Files
llm-intelligence/scripts/report_event_coverage.go

296 lines
6.8 KiB
Go
Raw Normal View History

//go:build llm_script
package main
import (
"database/sql"
"encoding/json"
"fmt"
"os"
"path/filepath"
"sort"
"strings"
"time"
_ "github.com/lib/pq"
)
type CoverageDay struct {
Date string
OfficialRelease bool
PromoCampaign bool
NewModel bool
PriceChange bool
}
func (d CoverageDay) HasTrueEvent() bool {
return d.OfficialRelease || d.PromoCampaign || d.NewModel || d.PriceChange
}
func (d CoverageDay) EventLabels() []string {
var labels []string
if d.OfficialRelease {
labels = append(labels, "official_release")
}
if d.PromoCampaign {
labels = append(labels, "promo_campaign")
}
if d.NewModel {
labels = append(labels, "new_model")
}
if d.PriceChange {
labels = append(labels, "price_change")
}
return labels
}
type CoverageSummary struct {
TotalDays int
CoveredDays int
CoveragePct float64
}
type CoveragePromoDefinition struct {
Date string `json:"date"`
}
func main() {
if len(os.Args) != 3 {
fmt.Fprintln(os.Stderr, "用法: go run -tags llm_script scripts/report_event_coverage.go YYYY-MM-DD YYYY-MM-DD")
os.Exit(2)
}
startDate, endDate, err := parseCoverageArgs(os.Args[1], os.Args[2])
if err != nil {
fmt.Fprintln(os.Stderr, "参数错误:", err)
os.Exit(2)
}
dsn := os.Getenv("DATABASE_URL")
if dsn == "" {
dsn = "postgres://long@/llm_intelligence?host=/var/run/postgresql"
}
db, err := sql.Open("postgres", dsn)
if err != nil {
fmt.Fprintln(os.Stderr, "连接数据库失败:", err)
os.Exit(1)
}
defer db.Close()
days, err := collectCoverageDays(db, startDate, endDate)
if err != nil {
fmt.Fprintln(os.Stderr, "统计覆盖率失败:", err)
os.Exit(1)
}
if len(days) == 0 {
fmt.Fprintln(os.Stderr, "指定范围内没有已生成的日报记录")
os.Exit(1)
}
summary := summarizeCoverage(days)
fmt.Printf("V2 事件覆盖率报告 [%s, %s]\n", startDate, endDate)
for _, day := range days {
status := "MISS"
if day.HasTrueEvent() {
status = "PASS"
}
fmt.Printf("%s\t%s\t%s\n", day.Date, status, strings.Join(day.EventLabels(), ", "))
}
fmt.Printf("\n总日报天数: %d\n", summary.TotalDays)
fmt.Printf("命中真实变化事件天数: %d\n", summary.CoveredDays)
fmt.Printf("覆盖率: %.2f%%\n", summary.CoveragePct)
if summary.CoveragePct < 80 {
os.Exit(1)
}
}
func parseCoverageArgs(start, end string) (string, string, error) {
startDate, err := time.Parse("2006-01-02", start)
if err != nil {
return "", "", fmt.Errorf("开始日期格式不正确")
}
endDate, err := time.Parse("2006-01-02", end)
if err != nil {
return "", "", fmt.Errorf("结束日期格式不正确")
}
if endDate.Before(startDate) {
return "", "", fmt.Errorf("结束日期不能早于开始日期")
}
return startDate.Format("2006-01-02"), endDate.Format("2006-01-02"), nil
}
func collectCoverageDays(db *sql.DB, startDate, endDate string) ([]CoverageDay, error) {
reportDates, err := loadGeneratedReportDates(db, startDate, endDate)
if err != nil {
return nil, err
}
promoDates, err := loadCoveragePromoDates()
if err != nil {
return nil, err
}
var days []CoverageDay
for _, date := range reportDates {
officialRelease, err := hasOfficialRelease(db, date)
if err != nil {
return nil, err
}
newModel, err := hasNewModel(db, date)
if err != nil {
return nil, err
}
priceChange, err := hasPriceChange(db, date)
if err != nil {
return nil, err
}
days = append(days, CoverageDay{
Date: date,
OfficialRelease: officialRelease,
PromoCampaign: promoDates[date],
NewModel: newModel,
PriceChange: priceChange,
})
}
sort.Slice(days, func(i, j int) bool {
return days[i].Date < days[j].Date
})
return days, nil
}
func loadGeneratedReportDates(db *sql.DB, startDate, endDate string) ([]string, error) {
rows, err := db.Query(`
SELECT DISTINCT report_date::text
FROM daily_report
WHERE report_date BETWEEN $1::date AND $2::date
AND status = 'generated'
ORDER BY report_date
`, startDate, endDate)
if err != nil {
return nil, err
}
defer rows.Close()
var dates []string
for rows.Next() {
var date string
if err := rows.Scan(&date); err != nil {
return nil, err
}
dates = append(dates, date)
}
return dates, rows.Err()
}
func hasOfficialRelease(db *sql.DB, date string) (bool, error) {
var exists bool
err := db.QueryRow(`
WITH latest_prices AS (
SELECT
rp.model_id,
COALESCE(o.type, 'reseller') AS operator_type,
ROW_NUMBER() OVER (
PARTITION BY rp.model_id
ORDER BY rp.effective_date DESC NULLS LAST, rp.id DESC
) AS rn
FROM region_pricing rp
LEFT JOIN operator o ON rp.operator_id = o.id
)
SELECT EXISTS (
SELECT 1
FROM models m
LEFT JOIN latest_prices lp ON lp.model_id = m.id AND lp.rn = 1
WHERE m.deleted_at IS NULL
AND m.release_date = $1::date
AND COALESCE(m.source_url, '') <> ''
AND COALESCE(lp.operator_type, 'reseller') IN ('official', 'cloud')
)
`, date).Scan(&exists)
return exists, err
}
func hasNewModel(db *sql.DB, date string) (bool, error) {
var exists bool
err := db.QueryRow(`
SELECT EXISTS (
SELECT 1
FROM models
WHERE deleted_at IS NULL
AND DATE(created_at) = $1::date
)
`, date).Scan(&exists)
return exists, err
}
func hasPriceChange(db *sql.DB, date string) (bool, error) {
var exists bool
err := db.QueryRow(`
SELECT EXISTS (
SELECT 1
FROM pricing_history
WHERE DATE(changed_at) = $1::date
AND (
COALESCE(old_input_price, 0) <> COALESCE(new_input_price, 0)
OR COALESCE(old_output_price, 0) <> COALESCE(new_output_price, 0)
)
)
`, date).Scan(&exists)
return exists, err
}
func loadCoveragePromoDates() (map[string]bool, error) {
path, err := resolveCoveragePromoDataPath()
if err != nil {
if os.IsNotExist(err) {
return map[string]bool{}, nil
}
return nil, err
}
body, err := os.ReadFile(path)
if err != nil {
return nil, err
}
var definitions []CoveragePromoDefinition
if err := json.Unmarshal(body, &definitions); err != nil {
return nil, err
}
dates := make(map[string]bool, len(definitions))
for _, definition := range definitions {
if strings.TrimSpace(definition.Date) != "" {
dates[definition.Date] = true
}
}
return dates, nil
}
func resolveCoveragePromoDataPath() (string, error) {
candidates := []string{
filepath.Join("scripts", "testdata", "report_promo_campaigns.json"),
filepath.Join("testdata", "report_promo_campaigns.json"),
}
for _, candidate := range candidates {
if _, err := os.Stat(candidate); err == nil {
return candidate, nil
}
}
return "", os.ErrNotExist
}
func summarizeCoverage(days []CoverageDay) CoverageSummary {
summary := CoverageSummary{TotalDays: len(days)}
for _, day := range days {
if day.HasTrueEvent() {
summary.CoveredDays++
}
}
if summary.TotalDays > 0 {
summary.CoveragePct = float64(summary.CoveredDays) * 100 / float64(summary.TotalDays)
}
return summary
}