fix: improve host name detection

Author: ldez

notwords.go

@@ -4,12 +4,17 @@ import (
 	"bytes"
 	"regexp"
 	"strings"
+	"unicode"
 )
 
 var (
-	reEmail     = regexp.MustCompile(`[a-zA-Z0-9_.%+-]+@[a-zA-Z0-9-.]+\.[a-zA-Z]{2,6}[^a-zA-Z]`)
-	reHost      = regexp.MustCompile(`[a-zA-Z0-9-.]+\.[a-zA-Z]+`)
-	reBackslash = regexp.MustCompile(`\\[a-z]`)
+	reEmail     = regexp.MustCompile(`[[:alnum:]_.%+-]+@[[:alnum:]-.]+\.[[:alpha:]]{2,6}[^[:alpha:]]`)
+	reBackslash = regexp.MustCompile(`\\[[:lower:]]`)
+
+	// reHost Host name regular expression.
+	// The length of any one label is limited between 1 and 63 octets. (https://www.ietf.org/rfc/rfc2181.txt)
+	// A TLD has at least 2 letters.
+	reHost = regexp.MustCompile(`([[:alnum:]-]+\.)+[[:alpha:]]{2,63}`)
 )
 
 // RemovePath attempts to strip away embedded file system paths, e.g.
@@ -62,14 +67,26 @@ func replaceWithBlanks(s string) string {
 	return strings.Repeat(" ", len(s))
 }
 
+// replaceHost same as replaceWithBlanks but if the string contains at least one uppercase letter returns the string.
+// Domain names are case-insensitive but browsers and DNS convert uppercase to lower case. (https://www.ietf.org/rfc/rfc4343.txt)
+func replaceHost(s string) string {
+	for _, r := range s {
+		if unicode.IsUpper(r) {
+			return s
+		}
+	}
+
+	return replaceWithBlanks(s)
+}
+
 // RemoveEmail remove email-like strings, e.g. "[email protected]", "[email protected]".
 func RemoveEmail(s string) string {
 	return reEmail.ReplaceAllStringFunc(s, replaceWithBlanks)
 }
 
 // RemoveHost removes host-like strings "foobar.com" "abc123.fo1231.biz".
 func RemoveHost(s string) string {
-	return reHost.ReplaceAllStringFunc(s, replaceWithBlanks)
+	return reHost.ReplaceAllStringFunc(s, replaceHost)
 }
 
 // RemoveBackslashEscapes removes characters that are preceded by a backslash.

notwords_test.go

@@ -14,8 +14,20 @@ func TestNotWords(t *testing.T) {
 		{word: "[/foo/bar] abc", want: "[        ] abc"},
 		{word: "/", want: "/"},
 		{word: "x [email protected] y", want: "x                   y"},
+		{word: "x fqdn.example.org. y", want: "x                 . y"},
 		{word: "x infinitie.net y", want: "x               y"},
-		{word: "(s.svc.GetObject(", want: "(               ("},
+		{word: "x infinitie.net ", want: "x               "},
+		{word: "x infinitie.net", want: "x              "},
+		{word: "x foo.example.com y", want: "x                 y"},
+		{word: "x foo.example.com ", want: "x                 "},
+		{word: "x foo.example.com", want: "x                "},
+		{word: "foo.example.com y", want: "                y"},
+		{word: "foo.example.com", want: "               "},
+		{word: "(s.svc.GetObject(", want: "(s.svc.GetObject("},
+		{word: "defer file.Close()", want: "defer file.Close()"},
+		{word: "defer file.c()", want: "defer file.c()"},
+		{word: "defer file.cl()", want: "defer        ()"},       // false negative
+		{word: "defer file.close()", want: "defer           ()"}, // false negative
 		{word: "\\nto", want: "  to"},
 	}