bugfix(41): fix XML format detection

pawelWritesCode · Sep 6, 2022 · b015def · b015def
1 parent 0b4735f
commit b015def
Show file tree

Hide file tree

Showing 2 changed files with 39 additions and 19 deletions.
diff --git a/pkg/format/format.go b/pkg/format/format.go
@@ -3,7 +3,6 @@ package format
 import (
 	"bytes"
 	"encoding/json"
-	"encoding/xml"
 	"strings"
 
 	"github.com/goccy/go-yaml"
@@ -59,15 +58,24 @@ func IsYAML(b []byte) bool {
 }
 
 // IsXML checks whether bytes are in XML format.
+// Function does not guarantee that standard xml.Unmarshal will work b, instead
+// it only looks for characteristics of XML formatted data.
 func IsXML(b []byte) bool {
-	var v any
-	err := xml.Unmarshal(b, &v)
-	if err == nil {
+	str := string(b)
+	idx := strings.Index(strings.TrimSpace(str), "<?xml version=")
+	if idx == 0 || idx == 1 {
 		return true
 	}
 
-	idx := strings.Index(strings.TrimSpace(string(b)), "<?xml version=")
-	return idx == 0
+	if !(strings.Contains(str, ">") && strings.Contains(str, "<")) {
+		return false
+	}
+
+	if strings.Count(str, "<") >= (strings.Count(str, "</") + strings.Count(str, "/>")) {
+		return true
+	}
+
+	return false
 }
 
 // IsHTML checks whether bytes are in HTML format.

diff --git a/pkg/format/format_test.go b/pkg/format/format_test.go
@@ -72,21 +72,9 @@ func TestIsXML(t *testing.T) {
 	}{
 		{name: "json", args: args{bytes: []byte(`{"name": "abc"}`)}, want: false},
 		{name: "plain text", args: args{bytes: []byte(`abcd efgh`)}, want: false},
-		{name: "xml", args: args{bytes: []byte(`<?xml version="1.0"?>
-<catalog>
-   <book id="bk101">
-      <author>Gambardella, Matthew</author>
-      <title>XML Developer's Guide</title>
-      <genre>Computer</genre>
-      <price>44.95</price>
-      <publish_date>2000-10-01</publish_date>
-      <description>An in-depth look at creating applications 
-      with XML.</description>
-   </book>
-</catalog>`)}, want: true},
 		{name: "yaml", args: args{bytes: []byte(`---
 name: "abc"`)}, want: false},
-		{name: "yaml", args: args{bytes: []byte(`<?xml version='1.0' encoding='us-ascii'?>
+		{name: "xml #1", args: args{bytes: []byte(`<?xml version='1.0' encoding='us-ascii'?>
 
 <!--  A SAMPLE set of slides  -->
 
@@ -111,6 +99,30 @@ name: "abc"`)}, want: false},
 
 </slideshow>
 `)}, want: true},
+		{name: "xml #2", args: args{bytes: []byte(`<?xml version="1.0"?>
+<catalog>
+   <book id="bk101">
+      <author>Gambardella, Matthew</author>
+      <title>XML Developer's Guide</title>
+      <genre>Computer</genre>
+      <price>44.95</price>
+      <publish_date>2000-10-01</publish_date>
+      <description>An in-depth look at creating applications 
+      with XML.</description>
+   </book>
+</catalog>`)}, want: true},
+		{name: "xml #3", args: args{bytes: []byte(`
+<catalog>
+   <book id="bk101">
+      <author>Gambardella, Matthew</author>
+      <title>XML Developer's Guide</title>
+      <genre>Computer</genre>
+      <price>44.95</price>
+      <publish_date>2000-10-01</publish_date>
+      <description>An in-depth look at creating applications 
+      with XML.</description>
+   </book>
+</catalog>`)}, want: true},
 	}
 	for _, tt := range tests {
 		t.Run(tt.name, func(t *testing.T) {