juntalis - 2011-07-17

The XML Tools plugin for Notepad++ is awesome and does almost everything I need when I'm working with XML files. There are times, however, when I'm working with markup-oriented files that aren't quite valid, or some return from a webapp that is not only not valid, but also extremely squished and ugly to read. Unfortunately, XML Tools will only format valid markup, so I threw this together. I used a script I found at http://xmlpp.codeplex.com/ as a base, then modified it to be a bit more customizable. It automatically grabs Notepad++'s settings to figure out how to format the markup, but the variables to change the settings are at the top if of the script if you'd like to do so. 

# Settings for the Notepad++ Script
# By default, it grabs your current notepad settings. If you'd prefer something else, however,
# you can set them below.
#
# Original script: Copyright (c) 2008, Fredrik Ekholdt, All rights reserved.
# Found at http://xmlpp.codeplex.com/
__indent_char = '\t' if editor.getUseTabs() else ' '
__indent_size = 1 if editor.getUseTabs() else editor.getIndent()
__width = 100 if editor.getEdgeMode() == EDGEVISUALSTYLE.NONE else editor.getEdgeColumn()
class xmlpp:
    import sys, re
    indent_char = ' '

    def __pprint_line(self, indent_level, line, width, output):
        if line.strip():
            start = ""
            number_chars = 0
            indent_width = 1
            if self.indent_char == '\t':
                indent_width=editor.getTabWidth()
            for l in range(indent_level):
                start += self.indent_char
                number_chars = number_chars + 1
            try:
                number_chars = number_chars
                elem_start = re.findall("(\<\W{0,1}\w+:\w+) ?", line)[0]
                elem_finished = re.findall("([?|\]\]/]*\>)", line)[0] 
                #should not have *
                attrs = re.findall("(\S*?\=\".*?\")", line)
                output.write(start + elem_start)
                number_chars = (len(start) * indent_width) + len(elem_start)
                for attr in attrs:
                    if (attrs.index(attr) + 1) == len(attrs):
                        number_chars = number_chars + len(elem_finished)
                    if (number_chars + len(attr) + 1) > width:
                        output.write("\n")
                        for i in range(len(start + elem_start) + 1):
                            output.write(self.indent_char)
                        number_chars = (len(start) * indent_width) + len(elem_start) + 1 
                    else:
                        output.write(self.indent_char)
                        number_chars = number_chars + 1
                    output.write(attr)
                    number_chars = number_chars + len(attr)
                output.write(elem_finished + "\n")
            except IndexError:
                #give up pretty print this line
                output.write(start + line + "\n")

    def __pprint_elem_content(self, indent_level, line, output=sys.stdout):
        if line.strip():
            for l in range(indent_level):
                output.write(self.indent_char)
            output.write(line + "\n")
    def __get_next_elem(self,data):
        start_pos = data.find("<")
        end_pos = data.find(">") + 1
        retval = data[start_pos:end_pos]
        stopper = retval.rfind("/") 
        if stopper < retval.rfind("\""):
            stopper = -1
        single = (stopper > -1 and ((retval.find(">") - stopper) < (stopper - retval.find("<"))))
        ignore_excl = retval.find("<!") > -1
        ignore_question =  retval.find("<?") > -1
        if ignore_excl:
            cdata = retval.find("<![CDATA[") > -1
            if cdata:
                end_pos = data.find("]]>")
                if end_pos > -1:
                    end_pos = end_pos + len("]]>")
        elif ignore_question:
            end_pos = data.find("?>") + len("?>")
        ignore = ignore_excl or ignore_question

        no_indent = ignore or single
        #print retval, end_pos, start_pos, stopper > -1, no_indent
        return start_pos, \
               end_pos, \
               stopper > -1, \
               no_indent
    def get_pprint(self, xml, indent=4, indentchar=' ', width=80):
        self.indent_char = indentchar
        """Returns the pretty printed xml """
        class out:
            output = ""
            def write(self, string): 
                self.output += string
        out = out()
        self.pprint(xml, output=out, indent=indent, width=width)

        docFormat = notepad.getFormatType(notepad.getCurrentBufferID())
        newLine = '\r\n' if docFormat == FORMATTYPE.WIN else '\n'
        return out.output.replace('\n', newLine)
    def pprint(self,xml, output=None, indent=4, width=80):
        """Pretty print xml. 
        Use output to select output stream. Default is sys.stdout
        Use indent to select indentation level. Default is 4   """
        data = xml
        indent_level = 0
        start_pos, end_pos, is_stop, no_indent  = self.__get_next_elem(data)
        while ((start_pos > -1 and end_pos > -1)):
            self.__pprint_elem_content(indent_level, data[:start_pos].strip(), 
                                 output=output)
            data = data[start_pos:]
            if is_stop and not no_indent:
                indent_level = indent_level - indent
            self.__pprint_line(indent_level, data[:end_pos - start_pos],
                        width=width, output=output)
            data = data[end_pos - start_pos:]
            if not is_stop and not no_indent :
                indent_level = indent_level + indent
            if not data:
                break
            else:
                start_pos, end_pos, is_stop, no_indent  = self.__get_next_elem(data)
if __name__=='__main__':
    editor.setText(
        xmlpp().get_pprint(
            editor.getText(),
            indent=__indent_size,
            indentchar=__indent_char,
            width=__width
        )
    )