libCAR 1.0.0

Python library to parse Catalog Asset Archives (CAR) of iOS Apps.

libCAR

Python library to parse Catalog Asset Archives (CAR) of iOS Apps.

this is a work in progress

Installation

Currently, this project does not have a PyPi installation candidate, so you have to clone the repository and run:

pip install .

Basic Usage

Info: All structures defined in python scripts are easy to understand as there are only field definitions (all classes are dataclasses).

BOM (Build of Materials)

As Catalog Asset Archvies are special BOM (Build Of Materials) files, this project also provides a parser for that file format. You can use the BOM class for almost all operations:

import io
from libcar import BOM

# Either load all bytes and use an in-memory buffer or
# operate in a 'with' statement.
with open("Assets.car", "rb") as fp:
    data = fp.read()

# Create the BOM instance (loads some details automatically)
bom = BOM(io.BytesIO(data))

header = bom.header # get bom-header
variables = bom.variables # get all variables
indexes = bom.indexes # list all indexes (warning: commonly a huge list)

index = bom.index_of("VARIABLE_NAME") # Get index by name
tree = bom.get_tree(index) or bom.tree_of("VARIABLE_NAME")

# Iterate over all tree elements (key, value)
for key, value in bom.iter_tree(tree):
    # key and value will be raw bytes
    ...

CAR (Catalog Asset Archive)

The creation of CAR objects may be handled easier compared to BOM objects as the class defines special parsing methods:

from libcar import CAR

# PArse CAR directly from file stream
car = CAR.parse_file("Assets.car")

header = car.header # get CAR header information
meta = car.metadata # get CAR extended metadata information
key_format = car.key_format # get key format of renditions

for name, facet in car.facets:
    # name will be string and facet CARFacetValue
    ...

# list all appearance keys (if present)
appearance_keys = dict(car.appearance_keys)

# get and parse all stored renditions:
for attributes, rendition in car.iter_renditions(key_format):
    # NOTE: it is recommended to NOT use print(randition) here as the object
    # will store all of its raw data as a blob, which then would be printed
    ...

CAR Scripts

CAR-Extract

Use the script car_extract.py to extract all raw data structures (some images will be ignored). Note that themes and colors won't be extracted.

CAR-Dump

The script car_dump.py offers a convienient way to inspect *.car files using the CLI. An example of a dumped rendition:

python3 car_dump.py Assets.car -T RENDITIONS:4,5

> RENDITIONS:
    - Key: '<00000000 00000001 00000000 00009703 00000055 0000002A>'
        - magic: ISTC
        - version: 1
        - flags: 4
        - width: 0
        - height: 0
        - scale_factor: 0
        - pixel_format: PDF
        - color_space_id: 0
        - reserved_1: '\x00\x00\x00'
        - modification_date: 0
        - layout: <Layout.vector: 9>
        - reserved_2: 0
        - name: 'Arrow Clockwise.pdf'
        - info_len: 28
        - bitmap_count: 1
        - reserved_3: 0
        - payload_size: 3882
        - data:
            - headers:
                [0] => <class 'libcar.rendition.CARRenditionInfoHeader'>
                    - magic: <Magic.composition: 1004>
                    - length: 8
                    - content:
                        - blend_mode: 0
                        - opacity: 1.0
                [1] => <class 'libcar.rendition.CARRenditionInfoHeader'>
                    - magic: <Magic.bitmap_info: 1006>
                    - length: 4
                    - content:
                        - exif_orientation: 1
            - value:
                - magic: DWAR
                - compressed: 0
                - length: 3870
                - data: b"..."