Import XML formations: filtre caractères de contrôles erronés

app/scodoc/sco_formations.py

@@ -222,6 +222,8 @@ def formation_export(
     """Get a formation, with UE, matieres, modules
     in desired format
     """
+    if fmt not in ("xml", "json"):
+        raise ScoValueError("Format invalide")
     formation = Formation.get_formation(formation_id)
     f_dict = formation_export_dict(
         formation,
@@ -296,7 +298,7 @@ def _formation_retreive_apc_niveau(
     return niveau.id if niveau is not None else None
 
 
-def formation_import_xml(doc: str, import_tags=True, use_local_refcomp=False):
+def formation_import_xml(doc: str | bytes, import_tags=True, use_local_refcomp=False):
     """Create a formation from XML representation
     (format dumped by formation_export( fmt='xml' ))
     XML may contain object (UE, modules) ids: this function returns two
@@ -311,11 +313,13 @@ def formation_import_xml(doc: str, import_tags=True, use_local_refcomp=False):
     """
     from app.scodoc import sco_edit_formation
 
+    if isinstance(doc, bytes):
+        doc = doc.decode(scu.SCO_ENCODING)
     try:
-        dom = xml.dom.minidom.parseString(doc)
+        dom = xml.dom.minidom.parseString(sco_xml.remove_control_characters(doc))
     except Exception as exc:
-        log("formation_import_xml: invalid XML data")
+        log(f"formation_import_xml: invalid XML data:\n{exc}")
-        raise ScoValueError("Fichier XML invalide") from exc
+        raise ScoValueError(f"Fichier XML invalide {exc}") from exc
 
     try:
         f = dom.getElementsByTagName("formation")[0]  # or dom.documentElement

app/scodoc/sco_utils.py

@@ -1117,8 +1117,9 @@ def sendXML(
     attached=False,
     quote=False,
     filename=None,
-):
+) -> Response:
-    if type(data) != list:
+    "Réponse XML: data est une liste d'objets"
+    if not isinstance(data, list):
         data = [data]  # always list-of-dicts
     if force_outer_xml_tag:
         data = [{tagname: data}]

app/scodoc/sco_xml.py

@@ -39,6 +39,16 @@ from app.scodoc.sco_vdi import ApoEtapeVDI
 XML_HEADER = """<?xml version="1.0" encoding="utf-8"?>"""
 
 
+def remove_control_characters(s: str) -> str:
+    """supprime tous les caractères de contrôle (code < 32)
+    sauf les tabs (9), les retours chariot (10 et les sauts de ligne (13).
+    """
+    if not hasattr(remove_control_characters, "control_chars"):
+        control_chars = dict.fromkeys(i for i in range(32) if i not in (9, 10, 13))
+        remove_control_characters.control_chars = control_chars
+    return s.translate(remove_control_characters.control_chars)
+
+
 def quote_xml_attr(data):
     """Escape &, <, >, quotes and double quotes"""
     return xml.sax.saxutils.escape(str(data), {"'": "&apos;", '"': "&quot;"})
@@ -93,7 +103,7 @@ def _dictlist2xml(dictlist, root=None, tagname=None, quote=False):
     for d in dictlist:
         elem = ElementTree.Element(tagname)
         root.append(elem)
-        if isinstance(d, scalar_types) or isinstance(d, ApoEtapeVDI):
+        if isinstance(d, (scalar_types, ApoEtapeVDI)):
             elem.set("code", _repr_as_xml(d))
         else:
             if quote: