go-enry/common_test.go

477 lines
21 KiB
Go
Raw Normal View History

2017-06-13 13:56:07 +02:00
package enry
2016-07-13 22:21:18 +02:00
import (
2017-05-29 10:05:16 +02:00
"fmt"
"io/ioutil"
"os"
"os/exec"
2017-05-29 10:05:16 +02:00
"path/filepath"
2016-07-13 22:21:18 +02:00
"testing"
"github.com/src-d/enry/v2/data"
2017-06-28 10:58:58 +02:00
2017-05-29 10:05:16 +02:00
"github.com/stretchr/testify/assert"
"github.com/stretchr/testify/require"
2017-05-29 10:05:16 +02:00
"github.com/stretchr/testify/suite"
2016-07-13 22:21:18 +02:00
)
const linguistURL = "https://github.com/github/linguist.git"
const linguistClonedEnvVar = "ENRY_TEST_REPO"
type EnryTestSuite struct {
2017-05-29 10:05:16 +02:00
suite.Suite
tmpLinguist string
needToClone bool
samplesDir string
}
func (s *EnryTestSuite) TestRegexpEdgeCases() {
var regexpEdgeCases = []struct {
lang string
filename string
}{
{lang: "ActionScript", filename: "FooBar.as"},
{lang: "Forth", filename: "asm.fr"},
{lang: "X PixMap", filename: "cc-public_domain_mark_white.pm"},
//{lang: "SQL", filename: "drop_stuff.sql"}, // https://github.com/src-d/enry/issues/194
{lang: "Fstar", filename: "Hacl.Spec.Bignum.Fmul.fst"},
{lang: "C++", filename: "Types.h"},
}
for _, r := range regexpEdgeCases {
filename := fmt.Sprintf("%s/samples/%s/%s", s.tmpLinguist, r.lang, r.filename)
content, err := ioutil.ReadFile(filename)
require.NoError(s.T(), err)
lang := GetLanguage(r.filename, content)
s.T().Logf("File:%s, lang:%s", filename, lang)
expLang, _ := data.LanguageByAlias(r.lang)
require.EqualValues(s.T(), expLang, lang)
}
}
func Test_EnryTestSuite(t *testing.T) {
suite.Run(t, new(EnryTestSuite))
}
func (s *EnryTestSuite) SetupSuite() {
var err error
s.tmpLinguist = os.Getenv(linguistClonedEnvVar)
s.needToClone = s.tmpLinguist == ""
if s.needToClone {
s.tmpLinguist, err = ioutil.TempDir("", "linguist-")
require.NoError(s.T(), err)
s.T().Logf("Cloning Linguist repo to '%s' as %s was not set\n",
s.tmpLinguist, linguistClonedEnvVar)
cmd := exec.Command("git", "clone", linguistURL, s.tmpLinguist)
err = cmd.Run()
require.NoError(s.T(), err)
}
s.samplesDir = filepath.Join(s.tmpLinguist, "samples")
s.T().Logf("using samples from %s", s.samplesDir)
cwd, err := os.Getwd()
assert.NoError(s.T(), err)
err = os.Chdir(s.tmpLinguist)
assert.NoError(s.T(), err)
cmd := exec.Command("git", "checkout", data.LinguistCommit)
err = cmd.Run()
assert.NoError(s.T(), err)
err = os.Chdir(cwd)
assert.NoError(s.T(), err)
2017-05-29 10:05:16 +02:00
}
func (s *EnryTestSuite) TearDownSuite() {
if s.needToClone {
err := os.RemoveAll(s.tmpLinguist)
assert.NoError(s.T(), err)
}
2017-05-29 10:05:16 +02:00
}
func (s *EnryTestSuite) TestGetLanguage() {
2017-05-29 10:05:16 +02:00
tests := []struct {
name string
filename string
content []byte
expected string
safe bool
2017-05-29 10:05:16 +02:00
}{
{name: "TestGetLanguage_0", filename: "foo.h", content: []byte{}, expected: "C"},
2017-05-29 10:05:16 +02:00
{name: "TestGetLanguage_1", filename: "foo.py", content: []byte{}, expected: "Python"},
{name: "TestGetLanguage_2", filename: "foo.m", content: []byte(":- module"), expected: "Mercury"},
{name: "TestGetLanguage_3", filename: "foo.m", content: nil, expected: "MATLAB"},
{name: "TestGetLanguage_4", filename: "foo.mo", content: []byte{0xDE, 0x12, 0x04, 0x95, 0x00, 0x00, 0x00, 0x00}, expected: OtherLanguage},
{name: "TestGetLanguage_5", filename: "", content: nil, expected: OtherLanguage},
2017-05-29 10:05:16 +02:00
}
for _, test := range tests {
language := GetLanguage(test.filename, test.content)
2017-05-31 12:07:46 +02:00
assert.Equal(s.T(), test.expected, language, fmt.Sprintf("%v: %v, expected: %v", test.name, language, test.expected))
2017-05-29 10:05:16 +02:00
}
}
func (s *EnryTestSuite) TestGetLanguagesByModelineLinguist() {
var modelinesDir = filepath.Join(s.tmpLinguist, "test/fixtures/Data/Modelines")
2017-05-29 10:05:16 +02:00
tests := []struct {
2017-06-12 13:42:20 +02:00
name string
filename string
candidates []string
expected []string
2017-05-29 10:05:16 +02:00
}{
// Emacs
2017-06-12 13:42:20 +02:00
{name: "TestGetLanguagesByModelineLinguist_1", filename: filepath.Join(modelinesDir, "example_smalltalk.md"), expected: []string{"Smalltalk"}},
{name: "TestGetLanguagesByModelineLinguist_2", filename: filepath.Join(modelinesDir, "fundamentalEmacs.c"), expected: []string{"Text"}},
{name: "TestGetLanguagesByModelineLinguist_3", filename: filepath.Join(modelinesDir, "iamphp.inc"), expected: []string{"PHP"}},
{name: "TestGetLanguagesByModelineLinguist_4", filename: filepath.Join(modelinesDir, "seeplusplusEmacs1"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_5", filename: filepath.Join(modelinesDir, "seeplusplusEmacs2"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_6", filename: filepath.Join(modelinesDir, "seeplusplusEmacs3"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_7", filename: filepath.Join(modelinesDir, "seeplusplusEmacs4"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_8", filename: filepath.Join(modelinesDir, "seeplusplusEmacs5"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_9", filename: filepath.Join(modelinesDir, "seeplusplusEmacs6"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_10", filename: filepath.Join(modelinesDir, "seeplusplusEmacs7"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_11", filename: filepath.Join(modelinesDir, "seeplusplusEmacs9"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_12", filename: filepath.Join(modelinesDir, "seeplusplusEmacs10"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_13", filename: filepath.Join(modelinesDir, "seeplusplusEmacs11"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_14", filename: filepath.Join(modelinesDir, "seeplusplusEmacs12"), expected: []string{"C++"}},
2017-05-29 10:05:16 +02:00
// Vim
2017-06-12 13:42:20 +02:00
{name: "TestGetLanguagesByModelineLinguist_15", filename: filepath.Join(modelinesDir, "seeplusplus"), expected: []string{"C++"}},
{name: "TestGetLanguagesByModelineLinguist_16", filename: filepath.Join(modelinesDir, "iamjs.pl"), expected: []string{"JavaScript"}},
{name: "TestGetLanguagesByModelineLinguist_17", filename: filepath.Join(modelinesDir, "iamjs2.pl"), expected: []string{"JavaScript"}},
{name: "TestGetLanguagesByModelineLinguist_18", filename: filepath.Join(modelinesDir, "not_perl.pl"), expected: []string{"Prolog"}},
{name: "TestGetLanguagesByModelineLinguist_19", filename: filepath.Join(modelinesDir, "ruby"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_20", filename: filepath.Join(modelinesDir, "ruby2"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_21", filename: filepath.Join(modelinesDir, "ruby3"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_22", filename: filepath.Join(modelinesDir, "ruby4"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_23", filename: filepath.Join(modelinesDir, "ruby5"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_24", filename: filepath.Join(modelinesDir, "ruby6"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_25", filename: filepath.Join(modelinesDir, "ruby7"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_26", filename: filepath.Join(modelinesDir, "ruby8"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_27", filename: filepath.Join(modelinesDir, "ruby9"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_28", filename: filepath.Join(modelinesDir, "ruby10"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_29", filename: filepath.Join(modelinesDir, "ruby11"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_30", filename: filepath.Join(modelinesDir, "ruby12"), expected: []string{"Ruby"}},
{name: "TestGetLanguagesByModelineLinguist_31", filename: filepath.Join(s.samplesDir, "C/main.c"), expected: nil},
{name: "TestGetLanguagesByModelineLinguist_32", filename: "", expected: nil},
2017-05-29 10:05:16 +02:00
}
for _, test := range tests {
var content []byte
var err error
if test.filename != "" {
content, err = ioutil.ReadFile(test.filename)
assert.NoError(s.T(), err)
}
2017-05-29 10:05:16 +02:00
2017-06-12 13:42:20 +02:00
languages := GetLanguagesByModeline(test.filename, content, test.candidates)
assert.Equal(s.T(), test.expected, languages, fmt.Sprintf("%v: languages = %v, expected: %v", test.name, languages, test.expected))
2017-05-29 10:05:16 +02:00
}
}
func (s *EnryTestSuite) TestGetLanguagesByModeline() {
2017-05-29 10:05:16 +02:00
const (
wrongVim = `# vim: set syntax=ruby ft =python filetype=perl :`
rightVim = `/* vim: set syntax=python ft =python filetype=python */`
noLangVim = `/* vim: set shiftwidth=4 softtabstop=0 cindent cinoptions={1s: */`
)
tests := []struct {
2017-06-12 13:42:20 +02:00
name string
filename string
content []byte
candidates []string
expected []string
2017-05-29 10:05:16 +02:00
}{
2017-06-12 13:42:20 +02:00
{name: "TestGetLanguagesByModeline_1", content: []byte(wrongVim), expected: nil},
{name: "TestGetLanguagesByModeline_2", content: []byte(rightVim), expected: []string{"Python"}},
{name: "TestGetLanguagesByModeline_3", content: []byte(noLangVim), expected: nil},
{name: "TestGetLanguagesByModeline_4", content: nil, expected: nil},
{name: "TestGetLanguagesByModeline_5", content: []byte{}, expected: nil},
2017-05-29 10:05:16 +02:00
}
for _, test := range tests {
2017-06-12 13:42:20 +02:00
languages := GetLanguagesByModeline(test.filename, test.content, test.candidates)
assert.Equal(s.T(), test.expected, languages, fmt.Sprintf("%v: languages = %v, expected: %v", test.name, languages, test.expected))
2017-05-29 10:05:16 +02:00
}
}
func (s *EnryTestSuite) TestGetLanguagesByFilename() {
2017-05-29 10:05:16 +02:00
tests := []struct {
2017-06-12 13:42:20 +02:00
name string
filename string
content []byte
candidates []string
expected []string
2017-05-29 10:05:16 +02:00
}{
2017-06-12 13:42:20 +02:00
{name: "TestGetLanguagesByFilename_1", filename: "unknown.interpreter", expected: nil},
{name: "TestGetLanguagesByFilename_2", filename: ".bashrc", expected: []string{"Shell"}},
{name: "TestGetLanguagesByFilename_3", filename: "Dockerfile", expected: []string{"Dockerfile"}},
{name: "TestGetLanguagesByFilename_4", filename: "Makefile.frag", expected: []string{"Makefile"}},
{name: "TestGetLanguagesByFilename_5", filename: "makefile", expected: []string{"Makefile"}},
{name: "TestGetLanguagesByFilename_6", filename: "Vagrantfile", expected: []string{"Ruby"}},
{name: "TestGetLanguagesByFilename_7", filename: "_vimrc", expected: []string{"Vim script"}},
{name: "TestGetLanguagesByFilename_8", filename: "pom.xml", expected: []string{"Maven POM"}},
{name: "TestGetLanguagesByFilename_9", filename: "", expected: nil},
2017-05-29 10:05:16 +02:00
}
for _, test := range tests {
2017-06-12 13:42:20 +02:00
languages := GetLanguagesByFilename(test.filename, test.content, test.candidates)
assert.Equal(s.T(), test.expected, languages, fmt.Sprintf("%v: languages = %v, expected: %v", test.name, languages, test.expected))
2017-05-29 10:05:16 +02:00
}
}
func (s *EnryTestSuite) TestGetLanguagesByShebang() {
2017-05-29 10:05:16 +02:00
const (
multilineExecHack = `#!/bin/sh
# Next line is comment in Tcl, but not in sh... \
exec tclsh "$0" ${1+"$@"}`
multilineNoExecHack = `#!/bin/sh
#<<<#
echo "A shell script in a zkl program ($0)"
echo "Now run zkl <this file> with Hello World as args"
zkl $0 Hello World!
exit
#<<<#
println("The shell script says ",vm.arglist.concat(" "));`
)
tests := []struct {
2017-06-12 13:42:20 +02:00
name string
filename string
content []byte
candidates []string
expected []string
2017-05-29 10:05:16 +02:00
}{
2017-06-12 13:42:20 +02:00
{name: "TestGetLanguagesByShebang_1", content: []byte(`#!/unknown/interpreter`), expected: nil},
{name: "TestGetLanguagesByShebang_2", content: []byte(`no shebang`), expected: nil},
{name: "TestGetLanguagesByShebang_3", content: []byte(`#!/usr/bin/env`), expected: nil},
{name: "TestGetLanguagesByShebang_4", content: []byte(`#!/usr/bin/python -tt`), expected: []string{"Python"}},
{name: "TestGetLanguagesByShebang_5", content: []byte(`#!/usr/bin/env python2.6`), expected: []string{"Python"}},
{name: "TestGetLanguagesByShebang_6", content: []byte(`#!/usr/bin/env perl`), expected: []string{"Perl", "Pod"}},
2017-06-12 13:42:20 +02:00
{name: "TestGetLanguagesByShebang_7", content: []byte(`#! /bin/sh`), expected: []string{"Shell"}},
{name: "TestGetLanguagesByShebang_8", content: []byte(`#!bash`), expected: []string{"Shell"}},
{name: "TestGetLanguagesByShebang_9", content: []byte(multilineExecHack), expected: []string{"Tcl"}},
{name: "TestGetLanguagesByShebang_10", content: []byte(multilineNoExecHack), expected: []string{"Shell"}},
{name: "TestGetLanguagesByShebang_11", content: []byte(`#!`), expected: nil},
2017-05-29 10:05:16 +02:00
}
for _, test := range tests {
2017-06-12 13:42:20 +02:00
languages := GetLanguagesByShebang(test.filename, test.content, test.candidates)
assert.Equal(s.T(), test.expected, languages, fmt.Sprintf("%v: languages = %v, expected: %v", test.name, languages, test.expected))
2017-05-29 10:05:16 +02:00
}
}
func (s *EnryTestSuite) TestGetLanguagesByExtension() {
2017-05-29 10:05:16 +02:00
tests := []struct {
2017-06-12 13:42:20 +02:00
name string
filename string
content []byte
candidates []string
expected []string
2017-05-29 10:05:16 +02:00
}{
{name: "TestGetLanguagesByExtension_0", filename: "foo.h", expected: []string{"C", "C++", "Objective-C"}},
2017-06-12 13:42:20 +02:00
{name: "TestGetLanguagesByExtension_1", filename: "foo.foo", expected: nil},
{name: "TestGetLanguagesByExtension_2", filename: "foo.go", expected: []string{"Go"}},
{name: "TestGetLanguagesByExtension_3", filename: "foo.go.php", expected: []string{"Hack", "PHP"}},
{name: "TestGetLanguagesByExtension_4", filename: "", expected: nil},
2017-05-29 10:05:16 +02:00
}
2016-07-13 22:21:18 +02:00
2017-05-29 10:05:16 +02:00
for _, test := range tests {
2017-06-12 13:42:20 +02:00
languages := GetLanguagesByExtension(test.filename, test.content, test.candidates)
assert.Equal(s.T(), test.expected, languages, fmt.Sprintf("%v: languages = %v, expected: %v", test.name, languages, test.expected))
2017-05-29 10:05:16 +02:00
}
}
func (s *EnryTestSuite) TestGetLanguagesByClassifier() {
2017-05-29 10:05:16 +02:00
test := []struct {
name string
filename string
2017-06-12 13:42:20 +02:00
candidates []string
2017-05-29 10:05:16 +02:00
expected string
}{
{name: "TestGetLanguagesByClassifier_1", filename: filepath.Join(s.samplesDir, "C/blob.c"), candidates: []string{"python", "ruby", "c", "c++"}, expected: "C"},
{name: "TestGetLanguagesByClassifier_2", filename: filepath.Join(s.samplesDir, "C/blob.c"), candidates: nil, expected: OtherLanguage},
{name: "TestGetLanguagesByClassifier_3", filename: filepath.Join(s.samplesDir, "C/main.c"), candidates: []string{}, expected: OtherLanguage},
{name: "TestGetLanguagesByClassifier_4", filename: filepath.Join(s.samplesDir, "C/blob.c"), candidates: []string{"python", "ruby", "c++"}, expected: "C++"},
{name: "TestGetLanguagesByClassifier_5", filename: filepath.Join(s.samplesDir, "C/blob.c"), candidates: []string{"ruby"}, expected: "Ruby"},
{name: "TestGetLanguagesByClassifier_6", filename: filepath.Join(s.samplesDir, "Python/django-models-base.py"), candidates: []string{"python", "ruby", "c", "c++"}, expected: "Python"},
{name: "TestGetLanguagesByClassifier_7", filename: "", candidates: []string{"python"}, expected: "Python"},
2017-05-29 10:05:16 +02:00
}
for _, test := range test {
var content []byte
var err error
if test.filename != "" {
content, err = ioutil.ReadFile(test.filename)
assert.NoError(s.T(), err)
}
2017-05-29 10:05:16 +02:00
2017-06-12 13:42:20 +02:00
languages := GetLanguagesByClassifier(test.filename, content, test.candidates)
var language string
if len(languages) == 0 {
language = OtherLanguage
} else {
language = languages[0]
}
assert.Equal(s.T(), test.expected, language, fmt.Sprintf("%v: language = %v, expected: %v", test.name, language, test.expected))
}
}
func (s *EnryTestSuite) TestGetLanguagesBySpecificClassifier() {
2017-06-12 13:42:20 +02:00
test := []struct {
name string
filename string
candidates []string
classifier Classifier
expected string
}{
{name: "TestGetLanguagesByClassifier_1", filename: filepath.Join(s.samplesDir, "C/blob.c"), candidates: []string{"python", "ruby", "c", "c++"}, classifier: DefaultClassifier, expected: "C"},
{name: "TestGetLanguagesByClassifier_2", filename: filepath.Join(s.samplesDir, "C/blob.c"), candidates: nil, classifier: DefaultClassifier, expected: "C"},
{name: "TestGetLanguagesByClassifier_3", filename: filepath.Join(s.samplesDir, "C/main.c"), candidates: []string{}, classifier: DefaultClassifier, expected: "C"},
{name: "TestGetLanguagesByClassifier_4", filename: filepath.Join(s.samplesDir, "C/blob.c"), candidates: []string{"python", "ruby", "c++"}, classifier: DefaultClassifier, expected: "C++"},
{name: "TestGetLanguagesByClassifier_5", filename: filepath.Join(s.samplesDir, "C/blob.c"), candidates: []string{"ruby"}, classifier: DefaultClassifier, expected: "Ruby"},
{name: "TestGetLanguagesByClassifier_6", filename: filepath.Join(s.samplesDir, "Python/django-models-base.py"), candidates: []string{"python", "ruby", "c", "c++"}, classifier: DefaultClassifier, expected: "Python"},
{name: "TestGetLanguagesByClassifier_7", filename: os.DevNull, candidates: nil, classifier: DefaultClassifier, expected: "XML"},
2017-06-12 13:42:20 +02:00
}
for _, test := range test {
content, err := ioutil.ReadFile(test.filename)
assert.NoError(s.T(), err)
languages := GetLanguagesBySpecificClassifier(content, test.candidates, test.classifier)
var language string
if len(languages) == 0 {
language = OtherLanguage
} else {
language = languages[0]
}
assert.Equal(s.T(), test.expected, language, fmt.Sprintf("%v: language = %v, expected: %v", test.name, language, test.expected))
2017-05-29 10:05:16 +02:00
}
}
func (s *EnryTestSuite) TestGetLanguageExtensions() {
2017-05-29 10:05:16 +02:00
tests := []struct {
name string
language string
expected []string
}{
{name: "TestGetLanguageExtensions_1", language: "foo", expected: nil},
{name: "TestGetLanguageExtensions_2", language: "COBOL", expected: []string{".cob", ".cbl", ".ccp", ".cobol", ".cpy"}},
{name: "TestGetLanguageExtensions_3", language: "Maven POM", expected: nil},
}
for _, test := range tests {
extensions := GetLanguageExtensions(test.language)
2017-05-31 12:07:46 +02:00
assert.EqualValues(s.T(), test.expected, extensions, fmt.Sprintf("%v: extensions = %v, expected: %v", test.name, extensions, test.expected))
2017-05-29 10:05:16 +02:00
}
}
func (s *EnryTestSuite) TestGetLanguageType() {
2017-05-29 10:05:16 +02:00
tests := []struct {
name string
language string
expected Type
}{
{name: "TestGetLanguageType_1", language: "BestLanguageEver", expected: Unknown},
{name: "TestGetLanguageType_2", language: "JSON", expected: Data},
{name: "TestGetLanguageType_3", language: "COLLADA", expected: Data},
{name: "TestGetLanguageType_4", language: "Go", expected: Programming},
{name: "TestGetLanguageType_5", language: "Brainfuck", expected: Programming},
{name: "TestGetLanguageType_6", language: "HTML", expected: Markup},
{name: "TestGetLanguageType_7", language: "Sass", expected: Markup},
{name: "TestGetLanguageType_8", language: "AsciiDoc", expected: Prose},
{name: "TestGetLanguageType_9", language: "Textile", expected: Prose},
}
for _, test := range tests {
langType := GetLanguageType(test.language)
2017-05-31 12:07:46 +02:00
assert.Equal(s.T(), test.expected, langType, fmt.Sprintf("%v: langType = %v, expected: %v", test.name, langType, test.expected))
2017-05-29 10:05:16 +02:00
}
}
2016-07-13 22:21:18 +02:00
func (s *EnryTestSuite) TestGetLanguageByAlias() {
2017-05-29 10:05:16 +02:00
tests := []struct {
name string
alias string
expectedLang string
expectedOk bool
}{
{name: "TestGetLanguageByAlias_1", alias: "BestLanguageEver", expectedLang: OtherLanguage, expectedOk: false},
{name: "TestGetLanguageByAlias_2", alias: "aspx-vb", expectedLang: "ASP", expectedOk: true},
{name: "TestGetLanguageByAlias_3", alias: "C++", expectedLang: "C++", expectedOk: true},
{name: "TestGetLanguageByAlias_4", alias: "c++", expectedLang: "C++", expectedOk: true},
{name: "TestGetLanguageByAlias_5", alias: "objc", expectedLang: "Objective-C", expectedOk: true},
{name: "TestGetLanguageByAlias_6", alias: "golang", expectedLang: "Go", expectedOk: true},
{name: "TestGetLanguageByAlias_7", alias: "GOLANG", expectedLang: "Go", expectedOk: true},
{name: "TestGetLanguageByAlias_8", alias: "bsdmake", expectedLang: "Makefile", expectedOk: true},
{name: "TestGetLanguageByAlias_9", alias: "xhTmL", expectedLang: "HTML", expectedOk: true},
{name: "TestGetLanguageByAlias_10", alias: "python", expectedLang: "Python", expectedOk: true},
}
2016-07-18 16:20:12 +02:00
2017-05-29 10:05:16 +02:00
for _, test := range tests {
lang, ok := GetLanguageByAlias(test.alias)
2017-05-31 12:07:46 +02:00
assert.Equal(s.T(), test.expectedLang, lang, fmt.Sprintf("%v: lang = %v, expected: %v", test.name, lang, test.expectedLang))
assert.Equal(s.T(), test.expectedOk, ok, fmt.Sprintf("%v: ok = %v, expected: %v", test.name, ok, test.expectedOk))
2017-05-29 10:05:16 +02:00
}
2016-07-18 16:20:12 +02:00
}
func (s *EnryTestSuite) TestLinguistCorpus() {
const filenamesDir = "filenames"
var cornerCases = map[string]bool{
"drop_stuff.sql": true, // https://github.com/src-d/enry/issues/194
// .es and .ice fail heuristics parsing, but do not fail any tests
}
var total, failed, ok, other int
var expected string
filepath.Walk(s.samplesDir, func(path string, f os.FileInfo, err error) error {
if f.IsDir() {
if f.Name() != filenamesDir {
expected, _ = data.LanguageByAlias(f.Name())
}
return nil
}
filename := filepath.Base(path)
content, _ := ioutil.ReadFile(path)
total++
obtained := GetLanguage(filename, content)
if obtained == OtherLanguage {
2017-06-12 13:42:20 +02:00
obtained = "Other"
other++
}
var status string
if expected == obtained {
status = "ok"
ok++
} else {
status = "failed"
failed++
}
if _, ok := cornerCases[filename]; ok {
s.T().Logf("\t\t[considered corner case] %s\texpected: %s\tobtained: %s\tstatus: %s\n", filename, expected, obtained, status)
} else {
2017-06-12 13:42:20 +02:00
assert.Equal(s.T(), expected, obtained, fmt.Sprintf("%s\texpected: %s\tobtained: %s\tstatus: %s\n", filename, expected, obtained, status))
}
return nil
})
s.T().Logf("\t\ttotal files: %d, ok: %d, failed: %d, other: %d\n", total, ok, failed, other)
}