"""Canonical Arrow schemas for every output table (single source of truth)."""
from __future__ import annotations

import pyarrow as pa

WORKS_BASE = pa.schema([
    ("work_id", pa.string()),
    ("doi", pa.string()),
    ("title", pa.string()),
    ("publication_date", pa.string()),
    ("publication_year", pa.int16()),
    ("publication_month", pa.int8()),
    ("type", pa.string()),
    ("primary_topic_id", pa.string()),
    ("primary_topic_name", pa.string()),
    ("primary_topic_score", pa.float32()),
    ("field_id", pa.string()),
    ("field_name", pa.string()),
    ("subfield_id", pa.string()),
    ("subfield_name", pa.string()),
    ("domain_id", pa.string()),
    ("domain_name", pa.string()),
    ("topic_ids", pa.list_(pa.string())),
    ("topic_names", pa.list_(pa.string())),
    ("topic_count", pa.int16()),
    ("author_count", pa.int32()),
    ("author_names", pa.list_(pa.string())),
    ("author_ids", pa.list_(pa.string())),
    ("author_orcids", pa.list_(pa.string())),
    ("institution_count", pa.int32()),
    ("institution_ids", pa.list_(pa.string())),
    ("institution_names", pa.list_(pa.string())),
    ("institution_countries", pa.list_(pa.string())),
    ("country_codes", pa.list_(pa.string())),
    ("country_count", pa.int16()),
    ("source_id", pa.string()),
    ("source_name", pa.string()),
    ("source_type", pa.string()),
    ("issn_l", pa.string()),
    ("issns", pa.list_(pa.string())),
    ("publisher", pa.string()),
    ("cited_by_count", pa.int32()),
    ("recent_citations_3y", pa.int32()),
    ("referenced_works_count", pa.int32()),
    ("referenced_work_ids", pa.list_(pa.string())),
    ("is_oa", pa.bool_()),
    ("oa_status", pa.string()),
    ("oa_url", pa.string()),
    ("license", pa.string()),
    ("license_id", pa.string()),
    ("language", pa.string()),
    ("abstract", pa.string()),
    ("has_abstract", pa.bool_()),
    ("is_retracted", pa.bool_()),
    ("is_paratext", pa.bool_()),
    ("fwci", pa.float32()),
    ("openalex_citation_percentile", pa.float32()),
    ("created_date", pa.string()),
    ("updated_date", pa.string()),
    ("mesh_terms", pa.list_(pa.string())),
    ("keywords", pa.list_(pa.string())),
])

WORKS_DERIVED = pa.schema([
    ("paper_age_years", pa.float32()),
    ("citations_per_year", pa.float32()),
    ("citation_velocity", pa.float32()),
    ("international_collaboration", pa.bool_()),
    ("multi_institutional", pa.bool_()),
    ("publication_recency_score", pa.float32()),
    ("citation_percentile", pa.float32()),
    ("high_impact_flag", pa.bool_()),
    ("research_growth_score", pa.float32()),
    ("topic_growth_score", pa.float32()),
    ("emerging_topic_flag", pa.bool_()),
])

WORKS_FULL = pa.schema(list(WORKS_BASE) + list(WORKS_DERIVED))

WORK_AUTHORS = pa.schema([
    ("work_id", pa.string()),
    ("author_id", pa.string()),
    ("author_name", pa.string()),
    ("orcid", pa.string()),
    ("author_position", pa.string()),
    ("is_corresponding", pa.bool_()),
    ("institution_ids", pa.list_(pa.string())),
    ("country_codes", pa.list_(pa.string())),
])

WORK_INSTITUTIONS = pa.schema([
    ("work_id", pa.string()),
    ("institution_id", pa.string()),
    ("institution_name", pa.string()),
    ("country_code", pa.string()),
    ("institution_type", pa.string()),
])

WORK_TOPICS = pa.schema([
    ("work_id", pa.string()),
    ("topic_id", pa.string()),
    ("topic_name", pa.string()),
    ("score", pa.float32()),
    ("is_primary", pa.bool_()),
    ("subfield_id", pa.string()),
    ("subfield_name", pa.string()),
    ("field_id", pa.string()),
    ("field_name", pa.string()),
    ("domain_id", pa.string()),
    ("domain_name", pa.string()),
])

CITATION_EDGES = pa.schema([
    ("citing_work_id", pa.string()),
    ("cited_work_id", pa.string()),
    ("cited_in_dataset", pa.bool_()),
])

AUTHORS_DIM = pa.schema([
    ("author_id", pa.string()),
    ("author_name", pa.string()),
    ("orcid", pa.string()),
    ("works_count", pa.int32()),
    ("first_year", pa.int16()),
    ("last_year", pa.int16()),
    ("total_citations", pa.int64()),
    ("mean_citations", pa.float32()),
    ("institution_ids", pa.list_(pa.string())),
    ("country_codes", pa.list_(pa.string())),
    ("primary_field_id", pa.string()),
    ("primary_field_name", pa.string()),
])

INSTITUTIONS_DIM = pa.schema([
    ("institution_id", pa.string()),
    ("institution_name", pa.string()),
    ("country_code", pa.string()),
    ("works_count", pa.int32()),
    ("first_year", pa.int16()),
    ("last_year", pa.int16()),
    ("total_citations", pa.int64()),
    ("mean_citations", pa.float32()),
    ("primary_field_id", pa.string()),
    ("primary_field_name", pa.string()),
])

TOPICS_DIM = pa.schema([
    ("topic_id", pa.string()),
    ("topic_name", pa.string()),
    ("subfield_id", pa.string()),
    ("subfield_name", pa.string()),
    ("field_id", pa.string()),
    ("field_name", pa.string()),
    ("domain_id", pa.string()),
    ("domain_name", pa.string()),
    ("works_count", pa.int64()),
    ("first_year", pa.int16()),
    ("last_year", pa.int16()),
    ("total_citations", pa.int64()),
    ("mean_citations", pa.float32()),
])

SOURCES_DIM = pa.schema([
    ("source_id", pa.string()),
    ("source_name", pa.string()),
    ("source_type", pa.string()),
    ("issn_l", pa.string()),
    ("issns", pa.list_(pa.string())),
    ("publisher", pa.string()),
    ("is_oa", pa.bool_()),
    ("works_count", pa.int64()),
    ("total_citations", pa.int64()),
    ("mean_citations", pa.float32()),
])

TOPIC_YEAR_COUNTS = pa.schema([
    ("topic_id", pa.string()),
    ("topic_name", pa.string()),
    ("publication_year", pa.int16()),
    ("works_count", pa.int64()),
])

TOPIC_GROWTH = pa.schema([
    ("topic_id", pa.string()),
    ("topic_name", pa.string()),
    ("first_year", pa.int16()),
    ("last_year", pa.int16()),
    ("early_count", pa.int64()),
    ("late_count", pa.int64()),
    ("growth_ratio", pa.float32()),
    ("cagr", pa.float32()),
    ("trend_slope", pa.float32()),
    ("emerging_topic_flag", pa.bool_()),
])

COUNTRY_YEAR_COUNTS = pa.schema([
    ("country_code", pa.string()),
    ("publication_year", pa.int16()),
    ("works_count", pa.int64()),
])

COUNTRY_COLLABORATION = pa.schema([
    ("country_a", pa.string()),
    ("country_b", pa.string()),
    ("collaboration_count", pa.int64()),
])

INSTITUTION_ACTIVITY = pa.schema([
    ("institution_id", pa.string()),
    ("institution_name", pa.string()),
    ("country_code", pa.string()),
    ("publication_year", pa.int16()),
    ("works_count", pa.int64()),
    ("total_citations", pa.int64()),
])