0
Fork 0
mirror of https://codeberg.org/forgejo/forgejo.git synced 2024-12-25 17:04:12 -05:00
forgejo/modules/git/grep_test.go
Radosław Piliszek f4a7bf6d2a git-grep: skip binary files
It is a waste of resources to scan them looking for matches
because they are never returned back - they appear as empty
lines in the current format.

Notably, even if they were returned, it is unlikely that matching
in binary files makes sense when the goal is "code search".
2024-08-11 19:10:23 +02:00

178 lines
5.2 KiB
Go

// Copyright 2024 The Gitea Authors. All rights reserved.
// SPDX-License-Identifier: MIT
package git
import (
"bytes"
"context"
"os"
"path"
"path/filepath"
"testing"
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
)
func TestGrepSearch(t *testing.T) {
repo, err := openRepositoryWithDefaultContext(filepath.Join(testReposDir, "language_stats_repo"))
require.NoError(t, err)
defer repo.Close()
res, err := GrepSearch(context.Background(), repo, "public", GrepOptions{})
require.NoError(t, err)
assert.Equal(t, []*GrepResult{
{
Filename: "java-hello/main.java",
LineNumbers: []int{1, 3},
LineCodes: []string{
"public class HelloWorld",
" public static void main(String[] args)",
},
HighlightedRanges: [][3]int{{0, 0, 6}, {1, 1, 7}},
},
{
Filename: "main.vendor.java",
LineNumbers: []int{1, 3},
LineCodes: []string{
"public class HelloWorld",
" public static void main(String[] args)",
},
HighlightedRanges: [][3]int{{0, 0, 6}, {1, 1, 7}},
},
}, res)
res, err = GrepSearch(context.Background(), repo, "void", GrepOptions{MaxResultLimit: 1, ContextLineNumber: 2})
require.NoError(t, err)
assert.Equal(t, []*GrepResult{
{
Filename: "java-hello/main.java",
LineNumbers: []int{1, 2, 3, 4, 5},
LineCodes: []string{
"public class HelloWorld",
"{",
" public static void main(String[] args)",
" {",
" System.out.println(\"Hello world!\");",
},
HighlightedRanges: [][3]int{{2, 15, 19}},
},
}, res)
res, err = GrepSearch(context.Background(), repo, "world", GrepOptions{MatchesPerFile: 1})
require.NoError(t, err)
assert.Equal(t, []*GrepResult{
{
Filename: "i-am-a-python.p",
LineNumbers: []int{1},
LineCodes: []string{"## This is a simple file to do a hello world"},
HighlightedRanges: [][3]int{{0, 39, 44}},
},
{
Filename: "java-hello/main.java",
LineNumbers: []int{1},
LineCodes: []string{"public class HelloWorld"},
HighlightedRanges: [][3]int{{0, 18, 23}},
},
{
Filename: "main.vendor.java",
LineNumbers: []int{1},
LineCodes: []string{"public class HelloWorld"},
HighlightedRanges: [][3]int{{0, 18, 23}},
},
{
Filename: "python-hello/hello.py",
LineNumbers: []int{1},
LineCodes: []string{"## This is a simple file to do a hello world"},
HighlightedRanges: [][3]int{{0, 39, 44}},
},
}, res)
res, err = GrepSearch(context.Background(), repo, "no-such-content", GrepOptions{})
require.NoError(t, err)
assert.Empty(t, res)
res, err = GrepSearch(context.Background(), &Repository{Path: "no-such-git-repo"}, "no-such-content", GrepOptions{})
require.Error(t, err)
assert.Empty(t, res)
}
func TestGrepNoBinary(t *testing.T) {
tmpDir := t.TempDir()
err := InitRepository(DefaultContext, tmpDir, false, Sha1ObjectFormat.Name())
require.NoError(t, err)
gitRepo, err := openRepositoryWithDefaultContext(tmpDir)
require.NoError(t, err)
defer gitRepo.Close()
require.NoError(t, os.WriteFile(path.Join(tmpDir, "BINARY"), []byte("I AM BINARY\n\x00\nYOU WON'T SEE ME"), 0o666))
require.NoError(t, os.WriteFile(path.Join(tmpDir, "TEXT"), []byte("I AM NOT BINARY\nYOU WILL SEE ME"), 0o666))
err = AddChanges(tmpDir, true)
require.NoError(t, err)
err = CommitChanges(tmpDir, CommitChangesOptions{Message: "Binary and text files"})
require.NoError(t, err)
res, err := GrepSearch(context.Background(), gitRepo, "BINARY", GrepOptions{})
require.NoError(t, err)
assert.Len(t, res, 1)
assert.Equal(t, "TEXT", res[0].Filename)
}
func TestGrepLongFiles(t *testing.T) {
tmpDir := t.TempDir()
err := InitRepository(DefaultContext, tmpDir, false, Sha1ObjectFormat.Name())
require.NoError(t, err)
gitRepo, err := openRepositoryWithDefaultContext(tmpDir)
require.NoError(t, err)
defer gitRepo.Close()
require.NoError(t, os.WriteFile(path.Join(tmpDir, "README.md"), bytes.Repeat([]byte{'a'}, 65*1024), 0o666))
err = AddChanges(tmpDir, true)
require.NoError(t, err)
err = CommitChanges(tmpDir, CommitChangesOptions{Message: "Long file"})
require.NoError(t, err)
res, err := GrepSearch(context.Background(), gitRepo, "a", GrepOptions{})
require.NoError(t, err)
assert.Len(t, res, 1)
assert.Len(t, res[0].LineCodes[0], 65*1024)
}
func TestGrepRefs(t *testing.T) {
tmpDir := t.TempDir()
err := InitRepository(DefaultContext, tmpDir, false, Sha1ObjectFormat.Name())
require.NoError(t, err)
gitRepo, err := openRepositoryWithDefaultContext(tmpDir)
require.NoError(t, err)
defer gitRepo.Close()
require.NoError(t, os.WriteFile(path.Join(tmpDir, "README.md"), []byte{'A'}, 0o666))
require.NoError(t, AddChanges(tmpDir, true))
err = CommitChanges(tmpDir, CommitChangesOptions{Message: "add A"})
require.NoError(t, err)
require.NoError(t, gitRepo.CreateTag("v1", "HEAD"))
require.NoError(t, os.WriteFile(path.Join(tmpDir, "README.md"), []byte{'A', 'B', 'C', 'D'}, 0o666))
require.NoError(t, AddChanges(tmpDir, true))
err = CommitChanges(tmpDir, CommitChangesOptions{Message: "add BCD"})
require.NoError(t, err)
res, err := GrepSearch(context.Background(), gitRepo, "a", GrepOptions{RefName: "v1"})
require.NoError(t, err)
assert.Len(t, res, 1)
assert.Equal(t, "A", res[0].LineCodes[0])
}