geo/server/internal/tenant/repository/article_repo.go

package repository

import (
	"context"
	"strings"

	"github.com/geo-platform/tenant-api/internal/shared/contentstats"
	"github.com/geo-platform/tenant-api/internal/tenant/repository/generated"
	"github.com/jackc/pgx/v5"
)

type CreateArticleInput struct {
	TenantID    int64
	SourceType  string
	TemplateID  *int64
	KolPromptID *int64
}

type CreateArticleVersionInput struct {
	ArticleID       int64
	VersionNo       int
	Title           string
	HTMLContent     string
	MarkdownContent string
	WordCount       int
	SourceLabel     string
}

type ArticleRepository interface {
	CreateArticle(ctx context.Context, input CreateArticleInput) (int64, error)
	CreateArticleVersion(ctx context.Context, input CreateArticleVersionInput) (int64, error)
	UpdateArticleCurrentVersion(ctx context.Context, articleID, tenantID, versionID int64) error
	UpdateArticleGenerateStatus(ctx context.Context, articleID, tenantID int64, status string) error
}

type articleRepository struct {
	db generated.DBTX
	q  generated.Querier
}

func NewArticleRepository(db generated.DBTX) ArticleRepository {
	return &articleRepository{db: db, q: newQuerier(db)}
}

func (r *articleRepository) CreateArticle(ctx context.Context, input CreateArticleInput) (int64, error) {
	return r.q.CreateArticle(ctx, generated.CreateArticleParams{
		TenantID:    input.TenantID,
		SourceType:  input.SourceType,
		TemplateID:  pgInt8(input.TemplateID),
		KolPromptID: pgInt8(input.KolPromptID),
	})
}

func NextArticleVersionNo(ctx context.Context, db generated.DBTX, articleID int64) (int, error) {
	var nextVersionNo int
	err := db.QueryRow(ctx, `
		SELECT COALESCE(MAX(version_no), 0) + 1
		FROM article_versions
		WHERE article_id = $1
	`, articleID).Scan(&nextVersionNo)
	return nextVersionNo, err
}

func (r *articleRepository) CreateArticleVersion(ctx context.Context, input CreateArticleVersionInput) (int64, error) {
	if input.VersionNo <= 0 {
		nextVersionNo, err := NextArticleVersionNo(ctx, r.db, input.ArticleID)
		if err != nil {
			return 0, err
		}
		input.VersionNo = nextVersionNo
	}

	previous, err := loadLatestArticleVersionContent(ctx, r.db, input.ArticleID)
	if err != nil {
		return 0, err
	}

	storedHTML, storedMarkdown := buildArticleVersionDiffStorage(*previous, ArticleVersionContent{
		HTMLContent:     stringPtr(input.HTMLContent),
		MarkdownContent: stringPtr(input.MarkdownContent),
	})

	var versionID int64
	err = r.db.QueryRow(ctx, `
		INSERT INTO article_versions (
			article_id,
			version_no,
			title,
			html_content,
			markdown_content,
			word_count,
			source_label,
			plaintext_snapshot
		)
		VALUES ($1, $2, $3, $4, $5, $6, $7, $8)
		RETURNING id
	`,
		input.ArticleID,
		input.VersionNo,
		pgText(&input.Title),
		pgText(storedHTML),
		pgText(storedMarkdown),
		input.WordCount,
		pgText(&input.SourceLabel),
		strings.TrimSpace(input.Title+"\n"+ExtractArticlePlaintext(input.MarkdownContent)),
	).Scan(&versionID)
	if err != nil {
		return 0, err
	}
	return versionID, nil
}

func ExtractArticlePlaintext(markdown string) string {
	markdown = strings.TrimSpace(markdown)
	if markdown == "" {
		return ""
	}

	var builder strings.Builder
	builder.Grow(len(markdown))
	inFence := false
	lines := strings.Split(markdown, "\n")
	for _, line := range lines {
		trimmed := strings.TrimSpace(line)
		if strings.HasPrefix(trimmed, "```") || strings.HasPrefix(trimmed, "~~~") {
			inFence = !inFence
			continue
		}
		if inFence {
			builder.WriteString(trimmed)
			builder.WriteByte('\n')
			continue
		}
		line = stripMarkdownLine(line)
		if strings.TrimSpace(line) == "" {
			continue
		}
		builder.WriteString(line)
		builder.WriteByte('\n')
	}
	return strings.TrimSpace(builder.String())
}

func stripMarkdownLine(line string) string {
	replacer := strings.NewReplacer(
		"#", " ",
		"*", " ",
		"_", " ",
		"`", " ",
		">", " ",
		"[", " ",
		"]", " ",
		"(", " ",
		")", " ",
		"!", " ",
		"|", " ",
	)
	line = replacer.Replace(line)
	line = strings.TrimLeft(line, "-+0123456789. \t")
	return strings.Join(strings.Fields(line), " ")
}

func CountArticlePlaintextWords(markdown string) int {
	return contentstats.CountWords(ExtractArticlePlaintext(markdown))
}

func (r *articleRepository) UpdateArticleCurrentVersion(ctx context.Context, articleID, tenantID, versionID int64) error {
	tag, err := r.db.Exec(ctx, `
		UPDATE articles
		SET current_version_id = $1,
		    updated_at = NOW()
		WHERE id = $2
		  AND tenant_id = $3
		  AND deleted_at IS NULL
	`, pgInt8(&versionID), articleID, tenantID)
	if err != nil {
		return err
	}
	if tag.RowsAffected() == 0 {
		return pgx.ErrNoRows
	}
	return nil
}

func (r *articleRepository) UpdateArticleGenerateStatus(ctx context.Context, articleID, tenantID int64, status string) error {
	tag, err := r.db.Exec(ctx, `
		UPDATE articles
		SET generate_status = $1,
		    updated_at = NOW()
		WHERE id = $2
		  AND tenant_id = $3
		  AND deleted_at IS NULL
	`, status, articleID, tenantID)
	if err != nil {
		return err
	}
	if tag.RowsAffected() == 0 {
		return pgx.ErrNoRows
	}
	return nil
}