Skip to main content

Standard and idiosyncratic schemata for Sanskrit data, with a library of validation, (de-)serialization and other utilities.

Project description

Introduction

This module defines: - shared standard schema for communicating and storing Sanskrit data. - various idiosyncratic notations used by various modules which deviate from the proposed standards.

Together with this, it provides python classes (corresponding to the schema) and shared libraries for validating, (de-)serializing and storing sanskrit data. Similar libraries in various other languages are being built: - Scala (likely compatible with Java): db-interface .

Motivation

  • Various sanskrit modules need to communicate data amongst each other (for example through a REST API or database stores or even function calls). Examples of the data being communicated could be:

  • Gramatical details of a given word

  • Sentences in a given book chapter

  • Annotations on a given phrase

  • When it comes to serialization formats - two distinct approaches present themselves to us:

  • One possible route is to have each project defining and using its own idiosyncratic notation. But this entails an additional burdens:

    • Each communicating module having to convert the data from one idiosyncratic notation to another.

    • Good schema design or notation is non trivial. Even if no external module is using the data, it is a waste to have to reinvent the wheel.

  • A superior route is to have a common, standard format for encoding various data-types for storage/ communication.

  • To the extant possible, we should take latter approach to data storage and communication.

  • Where idiosyncratic notations are adapted for various reasons, it is still desirable to collect such definitions in a single module - to facilitate conversion to the standard format.

For users

  • sudo pip2 install indic_transliteration -U

  • Web.

  • The sanskrit_data.schema module contains various python files describing various Python classes for storing Sanskrit data, and their corresponding schema.

  • At the base of every such class is the common.JsonObject class.

  • Design considerations for data containers corresponding to the various submodules (such as books and annotations) are given below - or in the corresponding source files.

For contributors

Contact

Have a problem or question? Please head to github.

Packaging

  • ~/.pypirc should have your pypi login credentials.

    python setup.py bdist_wheel
    twine upload dist/* --skip-existing

Design principles

General principles

  • We want data to be stored and communicated between programs in a popular, extensible format - we want to take advantage of existing technologies to the maximum possible extant and not waste time reinventing associated (de)serialization, validation and other libraries.

  • But this does not prevent the data from being presented in a different format for human consumption.

While designing the JSON data-model: - Type-hint in JSON should be jsonClass (a language-independent name we’ve picked). - Try to avoid field-names which conflict with programming language keywords. (Eg. Prefer “source_type” to “type”). - In general, use camelCase or underscore_case for field names - both are fine. Where romanized (potentially mixed case) sanskrit words are used, the latter is the superior convention. - Where field names and values are to be automatically rendered into various scripts, as in case of sanskrit vyAkarana jargon (eg: vibhakti, lakAra), we prefer SLP1 transliteration (“viBakti”, “lakAra”). - PS: Convenient transliteration modules are available in various languages: please see them listed here. - A transliteration map for reference. - When in doubt, keep fields optional.

Books and annotations

  • Basic principles

  • Books are stored as a hierarchy of BookPortion objects - book containing many chapters containing many lines etc..

  • Annotations are stored in a similar hierarchy, for example - a TextAnnotation having PadaAnnotations having SamaasaAnnotations.

    • Some Annotations (eg. SandhiAnnotation, TextAnnotation) can have multiple “targets” (ie. other objects being annotated).

    • Rather than a simple tree, we end up with a Directed Acyclic Graph (DAG) of Annotation objects.

  • JSON schema mindmap here (Updated as needed).

  • The data containers are in a separate sanskrit_data module - so that it can be extracted and used outside this server.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distribution

sanskrit_data-0.1.2-py2.py3-none-any.whl (17.4 kB view hashes)

Uploaded Python 2 Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page