XML parse VBA excel (function trip, & MSXML2.DOMDocument)
I need to parse hundreds of XML files having all the same structure as follows:
<?xml version="1.0" encoding="UTF-8"?> <Concepts> <ConceptModel name="food"> <Filters> <Filter type="CC"/> </Filters> <Queries> <Query lang="EN">(cheese, bread, wine)</Query> <Query lang="DE">(Käse, Brot, Wein)</Query> <Query lang="FR">(fromaige, pain, vin)</Query> </Queries> </ConceptModel> </Concepts>
I have read several articles and posts in internet like below but I could not come up with a solution:
So far I am doing:
Dim oXml As MSXML2.DOMDocument Set oXml = New MSXML2.DOMDocument oXml.LoadXML ("C:\folder\folder\name.xml") Dim Queries As IXMLDOMNodeList Dim Query As IXMLDOMNode ThisWorkbook.Sheets(3).Cells(i, 1) = "before loop" Set Queries = oXml.SelectNodes("/concepts/Queries") MsgBox "how many Queries " & Queries.Length For Each Query In Queries ThisWorkbook.Sheets(3).Cells(i, 1) = "Works" ThisWorkbook.Sheets(3).Cells(i, 2) = Query.SelectNodes("Query").iTem(0).Text i = i + 1 Next
This code seems to be understood by VBA but it does not read the contents. The loop does not get read, meaning (I guess) that Queries is not looped at all. Which is confirmed by the fact that the
Msgbox "how many queries" gives 0 as result. But actually there are three queries. Could someone give me a hand?
As second issue I would like to ask if
Dim oXml As MSXML2.DOMDocument
would be the same as
Dim oXml As MSXML2.DOMDocument60
Since I checked in tools/references "Microsof XML, v6.0"
I thought that the queries having a tag might cause a problem. and I added the follwoing lines:
Dim childs As IXMLDOMNodeList Set childs = oXml.SelectNodes("/concepts") MsgBox "childs " & childs.Length
which also gives 0 as result. I would expect 3, since concepts has three children, namely ConceptModel, Filter and Queries. So, I am even more puzzled.
As close as possible to your OP
I 'd draw your attention to several errors or misunderstandings:
What is then the difference between .LoadXML ("C:\folder\folder\name.xml") and .Load ("C:\folder\folder\name.xml") ?
Load expects a file path and then loads the file content into the oXML object.
LoadXML doesn't expect a file parameter, but its actual XML text content that has to be a well formed string.
XML distinguishes between lower and upper case, therefore nodes need to be addressed by their exact literal names: the
<Query>node wouldn't be identified by "query", "ConceptModel" isn't the same as "conceptmodel".
As second issue I would like to ask if
Dim oXml As MSXML2.DOMDocumentwould be the same as
Dim oXml As MSXML2.DOMDocument60, since I checked in tools/references "Microsof XML, v6.0"?
No, it isn't. - Please note that the former declaration would load version 3.0 by default. However it's absolutely preferrable to get the version 6.0 (any other versions are obsolete nowadays!)
As you are using so called early binding (referencing "Microsoft XML, v6.0"), I'll do the same but am referring to the current version 6.0:
Dim oXml As MSXML2.DOMDocument60 ' declare the xml doc object Set oXml = New MSXML2.DOMDocument60 ' set an instance of it to memory
misunderstanding some XPath expressions
A starting slash "/" in the XPath expression always refers to the DocumentElement (
you can add
.DocumentElement to your document object instead. A starting double slash "//xyz" would find any "xyz" node if existant.
returns the same childNodes number (here: 3) as
oXml.DocumentElement.SelectNodes("//Query").Length ' or oXml.SelectSingleNode("//Queries").ChildNodes.Length ' or even oXml.SelectNodes("/*/*/*/Query").Length`.
Code example with reference to XML version 6.0
Of course you'd have to loop over several xml files, the example only uses one (starting in row 2).
Just for the case of not well formed xml files I added a detailled error Routine that enables you to identify the presumed error location.
LoadXML both return a boolean value (True if loaded correctly, False if not).
Sub xmlTest() Dim ws As Worksheet: Set ws = ThisWorkbook.Sheets(3) Dim oXml As MSXML2.DOMDocument60 Set oXml = New MSXML2.DOMDocument60 With oXml .validateOnParse = True .setProperty "SelectionLanguage", "XPath" ' necessary in version 3.0, possibly redundant here .async = False If Not .Load(ThisWorkbook.Path & "\xml\" & "name.xml") Then Dim xPE As Object ' Set xPE = CreateObject("MSXML2.IXMLDOMParseError") Dim strErrText As String Set xPE = .parseError With xPE strErrText = "Load error " & .ErrorCode & " xml file " & vbCrLf & _ Replace(.URL, "file:///", "") & vbCrLf & vbCrLf & _ xPE.reason & _ "Source Text: " & .srcText & vbCrLf & vbCrLf & _ "Line No.: " & .Line & vbCrLf & _ "Line Pos.: " & .linepos & vbCrLf & _ "File Pos.: " & .filepos & vbCrLf & vbCrLf End With MsgBox strErrText, vbExclamation Set xPE = Nothing Exit Sub End If ' Debug.Print "|" & oXml.XML & "|" Dim Queries As IXMLDOMNodeList, Query As IXMLDOMNode Dim Searched As String Dim i&, ii& i = 2 ' start row ' start XPath Searched = "ConceptModel/Queries/Query" ' search string Set Queries = oXml.DocumentElement.SelectNodes(Searched) ' XPath ' ws.Cells(i, 1) = IIf(Queries.Length = 0, "No items", Queries.Length & " items") ii = 1 For Each Query In Queries ii = ii + 1 ws.Cells(i, ii) = Query.Text Next End With End Sub
You also might be interested in an example how to list all child nodes via XMLDOM and to obtain attribute names from XML using VBA.
I include a further hint due to later comment (thanks to @barrowc )
"A further issue with using MSXML, v3.0 is that the default selection language is XSLPatterns instead of XPath. Details on some of the differences between MSXML versions are here and the differences between the two selection languages are discussed here."
In the current MSXML2 version 6.0 XPath 1.0 is fully supported. So it seems XSL Patterns have been implemented by Microsoft in earlier days, basically it can be regarded as a simplified subset of XPath expressions before W3C standardisation of XPath.
MSXML2 Version 3.0 allows the integration of XPath 1.0 at least by explicit selection language setting:
oXML.setProperty "SelectionLanguage", "XPath" ' oXML being the DOMDocument object as used in original post