# -*- mode: python -*- # -*- coding: utf-8 -*- ############################################################################## # # Gestion scolarite IUT # # Copyright (c) 1999 - 2022 Emmanuel Viennet. All rights reserved. # # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License as published by # the Free Software Foundation; either version 2 of the License, or # (at your option) any later version. # # This program is distributed in the hope that it will be useful, # but WITHOUT ANY WARRANTY; without even the implied warranty of # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the # GNU General Public License for more details. # # You should have received a copy of the GNU General Public License # along with this program; if not, write to the Free Software # Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA # # Emmanuel Viennet emmanuel.viennet@viennet.net # ############################################################################## """ Exports XML """ from xml.etree import ElementTree import xml.sax.saxutils from xml.dom import minidom from app.scodoc import sco_utils as scu from app.scodoc.sco_vdi import ApoEtapeVDI XML_HEADER = """<?xml version="1.0" encoding="utf-8"?>""" def quote_xml_attr(data): """Escape &, <, >, quotes and double quotes""" return xml.sax.saxutils.escape(str(data), {"'": "'", '"': """}) # ScoDoc7 legacy function: def simple_dictlist2xml(dictlist, tagname=None, quote=False, pretty=True): """Represent a dict as XML data. All keys with string or numeric values are attributes (numbers converted to strings). All list values converted to list of childs (recursively). *** all other values are ignored ! *** Values (xml entities) are not quoted, except if requested by quote argument. Exemple: simple_dictlist2xml([ { 'id' : 1, 'ues' : [{'note':10},{}] } ], tagname='infos') <?xml version="1.0" encoding="utf-8"?> <infos id="1"> <ues note="10" /> <ues /> </infos> """ if not tagname: raise ValueError("invalid empty tagname !") elements = _dictlist2xml(dictlist, root=[], tagname=tagname, quote=quote) ans = XML_HEADER + b"\n".join([ElementTree.tostring(x) for x in elements]).decode( scu.SCO_ENCODING ) if pretty: # solution peu satisfaisante car on doit reparser le XML # de plus, on encode/decode pour avoir le tag <?xml version="1.0" encoding="utf-8"?> try: ans = ( minidom.parseString(ans) .toprettyxml(indent="\t", encoding="utf-8") .decode("utf-8") ) except xml.parsers.expat.ExpatError: pass return ans def _repr_as_xml(v): if isinstance(v, bool): return str(int(v)) # booleans as "0" / "1" return str(v) def _dictlist2xml(dictlist, root=None, tagname=None, quote=False): scalar_types = (bytes, str, int, float, bool) for d in dictlist: elem = ElementTree.Element(tagname) root.append(elem) if isinstance(d, scalar_types) or isinstance(d, ApoEtapeVDI): elem.set("code", _repr_as_xml(d)) else: if quote: d_scalar = dict( [ (k, quote_xml_attr(_repr_as_xml(v))) for (k, v) in d.items() if isinstance(v, scalar_types) ] ) else: d_scalar = dict( [ (k, _repr_as_xml(v)) for (k, v) in d.items() if isinstance(v, scalar_types) ] ) for k in d_scalar: elem.set(k, d_scalar[k]) d_list = dict([(k, v) for (k, v) in d.items() if isinstance(v, list)]) if d_list: for (k, v) in d_list.items(): _dictlist2xml(v, tagname=k, root=elem, quote=quote) return root ELEMENT_NODE = 1 TEXT_NODE = 3 def xml_to_dicts(element): """Represent dom element as a dict Example: <foo x="1" y="2"><bar z="2"/></foo> will give us: ('foo', {'y': '2', 'x': '1'}, [('bar', {'z': '2'}, [])]) """ d = {} # attributes if element.attributes: for i in range(len(element.attributes)): a = element.attributes.item(i).nodeName v = element.getAttribute(element.attributes.item(i).nodeName) d[a] = v # descendants childs = [] for child in element.childNodes: if child.nodeType == ELEMENT_NODE: childs.append(xml_to_dicts(child)) return (element.nodeName, d, childs)