pykeepass

This library allows you to write entries to a KeePass database.

Come chat at #pykeepass on Freenode or #pykeepass:matrix.org on Matrix.

Installation

sudo apt install python3-lxml
pip install pykeepass

Example

from pykeepass import PyKeePass

# load database
>>> kp = PyKeePass('db.kdbx', password='somePassw0rd')

# find any group by its name
>>> group = kp.find_groups(name='social', first=True)

# get the entries in a group
>>> group.entries
[Entry: "social/facebook (myusername)", Entry: "social/twitter (myusername)"]

# find any entry by its title
>>> entry = kp.find_entries(title='facebook', first=True)

# retrieve the associated password
>>> entry.password
's3cure_p455w0rd'

# update an entry
>>> entry.notes = 'primary facebook account'

# create a new group
>>> group = kp.add_group(kp.root_group, 'email')

# create a new entry
>>> kp.add_entry(group, 'gmail', 'myusername', 'myPassw0rdXX')
Entry: "email/gmail (myusername)"

# save database
>>> kp.save()

Finding Entries

find_entries (title=None, username=None, password=None, url=None, notes=None, otp=None, path=None, uuid=None, tags=None, string=None, group=None, recursive=True, regex=False, flags=None, history=False, first=False)

Returns entries which match all provided parameters, where title, username, password, url, notes, otp, autotype_window and autotype_sequence are strings, path is a list, string is a dict, autotype_enabled is a boolean, uuid is a uuid.UUID and tags is a list of strings. This function has optional regex boolean and flags string arguments, which means to interpret search strings as XSLT style regular expressions with flags.

The path list is a full path to an entry (ex. ['foobar_group', 'foobar_entry']). This implies first=True. All other arguments are ignored when this is given. This is useful for handling user input.

The string dict allows for searching custom string fields. ex. {'custom_field1': 'custom value', 'custom_field2': 'custom value'}

The group argument determines what Group to search under, and the recursive boolean controls whether to search recursively.

The history (default False) boolean controls whether history entries should be included in the search results.

The first (default False) boolean controls whether to return the first matched item, or a list of matched items.

• if first=False, the function returns a list of Entry s or [] if there are no matches
• if first=True, the function returns the first Entry match, or None if there are no matches

entries

a flattened list of all entries in the database

>>> kp.entries
[Entry: "foo_entry (myusername)", Entry: "foobar_entry (myusername)", Entry: "social/gmail (myusername)", Entry: "social/facebook (myusername)"]

>>> kp.find_entries(title='gmail', first=True)
Entry: "social/gmail (myusername)"

>>> kp.find_entries(title='foo.*', regex=True)
[Entry: "foo_entry (myusername)", Entry: "foobar_entry (myusername)"]

>>> entry = kp.find_entries(title='foo.*', url='.*facebook.*', regex=True, first=True)
>>> entry.url
'facebook.com'
>>> entry.title
'foo_entry'
>>> entry.title = 'hello'

>>> group = kp.find_group(name='social', first=True)
>>> kp.find_entries(title='facebook', group=group, recursive=False, first=True)
Entry: "social/facebook (myusername)"

>>> entry.otp
otpauth://totp/test:lkj?secret=TEST%3D%3D%3D%3D&period=30&digits=6&issuer=test

Finding Groups

find_groups (name=None, path=None, uuid=None, notes=None, group=None, recursive=True, regex=False, flags=None, first=False)

where name and notes are strings, path is a list, uuid is a uuid.UUID. This function has optional regex boolean and flags string arguments, which means to interpret search strings as XSLT style regular expressions with flags.

The path list is a full path to a group (ex. ['foobar_group', 'sub_group']). This implies first=True. All other arguments are ignored when this is given. This is useful for handling user input.

The group argument determines what Group to search under, and the recursive boolean controls whether to search recursively.

The first (default False) boolean controls whether to return the first matched item, or a list of matched items.

• if first=False, the function returns a list of Group s or [] if there are no matches
• if first=True, the function returns the first Group match, or None if there are no matches

root_group

the Root group to the database

groups

a flattened list of all groups in the database

>>> kp.groups
[Group: "foo", Group "foobar", Group: "social", Group: "social/foo_subgroup"]

>>> kp.find_groups(name='foo', first=True)
Group: "foo"

>>> kp.find_groups(name='foo.*', regex=True)
[Group: "foo", Group "foobar"]

>>> kp.find_groups(path=['social'], regex=True)
[Group: "social", Group: "social/foo_subgroup"]

>>> kp.find_groups(name='social', first=True).subgroups
[Group: "social/foo_subgroup"]

>>> kp.root_group
Group: "/"

Entry Functions and Properties

add_entry (destination_group, title, username, password, url=None, notes=None, tags=None, expiry_time=None, icon=None, force_creation=False)

delete_entry (entry)

trash_entry (entry)

move a group to the recycle bin. The recycle bin is created if it does not exit. entry must be an empty Entry.

move_entry (entry, destination_group)

atime

access time

ctime

creation time

mtime

modification time

where destination_group is a Group instance. entry is an Entry instance. title, username, password, url, notes, tags, icon are strings. expiry_time is a datetime instance.

If expiry_time is a naive datetime object (i.e. expiry_time.tzinfo is not set), the timezone is retrieved from dateutil.tz.gettz().

# add a new entry to the Root group
>>> kp.add_entry(kp.root_group, 'testing', 'foo_user', 'passw0rd')
Entry: "testing (foo_user)"

# add a new entry to the social group
>>> group = kp.find_groups(name='social', first=True)
>>> entry = kp.add_entry(group, 'testing', 'foo_user', 'passw0rd')
Entry: "testing (foo_user)"

# save the database
>>> kp.save()

# delete an entry
>>> kp.delete_entry(entry)

# move an entry
>>> kp.move_entry(entry, kp.root_group)

# save the database
>>> kp.save()

# change creation time
>>> from datetime import datetime, timezone
>>> entry.ctime = datetime(2023, 1, 1, tzinfo=timezone.utc)

# update modification or access time
>>> entry.touch(modify=True)

Group Functions and Properties

add_group (destination_group, group_name, icon=None, notes=None)

delete_group (group)

trash_group (group)

move a group to the recycle bin. The recycle bin is created if it does not exit. group must be an empty Group.

empty_group (group)

delete all entries and subgroups of a group. group is an instance of Group.

move_group (group, destination_group)

atime

access time

ctime

creation time

mtime

modification time

destination_group and group are instances of Group. group_name is a string

# add a new group to the Root group
>>> group = kp.add_group(kp.root_group, 'social')

# add a new group to the social group
>>> group2 = kp.add_group(group, 'gmail')
Group: "social/gmail"

# save the database
>>> kp.save()

# delete a group
>>> kp.delete_group(group)

# move a group
>>> kp.move_group(group2, kp.root_group)

# save the database
>>> kp.save()

# change creation time
>>> from datetime import datetime, timezone
>>> group.ctime = datetime(2023, 1, 1, tzinfo=timezone.utc)

# update modification or access time
>>> group.touch(modify=True)

Attachments

In this section, binary refers to the bytes of the attached data (stored at the root level of the database), while attachment is a reference to a binary (stored in an entry). A binary can be referenced by none, one or many attachments.

add_binary (data, compressed=True, protected=True)

where data is bytes. Adds a blob of data to the database. The attachment reference must still be added to an entry (see below). compressed only applies to KDBX3 and protected only applies to KDBX4 (no effect if used on wrong database version). Returns id of attachment.

delete_binary (id)

where id is an int. Removes binary data from the database and deletes any attachments that reference it. Since attachments reference binaries by their positional index, attachments that reference binaries with id > id will automatically be decremented.

find_attachments (id=None, filename=None, element=None, recursive=True, regex=False, flags=None, history=False, first=False)

where id is an int, filename is a string, and element is an Entry or Group to search under.

• if first=False, the function returns a list of Attachment s or [] if there are no matches
• if first=True, the function returns the first Attachment match, or None if there are no matches

binaries

list of bytestrings containing binary data. List index corresponds to attachment id

attachments

list containing all Attachment s in the database.

Entry.add_attachment (id, filename)

where id is an int and filename is a string. Creates a reference using the given filename to a database binary. The existence of a binary with the given id is not checked. Returns Attachment.

Entry.delete_attachment (attachment)

where attachment is an Attachment. Deletes a reference to a database binary.

Entry.attachments

list of Attachment s for this Entry.

Attachment.id

id of data that this attachment points to

Attachment.filename

string representing this attachment

Attachment.data

the data that this attachment points to. Raises BinaryError if data does not exist.

Attachment.entry

the entry that this attachment is attached to

>>> e = kp.add_entry(kp.root_group, title='foo', username='', password='')

# add attachment data to the db
>>> binary_id = kp.add_binary(b'Hello world')

>>> kp.binaries
[b'Hello world']

# add attachment reference to entry
>>> a = e.add_attachment(binary_id, 'hello.txt')
>>> a
Attachment: 'hello.txt' -> 0

# access attachments
>>> a
Attachment: 'hello.txt' -> 0
>>> a.id
0
>>> a.filename
'hello.txt'
>>> a.data
b'Hello world'
>>> e.attachments
[Attachment: 'hello.txt' -> 0]

# list all attachments in the database
>>> kp.attachments
[Attachment: 'hello.txt' -> 0]

# search attachments
>>> kp.find_attachments(filename='hello.txt')
[Attachment: 'hello.txt** -> 0]

# delete attachment reference
>>> e.delete_attachment(a)

# or, delete both attachment reference and binary
>>> kp.delete_binary(binary_id**

Credential Expiry

credchange_date

datetime object with date of last credentials change

credchange_required

boolean whether database credentials have expired and are required to change

credchange_recommended

boolean whether database credentials have expired and are recommended to change

credchange_required_days

days after credchange_date that credential update is required

credchange_recommended_days

days after credchange_date that credential update is recommended

Miscellaneous

read (filename=None, password=None, keyfile=None, transformed_key=None, decrypt=False)

where filename, password, and keyfile are strings ( filename and keyfile may also be file-like objects). filename is the path to the database, password is the master password string, and keyfile is the path to the database keyfile. At least one of password and keyfile is required. Alternatively, the derived key can be supplied directly through transformed_key. decrypt tells whether the file should be decrypted or not.

Can raise CredentialsError, HeaderChecksumError, or PayloadChecksumError.

reload ()

reload database from disk using previous credentials

save (filename=None)

where filename is the path of the file to save to (filename may also be file-like object). If filename is not given, the path given in read will be used.

password

string containing database password. Can also be set. Use None for no password.

filename

string containing path to database. Can also be set

keyfile

string containing path to the database keyfile. Can also be set. Use None for no keyfile.

version

tuple containing database version. e.g. (3, 1) is a KDBX version 3.1 database.

encryption_algorithm

string containing algorithm used to encrypt database. Possible values are aes256, chacha20, and twofish.

create_database (filename, password=None, keyfile=None, transformed_key=None)

create a new database at filename with supplied credentials. Returns PyKeePass object

tree

database lxml tree

xml

get database XML data as string

dump_xml (filename)

pretty print database XML to file

TOTP

Entry.otp

TOTP URI which can be passed to an OTP library to generate codes

# find an entry which has otp attribute
>>> e = kp.find_entries(otp='.*', regex=True, first=True)
>>> import pyotp
>>> pyotp.parse_uri(e.otp).now()
799270

Tests and Debugging

Run tests with python tests/tests.py or python tests/tests.py SomeSpecificTest

Enable debugging when doing tests in console:

>>> from pykeepass.pykeepass import debug_setup
>>> debug_setup()
>>> kp.entries[0]
DEBUG:pykeepass.pykeepass:xpath query: //Entry
DEBUG:pykeepass.pykeepass:xpath query: (ancestor::Group)[last()]
DEBUG:pykeepass.pykeepass:xpath query: (ancestor::Group)[last()]
DEBUG:pykeepass.pykeepass:xpath query: String/Key[text()="Title"]/../Value
DEBUG:pykeepass.pykeepass:xpath query: String/Key[text()="UserName"]/../Value
Entry: "root_entry (foobar_user)"