# Generated by Django 5.2.6 on 2026-05-27 13:58

import django.db.models.deletion
import uuid
from django.conf import settings
from django.db import migrations, models


class Migration(migrations.Migration):

    dependencies = [
        ("dms", "0008_phase10_records"),
        ("organizations", "0001_initial"),
        ("tenants", "0001_initial"),
        migrations.swappable_dependency(settings.AUTH_USER_MODEL),
    ]

    operations = [
        migrations.CreateModel(
            name="OCRResult",
            fields=[
                (
                    "id",
                    models.BigAutoField(
                        auto_created=True, primary_key=True, serialize=False, verbose_name="ID"
                    ),
                ),
                (
                    "is_deleted",
                    models.BooleanField(db_index=True, default=False, verbose_name="deleted"),
                ),
                (
                    "deleted_at",
                    models.DateTimeField(blank=True, null=True, verbose_name="deleted at"),
                ),
                (
                    "created_at",
                    models.DateTimeField(
                        auto_now_add=True, db_index=True, verbose_name="created at"
                    ),
                ),
                ("updated_at", models.DateTimeField(auto_now=True, verbose_name="updated at")),
                (
                    "public_id",
                    models.UUIDField(
                        db_index=True,
                        default=uuid.uuid4,
                        editable=False,
                        unique=True,
                        verbose_name="public id",
                    ),
                ),
                (
                    "provider",
                    models.CharField(
                        default="default",
                        help_text="Identifier of the OCR provider (e.g. 'tesseract', 'azure_ocr').",
                        max_length=100,
                        verbose_name="provider",
                    ),
                ),
                (
                    "status",
                    models.CharField(
                        choices=[
                            ("pending", "Pending"),
                            ("processing", "Processing"),
                            ("completed", "Completed"),
                            ("failed", "Failed"),
                        ],
                        db_index=True,
                        default="pending",
                        max_length=20,
                        verbose_name="status",
                    ),
                ),
                (
                    "language_code",
                    models.CharField(
                        blank=True,
                        default="",
                        help_text="BCP-47 language tag detected or requested (e.g. 'en', 'fa').",
                        max_length=20,
                        verbose_name="language code",
                    ),
                ),
                (
                    "full_text",
                    models.TextField(
                        blank=True,
                        default="",
                        help_text="Concatenated raw OCR output across all pages.",
                        verbose_name="full text",
                    ),
                ),
                (
                    "confidence_score",
                    models.DecimalField(
                        blank=True,
                        decimal_places=4,
                        help_text="Aggregate confidence score in the range [0.0000, 1.0000].",
                        max_digits=5,
                        null=True,
                        verbose_name="confidence score",
                    ),
                ),
                ("page_count", models.PositiveIntegerField(default=0, verbose_name="page count")),
                (
                    "processing_metadata",
                    models.JSONField(
                        blank=True,
                        default=dict,
                        help_text="Provider-specific metadata (timing, warnings, per-page results).",
                        verbose_name="processing metadata",
                    ),
                ),
                (
                    "error_message",
                    models.TextField(blank=True, default="", verbose_name="error message"),
                ),
                (
                    "submitted_at",
                    models.DateTimeField(auto_now_add=True, verbose_name="submitted at"),
                ),
                (
                    "completed_at",
                    models.DateTimeField(blank=True, null=True, verbose_name="completed at"),
                ),
                (
                    "created_by",
                    models.ForeignKey(
                        blank=True,
                        editable=False,
                        null=True,
                        on_delete=django.db.models.deletion.SET_NULL,
                        related_name="+",
                        to=settings.AUTH_USER_MODEL,
                        verbose_name="created by",
                    ),
                ),
                (
                    "deleted_by",
                    models.ForeignKey(
                        blank=True,
                        null=True,
                        on_delete=django.db.models.deletion.SET_NULL,
                        related_name="+",
                        to=settings.AUTH_USER_MODEL,
                        verbose_name="deleted by",
                    ),
                ),
                (
                    "document_version",
                    models.ForeignKey(
                        on_delete=django.db.models.deletion.CASCADE,
                        related_name="ocr_results",
                        to="dms.documentversion",
                        verbose_name="document version",
                    ),
                ),
                (
                    "organization_node",
                    models.ForeignKey(
                        on_delete=django.db.models.deletion.PROTECT,
                        related_name="+",
                        to="organizations.organizationnode",
                        verbose_name="organization node",
                    ),
                ),
                (
                    "tenant",
                    models.ForeignKey(
                        on_delete=django.db.models.deletion.CASCADE,
                        related_name="+",
                        to="tenants.tenant",
                        verbose_name="tenant",
                    ),
                ),
                (
                    "updated_by",
                    models.ForeignKey(
                        blank=True,
                        editable=False,
                        null=True,
                        on_delete=django.db.models.deletion.SET_NULL,
                        related_name="+",
                        to=settings.AUTH_USER_MODEL,
                        verbose_name="updated by",
                    ),
                ),
            ],
            options={
                "verbose_name": "OCR result",
                "verbose_name_plural": "OCR results",
                "ordering": ("-submitted_at",),
            },
        ),
        migrations.CreateModel(
            name="ExtractedEntity",
            fields=[
                (
                    "id",
                    models.BigAutoField(
                        auto_created=True, primary_key=True, serialize=False, verbose_name="ID"
                    ),
                ),
                (
                    "is_deleted",
                    models.BooleanField(db_index=True, default=False, verbose_name="deleted"),
                ),
                (
                    "deleted_at",
                    models.DateTimeField(blank=True, null=True, verbose_name="deleted at"),
                ),
                (
                    "created_at",
                    models.DateTimeField(
                        auto_now_add=True, db_index=True, verbose_name="created at"
                    ),
                ),
                ("updated_at", models.DateTimeField(auto_now=True, verbose_name="updated at")),
                (
                    "public_id",
                    models.UUIDField(
                        db_index=True,
                        default=uuid.uuid4,
                        editable=False,
                        unique=True,
                        verbose_name="public id",
                    ),
                ),
                (
                    "entity_type",
                    models.CharField(
                        choices=[
                            ("person", "Person"),
                            ("organization", "Organization"),
                            ("date", "Date"),
                            ("amount", "Amount"),
                            ("location", "Location"),
                            ("reference", "Reference Number"),
                            ("custom", "Custom"),
                        ],
                        db_index=True,
                        default="custom",
                        max_length=50,
                        verbose_name="entity type",
                    ),
                ),
                (
                    "value",
                    models.CharField(
                        help_text="The raw extracted entity value.",
                        max_length=500,
                        verbose_name="value",
                    ),
                ),
                (
                    "normalized_value",
                    models.CharField(
                        blank=True,
                        default="",
                        help_text="Standardized form of the value (e.g. ISO date, canonical name).",
                        max_length=500,
                        verbose_name="normalized value",
                    ),
                ),
                (
                    "confidence_score",
                    models.DecimalField(
                        blank=True,
                        decimal_places=4,
                        max_digits=5,
                        null=True,
                        verbose_name="confidence score",
                    ),
                ),
                (
                    "provider",
                    models.CharField(
                        blank=True,
                        default="",
                        help_text="Provider that extracted this entity.",
                        max_length=100,
                        verbose_name="provider",
                    ),
                ),
                (
                    "position_metadata",
                    models.JSONField(
                        blank=True,
                        default=dict,
                        help_text="Location hints in the source document: page number, bounding box, character offsets.  Format is provider-specific.",
                        verbose_name="position metadata",
                    ),
                ),
                (
                    "created_by",
                    models.ForeignKey(
                        blank=True,
                        editable=False,
                        null=True,
                        on_delete=django.db.models.deletion.SET_NULL,
                        related_name="+",
                        to=settings.AUTH_USER_MODEL,
                        verbose_name="created by",
                    ),
                ),
                (
                    "deleted_by",
                    models.ForeignKey(
                        blank=True,
                        null=True,
                        on_delete=django.db.models.deletion.SET_NULL,
                        related_name="+",
                        to=settings.AUTH_USER_MODEL,
                        verbose_name="deleted by",
                    ),
                ),
                (
                    "document_version",
                    models.ForeignKey(
                        on_delete=django.db.models.deletion.CASCADE,
                        related_name="extracted_entities",
                        to="dms.documentversion",
                        verbose_name="document version",
                    ),
                ),
                (
                    "organization_node",
                    models.ForeignKey(
                        on_delete=django.db.models.deletion.PROTECT,
                        related_name="+",
                        to="organizations.organizationnode",
                        verbose_name="organization node",
                    ),
                ),
                (
                    "tenant",
                    models.ForeignKey(
                        on_delete=django.db.models.deletion.CASCADE,
                        related_name="+",
                        to="tenants.tenant",
                        verbose_name="tenant",
                    ),
                ),
                (
                    "updated_by",
                    models.ForeignKey(
                        blank=True,
                        editable=False,
                        null=True,
                        on_delete=django.db.models.deletion.SET_NULL,
                        related_name="+",
                        to=settings.AUTH_USER_MODEL,
                        verbose_name="updated by",
                    ),
                ),
                (
                    "source",
                    models.ForeignKey(
                        blank=True,
                        help_text="OCR run that produced this entity, if any.",
                        null=True,
                        on_delete=django.db.models.deletion.SET_NULL,
                        related_name="entities",
                        to="dms.ocrresult",
                        verbose_name="OCR source",
                    ),
                ),
            ],
            options={
                "verbose_name": "extracted entity",
                "verbose_name_plural": "extracted entities",
                "ordering": ["entity_type", "value"],
            },
        ),
        migrations.CreateModel(
            name="AIClassification",
            fields=[
                (
                    "id",
                    models.BigAutoField(
                        auto_created=True, primary_key=True, serialize=False, verbose_name="ID"
                    ),
                ),
                (
                    "is_deleted",
                    models.BooleanField(db_index=True, default=False, verbose_name="deleted"),
                ),
                (
                    "deleted_at",
                    models.DateTimeField(blank=True, null=True, verbose_name="deleted at"),
                ),
                (
                    "created_at",
                    models.DateTimeField(
                        auto_now_add=True, db_index=True, verbose_name="created at"
                    ),
                ),
                ("updated_at", models.DateTimeField(auto_now=True, verbose_name="updated at")),
                (
                    "public_id",
                    models.UUIDField(
                        db_index=True,
                        default=uuid.uuid4,
                        editable=False,
                        unique=True,
                        verbose_name="public id",
                    ),
                ),
                (
                    "provider",
                    models.CharField(
                        default="default",
                        help_text="Identifier of the AI provider (e.g. 'openai', 'azure_ai').",
                        max_length=100,
                        verbose_name="provider",
                    ),
                ),
                (
                    "classification_type",
                    models.CharField(
                        choices=[
                            ("document_type", "Document Type"),
                            ("topic", "Topic"),
                            ("sentiment", "Sentiment"),
                            ("summary", "Summary"),
                            ("custom", "Custom"),
                        ],
                        db_index=True,
                        default="document_type",
                        max_length=50,
                        verbose_name="classification type",
                    ),
                ),
                (
                    "label",
                    models.CharField(
                        blank=True,
                        default="",
                        help_text="Primary classification output (e.g. 'contract', 'finance').",
                        max_length=255,
                        verbose_name="label",
                    ),
                ),
                (
                    "confidence_score",
                    models.DecimalField(
                        blank=True,
                        decimal_places=4,
                        help_text="Confidence score in the range [0.0000, 1.0000].",
                        max_digits=5,
                        null=True,
                        verbose_name="confidence score",
                    ),
                ),
                (
                    "summary",
                    models.TextField(
                        blank=True,
                        default="",
                        help_text="Free-text synopsis generated by the AI (used for summary type).",
                        verbose_name="summary",
                    ),
                ),
                (
                    "tags",
                    models.JSONField(
                        blank=True,
                        default=list,
                        help_text="Auto-generated tag list from AI classification.",
                        verbose_name="tags",
                    ),
                ),
                (
                    "processing_metadata",
                    models.JSONField(blank=True, default=dict, verbose_name="processing metadata"),
                ),
                (
                    "created_by",
                    models.ForeignKey(
                        blank=True,
                        editable=False,
                        null=True,
                        on_delete=django.db.models.deletion.SET_NULL,
                        related_name="+",
                        to=settings.AUTH_USER_MODEL,
                        verbose_name="created by",
                    ),
                ),
                (
                    "deleted_by",
                    models.ForeignKey(
                        blank=True,
                        null=True,
                        on_delete=django.db.models.deletion.SET_NULL,
                        related_name="+",
                        to=settings.AUTH_USER_MODEL,
                        verbose_name="deleted by",
                    ),
                ),
                (
                    "document_version",
                    models.ForeignKey(
                        on_delete=django.db.models.deletion.CASCADE,
                        related_name="ai_classifications",
                        to="dms.documentversion",
                        verbose_name="document version",
                    ),
                ),
                (
                    "organization_node",
                    models.ForeignKey(
                        on_delete=django.db.models.deletion.PROTECT,
                        related_name="+",
                        to="organizations.organizationnode",
                        verbose_name="organization node",
                    ),
                ),
                (
                    "tenant",
                    models.ForeignKey(
                        on_delete=django.db.models.deletion.CASCADE,
                        related_name="+",
                        to="tenants.tenant",
                        verbose_name="tenant",
                    ),
                ),
                (
                    "updated_by",
                    models.ForeignKey(
                        blank=True,
                        editable=False,
                        null=True,
                        on_delete=django.db.models.deletion.SET_NULL,
                        related_name="+",
                        to=settings.AUTH_USER_MODEL,
                        verbose_name="updated by",
                    ),
                ),
            ],
            options={
                "verbose_name": "AI classification",
                "verbose_name_plural": "AI classifications",
                "ordering": ("-created_at",),
                "indexes": [
                    models.Index(
                        fields=["document_version_id", "classification_type"],
                        name="dms_ai_version_type_idx",
                    ),
                    models.Index(fields=["tenant_id", "provider"], name="dms_ai_tenant_prov_idx"),
                ],
            },
        ),
        migrations.AddIndex(
            model_name="ocrresult",
            index=models.Index(
                fields=["document_version_id", "status"], name="dms_ocr_version_status_idx"
            ),
        ),
        migrations.AddIndex(
            model_name="ocrresult",
            index=models.Index(
                fields=["tenant_id", "provider"], name="dms_ocr_tenant_provider_idx"
            ),
        ),
        migrations.AddIndex(
            model_name="extractedentity",
            index=models.Index(
                fields=["document_version_id", "entity_type"], name="dms_entity_version_type_idx"
            ),
        ),
        migrations.AddIndex(
            model_name="extractedentity",
            index=models.Index(
                fields=["tenant_id", "entity_type"], name="dms_entity_tenant_type_idx"
            ),
        ),
    ]
