gitea-mcp/internal/gitea/code_search_test.go

package gitea_test

import (
	"context"
	"encoding/base64"
	"errors"
	"fmt"
	"net/http"
	"net/http/httptest"
	"strings"
	"testing"

	"git.d-ma.be/mathias/gitea-mcp/internal/gitea"
	"github.com/stretchr/testify/assert"
	"github.com/stretchr/testify/require"
)

func b64(s string) string { return base64.StdEncoding.EncodeToString([]byte(s)) }

// codeSearchFake serves GetRepo + GetTree + GetFileContents off an in-memory
// file map — the REAL endpoints SearchCode uses now that Gitea's REST API has
// no code-content-search endpoint (confirmed against a live 1.25.5 instance's
// swagger spec: only /repos/search, /repos/issues/search etc exist — the web
// UI's own code search falls back to server-side `git grep`, an HTML-only
// route, not JSON API). Records which paths were actually fetched, so tests
// can assert a file was SKIPPED (binary ext, oversized) without ever being
// read, not just absent from results.
type codeSearchFake struct {
	files   map[string]string // path -> content
	sizes   map[string]int64  // path -> tree-entry size (defaults to len(content))
	fetched []string
}

func (f *codeSearchFake) handler(t *testing.T, owner, repo, branch string) http.HandlerFunc {
	return func(w http.ResponseWriter, r *http.Request) {
		w.Header().Set("Content-Type", "application/json")
		p := r.URL.Path
		switch {
		case r.Method == http.MethodGet && p == "/api/v1/repos/"+owner+"/"+repo:
			_, _ = fmt.Fprintf(w, `{"name":%q,"full_name":"%s/%s","default_branch":%q}`, repo, owner, repo, branch)

		case r.Method == http.MethodGet && strings.HasPrefix(p, "/api/v1/repos/"+owner+"/"+repo+"/git/trees/"):
			var entries []string
			for path, content := range f.files {
				size := int64(len(content))
				if s, ok := f.sizes[path]; ok {
					size = s
				}
				entries = append(entries, fmt.Sprintf(`{"path":%q,"type":"blob","sha":"s","size":%d}`, path, size))
			}
			_, _ = fmt.Fprintf(w, `{"sha":"root","tree":[%s],"truncated":false}`, strings.Join(entries, ","))

		case r.Method == http.MethodGet && strings.HasPrefix(p, "/api/v1/repos/"+owner+"/"+repo+"/contents/"):
			path := strings.TrimPrefix(p, "/api/v1/repos/"+owner+"/"+repo+"/contents/")
			f.fetched = append(f.fetched, path)
			content, ok := f.files[path]
			if !ok {
				w.WriteHeader(http.StatusNotFound)
				_, _ = w.Write([]byte(`{"message":"not found"}`))
				return
			}
			_, _ = fmt.Fprintf(w, `{"path":%q,"sha":"s","size":%d,"content":%q,"encoding":"base64"}`, path, len(content), b64(content))

		default:
			t.Errorf("unexpected request: %s %s", r.Method, p)
			w.WriteHeader(http.StatusNotFound)
		}
	}
}

func TestSearchCode_FindsMatchInTree(t *testing.T) {
	f := &codeSearchFake{files: map[string]string{
		"internal/gitea/code_search.go": "func (c *Client) SearchCode(ctx context.Context) {}\n",
		"internal/gitea/repos.go":       "func (c *Client) ListRepos() {}\n",
	}}
	srv := httptest.NewServer(f.handler(t, "mathias", "infra", "main"))
	defer srv.Close()

	c := gitea.NewClient(srv.URL, "tok")
	hits, err := c.SearchCode(context.Background(), "mathias", "infra", "SearchCode", 1, 30)
	require.NoError(t, err)
	require.Len(t, hits, 1)
	assert.Equal(t, "internal/gitea/code_search.go", hits[0].Path)
	assert.Contains(t, hits[0].Snippet, "SearchCode")
	assert.Contains(t, hits[0].HTMLURL, "internal/gitea/code_search.go")
	assert.Equal(t, 1.0, hits[0].Score)
}

func TestSearchCode_CaseInsensitive(t *testing.T) {
	f := &codeSearchFake{files: map[string]string{
		"README.md": "# MyProject\n\nBuild instructions here.\n",
	}}
	srv := httptest.NewServer(f.handler(t, "mathias", "infra", "main"))
	defer srv.Close()

	c := gitea.NewClient(srv.URL, "tok")
	hits, err := c.SearchCode(context.Background(), "mathias", "infra", "myproject", 1, 30)
	require.NoError(t, err)
	require.Len(t, hits, 1)
}

func TestSearchCode_SkipsBinaryExtensionWithoutFetching(t *testing.T) {
	f := &codeSearchFake{files: map[string]string{
		"assets/logo.png": "SearchCode", // would match if fetched — must not be
		"main.go":         "package main\n",
	}}
	srv := httptest.NewServer(f.handler(t, "mathias", "infra", "main"))
	defer srv.Close()

	c := gitea.NewClient(srv.URL, "tok")
	hits, err := c.SearchCode(context.Background(), "mathias", "infra", "SearchCode", 1, 30)
	require.NoError(t, err)
	assert.Empty(t, hits)
	assert.NotContains(t, f.fetched, "assets/logo.png", "binary extension must be skipped before fetch")
}

func TestSearchCode_SkipsOversizedFileWithoutFetching(t *testing.T) {
	f := &codeSearchFake{
		files: map[string]string{"vendor/bundle.txt": "SearchCode"},
		sizes: map[string]int64{"vendor/bundle.txt": 10 * 1024 * 1024}, // 10MB per the tree listing
	}
	srv := httptest.NewServer(f.handler(t, "mathias", "infra", "main"))
	defer srv.Close()

	c := gitea.NewClient(srv.URL, "tok")
	hits, err := c.SearchCode(context.Background(), "mathias", "infra", "SearchCode", 1, 30)
	require.NoError(t, err)
	assert.Empty(t, hits)
	assert.NotContains(t, f.fetched, "vendor/bundle.txt", "oversized file must be skipped before fetch")
}

func TestSearchCode_SkipsBinaryContentNullByte(t *testing.T) {
	f := &codeSearchFake{files: map[string]string{
		"data.bin": "SearchCode\x00binary-marker",
	}}
	srv := httptest.NewServer(f.handler(t, "mathias", "infra", "main"))
	defer srv.Close()

	c := gitea.NewClient(srv.URL, "tok")
	hits, err := c.SearchCode(context.Background(), "mathias", "infra", "SearchCode", 1, 30)
	require.NoError(t, err)
	assert.Empty(t, hits, "content with a null byte must be treated as binary even past the extension filter")
}

func TestSearchCode_Pagination(t *testing.T) {
	files := map[string]string{}
	for i := 0; i < 5; i++ {
		files[fmt.Sprintf("file%d.go", i)] = strings.Repeat("hit ", i+1) // increasing occurrence count => increasing score
	}
	f := &codeSearchFake{files: files}
	srv := httptest.NewServer(f.handler(t, "mathias", "infra", "main"))
	defer srv.Close()

	c := gitea.NewClient(srv.URL, "tok")
	page1, err := c.SearchCode(context.Background(), "mathias", "infra", "hit", 1, 2)
	require.NoError(t, err)
	require.Len(t, page1, 2)
	page2, err := c.SearchCode(context.Background(), "mathias", "infra", "hit", 2, 2)
	require.NoError(t, err)
	require.Len(t, page2, 2)

	assert.NotEqual(t, page1[0].Path, page2[0].Path, "pages must not overlap")
	assert.True(t, page1[0].Score >= page1[1].Score, "page1 sorted desc by score")
	assert.True(t, page1[1].Score >= page2[0].Score, "page1's worst must rank >= page2's best")
}

func TestSearchCode_EmptyQueryErrors(t *testing.T) {
	c := gitea.NewClient("http://unused", "tok")
	_, err := c.SearchCode(context.Background(), "mathias", "infra", "", 1, 30)
	require.Error(t, err)
	assert.True(t, errors.Is(err, gitea.ErrValidation))
}