Use the decoder as a Python package

Installation

The distribution name is sandborg-petersen-decoder; the Python import name is sp_morph_decoder.

Install it from PyPI:

python -m pip install sandborg-petersen-decoder

To use the latest development revision directly from GitHub:

python -m pip install "git+https://github.com/tonyjurg/Sandborg-Petersen-decoder.git"

Quick start

from sp_morph_decoder import decode_tag

result = decode_tag("V-PAI-3S")
print(result)

The result is a dictionary containing the decoded fields:

{
    'Part of Speech': 'Verb',
    'Tense': 'Present',
    'Voice': 'Active',
    'Mood': 'Indicative',
    'Person': 'Third Person',
    'Number': 'Singular'
}

Input is normalized by trimming surrounding whitespace and converting letters to uppercase.

Permissive and strict modes

Permissive mode

Permissive mode is the default. It returns every field that can be recovered and adds an Errors list when structural validation finds a problem.

from sp_morph_decoder import decode_tag

result = decode_tag("N-XYZ")
for message in result.get("Errors", []):
    print(message)

Strict mode

Strict mode raises MorphologyDecodeError when a tag cannot be decoded cleanly. The exception retains the normalized partial result and individual validation messages.

from sp_morph_decoder import MorphologyDecodeError, decode_tag

try:
    result = decode_tag("N-XYZ", mode="strict")
except MorphologyDecodeError as error:
    print(error.tag)       # original input
    print(error.errors)    # tuple of messages
    print(error.result)    # partial decoded dictionary

Public API

Command-line interface

The package installs the sp-morph-decode command:

sp-morph-decode V-PAI-3S
sp-morph-decode N-XYZ --mode strict

It can also be invoked as a Python module:

python -m sp_morph_decoder V-PAI-3S --mode permissive

Results are written as formatted JSON. Successful and permissive decodes exit with status 0; a strict validation failure exits with status 2.

Example tags