Add ASCII-only option, to mimic default RE2 behavior

.travis.yml

@@ -1,4 +1,9 @@
+dist: trusty
 language: go
+go:
+  - '1.11.x'
+  - '1.12.x'
+
 env:
   global:
     - LD_LIBRARY_PATH="/usr/local/lib":${LD_LIBRARY_PATH}
@@ -8,8 +13,11 @@ addons:
     packages:
       - libonig-dev
 
-jobs:
-  include:
-    - go: 1.11.x
-      script:
-        - go test -v --cover -race
+before_install:
+  - sudo apt-get install -y dpkg  # dpkg >= 1.17.5ubuntu5.8, which fixes https://bugs.launchpad.net/ubuntu/+source/dpkg/+bug/1730627- sudo dpkg -i libonig-dev_6.9.1-1_amd64.deb
+  - wget http://archive.ubuntu.com/ubuntu/pool/universe/libo/libonig/libonig5_6.9.1-1_amd64.deb
+  - sudo dpkg -i libonig5_6.9.1-1_amd64.deb
+  - wget http://archive.ubuntu.com/ubuntu/pool/universe/libo/libonig/libonig-dev_6.9.1-1_amd64.deb
+  - sudo dpkg -i libonig-dev_6.9.1-1_amd64.deb
+script:
+  - go test -v --cover -race

chelper.c

@@ -17,7 +17,8 @@ int NewOnigRegex( char *pattern, int pattern_length, int option,
     *error_info = (OnigErrorInfo *) malloc(sizeof(OnigErrorInfo));
     memset(*error_info, 0, sizeof(OnigErrorInfo));
 
-    *encoding = (void*)ONIG_ENCODING_UTF8;
+    OnigEncoding use_encs[] = { *encoding };
+    onig_initialize(use_encs, sizeof(use_encs)/sizeof(use_encs[0]));
 
     *error_buffer = (char*) malloc(ONIG_MAX_ERROR_MESSAGE_LEN * sizeof(char));
 

regex.go

@@ -48,13 +48,23 @@ type Regexp struct {
 }
 
 func NewRegexp(pattern string, option int) (re *Regexp, err error) {
-	re = &Regexp{pattern: pattern}
-	patternCharPtr := C.CString(pattern)
-	defer C.free(unsafe.Pointer(patternCharPtr))
+	re = &Regexp{pattern: pattern, encoding: C.ONIG_ENCODING_UTF8}
+	return initRegexp(re, option)
+}
+
+// NewRegexpASCII is equivalent of NewRegexp but matching only ASCII.
+func NewRegexpASCII(pattern string, option int) (re *Regexp, err error) {
+	re = &Regexp{pattern: pattern, encoding: C.ONIG_ENCODING_ASCII}
+	return initRegexp(re, option)
+}
 
+func initRegexp(re *Regexp, option int) (*Regexp, error) {
+	var err error
+	patternCharPtr := C.CString(re.pattern)
+	defer C.free(unsafe.Pointer(patternCharPtr))
 	mutex.Lock()
 	defer mutex.Unlock()
-	error_code := C.NewOnigRegex(patternCharPtr, C.int(len(pattern)), C.int(option), &re.regex, &re.region, &re.encoding, &re.errorInfo, &re.errorBuf)
+	error_code := C.NewOnigRegex(patternCharPtr, C.int(len(re.pattern)), C.int(option), &re.regex, &re.region, &re.encoding, &re.errorInfo, &re.errorBuf)
 	if error_code != C.ONIG_NORMAL {
 		err = errors.New(C.GoString(re.errorBuf))
 	} else {
@@ -95,6 +105,15 @@ func MustCompileWithOption(str string, option int) *Regexp {
 	return regexp
 }
 
+// MustCompileASCII equivalent of MustCompile but with char matching only ASCII.
+func MustCompileASCII(str string) *Regexp {
+	regexp, error := NewRegexpASCII(str, ONIG_OPTION_DEFAULT)
+	if error != nil {
+		panic("regexp: compiling " + str + ": " + error.Error())
+	}
+	return regexp
+}
+
 func (re *Regexp) Free() {
 	mutex.Lock()
 	if re.regex != nil {