Merge pull request #227 from bzz/ci-fix-cgo-tests

CI: fix cgo profiles

.travis.yml

@@ -1,7 +1,5 @@
 dist: trusty
-
 language: go
-
 go:
   - '1.12.x'
   - '1.11.x'
@@ -10,17 +8,13 @@ env:
     - GO_VERSION_FOR_JVM='1.11.x'
     - CGO_ENABLED=0
     - GO111MODULE=on
+    - ONIGURUMA_VERSION='6.9.1'
   matrix:
     - ONIGURUMA=0
     - ONIGURUMA=1
 matrix:
   fast_finish: true
 
-addons:
-  apt:
-    packages:
-      - libonig-dev
-
 stages:
   - name: test
   - name: release
@@ -32,8 +26,14 @@ stage: test
 install:
   - >
     if [[ "${ONIGURUMA}" -gt 0 ]]; then
-      export CGO_ENABLED=1;
-      export GO_TAGS='oniguruma';
+      export CGO_ENABLED=1
+      export GO_TAGS='oniguruma'
+      # install oniguruma manually as trusty has only ancient 5.x
+      sudo apt-get install -y dpkg # dpkg >= 1.17.5ubuntu5.8 fixes https://bugs.launchpad.net/ubuntu/+source/dpkg/+bug/1730627
+      wget "http://archive.ubuntu.com/ubuntu/pool/universe/libo/libonig/libonig5_${ONIGURUMA_VERSION}-1_amd64.deb"
+      sudo dpkg -i "libonig5_${ONIGURUMA_VERSION}-1_amd64.deb"
+      wget "http://archive.ubuntu.com/ubuntu/pool/universe/libo/libonig/libonig-dev_${ONIGURUMA_VERSION}-1_amd64.deb"
+      sudo dpkg -i "libonig-dev_${ONIGURUMA_VERSION}-1_amd64.deb"
     fi;
 script:
   - make test-coverage

go.mod

@@ -3,7 +3,7 @@ module github.com/src-d/enry/v2
 go 1.12
 
 require (
-	github.com/src-d/go-oniguruma v1.0.0
+	github.com/src-d/go-oniguruma v1.1.0
 	github.com/stretchr/testify v1.3.0
 	github.com/toqueteos/trie v1.0.0 // indirect
 	gopkg.in/toqueteos/substring.v1 v1.0.2

go.sum

@@ -2,8 +2,8 @@ github.com/davecgh/go-spew v1.1.0 h1:ZDRjVQ15GmhC3fiQ8ni8+OwkZQO4DARzQgrnXU1Liz8
 github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
 github.com/pmezard/go-difflib v1.0.0 h1:4DBwDE0NGyQoBHbLQYPwSUPoCMWR5BEzIk/f1lZbAQM=
 github.com/pmezard/go-difflib v1.0.0/go.mod h1:iKH77koFhYxTK1pcRnkKkqfTogsbg7gZNVY4sRDYZ/4=
-github.com/src-d/go-oniguruma v1.0.0 h1:JDk5PUAjreGsGAKLsoDLNmrsaryjJ5RqT3h+Si6aw/E=
-github.com/src-d/go-oniguruma v1.0.0/go.mod h1:chVbff8kcVtmrhxtZ3yBVLLquXbzCS6DrxQaAK/CeqM=
+github.com/src-d/go-oniguruma v1.1.0 h1:EG+Nm5n2JqWUaCjtM0NtutPxU7ZN5Tp50GWrrV8bTww=
+github.com/src-d/go-oniguruma v1.1.0/go.mod h1:chVbff8kcVtmrhxtZ3yBVLLquXbzCS6DrxQaAK/CeqM=
 github.com/stretchr/objx v0.1.0/go.mod h1:HFkY916IF+rwdDfMAkV7OtwuqBVzrE8GR6GFx+wExME=
 github.com/stretchr/testify v1.3.0 h1:TivCn/peBQ7UY8ooIcPgZFpTNSz0Q2U6UrFlUfqbe0Q=
 github.com/stretchr/testify v1.3.0/go.mod h1:M5WIy9Dh21IEIfnGCwXGc5bZfKNJtfHm1UVUgZn+9EI=

internal/tokenizer/tokenize_c.go

@@ -2,7 +2,7 @@
 
 package tokenizer
 
-import "gopkg.in/src-d/enry.v1/internal/tokenizer/flex"
+import "github.com/src-d/enry/v2/internal/tokenizer/flex"
 
 // Tokenize returns lexical tokens from content. The tokens returned match what
 // the Linguist library returns. At most the first ByteLimit bytes of content are tokenized.

internal/tokenizer/tokenize_test.go

@@ -115,6 +115,45 @@ func TestTokenize(t *testing.T) {
 	}
 }
 
+func TestTokenizerLatin1AsUtf8(t *testing.T) {
+	content := []byte("th\xe5 filling") // `th<74> filling`
+	t.Logf("%v - %q", content, string(content))
+	tokens := Tokenize(content)
+	for i, token := range tokens {
+		t.Logf("token %d, %s", i+1, token)
+	}
+	require.Equal(t, 3, len(tokens))
+}
+
+func TestRegexpOnInvalidUtf8(t *testing.T) {
+	origContent := []struct {
+		text   string
+		tokens []string
+	}{
+		{"th\xe0 filling", []string{"th", "filling"}},   // `th<74> filling`
+		{"th\u0100 filling", []string{"th", "filling"}}, // `thĀ filling`
+		{"привет, как дела?", []string{}},               // empty, no ASCII tokens
+	}
+	re := reRegularToken
+
+	for _, content := range origContent {
+		t.Run("", func(t *testing.T) {
+			t.Logf("%v - %q", content, content.text)
+			input := []byte(content.text)
+			tokens := re.FindAll(input, -1)
+			require.Equal(t, len(content.tokens), len(tokens))
+
+			newContent := re.ReplaceAll(input, []byte(` `))
+			t.Logf("content:%q, tokens:[", newContent)
+			for i, token := range tokens {
+				t.Logf("\t%q,", string(token))
+				require.Equal(t, content.tokens[i], string(token))
+			}
+			t.Logf(" ]\n")
+		})
+	}
+}
+
 func BenchmarkTokenizer_BaselineCopy(b *testing.B) {
 	b.ReportAllocs()
 	for i := 0; i < b.N; i++ {

regex/oniguruma.go

@@ -9,7 +9,7 @@ import (
 type EnryRegexp = *rubex.Regexp
 
 func MustCompile(str string) EnryRegexp {
-	return rubex.MustCompile(str)
+	return rubex.MustCompileASCII(str)
 }
 
 func QuoteMeta(s string) string {