an
/
go-pkg-xmlx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
go-pkg-xmlx/xmlx_test.go

401 lines
9.7 KiB
Go
Raw Permalink Normal View History

Switch to more liberay CC0 public domain license
2011-03-19 13:50:46 +00:00
// This work is subject to the CC0 1.0 Universal (CC0 1.0) Public Domain Dedication
// license. Its contents can be found at:
// http://creativecommons.org/publicdomain/zero/1.0/
Reworked README and xmlx/entitymap.go. Now includes conversion of named xml entities instead of just numeric ones. Eg: π and π to a unicode character and vice-versa. It also came to my attention that lack of a proper opensource license might put people off of using this package. I therefor added a 1-clause BSD license under which this code now falls.
2010-09-26 20:59:14 +00:00
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
package xmlx
Fixed a needless conversion to byte and string. Added comment to show the purpose. Added testcase.
2015-01-05 19:54:53 +00:00
import (
"encoding/xml"
added SelectNodesDirect function
2015-02-03 21:30:24 +00:00
"testing"
Fixed a needless conversion to byte and string. Added comment to show the purpose. Added testcase.
2015-01-05 19:54:53 +00:00
)
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
Added *document.LoadUri() method to load documents from the web (over http)
2009-11-25 01:50:06 +00:00
func TestLoadLocal(t *testing.T) {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
doc := New()
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
Remove CharsetFunc as field for Document type and instead supply it as a parameter for all Document.LoadXXX methods. There should be no need to store the function pointer in the Document struct.
2012-02-29 10:21:35 +00:00
if err := doc.LoadFile("test.xml", nil); err != nil {
gofix error
2011-11-02 15:50:45 +00:00
t.Error(err.Error())
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
return
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
}
if len(doc.Root.Children) == 0 {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("Root node has no children.")
return
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
}
}
Add wildcard support for node names. Allows selecting of all child nodes with a specific namespace (See issue #4).
2011-07-13 03:17:00 +00:00
func TestWildcard(t *testing.T) {
doc := New()
Remove CharsetFunc as field for Document type and instead supply it as a parameter for all Document.LoadXXX methods. There should be no need to store the function pointer in the Document struct.
2012-02-29 10:21:35 +00:00
if err := doc.LoadFile("test2.xml", nil); err != nil {
gofix error
2011-11-02 15:50:45 +00:00
t.Error(err.Error())
Add wildcard support for node names. Allows selecting of all child nodes with a specific namespace (See issue #4).
2011-07-13 03:17:00 +00:00
return
}
Rolled back mistaken bugfix
2014-01-29 21:27:00 +00:00
list := doc.SelectNodes("ns", "*")
Reworked the tests just a bit
2013-10-25 21:31:10 +00:00
if len(list) != 1 {
t.Errorf("Wrong number of child elements. Expected 1, got %d.", len(list))
Making the recursive SelectNodes a separate function and adding tests
2012-07-29 19:28:03 +00:00
return
}
}
func TestWildcardRecursive(t *testing.T) {
doc := New()
if err := doc.LoadFile("test2.xml", nil); err != nil {
t.Error(err.Error())
return
}
list := doc.SelectNodesRecursive("ns", "*")
if len(list) != 7 {
t.Errorf("Wrong number of child elements. Expected 7, got %d.", len(list))
Add wildcard support for node names. Allows selecting of all child nodes with a specific namespace (See issue #4).
2011-07-13 03:17:00 +00:00
return
}
}
func _TestLoadRemote(t *testing.T) {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
doc := New()
Added *document.LoadUri() method to load documents from the web (over http)
2009-11-25 01:50:06 +00:00
Remove CharsetFunc as field for Document type and instead supply it as a parameter for all Document.LoadXXX methods. There should be no need to store the function pointer in the Document struct.
2012-02-29 10:21:35 +00:00
if err := doc.LoadUri("http://blog.golang.org/feeds/posts/default", nil); err != nil {
gofix error
2011-11-02 15:50:45 +00:00
t.Error(err.Error())
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
return
Added *document.LoadUri() method to load documents from the web (over http)
2009-11-25 01:50:06 +00:00
}
if len(doc.Root.Children) == 0 {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("Root node has no children.")
return
Added *document.LoadUri() method to load documents from the web (over http)
2009-11-25 01:50:06 +00:00
}
}
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
func TestSave(t *testing.T) {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
doc := New()
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
Remove CharsetFunc as field for Document type and instead supply it as a parameter for all Document.LoadXXX methods. There should be no need to store the function pointer in the Document struct.
2012-02-29 10:21:35 +00:00
if err := doc.LoadFile("test.xml", nil); err != nil {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("LoadFile(): %s", err)
return
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
}
Adds optional indented output for the Node.String() and Node.Bytes() methods. Set the global `IndentPrefix` var to a tab or spaces to enable indented output. The existing API has not been changed. This addresses issue #7.
2012-11-25 22:46:56 +00:00
IndentPrefix = "\t"
Updated code to use more recent go idioms. Yielding more compact xml processing. Hopefully better performant.
2010-05-26 00:24:44 +00:00
if err := doc.SaveFile("test1.xml"); err != nil {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("SaveFile(): %s", err)
return
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
}
}
func TestNodeSearch(t *testing.T) {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
doc := New()
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
Remove CharsetFunc as field for Document type and instead supply it as a parameter for all Document.LoadXXX methods. There should be no need to store the function pointer in the Document struct.
2012-02-29 10:21:35 +00:00
if err := doc.LoadFile("test1.xml", nil); err != nil {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("LoadFile(): %s", err)
return
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
}
Updated code to use more recent go idioms. Yielding more compact xml processing. Hopefully better performant.
2010-05-26 00:24:44 +00:00
if node := doc.SelectNode("", "item"); node == nil {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("SelectNode(): No node found.")
return
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
}
Rolled back mistaken bugfix
2014-01-29 21:27:00 +00:00
nodes := doc.SelectNodes("", "item")
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
if len(nodes) == 0 {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("SelectNodes(): no nodes found.")
return
new file: README new file: src/Makefile new file: src/document.go new file: src/io.go new file: src/node.go new file: src/test.xml new file: src/xmlx_test.go
2009-11-23 04:16:27 +00:00
}
Fixed inadvertent recursion in SelectNodes. Updated/add tests for this behavior
2013-10-25 20:44:32 +00:00
ch := doc.SelectNode("", "channel")
Rolled back mistaken bugfix
2014-01-29 21:27:00 +00:00
// Test that SelectNodes properly selects multiple nodes
Reworked the tests just a bit
2013-10-25 21:31:10 +00:00
links := ch.SelectNodes("", "link")
if len(links) != 8 {
Rolled back mistaken bugfix
2014-01-29 21:27:00 +00:00
t.Errorf("SelectNodes(): Expected 8, Got %d", len(links))
Fixed inadvertent recursion in SelectNodes. Updated/add tests for this behavior
2013-10-25 20:44:32 +00:00
return
}
}
Added *node.Unmarshal() method
2009-12-02 18:38:35 +00:00
type Image struct {
Run gofmt and fix the Unmarshal test.
2012-02-29 10:42:39 +00:00
Title string `xml:"title"`
Url string `xml:"url"`
Link string `xml:"link"`
Description string `xml:"description"`
Width int `xml:"width"`
Height int `xml:"height"`
Added *node.Unmarshal() method
2009-12-02 18:38:35 +00:00
}
func TestUnmarshal(t *testing.T) {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
doc := New()
Remove CharsetFunc as field for Document type and instead supply it as a parameter for all Document.LoadXXX methods. There should be no need to store the function pointer in the Document struct.
2012-02-29 10:21:35 +00:00
err := doc.LoadFile("test1.xml", nil)
Added *node.Unmarshal() method
2009-12-02 18:38:35 +00:00
if err != nil {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("LoadFile(): %s", err)
return
Added *node.Unmarshal() method
2009-12-02 18:38:35 +00:00
}
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
node := doc.SelectNode("", "image")
Added *node.Unmarshal() method
2009-12-02 18:38:35 +00:00
if node == nil {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("SelectNode(): No node found.")
return
Added *node.Unmarshal() method
2009-12-02 18:38:35 +00:00
}
Run gofmt and fix the Unmarshal test.
2012-02-29 10:42:39 +00:00
var img Image
Refactored some code to be faster and a little more elegant.
2010-08-22 03:07:38 +00:00
if err = node.Unmarshal(&img); err != nil {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("Unmarshal(): %s", err)
return
Added *node.Unmarshal() method
2009-12-02 18:38:35 +00:00
}
if img.Title != "WriteTheWeb" {
fixed some slice copy routines
2010-05-06 03:36:48 +00:00
t.Errorf("Image.Title has incorrect value. Got '%s', expected 'WriteTheWeb'.", img.Title)
return
Added *node.Unmarshal() method
2009-12-02 18:38:35 +00:00
}
}
fix: String() handled ns prefixes incorrectly Previously parsing a document like this: <root xmlns:foo='http:/example.org/foo' foo:bar='1'/> Would produce an incorrect document when turning it back into a string: <root xmlns:foo='http:/example.org/foo' http:/example.org/foo:bar='1'/> This patch fixes this by implementing a proper lookup for namespace prefixes.
2013-09-03 22:35:22 +00:00
Implement text nodes / content escaping This removes the support for re-indenting XML, but allows to faithfully parse / stringify a document. Beautification could be added again later on if desired.
2013-09-10 21:18:13 +00:00
func TestStringNamespaces(t *testing.T) {
fix: String() handled ns prefixes incorrectly Previously parsing a document like this: <root xmlns:foo='http:/example.org/foo' foo:bar='1'/> Would produce an incorrect document when turning it back into a string: <root xmlns:foo='http:/example.org/foo' http:/example.org/foo:bar='1'/> This patch fixes this by implementing a proper lookup for namespace prefixes.
2013-09-03 22:35:22 +00:00
doc := New()
err := doc.LoadFile("test3.xml", nil)
if err != nil {
t.Errorf("LoadFile(): %s", err)
return
}
expected := `<root xmlns:foo="http:/example.org/foo">
Implement text nodes / content escaping This removes the support for re-indenting XML, but allows to faithfully parse / stringify a document. Beautification could be added again later on if desired.
2013-09-10 21:18:13 +00:00
<child foo:bar="1">
<grandchild xmlns:foo="">
<great-grandchild bar="2">&#xA; </great-grandchild>
</grandchild>
</child>
fix: String() handled ns prefixes incorrectly Previously parsing a document like this: <root xmlns:foo='http:/example.org/foo' foo:bar='1'/> Would produce an incorrect document when turning it back into a string: <root xmlns:foo='http:/example.org/foo' http:/example.org/foo:bar='1'/> This patch fixes this by implementing a proper lookup for namespace prefixes.
2013-09-03 22:35:22 +00:00
</root>
`
if got := doc.Root.String(); got != expected {
t.Fatalf("expected: %s\ngot: %s\n", expected, got)
}
}
Implement text nodes / content escaping This removes the support for re-indenting XML, but allows to faithfully parse / stringify a document. Beautification could be added again later on if desired.
2013-09-10 21:18:13 +00:00
func TestStringEscaping(t *testing.T) {
doc := New()
err := doc.LoadFile("test4.xml", nil)
if err != nil {
t.Errorf("LoadFile(): %s", err)
return
}
expected := `<body>  &lt;https://example.com/file/fm/SU0vRk0xLzIwMTMwOTEwLzA1MDA0MS5ybXdhdGVzdEByZXV0ZXJzLmNvbTEzNzg4NDU1OTk4OTA/Screen%20Shot%202013-09-10%20at%2021.33.54.png&gt; File Attachment:-Screen Shot 2013-09-10 at 21.33.54.png  </body>
`
if got := doc.Root.String(); got != expected {
t.Fatalf("expected: %s\ngot: %s\n", expected, got)
}
}
Fix the value fetcher to deal with NT_TEXT nodes. Due to the introduction of the NT_TEXT nodes, all of the simple value fetchers broke. To fix, make a GetValue function fetch all NT_TEXT nodes for NT_ELEMENT nodes, and put them together, using string.TrimSpace to remove extra, maintaining compatibility. Then change the .S function to use GetValue to fetch a node's content. Also, make all the other value fetchers use .S to get a string representation to convert, centralizing the implementation. Use .S to centralize the handling of a not found node. Also, implement a series of tests to verify all the fetcher's functionality. The test is based on the test produced by Tim Jurcka.
2013-09-30 05:09:47 +00:00
func TestElementNodeValueFetch(t *testing.T) {
data := `<car><color>
r<cool />
ed</color><brand>BMW</brand><price>50
<cheap />.25</price><count>6
<small />2
</count><available>
Tr
<found />
ue</available></car>`
doc := New()
if err := doc.LoadString(data, nil); nil != err {
t.Fatalf("LoadString(): %s", err)
}
carN := doc.SelectNode("", "car")
if v := carN.S("", "brand"); v != "BMW" {
t.Errorf("Failed to get brand as string, got: '%s', wanted: 'BMW'", v)
}
if v := carN.S("", "color"); v != "red" {
t.Errorf("Failed to get color as string, got: '%s', wanted: 'red'", v)
}
if v := carN.I("", "count"); v != 62 {
t.Errorf("Failed to get count using I, got: %v, wanted: 62", v)
}
if v := carN.I8("", "count"); v != 62 {
t.Errorf("Failed to get count using I8, got: %v, wanted: 62", v)
}
if v := carN.I16("", "count"); v != 62 {
t.Errorf("Failed to get count using I16, got: %v, wanted: 62", v)
}
if v := carN.I32("", "count"); v != 62 {
t.Errorf("Failed to get count using I32, got: %v, wanted: 62", v)
}
if v := carN.I64("", "count"); v != 62 {
t.Errorf("Failed to get count using I64, got: %v, wanted: 62", v)
}
if v := carN.U("", "count"); v != 62 {
t.Errorf("Failed to get count using U, got: %v, wanted: 62", v)
}
if v := carN.U8("", "count"); v != 62 {
t.Errorf("Failed to get count using U8, got: %v, wanted: 62", v)
}
if v := carN.U16("", "count"); v != 62 {
t.Errorf("Failed to get count using U16, got: %v, wanted: 62", v)
}
if v := carN.U32("", "count"); v != 62 {
t.Errorf("Failed to get count using U32, got: %v, wanted: 62", v)
}
if v := carN.U64("", "count"); v != 62 {
t.Errorf("Failed to get count using U64, got: %v, wanted: 62", v)
}
if v := carN.F32("", "price"); v != 50.25 {
t.Errorf("Failed to get price using F32, got: %v, wanted: 50.25", v)
}
if v := carN.F64("", "price"); v != 50.25 {
t.Errorf("Failed to get price using F64, got: %v, wanted: 50.25", v)
}
if v := carN.B("", "available"); v != true {
t.Errorf("Failed to get availability using B, got: %v, wanted: true", v)
}
}
Added Node.SetValue(string) method. SetValue replaces all children of the current node with a single NT_TEXT node and sets its Value to the given parameter. This lets us replace node values before writing out the document.
2015-01-05 16:13:46 +00:00
Fixed a needless conversion to byte and string. Added comment to show the purpose. Added testcase.
2015-01-05 19:54:53 +00:00
// node.SetValue(x); x == node.GetValue
added SelectNodesDirect function
2015-02-03 21:30:24 +00:00
func TestElementNodeValueFetchAndSetIdentity(t *testing.T) {
Fixed a needless conversion to byte and string. Added comment to show the purpose. Added testcase.
2015-01-05 19:54:53 +00:00
// Setup: <root><text>xyzzy</text></root>
added SelectNodesDirect function
2015-02-03 21:30:24 +00:00
// The xmlx parser creates a nameless NT_TEXT node containing the value 'xyzzy'
Fixed a needless conversion to byte and string. Added comment to show the purpose. Added testcase.
2015-01-05 19:54:53 +00:00
rootN := NewNode(NT_ROOT)
rootN.Name = xml.Name{Space: "", Local: "root"}
textN := NewNode(NT_ELEMENT)
textN.Name = xml.Name{Space: "", Local: "text"}
namelessN := NewNode(NT_TEXT)
namelessN.Value = "xyzzy"
rootN.AddChild(textN)
textN.AddChild(namelessN)
targetN := rootN.SelectNode("", "text") // selects textN
added SelectNodesDirect function
2015-02-03 21:30:24 +00:00
if targetN != textN {
Fixed a needless conversion to byte and string. Added comment to show the purpose. Added testcase.
2015-01-05 19:54:53 +00:00
t.Errorf("Failed to get the correct textN, got %#v", targetN)
}
added SelectNodesDirect function
2015-02-03 21:30:24 +00:00
Fixed a needless conversion to byte and string. Added comment to show the purpose. Added testcase.
2015-01-05 19:54:53 +00:00
// targetN.Value is empty (as the value lives in the childNode)
added SelectNodesDirect function
2015-02-03 21:30:24 +00:00
if targetN.Value != "" {
Fixed a needless conversion to byte and string. Added comment to show the purpose. Added testcase.
2015-01-05 19:54:53 +00:00
t.Errorf("Failed to prepare correctly, TargetN.Value is not empty, it contains %#v", targetN.Value)
}
// Test correct retrieval
if v := rootN.S("", "text"); v != "xyzzy" {
t.Errorf("Failed to get value as string, got: '%s', wanted: 'xyzzy'", v)
}
// Set the value of the nameless child
targetN.SetValue("plugh")
// Test correct retrieval
if v := rootN.S("", "text"); v != "plugh" {
t.Errorf("Failed to get value as string, got: '%s', wanted: 'plugh'", v)
}
}
// Test as it could be used to read in a XML file, change some values and write it out again.
// For example, a HTML/XML proxy service.
Added Node.SetValue(string) method. SetValue replaces all children of the current node with a single NT_TEXT node and sets its Value to the given parameter. This lets us replace node values before writing out the document.
2015-01-05 16:13:46 +00:00
func TestElementNodeValueFetchAndSet(t *testing.T) {
IndentPrefix = ""
data := `<car><color>
r<cool />
ed</color><brand>BM<awesome />W</brand><price>50
<cheap />.25</price><count>6
<small />2
</count><available>
Tr
<found />
ue</available></car>`
doc := New()
if err := doc.LoadString(data, nil); nil != err {
t.Fatalf("LoadString(): %s", err)
}
carN := doc.SelectNode("", "car")
if carN == nil {
t.Fatalf("Failed to get the car, got nil, wanted Node{car}")
}
added SelectNodesDirect function
2015-02-03 21:30:24 +00:00
Added Node.SetValue(string) method. SetValue replaces all children of the current node with a single NT_TEXT node and sets its Value to the given parameter. This lets us replace node values before writing out the document.
2015-01-05 16:13:46 +00:00
colorNode := carN.SelectNode("", "color")
if colorNode == nil {
t.Fatalf("Failed to get the color, got nil, wanted Node{color}")
}
colorVal := colorNode.GetValue()
if colorVal != "red" {
t.Fatalf("Failed to get the color, got %v, wanted 'red'", colorVal)
}
colorNode.SetValue("blue")
expected := `<car><color>blue</color><brand>BM<awesome />W</brand><price>50
<cheap />.25</price><count>6
<small />2
</count><available>
Tr
<found />
ue</available></car>`
if got := doc.Root.String(); got != expected {
t.Fatalf("expected: \n%s\ngot: \n%s\n", expected, got)
}
// now set the brand
brandNode := carN.SelectNode("", "brand")
if brandNode == nil {
t.Fatalf("Failed to get the color, got nil, wanted Node{brand}")
}
brandVal := brandNode.GetValue()
if brandVal != "BMW" {
t.Fatalf("Failed to get the brand, got %v, wanted 'BMW'", brandVal)
}
brandNode.SetValue("Trabant")
// Notice, we lose the <awesome /> tag in BMW, that's intentional
expected = `<car><color>blue</color><brand>Trabant</brand><price>50
<cheap />.25</price><count>6
<small />2
</count><available>
Tr
<found />
ue</available></car>`
if got := doc.Root.String(); got != expected {
t.Fatalf("expected: \n%s\ngot: \n%s\n", expected, got)
}
}
added SelectNodesDirect function
2015-02-03 21:30:24 +00:00
func TestSelectNodesDirect(t *testing.T) {
data := `<root><wrapper><hidden></hidden>
<hidden></hidden></wrapper></root>`
doc := New()
if err := doc.LoadString(data, nil); nil != err {
t.Fatalf("LoadString(): %s", err)
}
root := doc.SelectNode("*", "root")
if root == nil {
t.Fatalf("Failed to get root, got nil, wanted Node{root}")
}
nodes := root.SelectNodesDirect("*", "hidden")
if len(nodes) != 0 {
t.Errorf("SelectDirectNodes should not select children of children.")
}
wrapper := root.SelectNode("*", "wrapper")
if wrapper == nil {
t.Fatalf("Failed to get wrapper, got nil, wanted Node{wrapper}")
}
nodes = wrapper.SelectNodesDirect("*", "hidden")
if len(nodes) != 2 {
t.Errorf("Unexcepted hidden nodes found. Expected: 2, Got: %d", len(nodes))
}
}