sow-tools/internal/topdata/parts_manifest.go

package topdata

import (
	"bytes"
	"encoding/json"
	"fmt"
	"io"
	"net/http"
	"net/url"
	"os"
	"path/filepath"
	"strings"
	"time"

	"gitea.westgate.pw/ShadowsOverWestgate/sow-tools/internal/project"
)

const (
	defaultSowAssetsRepoOwner   = "ShadowsOverWestgate"
	defaultSowAssetsRepoName    = "sow-assets"
	defaultGiteaBaseURL         = "https://gitea.westgate.pw"
	partsManifestReleaseTag     = "parts-manifest-current"
	partsManifestAssetName      = "sow-parts-manifest.json"
	partsManifestCacheFileName  = "sow-parts-manifest.json"
	partsManifestCacheMaxAge    = time.Hour
	partsManifestRequestTimeout = 30 * time.Second
)

type partsManifest struct {
	Repo        string           `json:"repo"`
	Ref         string           `json:"ref"`
	GeneratedAt string           `json:"generated_at"`
	Categories  map[string][]int `json:"categories"`
}

type giteaRepoSpec struct {
	BaseURL string
	Owner   string
	Repo    string
}

var partsManifestHTTPClient = &http.Client{Timeout: partsManifestRequestTimeout}

func resolveReleasedPartsManifest(p *project.Project, progress func(string)) (map[string]map[int]struct{}, error) {
	if progress == nil {
		progress = func(string) {}
	}
	cachePath := p.AutogenCachePath(partsManifestCacheFileName)
	if strings.TrimSpace(os.Getenv("SOW_PARTS_MANIFEST_REFRESH")) == "" {
		manifest, fresh, err := readFreshPartsManifestCache(cachePath, partsManifestCacheMaxAge)
		if err != nil {
			progress(fmt.Sprintf("Ignoring cached parts manifest: %v", err))
		} else if fresh {
			progress(fmt.Sprintf("Using cached released parts manifest from %s...", cachePath))
			return inventoryFromPartsManifest(manifest), nil
		}
	}
	spec, err := deriveSowAssetsRepoSpec(p.Root)
	if err != nil {
		return nil, err
	}
	manifestURL, err := resolvePartsManifestAssetURL(spec)
	if err != nil {
		return nil, err
	}
	progress(fmt.Sprintf("Fetching released parts manifest from %s...", manifestURL))
	manifest, err := fetchPartsManifest(manifestURL)
	if err != nil {
		return nil, err
	}
	if err := writePartsManifestCache(cachePath, manifest); err != nil {
		return nil, err
	}
	return inventoryFromPartsManifest(manifest), nil
}

func deriveSowAssetsRepoSpec(root string) (giteaRepoSpec, error) {
	serverOverride := strings.TrimSpace(os.Getenv("SOW_ASSETS_SERVER_URL"))
	repoOverride := strings.TrimSpace(os.Getenv("SOW_ASSETS_REPO"))
	remote, err := gitOutput(root, "remote", "get-url", "origin")
	spec := giteaRepoSpec{BaseURL: defaultGiteaBaseURL, Owner: defaultSowAssetsRepoOwner, Repo: defaultSowAssetsRepoName}
	if err == nil && strings.TrimSpace(remote) != "" {
		if derived, ok := parseRepoSpec(strings.TrimSpace(remote)); ok {
			derived.Repo = defaultSowAssetsRepoName
			spec = derived
		}
	}

	if serverOverride != "" {
		spec.BaseURL = strings.TrimRight(serverOverride, "/")
	}
	if repoOverride != "" {
		owner, repo, ok := parseOwnerRepo(repoOverride)
		if !ok {
			return giteaRepoSpec{}, fmt.Errorf("invalid SOW_ASSETS_REPO %q; expected owner/repo", repoOverride)
		}
		spec.Owner = owner
		spec.Repo = repo
	}
	return spec, nil
}

func parseOwnerRepo(raw string) (string, string, bool) {
	parts := strings.Split(strings.Trim(strings.TrimSuffix(raw, ".git"), "/"), "/")
	if len(parts) < 2 {
		return "", "", false
	}
	owner := strings.TrimSpace(parts[len(parts)-2])
	repo := strings.TrimSpace(parts[len(parts)-1])
	if owner == "" || repo == "" {
		return "", "", false
	}
	return owner, repo, true
}

func parseRepoSpec(raw string) (giteaRepoSpec, bool) {
	if strings.HasPrefix(raw, "ssh://") || strings.HasPrefix(raw, "http://") || strings.HasPrefix(raw, "https://") {
		u, err := url.Parse(raw)
		if err != nil || u.Host == "" {
			return giteaRepoSpec{}, false
		}
		parts := strings.Split(strings.Trim(strings.TrimSuffix(u.Path, ".git"), "/"), "/")
		if len(parts) < 2 {
			return giteaRepoSpec{}, false
		}
		base := (&url.URL{Scheme: "https", Host: normalizeGiteaHTTPHost(u.Hostname())}).String()
		if u.Scheme == "http" || u.Scheme == "https" {
			base = (&url.URL{Scheme: u.Scheme, Host: u.Host}).String()
		}
		return giteaRepoSpec{BaseURL: strings.TrimRight(base, "/"), Owner: parts[len(parts)-2], Repo: parts[len(parts)-1]}, true
	}
	if strings.HasPrefix(raw, "git@") {
		withoutUser := strings.TrimPrefix(raw, "git@")
		hostAndPath := strings.SplitN(withoutUser, ":", 2)
		if len(hostAndPath) != 2 {
			return giteaRepoSpec{}, false
		}
		parts := strings.Split(strings.Trim(strings.TrimSuffix(hostAndPath[1], ".git"), "/"), "/")
		if len(parts) < 2 {
			return giteaRepoSpec{}, false
		}
		base := (&url.URL{Scheme: "https", Host: normalizeGiteaHTTPHost(hostAndPath[0])}).String()
		return giteaRepoSpec{BaseURL: strings.TrimRight(base, "/"), Owner: parts[len(parts)-2], Repo: parts[len(parts)-1]}, true
	}
	return giteaRepoSpec{}, false
}

func normalizeGiteaHTTPHost(host string) string {
	host = strings.TrimSpace(host)
	if strings.HasPrefix(host, "git-ssh.") {
		return "gitea." + strings.TrimPrefix(host, "git-ssh.")
	}
	return host
}

func resolvePartsManifestAssetURL(spec giteaRepoSpec) (string, error) {
	apiURL := fmt.Sprintf("%s/api/v1/repos/%s/%s/releases/tags/%s", strings.TrimRight(spec.BaseURL, "/"), url.PathEscape(spec.Owner), url.PathEscape(spec.Repo), url.PathEscape(partsManifestReleaseTag))
	var release struct {
		Assets []struct {
			Name               string `json:"name"`
			BrowserDownloadURL string `json:"browser_download_url"`
			DownloadURL        string `json:"download_url"`
		} `json:"assets"`
	}
	if err := fetchJSON(apiURL, &release); err != nil {
		return "", fmt.Errorf("fetch parts manifest release metadata: %w", err)
	}
	for _, asset := range release.Assets {
		if asset.Name != partsManifestAssetName {
			continue
		}
		if asset.DownloadURL != "" {
			return asset.DownloadURL, nil
		}
		if asset.BrowserDownloadURL != "" {
			return asset.BrowserDownloadURL, nil
		}
	}
	return "", fmt.Errorf("parts manifest asset %s not found in release %s/%s:%s", partsManifestAssetName, spec.Owner, spec.Repo, partsManifestReleaseTag)
}

func fetchPartsManifest(manifestURL string) (*partsManifest, error) {
	var manifest partsManifest
	if err := fetchJSON(manifestURL, &manifest); err != nil {
		return nil, fmt.Errorf("download parts manifest: %w", err)
	}
	if len(manifest.Categories) == 0 {
		return nil, fmt.Errorf("download parts manifest: categories is empty")
	}
	return &manifest, nil
}

func readFreshPartsManifestCache(path string, maxAge time.Duration) (*partsManifest, bool, error) {
	info, err := os.Stat(path)
	if err != nil {
		if os.IsNotExist(err) {
			return nil, false, nil
		}
		return nil, false, err
	}
	if maxAge > 0 && time.Since(info.ModTime()) > maxAge {
		return nil, false, nil
	}
	raw, err := os.ReadFile(path)
	if err != nil {
		return nil, false, err
	}
	var manifest partsManifest
	if err := json.Unmarshal(raw, &manifest); err != nil {
		return nil, false, err
	}
	if len(manifest.Categories) == 0 {
		return nil, false, fmt.Errorf("categories is empty")
	}
	return &manifest, true, nil
}

func fetchJSON(target string, out any) error {
	req, err := http.NewRequest(http.MethodGet, target, nil)
	if err != nil {
		return err
	}
	if token := strings.TrimSpace(os.Getenv("SOW_ASSETS_TOKEN")); token != "" {
		req.Header.Set("Authorization", "token "+token)
	} else if token := strings.TrimSpace(os.Getenv("GITEA_TOKEN")); token != "" {
		req.Header.Set("Authorization", "token "+token)
	}
	resp, err := partsManifestHTTPClient.Do(req)
	if err != nil {
		return err
	}
	defer resp.Body.Close()
	if resp.StatusCode < 200 || resp.StatusCode >= 300 {
		body, _ := io.ReadAll(io.LimitReader(resp.Body, 4096))
		if resp.StatusCode == http.StatusUnauthorized || resp.StatusCode == http.StatusForbidden {
			return fmt.Errorf("HTTP %d from %s; set SOW_ASSETS_TOKEN or GITEA_TOKEN if the repo is private", resp.StatusCode, target)
		}
		return fmt.Errorf("HTTP %d from %s: %s", resp.StatusCode, target, strings.TrimSpace(string(body)))
	}
	decoder := json.NewDecoder(resp.Body)
	if err := decoder.Decode(out); err != nil {
		return err
	}
	return nil
}

func writePartsManifestCache(path string, manifest *partsManifest) error {
	if err := os.MkdirAll(filepath.Dir(path), 0o755); err != nil {
		return fmt.Errorf("create parts manifest cache dir: %w", err)
	}
	payload, err := json.MarshalIndent(manifest, "", "  ")
	if err != nil {
		return fmt.Errorf("marshal parts manifest cache: %w", err)
	}
	payload = append(payload, '\n')
	current, err := os.ReadFile(path)
	if err == nil && bytes.Equal(current, payload) {
		now := time.Now()
		if err := os.Chtimes(path, now, now); err != nil {
			return fmt.Errorf("refresh parts manifest cache timestamp: %w", err)
		}
		return nil
	}
	if err != nil && !os.IsNotExist(err) {
		return fmt.Errorf("read parts manifest cache: %w", err)
	}
	tmpPath := path + ".tmp"
	if err := os.WriteFile(tmpPath, payload, 0o644); err != nil {
		return fmt.Errorf("write parts manifest cache: %w", err)
	}
	if err := os.Rename(tmpPath, path); err != nil {
		return fmt.Errorf("replace parts manifest cache: %w", err)
	}
	return nil
}

func inventoryFromPartsManifest(manifest *partsManifest) map[string]map[int]struct{} {
	result := make(map[string]map[int]struct{}, len(supportedPartCategories))
	for _, category := range supportedPartCategories {
		ids := make(map[int]struct{})
		for _, rowID := range manifest.Categories[category] {
			ids[rowID] = struct{}{}
		}
		result[category] = ids
	}
	return result
}