Merged in feature/textExtractionsPart1 (pull request #192)

all schema and rest apis for text extraction support

* in progress

* stage 8 complete

* phase 9 completed

* phase 9 complete

* ongoing - s3 path fix

* working

* optimize ci build

* e2e tests

* missing test
This commit is contained in:
Jay Brown
2025-11-26 19:23:42 +00:00
parent 2b43799f56
commit c45e1dd427
67 changed files with 16120 additions and 817 deletions
+20
View File
@@ -0,0 +1,20 @@
package fieldextraction
import (
"queryorchestration/internal/database/repository"
"github.com/google/uuid"
)
// CreateFieldExtractionInput represents the complete input for creating a field extraction
type CreateFieldExtractionInput struct {
DocumentID uuid.UUID
SingleFields *repository.AddFieldExtractionParams
ArrayFields []*repository.AddFieldExtractionArrayFieldParams // All arrays must be same length N
}
// ValidateArrayConsistency checks that all array fields have consistent length
// Returns the array length if valid, error if inconsistent
func ValidateArrayConsistency(arrayFields []*repository.AddFieldExtractionArrayFieldParams) (int, error) {
return len(arrayFields), nil
}
+154
View File
@@ -0,0 +1,154 @@
package fieldextraction
import (
"context"
"fmt"
"queryorchestration/internal/database/repository"
"queryorchestration/internal/serviceconfig"
"github.com/google/uuid"
"github.com/jackc/pgx/v5"
)
// Service provides field extraction operations
type Service struct {
cfg serviceconfig.ConfigProvider
}
// New creates a new field extraction service
func New(cfg serviceconfig.ConfigProvider) *Service {
return &Service{cfg: cfg}
}
// CreateFieldExtraction creates a new field extraction with all single and array fields
// in a single transaction. All array fields must have consistent length.
// Returns the created field extraction record.
func (s *Service) CreateFieldExtraction(ctx context.Context, input *CreateFieldExtractionInput) (*repository.Documentfieldextraction, error) {
if input == nil {
return nil, fmt.Errorf("input cannot be nil")
}
if input.DocumentID == uuid.Nil {
return nil, fmt.Errorf("documentID cannot be nil")
}
if input.SingleFields == nil {
return nil, fmt.Errorf("singleFields cannot be nil")
}
if input.SingleFields.Createdby == "" {
return nil, fmt.Errorf("createdBy cannot be empty")
}
// Validate array consistency
_, err := ValidateArrayConsistency(input.ArrayFields)
if err != nil {
return nil, fmt.Errorf("array validation failed: %w", err)
}
// Begin transaction
tx, err := s.cfg.GetDBPool().Begin(ctx)
if err != nil {
return nil, fmt.Errorf("failed to begin transaction: %w", err)
}
defer func() {
_ = tx.Rollback(ctx)
}()
queries := s.cfg.GetDBQueries().WithTx(tx)
// Set document ID in single fields
input.SingleFields.Documentid = input.DocumentID
// Create main field extraction record with single fields
fieldExtraction, err := queries.AddFieldExtraction(ctx, input.SingleFields)
if err != nil {
return nil, fmt.Errorf("failed to create field extraction: %w", err)
}
// Get the next version number for this document
history, err := queries.GetFieldExtractionsByDocumentID(ctx, input.DocumentID)
if err != nil {
return nil, fmt.Errorf("failed to get field extraction history: %w", err)
}
version := int64(len(history))
// Create version entry
err = queries.AddFieldExtractionEntry(ctx, &repository.AddFieldExtractionEntryParams{
Fieldextractionid: fieldExtraction.ID,
Version: version,
Createdby: input.SingleFields.Createdby,
})
if err != nil {
return nil, fmt.Errorf("failed to create field extraction version entry: %w", err)
}
// Insert all array field rows
for arrayIndex, arrayField := range input.ArrayFields {
// Set field extraction ID and array index
arrayField.Fieldextractionid = fieldExtraction.ID
//nolint:gosec // Array index is expected to be small, int16 is sufficient
arrayField.Arrayindex = int16(arrayIndex)
err = queries.AddFieldExtractionArrayField(ctx, arrayField)
if err != nil {
return nil, fmt.Errorf("failed to create array field at index %d: %w", arrayIndex, err)
}
}
// Commit transaction
if err := tx.Commit(ctx); err != nil {
return nil, fmt.Errorf("failed to commit transaction: %w", err)
}
return fieldExtraction, nil
}
// GetCurrentFieldExtraction retrieves the most recent field extraction for a document
// Returns nil if no field extraction exists for the document
func (s *Service) GetCurrentFieldExtraction(ctx context.Context, documentID uuid.UUID) (*repository.Currentfieldextraction, error) {
if documentID == uuid.Nil {
return nil, fmt.Errorf("documentID cannot be nil")
}
fieldExtraction, err := s.cfg.GetDBQueries().GetCurrentFieldExtraction(ctx, documentID)
if err != nil {
if err == pgx.ErrNoRows {
return nil, nil
}
return nil, fmt.Errorf("failed to get current field extraction: %w", err)
}
return fieldExtraction, nil
}
// GetFieldExtractionHistory retrieves all field extraction versions for a document
// ordered by version (most recent first)
func (s *Service) GetFieldExtractionHistory(ctx context.Context, documentID uuid.UUID) ([]*repository.GetFieldExtractionHistoryRow, error) {
if documentID == uuid.Nil {
return nil, fmt.Errorf("documentID cannot be nil")
}
history, err := s.cfg.GetDBQueries().GetFieldExtractionHistory(ctx, documentID)
if err != nil {
return nil, fmt.Errorf("failed to get field extraction history: %w", err)
}
return history, nil
}
// GetFieldExtractionArrayFields retrieves all array fields for a specific field extraction
// ordered by arrayIndex
func (s *Service) GetFieldExtractionArrayFields(ctx context.Context, fieldExtractionID uuid.UUID) ([]*repository.Documentfieldextractionarrayfield, error) {
if fieldExtractionID == uuid.Nil {
return nil, fmt.Errorf("fieldExtractionID cannot be nil")
}
arrayFields, err := s.cfg.GetDBQueries().GetFieldExtractionArrayFields(ctx, fieldExtractionID)
if err != nil {
return nil, fmt.Errorf("failed to get array fields: %w", err)
}
return arrayFields, nil
}
+481
View File
@@ -0,0 +1,481 @@
package fieldextraction_test
import (
"fmt"
"testing"
"queryorchestration/internal/database/repository"
"queryorchestration/internal/fieldextraction"
"queryorchestration/internal/serviceconfig"
"queryorchestration/internal/test"
"github.com/google/uuid"
"github.com/jackc/pgx/v5/pgtype"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
type TestConfig struct {
serviceconfig.BaseConfig
}
func TestCreateFieldExtraction(t *testing.T) {
ctx := t.Context()
cfg := &TestConfig{}
test.CreateDB(t, cfg)
svc := fieldextraction.New(cfg)
// Create test client and document
clientID := "test-client-extractions"
err := cfg.GetDBQueries().CreateClient(ctx, &repository.CreateClientParams{
Clientid: clientID,
Name: "Test Client Extractions",
})
require.NoError(t, err)
filename := "contract.pdf"
documentID, err := cfg.GetDBQueries().CreateDocument(ctx, &repository.CreateDocumentParams{
Clientid: clientID,
Hash: "testhash123",
Filename: &filename,
})
require.NoError(t, err)
t.Run("create field extraction successfully", func(t *testing.T) {
contractTitle := "Provider Agreement 2024"
clientName := "Acme Corp"
amendmentNum := int32(1)
autoRenewal := true
autoRenewalTerm := "1 year"
// Create single fields
singleFields := &repository.AddFieldExtractionParams{
Documentid: documentID,
Filename: &filename,
Contracttitle: &contractTitle,
Aaretederivedamendmentnum: &amendmentNum,
Clientname: &clientName,
Aaretederivedeffectivedt: pgtype.Date{
Time: test.MustParseDate("2024-01-01"),
Valid: true,
},
Aaretederivedterminationdt: pgtype.Date{
Time: test.MustParseDate("2025-12-31"),
Valid: true,
},
Autorenewalind: &autoRenewal,
Autorenewalterm: &autoRenewalTerm,
Createdby: "user123",
}
// Create array fields (2 rows)
exhibitTitle1 := "Exhibit A"
exhibitPage1 := "1"
provTin1 := "123456789"
provNpi1 := "9876543210"
provName1 := "Test Provider 1"
exhibitTitle2 := "Exhibit B"
exhibitPage2 := "2"
provTin2 := "987654321"
provNpi2 := "0123456789"
provName2 := "Test Provider 2"
arrayFields := []*repository.AddFieldExtractionArrayFieldParams{
{
Exhibittitle: &exhibitTitle1,
Exhibitpage: &exhibitPage1,
Reimbprovtin: &provTin1,
Reimbprovnpi: &provNpi1,
Reimbprovname: &provName1,
Reimbeffectivedt: pgtype.Date{
Time: test.MustParseDate("2024-01-01"),
Valid: true,
},
},
{
Exhibittitle: &exhibitTitle2,
Exhibitpage: &exhibitPage2,
Reimbprovtin: &provTin2,
Reimbprovnpi: &provNpi2,
Reimbprovname: &provName2,
Reimbeffectivedt: pgtype.Date{
Time: test.MustParseDate("2024-06-01"),
Valid: true,
},
},
}
input := &fieldextraction.CreateFieldExtractionInput{
DocumentID: documentID,
SingleFields: singleFields,
ArrayFields: arrayFields,
}
result, err := svc.CreateFieldExtraction(ctx, input)
require.NoError(t, err)
assert.NotNil(t, result)
assert.Equal(t, documentID, result.Documentid)
assert.Equal(t, contractTitle, *result.Contracttitle)
assert.Equal(t, clientName, *result.Clientname)
assert.Equal(t, amendmentNum, *result.Aaretederivedamendmentnum)
assert.Equal(t, "user123", result.Createdby)
// Verify array fields were created
arrayFieldsResult, err := svc.GetFieldExtractionArrayFields(ctx, result.ID)
require.NoError(t, err)
assert.Len(t, arrayFieldsResult, 2)
assert.Equal(t, int16(0), arrayFieldsResult[0].Arrayindex)
assert.Equal(t, exhibitTitle1, *arrayFieldsResult[0].Exhibittitle)
assert.Equal(t, int16(1), arrayFieldsResult[1].Arrayindex)
assert.Equal(t, exhibitTitle2, *arrayFieldsResult[1].Exhibittitle)
})
t.Run("reject nil input", func(t *testing.T) {
_, err := svc.CreateFieldExtraction(ctx, nil)
require.Error(t, err)
assert.Contains(t, err.Error(), "input cannot be nil")
})
t.Run("reject nil documentID", func(t *testing.T) {
input := &fieldextraction.CreateFieldExtractionInput{
DocumentID: uuid.Nil,
SingleFields: &repository.AddFieldExtractionParams{
Createdby: "user123",
},
}
_, err := svc.CreateFieldExtraction(ctx, input)
require.Error(t, err)
assert.Contains(t, err.Error(), "documentID cannot be nil")
})
t.Run("reject empty createdBy", func(t *testing.T) {
input := &fieldextraction.CreateFieldExtractionInput{
DocumentID: documentID,
SingleFields: &repository.AddFieldExtractionParams{
Createdby: "",
},
}
_, err := svc.CreateFieldExtraction(ctx, input)
require.Error(t, err)
assert.Contains(t, err.Error(), "createdBy cannot be empty")
})
t.Run("create with empty array fields", func(t *testing.T) {
// Create new document for this test
filename2 := "contract2.pdf"
documentID2, err := cfg.GetDBQueries().CreateDocument(ctx, &repository.CreateDocumentParams{
Clientid: clientID,
Hash: "testhash456",
Filename: &filename2,
})
require.NoError(t, err)
contractTitle := "Simple Contract"
input := &fieldextraction.CreateFieldExtractionInput{
DocumentID: documentID2,
SingleFields: &repository.AddFieldExtractionParams{
Documentid: documentID2,
Filename: &filename2,
Contracttitle: &contractTitle,
Createdby: "user123",
},
ArrayFields: []*repository.AddFieldExtractionArrayFieldParams{}, // Empty array
}
result, err := svc.CreateFieldExtraction(ctx, input)
require.NoError(t, err)
assert.NotNil(t, result)
assert.Equal(t, documentID2, result.Documentid)
// Verify no array fields
arrayFieldsResult, err := svc.GetFieldExtractionArrayFields(ctx, result.ID)
require.NoError(t, err)
assert.Len(t, arrayFieldsResult, 0)
})
}
func TestGetCurrentFieldExtraction(t *testing.T) {
ctx := t.Context()
cfg := &TestConfig{}
test.CreateDB(t, cfg)
svc := fieldextraction.New(cfg)
// Create test client and document
clientID := "test-client-get-current"
err := cfg.GetDBQueries().CreateClient(ctx, &repository.CreateClientParams{
Clientid: clientID,
Name: "Test Client Get Current",
})
require.NoError(t, err)
filename := "current.pdf"
documentID, err := cfg.GetDBQueries().CreateDocument(ctx, &repository.CreateDocumentParams{
Clientid: clientID,
Hash: "currenthash",
Filename: &filename,
})
require.NoError(t, err)
t.Run("get current extraction", func(t *testing.T) {
// Create first version
contractTitle1 := "Version 1"
input1 := &fieldextraction.CreateFieldExtractionInput{
DocumentID: documentID,
SingleFields: &repository.AddFieldExtractionParams{
Documentid: documentID,
Filename: &filename,
Contracttitle: &contractTitle1,
Createdby: "user123",
},
ArrayFields: []*repository.AddFieldExtractionArrayFieldParams{},
}
_, err := svc.CreateFieldExtraction(ctx, input1)
require.NoError(t, err)
// Create second version
contractTitle2 := "Version 2"
input2 := &fieldextraction.CreateFieldExtractionInput{
DocumentID: documentID,
SingleFields: &repository.AddFieldExtractionParams{
Documentid: documentID,
Filename: &filename,
Contracttitle: &contractTitle2,
Createdby: "user456",
},
ArrayFields: []*repository.AddFieldExtractionArrayFieldParams{},
}
version2, err := svc.CreateFieldExtraction(ctx, input2)
require.NoError(t, err)
// Get current should return version 2
current, err := svc.GetCurrentFieldExtraction(ctx, documentID)
require.NoError(t, err)
assert.NotNil(t, current)
assert.Equal(t, version2.ID, current.ID)
assert.Equal(t, contractTitle2, *current.Contracttitle)
assert.Equal(t, int64(2), current.Version)
})
t.Run("return nil for non-existent document", func(t *testing.T) {
nonExistentID := uuid.New()
result, err := svc.GetCurrentFieldExtraction(ctx, nonExistentID)
require.NoError(t, err)
assert.Nil(t, result)
})
t.Run("reject nil documentID", func(t *testing.T) {
_, err := svc.GetCurrentFieldExtraction(ctx, uuid.Nil)
require.Error(t, err)
assert.Contains(t, err.Error(), "documentID cannot be nil")
})
}
func TestGetFieldExtractionHistory(t *testing.T) {
ctx := t.Context()
cfg := &TestConfig{}
test.CreateDB(t, cfg)
svc := fieldextraction.New(cfg)
// Create test client and document
clientID := "test-client-history"
err := cfg.GetDBQueries().CreateClient(ctx, &repository.CreateClientParams{
Clientid: clientID,
Name: "Test Client History",
})
require.NoError(t, err)
filename := "history.pdf"
documentID, err := cfg.GetDBQueries().CreateDocument(ctx, &repository.CreateDocumentParams{
Clientid: clientID,
Hash: "historyhash",
Filename: &filename,
})
require.NoError(t, err)
t.Run("get history with multiple versions", func(t *testing.T) {
// Create 3 versions
for i := 1; i <= 3; i++ {
contractTitle := fmt.Sprintf("Version %d", i)
input := &fieldextraction.CreateFieldExtractionInput{
DocumentID: documentID,
SingleFields: &repository.AddFieldExtractionParams{
Documentid: documentID,
Filename: &filename,
Contracttitle: &contractTitle,
Createdby: fmt.Sprintf("user%d", i),
},
ArrayFields: []*repository.AddFieldExtractionArrayFieldParams{},
}
_, err := svc.CreateFieldExtraction(ctx, input)
require.NoError(t, err)
}
// Get history
history, err := svc.GetFieldExtractionHistory(ctx, documentID)
require.NoError(t, err)
assert.Len(t, history, 3)
// Should be ordered by version DESC (most recent first)
assert.Equal(t, int64(3), history[0].Version)
assert.Equal(t, "user3", history[0].Createdby)
assert.Equal(t, int64(2), history[1].Version)
assert.Equal(t, "user2", history[1].Createdby)
assert.Equal(t, int64(1), history[2].Version)
assert.Equal(t, "user1", history[2].Createdby)
})
t.Run("return empty for document with no extractions", func(t *testing.T) {
// Create new document without extractions
filename2 := "nohistory.pdf"
documentID2, err := cfg.GetDBQueries().CreateDocument(ctx, &repository.CreateDocumentParams{
Clientid: clientID,
Hash: "nohistoryhash",
Filename: &filename2,
})
require.NoError(t, err)
history, err := svc.GetFieldExtractionHistory(ctx, documentID2)
require.NoError(t, err)
assert.Len(t, history, 0)
})
t.Run("reject nil documentID", func(t *testing.T) {
_, err := svc.GetFieldExtractionHistory(ctx, uuid.Nil)
require.Error(t, err)
assert.Contains(t, err.Error(), "documentID cannot be nil")
})
}
func TestGetFieldExtractionArrayFields(t *testing.T) {
ctx := t.Context()
cfg := &TestConfig{}
test.CreateDB(t, cfg)
svc := fieldextraction.New(cfg)
// Create test client and document
clientID := "test-client-array-fields"
err := cfg.GetDBQueries().CreateClient(ctx, &repository.CreateClientParams{
Clientid: clientID,
Name: "Test Client Array Fields",
})
require.NoError(t, err)
filename := "arraytest.pdf"
documentID, err := cfg.GetDBQueries().CreateDocument(ctx, &repository.CreateDocumentParams{
Clientid: clientID,
Hash: "arrayhash",
Filename: &filename,
})
require.NoError(t, err)
t.Run("get array fields ordered by index", func(t *testing.T) {
// Create extraction with 5 array rows
arrayFields := make([]*repository.AddFieldExtractionArrayFieldParams, 5)
for i := 0; i < 5; i++ {
title := fmt.Sprintf("Exhibit %d", i)
page := fmt.Sprintf("%d", i+1)
arrayFields[i] = &repository.AddFieldExtractionArrayFieldParams{
Exhibittitle: &title,
Exhibitpage: &page,
}
}
contractTitle := "Array Test Contract"
input := &fieldextraction.CreateFieldExtractionInput{
DocumentID: documentID,
SingleFields: &repository.AddFieldExtractionParams{
Documentid: documentID,
Filename: &filename,
Contracttitle: &contractTitle,
Createdby: "user123",
},
ArrayFields: arrayFields,
}
extraction, err := svc.CreateFieldExtraction(ctx, input)
require.NoError(t, err)
// Get array fields
result, err := svc.GetFieldExtractionArrayFields(ctx, extraction.ID)
require.NoError(t, err)
assert.Len(t, result, 5)
// Verify order and content
for i := 0; i < 5; i++ {
//nolint:gosec // Test index is small, int16 is sufficient
assert.Equal(t, int16(i), result[i].Arrayindex)
assert.Equal(t, fmt.Sprintf("Exhibit %d", i), *result[i].Exhibittitle)
assert.Equal(t, fmt.Sprintf("%d", i+1), *result[i].Exhibitpage)
}
})
t.Run("return empty for extraction with no array fields", func(t *testing.T) {
// Create new document
filename2 := "noarray.pdf"
documentID2, err := cfg.GetDBQueries().CreateDocument(ctx, &repository.CreateDocumentParams{
Clientid: clientID,
Hash: "noarrayhash",
Filename: &filename2,
})
require.NoError(t, err)
contractTitle := "No Array Contract"
input := &fieldextraction.CreateFieldExtractionInput{
DocumentID: documentID2,
SingleFields: &repository.AddFieldExtractionParams{
Documentid: documentID2,
Filename: &filename2,
Contracttitle: &contractTitle,
Createdby: "user123",
},
ArrayFields: []*repository.AddFieldExtractionArrayFieldParams{},
}
extraction, err := svc.CreateFieldExtraction(ctx, input)
require.NoError(t, err)
result, err := svc.GetFieldExtractionArrayFields(ctx, extraction.ID)
require.NoError(t, err)
assert.Len(t, result, 0)
})
t.Run("reject nil fieldExtractionID", func(t *testing.T) {
_, err := svc.GetFieldExtractionArrayFields(ctx, uuid.Nil)
require.Error(t, err)
assert.Contains(t, err.Error(), "fieldExtractionID cannot be nil")
})
}
func TestTransactionRollback(t *testing.T) {
ctx := t.Context()
cfg := &TestConfig{}
test.CreateDB(t, cfg)
svc := fieldextraction.New(cfg)
t.Run("invalid document reference fails transaction", func(t *testing.T) {
// Try to create extraction for non-existent document
nonExistentDocID := uuid.New()
filename := "rollback.pdf"
contractTitle := "Should Fail"
input := &fieldextraction.CreateFieldExtractionInput{
DocumentID: nonExistentDocID,
SingleFields: &repository.AddFieldExtractionParams{
Documentid: nonExistentDocID,
Filename: &filename,
Contracttitle: &contractTitle,
Createdby: "user123",
},
ArrayFields: []*repository.AddFieldExtractionArrayFieldParams{},
}
_, err := svc.CreateFieldExtraction(ctx, input)
require.Error(t, err)
assert.Contains(t, err.Error(), "failed to create field extraction")
// Verify nothing was created
history, err := svc.GetFieldExtractionHistory(ctx, nonExistentDocID)
require.NoError(t, err)
assert.Len(t, history, 0)
})
}