Python3.8 DocumentReader bug

[mathijswesterhof]

in the dowloader of the suds reader there is an issue where the documentReader opens a new url that is not cached. It tries to download the url data, however the open function of the documentStore resolves that url and returns content instead of the file-handle. This is a bytestream. this bytestream has no read function. therefore when fp is not empty it generates an exception.

current code

    def download(self, url):
        store = DocumentStore()
        fp = store.open(url)
        if fp is None:
            fp = self.options.transport.open(Request(url))
        content = fp.read()
        fp.close()
        ctx = self.plugins.document.loaded(url=url, document=content)
        content = ctx.document
        sax = Parser()
        return sax.parse(string=content)

working solution

    def download(self, url):
        store = DocumentStore()
        fp = store.open(url)
        if fp is None:
            fp = self.options.transport.open(Request(url))
            content = fp.read()
            fp.close()
        else:
            content = fp
        ctx = self.plugins.document.loaded(url=url, document=content)
        content = ctx.document
        sax = Parser()
        return sax.parse(string=content)

The store.open() function can either return the bytestream (can be empty) or generate an exception. it has no option to return a file-handle. I am by far no expert on suds, can anyone verify this problem and help with a proper solution?

[cackharot]

Can you create a PR if this is required for all Python 3.x versions?