Hi,

I installed the latest Xcode with Swift 3.0, and AEXML generates numerous errors. Would appreciate a new release which supports.

Thanks much, Brian.

line 223 of AEXML.swift has:

          if let parser = NSXMLParser(data: xmlData) {

Though, it appears as of the most recent Xcode, the following is required:

if let parser = NSXMLParser(data: xmlData) as NSXMLParser! {

Can you please validate my change is correct, and why you are not seeing this issue.

Thanks, Brian

It seems that AEXML's xmlstring property does not escape characters like "greater than", "less than" and "ampersand". This in turn leads to invalid XML being returned if an element's value contains such a character. Alternatively escaping could also be done when adding an element. Nor sure which is the better way...

Let's say I've created an AEXMLElement object from API data and I assume its value is a double but I'm obviously not sure, since it's coming from an external source.

If I then use the AEXMLElement object's double property, I have no way of knowing if the data was invalid, I will just get 0.0 if it's invalid, but that may be valid data. Why isn't double an optional, so I can check if it's actually valid? As far as I can see I would have to check if value is non-nil and can be cast to Double myself.

Actually, your parser consume a lot of memory in Xcode 9 beta (both 1 and 2) with Swift 4. The problem is with appending string during parsing XML in Parser.swift. Appending founded string into currentValue allocate new memory, but still hold old one. I'm not sure if it's problem of implementation of calling delegate from XMLParser or they changed something and transfer responsibility of management into implementation of the delegate. Solution is fold processing of func parser(_ parser: XMLParser, foundCharacters string: String) into autoreleasepool closure.

Does have anybody else that problem with Swift 4?

I'm parsing some Shakespeare xml files with AEXML (which is working great for me otherwise BTW) when I experience some problems getting the correct values out of elements from the following xml:

I'm running through the SPEECH element (speechElement:AEXMLElement)

for child in speechElement.children { // when child equals the the first LINE element child.value // should equal "Is the law of our side, if I say" but is nil }

The LINE element contains the STAGEDIR element correctly, but loses anything after the STAGEDIR.

I've also moved the STAGEDIR element in the line, like so: Is the law of Aside to GREGORY our side, if I say and " our side, if I say" is lost.

I've worked around it by modifying the xml files, but I wanted to file a bug and let you know.

I'm not sure if this is somehow a problem with AEXML or a bug in XCode, or some other weird thing, so pardon me if it has nothing to do with this library.

Somehow, I'm getting an error on a file that doesn't use AEXML that there is a cast from AEXMLElement. If I comment out the #import AEXML line (in a totally unrelated file) then the project builds. Then when I uncomment the import, it builds fine.

I put together a simple sample project to illustrate this.

https://github.com/aaronpk/AEXML_Test

It cramps my style. :smile_cat: It also feels awkward having to type:

let element = elementList["MyElement"]
if element.name != AEXMLElement.errorElementName {
    let string = element["MyString"].stringValue
}

as opposed to something like:

let element = elementList["MyElement"]
let string = element["MyString"].stringValue // I would expect string == ""

because 99% of the time I don't want the string, "element <MyString> not found", because I use stringValue. Was under the assumption that stringValue would return an empty string upon lookup error, not an error string.

This is coming from having used SwiftyJSON. So that is where my expectations are coming from. Feel free to disagree if that's how you feel.

Thoughts?

Sources/Parser.swift:18:20: error: instance member 'document' cannot be used on type 'AEXMLParser'
        let trim = document.options.parserSettings.shouldTrimWhitespace
                   ^~~~~~~~

It appears this regression was added recently in this commit: https://github.com/tadija/AEXML/commit/506c15df3160f720b58b543fa7b42226e21e1463

This fixes the problem where Xcode pops a warning that it needs to convert the project source into Swift 3, since it isn't sure that the code's been converted already.

I added the Package.swift configuration file in order to work with the new Swift Package Manager.

I only had to do one change in the code and that was to remove the var keyword in the indentation function declaration. All tests kept on running, so this shouldn't be a problem.

Hello, I got an issue when parsing large files on linux. I got the following issue : Error Domain=NSXMLParserErrorDomain Code=1 "(null)"

When I compile and run on MacOS everything runs fine

Hello! I have a question When I parse an e-book, I get a mass of XML files Then I do as in your example and combine everything in one line

let document = try AEXMLDocument (xml: data) var parsedText = String () // parse unknown structure for child in document.root.children { parsedText + = child.xml + "\ n" } After that I get one long line of code, realizing that this is XML, how can I convert it to HTML to display it in the WebView, or how to convert it to tex to display it in TextView?

I would like to specify the correct DTD for the XML file used. How should I do this? Right now I consider enhancing AEXMLOptions.DocumentHeader with 3 additional optional values which would allow specifying the correct settings for the DTD (see eg. https://xmlwriter.net/xml_guide/doctype_declaration.shtml). Any better idea?

Many thanks in advance!

Didn't see this issue in the documentation but it's only hitting the first node in my XML document. For example I have multiple "Alert" nodes

<CTAAlerts>
<TimeStamp>20171127 11:47</TimeStamp>
<ErrorCode>0</ErrorCode>
<ErrorMessage/>
<Alert>...</Alert>
<Alert>...</Alert>
<Alert>...</Alert>

Even though I'm using a for loop it only returns the value I want from the first node and ignores the rest. Am I missing a feature that's not mentioned?

I'm trying to understand the preferred way to access declared namespaces. I can get a dictionary of the declared namespaces from the root element's attributes. Is there a more direct way?

Also, if process namespaces is set to true, I can't find anywhere where the namespace is saved. #34 covers this; am I correct in understanding that the namespace is just removed and discarded?