mesh-controller/cmd/mesh-controller/delivery_changes.go

package main

import (
	"context"
	"encoding/base64"
	"encoding/json"
	"fmt"
	"net/url"
	"path"
	"slices"
	"strings"
	"sync"
	"time"

	"github.com/novox/mesh-controller/internal/catalogue"
	"github.com/novox/mesh-controller/internal/link"
)

// What a change says of the manifests it touches, read from the forge (novox/hq ADR 0419 clause 3, the lead's settlement
// of issue 977 question 1): the controller asks the forge for each changed module.json at the commit and at what it
// replaced, once per commit. The composed check's result is not used: the plan is cut after the merge.

// changeSource is one change to a repository whose manifests are read: the commit after it and, when known, what it
// replaced (a pull request's base commit or branch). An empty Old is the first parent of New: the merge's base.
type changeSource struct {
	Owner, Repo string
	New, Old    string
	Paths       []string
	Truncated   bool
	Removed     []string
}

// manifestAt reads a repository's module.json at a ref from the forge; the bool says it exists there. A variable so a
// test needs no forge.
var manifestAt = func(ctx context.Context, owner, repo, file, ref string) (catalogue.Manifest, bool, error) {
	var found struct {
		Result struct {
			Content  string `json:"content"`
			Encoding string `json:"encoding"`
		} `json:"result"`
	}
	missing, err := askForgeApi(ctx, "/repos/"+owner+"/"+repo+"/contents/"+escapedPath(file)+"?ref="+url.QueryEscape(ref), &found)
	if err != nil {
		return catalogue.Manifest{}, false, err
	}
	if missing {
		return catalogue.Manifest{}, false, nil
	}
	body := []byte(found.Result.Content)
	if found.Result.Encoding == "base64" {
		if body, err = base64.StdEncoding.DecodeString(strings.ReplaceAll(found.Result.Content, "\n", "")); err != nil {
			return catalogue.Manifest{}, false, fmt.Errorf("%s/%s %s at %s cannot be read: %w", owner, repo, file, ref, err)
		}
	}
	var m catalogue.Manifest
	if err := json.Unmarshal(body, &m); err != nil {
		return catalogue.Manifest{}, false, fmt.Errorf("%s/%s %s at %s is not a manifest: %w", owner, repo, file, ref, err)
	}
	return m, true, nil
}

// parentOf is the first parent of a commit, from the forge. A variable so a test needs no forge.
var parentOf = func(ctx context.Context, owner, repo, commit string) (string, error) {
	var found struct {
		Result struct {
			Parents []struct {
				SHA string `json:"sha"`
			} `json:"parents"`
		} `json:"result"`
	}
	if _, err := askForgeApi(ctx, "/repos/"+owner+"/"+repo+"/git/commits/"+url.PathEscape(commit), &found); err != nil {
		return "", err
	}
	if len(found.Result.Parents) == 0 {
		return "", nil
	}
	return found.Result.Parents[0].SHA, nil
}

func escapedPath(p string) string {
	parts := strings.Split(p, "/")
	for i := range parts {
		parts[i] = url.PathEscape(parts[i])
	}
	return strings.Join(parts, "/")
}

// askForgeApi makes one read of the forge's API through its module's tool on the bus. It answers true when the forge
// says there is no such thing (404), and an error for anything else it cannot answer.
func askForgeApi(ctx context.Context, apiPath string, into any) (bool, error) {
	js, err := aBus()
	if err != nil {
		return false, err
	}
	defer js.Close()
	raw, _ := json.Marshal(map[string]any{"path": apiPath})
	answer, err := link.Ask(ctx, link.OverNATS{Conn: js.Conn()}, "gitea", "gitea_api", raw, 30*time.Second)
	if err != nil {
		return false, err
	}
	if answer.Error != "" {
		if strings.Contains(answer.Error, " 404 ") {
			return true, nil
		}
		return false, fmt.Errorf("the forge did not answer %s: %s", apiPath, answer.Error)
	}
	if err := json.Unmarshal(answer.Result, into); err != nil {
		return false, fmt.Errorf("the forge's answer to %s cannot be read: %w", apiPath, err)
	}
	return false, nil
}

// manifestsRead keeps what the forge said of a manifest at a commit, which never changes: one ask per commit.
var (
	manifestsRead   = map[string]manifestRead{}
	manifestsReadMu sync.Mutex
)

type manifestRead struct {
	manifest catalogue.Manifest
	exists   bool
}

func manifestCached(ctx context.Context, owner, repo, file, ref string) (catalogue.Manifest, bool, error) {
	key := strings.Join([]string{owner, repo, file, ref}, "\x00")
	if len(ref) == 40 {
		manifestsReadMu.Lock()
		got, ok := manifestsRead[key]
		manifestsReadMu.Unlock()
		if ok {
			return got.manifest, got.exists, nil
		}
	}
	m, exists, err := manifestAt(ctx, owner, repo, file, ref)
	if err == nil && len(ref) == 40 {
		manifestsReadMu.Lock()
		if len(manifestsRead) > 512 {
			clear(manifestsRead)
		}
		manifestsRead[key] = manifestRead{m, exists}
		manifestsReadMu.Unlock()
	}
	return m, exists, err
}

// verbsServed are the seat verbs a manifest serves, as "<seat>.<verb>": those of the seats it defines, those of the
// seats it claims, and its own tools as "<module>.<tool>".
func verbsServed(m catalogue.Manifest) []string {
	var out []string
	for _, s := range m.DefinesSeats {
		for _, v := range s.Serves {
			out = appendNew(out, s.Name+"."+v.Name)
		}
	}
	for _, c := range m.Claims {
		for _, v := range c.ServesFor(m) {
			out = appendNew(out, c.Name+"."+v)
		}
	}
	for _, t := range m.Tools {
		out = appendNew(out, m.Module+"."+t)
	}
	return out
}

// verbsInvoked are the verbs a manifest calls, as "<seat>.<verb>" or "<module>.<tool>".
func verbsInvoked(m catalogue.Manifest) []string {
	var out []string
	for _, i := range m.Invokes {
		if i = strings.TrimPrefix(i, "seat:"); i != "*" && i != "" {
			out = appendNew(out, i)
		}
	}
	return out
}

func added(now, was []string) []string {
	var out []string
	for _, v := range now {
		if !slices.Contains(was, v) {
			out = append(out, v)
		}
	}
	return out
}

// changesOf reads, for each manifest the changes touch, what they add to the verbs its module serves and invokes, by
// module. A change whose paths are unsaid or truncated names no manifest: nothing is read for it. An error is the
// forge's, never read as "no change": the plan is not cut on a guess.
func changesOf(ctx context.Context, sources []changeSource) (map[string]moduleChange, error) {
	out := map[string]moduleChange{}
	for _, s := range sources {
		if s.Truncated {
			continue
		}
		for _, p := range s.Paths {
			if path.Base(p) != "module.json" || slices.Contains(s.Removed, p) {
				continue
			}
			now, exists, err := manifestCached(ctx, s.Owner, s.Repo, p, s.New)
			if err != nil {
				return nil, err
			}
			if !exists || now.Module == "" {
				continue
			}
			old := s.Old
			if old == "" {
				if old, err = parentOf(ctx, s.Owner, s.Repo, s.New); err != nil {
					return nil, err
				}
			}
			var was catalogue.Manifest
			if old != "" {
				if was, _, err = manifestCached(ctx, s.Owner, s.Repo, p, old); err != nil {
					return nil, err
				}
			}
			c := out[now.Module]
			c.Manifest = true
			for _, v := range added(verbsServed(now), verbsServed(was)) {
				c.Serves = appendNew(c.Serves, v)
			}
			for _, v := range added(verbsInvoked(now), verbsInvoked(was)) {
				c.Invokes = appendNew(c.Invokes, v)
			}
			out[now.Module] = c
		}
	}
	return out, nil
}

// changeSourcesOf are a group's members as changes to read: each head over what its pull request merges into.
func changeSourcesOf(members []orderMember) []changeSource {
	var out []changeSource
	for _, m := range members {
		p := m.pull()
		old := p.BaseSHA
		if old == "" {
			old = p.Base
		}
		out = append(out, changeSource{Owner: p.Owner, Repo: p.Repo, New: p.Commit, Old: old, Paths: p.Paths,
			Truncated: p.PathsTruncated, Removed: p.Removed})
	}
	return out
}