diple/cache.go

package main

import (
	"context"
	"crypto/sha256"
	"encoding/hex"
	"encoding/json"
	"errors"
	"fmt"
	"os"
	"path/filepath"
	"sort"
	"time"
)

type cacheEnvelope[T any] struct {
	Version     int       `json:"version,omitempty"`
	SavedAt     time.Time `json:"saved_at"`
	ContentHash string    `json:"content_hash,omitempty"`
	Value       T         `json:"value"`
}

type CachedGitHubService struct {
	remote     GitHubService
	dir        string
	maxAge     time.Duration
	maxEntries int
	health     healthTracker
}

type cachedSnapshotService interface {
	CachedPullRequests(string, string, int, bool) ([]PullRequest, error)
	CachedPullRequest(string, string, int) (PRDetails, error)
}

type liveGitHubService interface {
	LivePullRequests(context.Context, string, string, int, bool) ([]PullRequest, error)
	LivePullRequest(context.Context, string, string, int) (PRDetails, error)
}

const cacheSchemaVersion = 1

func NewCachedGitHubService(
	remote GitHubService, dir string, maxAge time.Duration, configuredMaxEntries ...int,
) *CachedGitHubService {
	maxEntries := 200
	if len(configuredMaxEntries) > 0 && configuredMaxEntries[0] > 0 {
		maxEntries = configuredMaxEntries[0]
	}
	return &CachedGitHubService{
		remote: remote, dir: dir, maxAge: maxAge, maxEntries: maxEntries,
	}
}

func (c *CachedGitHubService) CachedPullRequests(
	owner, repo string, limit int, showAll bool,
) ([]PullRequest, error) {
	var cached cacheEnvelope[[]PullRequest]
	savedAt, err := c.read(c.pullRequestsKey(owner, repo, limit, showAll), &cached)
	if err != nil {
		return nil, err
	}
	cached.SavedAt = savedAt
	for i := range cached.Value {
		cached.Value[i].FromCache, cached.Value[i].CachedAt = true, cached.SavedAt
	}
	return cached.Value, nil
}

func (c *CachedGitHubService) CachedPullRequest(owner, repo string, number int) (PRDetails, error) {
	var cached cacheEnvelope[PRDetails]
	savedAt, err := c.read(c.pullRequestKey(owner, repo, number), &cached)
	if err != nil {
		return PRDetails{}, err
	}
	cached.SavedAt = savedAt
	cached.Value.FromCache, cached.Value.CachedAt = true, cached.SavedAt
	return cached.Value, nil
}

func (c *CachedGitHubService) ListPullRequests(
	ctx context.Context, owner, repo string, limit int, showAll bool,
) ([]PullRequest, error) {
	prs, err := c.LivePullRequests(ctx, owner, repo, limit, showAll)
	if err == nil {
		return prs, nil
	}
	cached, cacheErr := c.CachedPullRequests(owner, repo, limit, showAll)
	if cacheErr != nil {
		return nil, fmt.Errorf("%w (cache unavailable: %v)", err, cacheErr)
	}
	return cached, nil
}

func (c *CachedGitHubService) GetPullRequest(
	ctx context.Context, owner, repo string, number int,
) (PRDetails, error) {
	details, err := c.LivePullRequest(ctx, owner, repo, number)
	if err == nil {
		return details, nil
	}
	cached, cacheErr := c.CachedPullRequest(owner, repo, number)
	if cacheErr != nil {
		return PRDetails{}, fmt.Errorf("%w (cache unavailable: %v)", err, cacheErr)
	}
	return cached, nil
}

func (c *CachedGitHubService) LivePullRequests(
	ctx context.Context, owner, repo string, limit int, showAll bool,
) ([]PullRequest, error) {
	prs, err := c.remote.ListPullRequests(ctx, owner, repo, limit, showAll)
	if err != nil {
		return nil, err
	}
	for i := range prs {
		prs[i].FromCache, prs[i].CachedAt = false, time.Time{}
	}
	_ = c.write(c.pullRequestsKey(owner, repo, limit, showAll), prs)
	return prs, nil
}

func (c *CachedGitHubService) LivePullRequest(
	ctx context.Context, owner, repo string, number int,
) (PRDetails, error) {
	details, err := c.remote.GetPullRequest(ctx, owner, repo, number)
	if err != nil {
		return PRDetails{}, err
	}
	details.FromCache, details.CachedAt = false, time.Time{}
	_ = c.write(c.pullRequestKey(owner, repo, number), details)
	return details, nil
}

func (c *CachedGitHubService) SetThreadResolved(
	ctx context.Context, threadID string, resolved bool,
) (ReviewThread, error) {
	writer, ok := c.remote.(GitHubWriteService)
	if !ok {
		return ReviewThread{}, errors.New("GitHub service does not support write actions")
	}
	return writer.SetThreadResolved(ctx, threadID, resolved)
}

func (c *CachedGitHubService) ReplyToThread(
	ctx context.Context, threadID, body string,
) (ReviewComment, error) {
	writer, ok := c.remote.(GitHubWriteService)
	if !ok {
		return ReviewComment{}, errors.New("GitHub service does not support write actions")
	}
	return writer.ReplyToThread(ctx, threadID, body)
}

func (c *CachedGitHubService) UpdatePullRequest(
	ctx context.Context,
	pullRequestID string,
	update PullRequestMetadata,
) (PullRequestMetadata, error) {
	writer, ok := c.remote.(GitHubPullRequestWriteService)
	if !ok {
		return PullRequestMetadata{}, errors.New("GitHub service does not support pull request updates")
	}
	return writer.UpdatePullRequest(ctx, pullRequestID, update)
}

func (c *CachedGitHubService) ListBranches(
	ctx context.Context, owner, repo string,
) ([]RepositoryBranch, error) {
	service, ok := c.remote.(GitHubBranchService)
	if !ok {
		return nil, errors.New("GitHub service does not support listing branches")
	}
	branches, err := service.ListBranches(ctx, owner, repo)
	if err == nil {
		_ = c.write(c.branchesKey(owner, repo), branches)
		return branches, nil
	}
	var cached cacheEnvelope[[]RepositoryBranch]
	if _, cacheErr := c.read(c.branchesKey(owner, repo), &cached); cacheErr == nil {
		c.health.set(HealthComponent{
			Name: "branch cache", Level: healthWarning,
			Summary: "using cached branches", Detail: err.Error(), UpdatedAt: time.Now(),
		})
		return cached.Value, nil
	}
	return nil, err
}

func (c *CachedGitHubService) EnrichPullRequest(
	ctx context.Context, details PRDetails,
) PRDetailsEnrichment {
	service, ok := c.remote.(GitHubEnrichmentService)
	if !ok {
		return PRDetailsEnrichment{
			Owner: details.Owner, Repository: details.Repository, Number: details.Number,
			HeadOID: details.HeadOID,
			Issues: []DataIssue{{
				Component: "PR enrichment",
				Message:   "GitHub service does not support secondary PR data",
			}},
		}
	}
	return service.EnrichPullRequest(ctx, details)
}

func (c *CachedGitHubService) pullRequestsKey(owner, repo string, limit int, showAll bool) string {
	return fmt.Sprintf("prs:%s/%s:%d:%t", owner, repo, limit, showAll)
}

func (c *CachedGitHubService) pullRequestKey(owner, repo string, number int) string {
	return fmt.Sprintf("pr:%s/%s:%d", owner, repo, number)
}

func (c *CachedGitHubService) branchesKey(owner, repo string) string {
	return fmt.Sprintf("branches:%s/%s", owner, repo)
}

func (c *CachedGitHubService) file(key string) string {
	sum := sha256.Sum256([]byte(key))
	return filepath.Join(c.dir, hex.EncodeToString(sum[:])+".json")
}

func (c *CachedGitHubService) write(key string, value any) (resultErr error) {
	defer func() {
		component := HealthComponent{
			Name: "disk cache", Level: healthOK, Summary: "cache write succeeded",
			Detail: c.dir, UpdatedAt: time.Now(),
		}
		if resultErr != nil {
			component.Level = healthWarning
			component.Summary = resultErr.Error()
		}
		c.health.set(component)
	}()
	if err := os.MkdirAll(c.dir, 0o700); err != nil {
		return err
	}
	valueData, err := json.Marshal(value)
	if err != nil {
		return err
	}
	sum := sha256.Sum256(valueData)
	contentHash := hex.EncodeToString(sum[:])
	target := c.file(key)
	if existing, err := os.ReadFile(target); err == nil {
		var metadata struct {
			ContentHash string `json:"content_hash"`
		}
		if json.Unmarshal(existing, &metadata) == nil && metadata.ContentHash == contentHash {
			if info, statErr := os.Stat(target); statErr == nil &&
				time.Since(info.ModTime()) >= c.cacheTouchInterval() {
				now := time.Now()
				_ = os.Chtimes(target, now, now)
			}
			return nil
		}
	}
	data, err := json.Marshal(cacheEnvelope[any]{
		Version: cacheSchemaVersion, SavedAt: time.Now(),
		ContentHash: contentHash, Value: value,
	})
	if err != nil {
		return err
	}
	temp, err := os.CreateTemp(c.dir, ".cache-*")
	if err != nil {
		return err
	}
	name := temp.Name()
	defer os.Remove(name)
	if err := temp.Chmod(0o600); err != nil {
		temp.Close()
		return err
	}
	if _, err := temp.Write(data); err != nil {
		temp.Close()
		return err
	}
	if err := temp.Close(); err != nil {
		return err
	}
	if err := os.Rename(name, target); err != nil {
		return err
	}
	return c.prune()
}

func (c *CachedGitHubService) prune() error {
	if c.maxEntries <= 0 {
		return nil
	}
	entries, err := os.ReadDir(c.dir)
	if err != nil {
		return err
	}
	type cacheFile struct {
		path    string
		modTime time.Time
	}
	var files []cacheFile
	for _, entry := range entries {
		if entry.IsDir() || filepath.Ext(entry.Name()) != ".json" {
			continue
		}
		info, err := entry.Info()
		if err != nil {
			continue
		}
		files = append(files, cacheFile{
			path: filepath.Join(c.dir, entry.Name()), modTime: info.ModTime(),
		})
	}
	sort.Slice(files, func(i, j int) bool { return files[i].modTime.Before(files[j].modTime) })
	for len(files) > c.maxEntries {
		if err := os.Remove(files[0].path); err != nil && !errors.Is(err, os.ErrNotExist) {
			return err
		}
		files = files[1:]
	}
	return nil
}

func (c *CachedGitHubService) cacheTouchInterval() time.Duration {
	interval := 24 * time.Hour
	if c.maxAge > 0 && c.maxAge/2 < interval {
		interval = c.maxAge / 2
	}
	return interval
}

func (c *CachedGitHubService) read(key string, target any) (
	saved time.Time, resultErr error,
) {
	defer func() {
		component := HealthComponent{
			Name: "disk cache", Level: healthOK, Summary: "cache read succeeded",
			Detail: c.dir, UpdatedAt: time.Now(),
		}
		if resultErr != nil {
			component.Level = healthWarning
			component.Summary = resultErr.Error()
		}
		c.health.set(component)
	}()
	path := c.file(key)
	data, err := os.ReadFile(path)
	if err != nil {
		return time.Time{}, err
	}
	if err := json.Unmarshal(data, target); err != nil {
		return time.Time{}, err
	}
	var metadata struct {
		Version int       `json:"version"`
		SavedAt time.Time `json:"saved_at"`
	}
	if err := json.Unmarshal(data, &metadata); err != nil {
		return time.Time{}, err
	}
	if metadata.Version != 0 && metadata.Version != cacheSchemaVersion {
		return time.Time{}, fmt.Errorf(
			"unsupported cache schema version %d", metadata.Version,
		)
	}
	savedAt := metadata.SavedAt
	if info, statErr := os.Stat(path); statErr == nil && info.ModTime().After(savedAt) {
		savedAt = info.ModTime()
	}
	if c.maxAge > 0 && time.Since(savedAt) > c.maxAge {
		return time.Time{}, errors.New("cached data expired")
	}
	return savedAt, nil
}

func (c *CachedGitHubService) HealthReport() []HealthComponent {
	components := c.health.report()
	if provider, ok := c.remote.(healthProvider); ok {
		components = append(components, provider.HealthReport()...)
	}
	return components
}

func (c *CachedGitHubService) RateLimit() RateLimitSnapshot {
	if provider, ok := c.remote.(healthProvider); ok {
		return provider.RateLimit()
	}
	return RateLimitSnapshot{}
}

type readStateStore struct {
	path    string
	Data    map[string]readPRState `json:"pull_requests"`
	loadErr error
}

const readStateSchemaVersion = 1

type readStateEnvelope struct {
	Version      int                    `json:"version"`
	PullRequests map[string]readPRState `json:"pull_requests"`
}

type readPRState struct {
	Initialized bool            `json:"initialized"`
	Threads     map[string]bool `json:"threads"`
	Comments    map[string]bool `json:"comments"`
}

func loadReadState(path string) *readStateStore {
	store := &readStateStore{path: path, Data: make(map[string]readPRState)}
	data, err := os.ReadFile(path)
	if err == nil {
		var envelope readStateEnvelope
		if json.Unmarshal(data, &envelope) == nil &&
			envelope.Version == readStateSchemaVersion &&
			envelope.PullRequests != nil {
			store.Data = envelope.PullRequests
		} else {
			// Backward-compatible migration from the original unversioned map.
			if migrationErr := json.Unmarshal(data, &store.Data); migrationErr != nil {
				store.Data = make(map[string]readPRState)
				store.loadErr = fmt.Errorf("read state is corrupt: %w", migrationErr)
			}
		}
	} else if !errors.Is(err, os.ErrNotExist) {
		store.loadErr = err
	}
	return store
}

func (s *readStateStore) save() error {
	if s == nil || s.path == "" {
		return nil
	}
	if err := os.MkdirAll(filepath.Dir(s.path), 0o700); err != nil {
		return err
	}
	return atomicWriteJSON(s.path, readStateEnvelope{
		Version: readStateSchemaVersion, PullRequests: s.Data,
	}, 0o600)
}