------
def jid(prefix: str, *parts: str) -> str:
"""Deterministicke 32-hex Joplin ID odvozene z Evernote identifikatoru."""
h = hashlib.md5((prefix + ":" + ":".join(parts)).encode("utf-8"))
return h.hexdigest()
def iso(ms) -> str:
"""ms timestamp (int) -> Joplin ISO 8601 'YYYY-MM-DDTHH:MM:SS.000Z'."""
if not ms:
ms = 0
dt = datetime.fromtimestamp(ms / 1000, tz=timezone.utc)
return dt.strftime("%Y-%m-%dT%H:%M:%S.000Z")
def now_iso() -> str:
return datetime.now(tz=timezone.utc).strftime("%Y-%m-%dT%H:%M:%S.000Z")
# --- prevod ENML -> Joplin HTML telo ----------------------------------------
EN_NOTE_RE = re.compile(r"<en-note[^>]*>(.*)</en-note>", re.DOTALL)
EN_MEDIA_RE = re.compile(r"<en-media\b[^>]*?/?>", re.DOTALL)
EN_TODO_RE = re.compile(r"<en-todo\b([^>]*?)/?>", re.DOTALL)
HASH_ATTR_RE = re.compile(r'hash="([0-9a-fA-F]+)"')
def enml_to_html(content: str, hash_to_res: dict) -> str:
"""Vytahne vnitrek <en-note> a prevede en-media/en-todo na Joplin HTML."""
m = EN_NOTE_RE.search(content or "")
body = m.group(1) if m else (content or "")
def repl_media(mt):
tag = mt.group(0)
hm = HASH_ATTR_RE.search(tag)
if not hm:
return ""
res = hash_to_res.get(hm.group(1).lower())
if not res:
return ""
rid, mime, fname = res
if mime.startswith(IMG_MIMES):
return f'<img src=":/{rid}"/>'
label = fname or "priloha"
return f'<a href=":/{rid}">{label}</a>'
body = EN_MEDIA_RE.sub(repl_media, body)
def repl_todo(tt):
checked = "checked" in tt.group(1).lower() and "true" in tt.group(1).lower()
return "&#9745; " if checked else "&#9744; "
body = EN_TODO_RE.sub(repl_todo, body)
return body.strip()
# --- Joplin sync API klient -------------------------------------------------
class Joplin:
def __init__(self, base, email, password, dry_run=False):
self.base = base.rstrip("/")
self.dry_run = dry_run
self.token = None
if not dry_run:
self._login(email, password)
def _login(self, email, password):
req = urllib.request.Request(
f"{self.base}/api/sessions",
data=json.dumps({"email": email, "password": password}).encode(),
headers={"Content-Type": "application/json"},
)
r = urllib.request.urlopen(req, timeout=30)
self.token = json.loads(r.read().decode())["id"]
def put_item(self, path: str, body: bytes):
"""PUT na /api/items/root:/<path>:/content."""
if self.dry_run:
return None
url = f"{self.base}/api/items/root:/{path}:/content"
req = urllib.request.Request(
url, data=body, method="PUT",
headers={"X-API-AUTH": self.token,
"Content-Type": "application/octet-stream"},
)
r = urllib.request.urlopen(req, timeout=120)
return json.loads(r.read().decode())
# --- skladani Joplin sync polozek -------------------------------------------
def folder_item(fid, title, parent_id=""):
t = now_iso()
return (
f"{title}\n\n"
f"id: {fid}\n"
f"created_time: {t}\n"
f"updated_time: {t}\n"
f"user_created_time: {t}\n"
f"user_updated_time: {t}\n"
f"encryption_cipher_text: \n"
f"encryption_applied: 0\n"
f"parent_id: {parent_id}\n"
f"is_shared: 0\n"
f"share_id: \n"
f"master_key_id: \n"
f"icon: \n"
f"user_data: \n"
f"deleted_time: 0\n"
f"type_: {T_FOLDER}"
).encode("utf-8")
def note_item(nid, parent_id, title, body_html, created, updated, author):
ct, ut = iso(created), iso(updated)
safe_title = (title or "(bez nazvu)").replace("\n", " ")
return (
f"{safe_title}\n\n"
f"{body_html}\n\n"
f"id: {nid}\n"
f"parent_id: {parent_id}\n"
f"created_time: {ct}\n"
f"updated_time: {ut}\n"
f"user_created_time: {ct}\n"
f"user_updated_time: {ut}\n"
f"is_conflict: 0\n"
f"latitude: 0.00000000\n"
f"longitude: 0.00000000\n"
f"altitude: 0.0000\n"
f"author: {author or ''}\n"
f"source_url: \n"
f"is_todo: 0\n"
f"todo_due: 0\n"
f"todo_completed: 0\n"
f"source: evernote-mirror\n"
f"source_application: evernote_to_joplin_mirror\n"
f"application_data: \n"
f"order: 0\n"
f"encryption_cipher_text: \n"
f"encryption_applied: 0\n"
f"markup_language: 2\n" # 2 = HTML
f"is_shared: 0\n"
f"share_id: \n"
f"conflict_original_id: \n"
f"master_key_id: \n"
f"user_data: \n"
f"deleted_time: 0\n"
f"type_: {T_NOTE}"
).encode("utf-8")
def resource_item(rid, title, mime, filename, size, created, updated):
ct, ut = iso(created), iso(updated)
ext = ""
if filename and "." in filename:
ext = filename.rsplit(".", 1)[1].lower()
elif "/" in mime:
ext = mime.split("/", 1)[1]
return (
f"{title}\n\n"
f"id: {rid}\n"
f"mime: {mime}\n"
f"filename: {filename or ''}\n"
f"created_time: {ct}\n"
f"updated_time: {ut}\n"
f"user_created_time: {ct}\n"
f"user_updated_time: {ut}\n"
f"file_extension: {ext}\n"
f"encryption_cipher_text: \n"
f"encryption_applied: 0\n"
f"encryption_blob_encrypted: 0\n"
f"size: {size}\n"
f"is_shared: 0\n"
f"share_id: \n"
f"master_key_id: \n"
f"user_data: \n"
f"blob_updated_time: {ut}\n"
f"ocr_text: \n"
f"ocr_details: \n"
f"ocr_status: 0\n"
f"ocr_error: \n"
f"type_: {T_RESOURCE}"
).encode("utf-8")
# --- hlavni logika ----------------------------------------------------------
def main():
ap = argparse.ArgumentParser(description="Evernote -> Joplin mirror v1.0")
ap.add_argument("--notebook", action="append", default=[],
help="nazev notebooku k zrcadleni (lze opakovat)")
ap.add_argument("--all", action="store_true", help="zrcadlit vsechny notebooky")
ap.add_argument("--limit", type=int, default=0,
help="max poznamek na notebook (0 = bez limitu, pro test)")
ap.add_argument("--dry-run", action="store_true",
help="nic nezapisovat, jen vypsat")
ap.add_argument("--db", default=str(DB_PATH), help="cesta k en_backup.db")
args = ap.parse_args()
if not args.all and not args.notebook:
ap.error("zadej --notebook NAZEV nebo --all")
from evernote_backup.note_storage import SqliteStorage
storage = SqliteStorage(Path(args.db))
jop = Joplin(JOPLIN_BASE, JOPLIN_EMAIL, JOPLIN_PASSWORD, dry_run=args.dry_run)
mode = "DRY-RUN" if args.dry_run else "ZAPIS"
print(f"[{mode}] Joplin {JOPLIN_BASE} | db {args.db}")
notebooks = list(storage.notebooks.iter_notebooks())
if not args.all:
wanted = set(args.notebook)
notebooks = [nb for nb in notebooks if nb.name in wanted]
if not notebooks:
print("Zadny odpovidajici notebook nenalezen.", file=sys.stderr)
sys.exit(1)
# stack -> folder id (vytvarime nadrazene foldery podle stacku)
stack_ids = {}
n_nb = n_notes = n_res = n_err = 0
for nb in notebooks:
parent_id = ""
if nb.stack:
sid = stack_ids.get(nb.stack)
if sid is None:
sid = jid("evernote-stack", nb.stack)
stack_ids[nb.stack] = sid
try:
jop.put_item(f"{sid}.md", folder_item(sid, nb.stack))
except Exception as e:
print(f" ! stack '{nb.stack}': {e}", file=sys.stderr)
parent_id = sid
fid = jid("evernote-notebook", nb.guid)
try:
jop.put_item(f"{fid}.md", folder_item(fid, nb.name, parent_id))
n_nb += 1
except Exception as e:
print(f" ! notebook '{nb.name}': {e}", file=sys.stderr)
continue
print(f"[notebook] {nb.name}")
count = 0
for note in storage.notes.iter_notes(nb.guid):
if args.limit and count >= args.limit:
break
count += 1
# priprav resources + mapu hash->resid
hash_to_res = {}
for res in (note.resources or []):
if not (res.data and res.data.body):
continue
bh = res.data.bodyHash
hexhash = bh.hex() if isinstance(bh, (bytes, bytearray)) else str(bh)
fname = res.attributes.fileName if res.attributes else None
rid = jid("evernote-resource", note.guid, hexhash)
hash_to_res[hexhash.lower()] = (rid, res.mime or "application/octet-stream", fname)
try:
jop.put_item(f"{rid}.md", resource_item(
rid, fname or "priloha", res.mime or "application/octet-stream",
fname, len(res.data.body), note.created, note.updated))
jop.put_item(f".resource/{rid}", res.data.body)
n_res += 1
except Exception as e:
n_err += 1
print(f" ! resource {fname}: {e}", file=sys.stderr)
body = enml_to_html(note.content, hash_to_res)
nid = jid("evernote-note", note.guid)
author = note.attributes.author if note.attributes else None
try:
jop.put_item(f"{nid}.md", note_item(
nid, fid, note.title, body, note.created, note.updated, author))
n_notes += 1
except Exception as e:
n_err += 1
print(f" ! note '{note.title}': {e}", file=sys.stderr)
print(f" ({count} poznamek)")
print(f"\nHOTOVO: notebooku={n_nb} poznamek={n_notes} priloh={n_res} chyb={n_err}")
if not args.dry_run:
print("V Joplin klientovi spust synchronizaci, aby se polozky stahly.")
if __name__ == "__main__":
main()
+35
View File
@@ -0,0 +1,35 @@
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
# Smaze VSECHNY polozky z Joplin Serveru (pro cisty full-sync). Jednorazova pomucka.
import json, urllib.request
BASE = "https://joplin.buzalka.cz"
EMAIL = "vladimir.buzalka@buzalka.cz"
PASSWORD = "Vlado7309208104++"
tok = json.loads(urllib.request.urlopen(urllib.request.Request(
BASE + "/api/sessions",
data=json.dumps({"email": EMAIL, "password": PASSWORD}).encode(),
headers={"Content-Type": "application/json"})).read())["id"]
def req(method, path, data=None):
r = urllib.request.Request(BASE + path, data=data, method=method,
headers={"X-API-AUTH": tok})
return urllib.request.urlopen(r, timeout=60).read().decode()
deleted = 0
while True:
page = json.loads(req("GET", "/api/items/root/children"))
items = page.get("items", [])
if not items:
break
for it in items:
req("DELETE", "/api/items/root:/{}:".format(it["name"]))
deleted += 1
print(" smazano {} (has_more={})".format(deleted, page.get("has_more")))
print("CELKEM smazano polozek:", deleted)
chk = json.loads(req("GET", "/api/items/root/children"))
print("zbyva polozek na serveru:", len(chk.get("items", [])))