notes2bear

#!/usr/bin/env python3
import zlib, os, sqlite3, re, zipfile, json
from struct import unpack_from
from datetime import datetime

# Simple script to export Notes.app to Bear.app backup format
# I have code to decode the tables and drawings (to svg), but not necessary for Bear

def uvarint(data,pos):
    x = s = 0
    while True:
        b = data[pos]
        pos += 1
        x = x | ((b&0x7f)<<s)
        if b < 0x80: return x,pos
        s += 7

def readbytes(data,pos):
    l,pos = uvarint(data,pos)
    return data[pos:pos+l], pos+l

def readstruct(fmt,l):
    return lambda data,pos: (unpack_from(fmt,data,pos)[0],pos+l)

readers = [ uvarint, readstruct('<d',8), readbytes, None, None, readstruct('<f',4) ]

def parse(data, schema):
    "parses a protobuf"
    obj = {}
    pos = 0
    while pos < len(data):
        val,pos = uvarint(data,pos)
        typ = val & 7
        key = val >> 3
        val, pos = readers[typ](data,pos)
        if key not in schema: 
            continue
        name, repeated, typ = schema[key]
        if isinstance(typ, dict):
            val = parse(val, typ)
        if typ == 'string':
            val = val.decode('utf8')
        if repeated:
            val = obj.get(name,[]) + [val]
        obj[name] = val
    return obj

def translate(data, media):
    styles = {0: '# ', 1: '## ',100: '* ', 101: '* ', 102: '1. ', 103: '- '}
    rval = []
    refs = []
    txt = data['string']
    pos = 0
    acc = None
    pre = False
    for run in data['attributeRun']:
        l = run['length']
        for frag in re.findall(r'\n|[^\n]+',txt[pos:pos+l]):
            if acc is None: # start paragraph
                pstyle = run.get('paragraphStyle',{}).get('style')
                indent = run.get('paragraphStyle',{}).get('indent',0)
                acc = "  "*indent+styles.get(pstyle,"")
                if pstyle == 103 and run['paragraphStyle']['todo']['done']:
                    acc = "  "*indent+"+ "
                if pstyle == 4:
                    if not pre: 
                        rval.append("```")
                elif pre:
                    rval.append("```")
                pre = pstyle == 4
            if frag == '\n': # end paragraph
                rval.append(acc)
                acc = None
            else: # accumulate and handle inline styles - although bear doesn't seem to support nested ones. 
                link = run.get('link')
                info = run.get('attachmentInfo')
                style = run.get('fontHints',0) + 4*run.get('underline',0) + 8*run.get('strikethrough',0)
                if style & 1: frag = f'*{frag}*'
                if style & 2: frag = f'/{frag}/'
                if style & 4: frag = f'_{frag}_'
                if style & 8: frag = f'~{frag}~'
                if link: frag = f'[{frag}]({link})'
                if info:
                    id = info.get('attachmentIdentifier')
                    fn = media.get(id)
                    if fn:
                        _,e = os.path.splitext(fn)
                        acc += f'[assets/{id}{e}]'
                        refs.append(id)
                    else:
                        acc += f"ATTACH {info}"
                else:
                    acc += frag
        pos += l
    if acc: rval.append(acc)
    rval = '\n'.join(rval)+"\n"
    return rval,refs

# The schema subset needed for bear export
docschema = { 
    2: [ "version", 1, { 
        3: [ "data", 0, {
            2: [ "string", 0, "string"],
            5: [ "attributeRun", 1, {
                1: ["length",0,0],
                2: ["paragraphStyle", 0, {
                    1: ["style", 0,0],
                    4: ["indent",0,0],
                    5: ["todo",0,{ 
                        1: ["todoUUID", 0, "bytes"],
                        2: ["done",0,0]
                    }]
                }],
                5: ["fontHints",0,0],
                6: ["underline",0,0],
                7: ["strikethrough",0,0],
                9: [ "link", 0, "string" ],
                12: [ "attachmentInfo", 0, {
                    1: [ "attachmentIdentifier", 0, "string"],
                    2: [ "typeUTI", 0, "string"]
                }]
            }]
        }]
    }]
}

if __name__ == '__main__':
    root = os.path.expanduser("~/Library/Group Containers/group.com.apple.notes")
    db = sqlite3.Connection(os.path.join(root,'NoteStore.sqlite'))
    media = {} # there is some indirection for attachments
    for a,b,fn in db.execute('select a.zidentifier, b.zidentifier, b.zfilename from ziccloudsyncingobject a left join ziccloudsyncingobject b on a.zmedia = b.z_pk'):
        if fn:
            full = os.path.join(root,'Media',b,fn)
        else:
            full = os.path.join(root,'FallbackImages',a+".jpg")
        if os.path.exists(full):
            media[a] = full

    count = 0
    with zipfile.ZipFile('notes.bearbk','w') as zip:
        for id, title, data, cdate, mdate in db.execute('select o.zidentifier, ztitle1, zdata, o.zcreationdate1, o.zmodificationdate1 from zicnotedata join ziccloudsyncingobject o on znotedata = zicnotedata.z_pk where zicnotedata.zcryptotag is null'):
            ze = f'notes.bearbk/{id}.textbundle'
            if data:
                pb = zlib.decompress(data, 47)
                doc = parse(pb, docschema)['version'][0]['data']
                if not doc['string']: continue # some are blank
                text,refs = translate(doc, media)
                info = {
                    "type":"public.plain-text", 
                    "creatorIdentifier": "net.shinyfrog.bear",
                    "net.shinyfrog.bear": { 
                        "modificationDate": datetime.fromtimestamp(int(mdate)+978307200).isoformat(),
                        "creationDate": datetime.fromtimestamp(int(cdate)+978307200).isoformat()
                    },
                    "version":2
                }
                zip.writestr(f'{ze}/info.json',json.dumps(info))
                zip.writestr(f'{ze}/text.txt',text)
                for ref in refs:
                    _,e = os.path.splitext(media[ref])
                    fn = f'{ze}/assets/{ref}{e}'
                    zip.write(media[ref],fn)
                count += 1
    print(f"wrote {count} notes to notes.bearbk")