structa

structa is a hidden gem. It's one of many great utilities created by Dave Jones. It's a command line tool for analysing JSON files. Sometimes you need to inspect the structure of a large nested JSON file, and it's too unwieldy to work out what kind of data it contains. structa is perfect for showing you.

The People in Space API shows the number of people currently in space, and their names and craft name:

wget http://api.open-notify.org/astros.json structa astros.json

Output:

├─ 'number': 5 ├─ 'message': success └─ 'people': [ ├─ 'craft': ISS └─ 'name': {Doug Hurley|Bob Behnken|Chris Cassidy|Anatoly...} ]

As you can see, most of this data is static – so the values are just given as is. But the people value is an array, so it shows the keys and the data types. Again, since craft is the same every time, it's just given, but name varies, so the value is given as a set of the variations.

The Python Package Index (PyPI) provides a JSON API for packages:

wget https://pypi.org/pypi/numpy/json -O numpy.json structa numpy.json

Output:

├─ 'last_serial': 7391399 ├─ 'urls': [  ├─ 'requires_python': >=3.5  ├─ 'filename': <str>  ├─ 'yanked_reason': None  ├─ 'comment_text':  ├─ 'python_version': {cp36|source|cp35|cp38|cp37}  ├─ 'md5_digest': <str 'hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh'>  ├─ 'size': <int 5.2M..19.6M>  ├─ 'downloads': -1  ├─ 'url': <URL>  ├─ 'has_sig': False  ├─ 'upload_time': <datetime '%Y-%m-%dT%H:%M:%S' 2020-06-04 00:10:51..2020-06-04 00:27:58>  ├─ 'upload_time_iso_8601': <str '2020-06-04T00:dd:dd.ddddddZ'>  ├─ 'digests':  │ ├─ 'md5': <str 'hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh'>  │ └─ 'sha256': <str ...>  ├─ 'packagetype': {sdist|bdist_wheel}  └─ 'yanked': False ] ├─ 'info': │ └─ <str>: <value> └─ 'releases':  └─ <str>: [  ├─ 'requires_python': {>=3.5|>=2.7,!=3.0.*,!=3.1.*,!=3.2.*,!=3.3.*|>=3.6|None}  ├─ 'filename': <str>  ├─ 'yanked_reason': None  ├─ 'comment_text': {|Simple installer, no SSE instructions. |Simple windows...}  ├─ 'python_version': <str>  ├─ 'md5_digest': <str 'hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh'>  ├─ 'size': <int 1.8M..23.4M>  ├─ 'downloads': -1  ├─ 'url': <URL>  ├─ 'has_sig': {False|True}  ├─ 'upload_time': <datetime '%Y-%m-%dT%H:%M:%S' 2006-12-02 02:07:43..2020-06-04 00:35:35>  ├─ 'upload_time_iso_8601': <str>  ├─ 'digests':  │ ├─ 'md5': <str 'hhhhhhhhhhhhhhhhhhhhhhhhhhhhhhhh'>  │ └─ 'sha256': <str ...>  ├─ 'packagetype': {bdist_wheel|bdist_wininst|sdist}  └─ 'yanked': False  ]

Here you can see a wide variety of helpful information inferred from the data, including:

Under urls:
filename is a string (too many options to show as a set)
python_version is always one of: {cp36|source|cp35|cp38|cp37}
size is an integer between 5.2 million and 19.6 million (bytes)
md5_digest is a 32-character hex string
url is detected as being a URL
upload_time is a datetime string in the format '%Y-%m-%dT%H:%M:%S' ranging between 2020-06-04 00:10:51 and 2020-06-04 00:27:58
packagetype is either sdist or bdist_wheel
Under info:
There's a variable key name, which contains a variety of values
Under releases:
There's a variable key name, which contains an array...

Unfortunately, structa is not published on PyPI yet, so for now you have to install it from GitHub:

git clone https://github.com/waveform80/structa cd structa pip install .