gitea-mcp/internal/tools/code_search.go

package tools

import (
	"context"
	"encoding/json"
	"fmt"
	"sort"
	"sync"
	"time"

	"gitea.d-ma.be/mathias/gitea-mcp/internal/allowlist"
	"gitea.d-ma.be/mathias/gitea-mcp/internal/gitea"
	"gitea.d-ma.be/mathias/gitea-mcp/internal/registry"
)

type semaphore chan struct{}

func newSem(n int) semaphore { return make(semaphore, n) }
func (s semaphore) acquire() { s <- struct{}{} }
func (s semaphore) release() { <-s }

type CodeSearch struct {
	c *gitea.Client
	a *allowlist.Allowlist
}

func NewCodeSearch(c *gitea.Client, a *allowlist.Allowlist) *CodeSearch {
	return &CodeSearch{c: c, a: a}
}

func (t *CodeSearch) Descriptor() registry.ToolDescriptor {
	return registry.ToolDescriptor{
		Name:        "code_search",
		Description: "Search code across one repo or fan out across an owner's repos.",
		InputSchema: json.RawMessage(`{
			"type":"object",
			"properties":{
				"q":{"type":"string"},
				"owner":{"type":"string"},
				"repo":{"type":"string"},
				"page":{"type":"integer","minimum":1},
				"limit":{"type":"integer","minimum":1,"maximum":50}
			},
			"required":["q","owner"]
		}`),
	}
}

type codeSearchArgs struct {
	Q     string `json:"q"`
	Owner string `json:"owner"`
	Repo  string `json:"repo"`
	Page  int    `json:"page"`
	Limit int    `json:"limit"`
}

type codeSearchResult struct {
	Repo    string  `json:"repo"`
	Path    string  `json:"path"`
	Snippet string  `json:"snippet"`
	Score   float64 `json:"score"`
	HTMLURL string  `json:"html_url"`
}

func (t *CodeSearch) Call(ctx context.Context, raw json.RawMessage) (json.RawMessage, error) {
	var args codeSearchArgs
	if err := parseArgs(raw, &args); err != nil {
		return nil, err
	}
	if args.Q == "" {
		return nil, fmt.Errorf("q is required: %w", gitea.ErrValidation)
	}
	if err := t.a.Check(args.Owner); err != nil {
		return nil, err
	}
	if args.Page < 1 {
		args.Page = 1
	}
	if args.Limit < 1 || args.Limit > 50 {
		args.Limit = 30
	}

	if args.Repo != "" {
		return t.singleRepo(ctx, args)
	}
	return t.fanOut(ctx, args)
}

func (t *CodeSearch) singleRepo(ctx context.Context, args codeSearchArgs) (json.RawMessage, error) {
	hits, err := t.c.SearchCode(ctx, args.Owner, args.Repo, args.Q, args.Page, args.Limit)
	if err != nil {
		return nil, err
	}

	results := make([]codeSearchResult, 0, len(hits))
	repoFull := args.Owner + "/" + args.Repo
	for _, h := range hits {
		score := h.Score
		if score == 0 {
			score = 1.0
		}
		results = append(results, codeSearchResult{
			Repo:    repoFull,
			Path:    h.Path,
			Snippet: h.Snippet,
			Score:   score,
			HTMLURL: h.HTMLURL,
		})
	}
	out := map[string]any{"results": results}
	if len(hits) == args.Limit {
		out["next_page"] = args.Page + 1
	}
	return textOK(out)
}

func (t *CodeSearch) fanOut(ctx context.Context, args codeSearchArgs) (json.RawMessage, error) {
	repos, err := t.c.ListRepos(ctx, args.Owner, 1, 50)
	if err != nil {
		return nil, err
	}

	type repoResult struct {
		repo string
		hits []gitea.CodeSearchHit
		err  error
	}
	resultsCh := make(chan repoResult, len(repos))
	sem := newSem(5)
	var wg sync.WaitGroup

	for _, r := range repos {
		repo := r // capture
		wg.Add(1)
		go func() {
			defer wg.Done()
			sem.acquire()
			defer sem.release()

			rctx, cancel := context.WithTimeout(ctx, 5*time.Second)
			defer cancel()

			hits, err := t.c.SearchCode(rctx, args.Owner, repo.Name, args.Q, 1, args.Limit)
			resultsCh <- repoResult{repo: args.Owner + "/" + repo.Name, hits: hits, err: err}
		}()
	}
	wg.Wait()
	close(resultsCh)

	merged := make([]codeSearchResult, 0)
	var partialRepos []string
	for rr := range resultsCh {
		if rr.err != nil {
			partialRepos = append(partialRepos, rr.repo)
			continue
		}
		for _, h := range rr.hits {
			score := h.Score
			if score == 0 {
				score = 1.0
			}
			merged = append(merged, codeSearchResult{
				Repo: rr.repo, Path: h.Path, Snippet: h.Snippet, Score: score, HTMLURL: h.HTMLURL,
			})
		}
	}

	// Sort by score desc, then by repo+path for determinism.
	sort.Slice(merged, func(i, j int) bool {
		if merged[i].Score != merged[j].Score {
			return merged[i].Score > merged[j].Score
		}
		if merged[i].Repo != merged[j].Repo {
			return merged[i].Repo < merged[j].Repo
		}
		return merged[i].Path < merged[j].Path
	})
	if len(merged) > args.Limit {
		merged = merged[:args.Limit]
	}

	out := map[string]any{
		"results": merged,
		"partial": len(partialRepos) > 0,
	}
	if len(partialRepos) > 0 {
		sort.Strings(partialRepos)
		out["partial_repos"] = partialRepos
	}
	return textOK(out)
}