Updates for processing inline html attributes without <base>

Disclosure system update

arelle/DisclosureSystem.py

@@ -61,6 +61,7 @@ def clear(self):
                 setattr(self, typeTestVariable, False)
                 self.pluginTypes.add(typeName)
         self.validateFileText = False
+        self.allowedExternalHrefPattern = None
         self.schemaValidateSchema = None
         self.blockDisallowedReferences = False
         self.maxSubmissionSubdirectoryEntryNesting = 0
@@ -172,6 +173,8 @@ def select(self, name):
                                     for typeName, typeTestVariable in pluginXbrlMethod(self):
                                         setattr(self, typeTestVariable, self.validationType == typeName)
                                 self.validateFileText = dsElt.get("validateFileText") == "true"
+                                if dsElt.get("allowedExternalHrefPattern"):
+                                    self.allowedExternalHrefPattern = re.compile(dsElt.get("allowedExternalHrefPattern"))
                                 self.blockDisallowedReferences = dsElt.get("blockDisallowedReferences") == "true"
                                 try:
                                     self.maxSubmissionSubdirectoryEntryNesting = int(dsElt.get("maxSubmissionSubdirectoryEntryNesting"))

arelle/ModelDocument.py

@@ -1211,8 +1211,10 @@ def inlineXbrlDiscover(self, htmlElement):
                 if _ns in XbrlConst.ixbrlAll:
                     ixNS = _ns
                     break
+        # required by 12.4.1 of [HTML] bullet 3
+        # use of document base is commented out because it discloses/uses absolute server directory and defeats URI redirection
         if htmlBase is None:
-            htmlBase = os.path.dirname(self.uri) + "/"
+            htmlBase = "" # os.path.dirname(self.uri) + "/"
         if conflictingNSelts:
             self.modelXbrl.error("ix:multipleIxNamespaces",
                     _("Multiple ix namespaces were found"),

arelle/ValidateXbrlDTS.py

@@ -372,6 +372,7 @@ def linkbaseTopElts():
                                 if l > 255:
                                     val.modelXbrl.error("EFM.6.07.30",
                                         _("Schema targetNamespace length (%(length)s) is over 255 bytes long in utf-8 %(targetNamespace)s"),
+                                        edgarCode="du-0730-Uri-Length-Limit",
                                         modelObject=elt, length=l, targetNamespace=targetNamespace, value=targetNamespace)
                         if val.validateSBRNL:
                             if elt.get("targetNamespace") is None:
@@ -533,6 +534,7 @@ def linkbaseTopElts():
                             if l > 200:
                                 val.modelXbrl.error("EFM.6.07.29",
                                     _("Schema %(element)s has a name length (%(length)s) over 200 bytes long in utf-8, %(name)s."),
+                                    edgarCode="du-0729-Name-Length-Limit",
                                     modelObject=elt, element=localName, name=name, length=l)
 
                     if val.validateSBRNL and localName in {"all", "documentation", "any", "anyAttribute", "attributeGroup",

arelle/XhtmlValidate.py

@@ -11,7 +11,7 @@
 from arelle.ModelObject import ModelObject
 from arelle.PythonUtil import normalizeSpace
 from lxml import etree
-import os, re
+import os, re, posixpath
 
 EMPTYDICT = {}
 
@@ -562,13 +562,11 @@ def resolveHtmlUri(elt, name, value):
         if elt.modelDocument.htmlBase is not None:
             value = elt.modelDocument.htmlBase + value
     # canonicalize ../ and ./
-    authority, sep, path = value.rpartition("://")
-    inpaths = path.split("/")
-    outpaths = []
-    for path in inpaths:
-        if path == "..":
-            if len(outpaths) > 1:
-                outpaths.pop()
-        elif path != "." and (path != "" or len(outpaths) == 0):
-            outpaths.append(path.replace(" ", "%20"))
-    return authority + sep + "/".join(outpaths)
+    scheme, sep, pathpart = value.rpartition("://")
+    if sep:
+        pathpart = pathpart.replace('\\','/')
+        endingSep = '/' if pathpart[-1] == '/' else ''  # normpath drops ending directory separator
+        _uri = scheme + "://" + posixpath.normpath(pathpart) + endingSep
+    else:
+        _uri = os.path.normpath(value)
+    return _uri.replace(" ", "%20")

arelle/config/disclosuresystems.xsd

@@ -60,6 +60,7 @@
       <xs:attribute name="mappingsUrl" />
       <xs:attribute name="utrUrl" />
       <xs:attribute name="validateFileText" type="xs:boolean"/>
+      <xs:attribute name="allowedExternalHrefPattern" type="xs:string"/>
       <xs:attribute name="blockDisallowedReferences" type="xs:boolean"/>
       <xs:attribute name="maxSubmissionSubdirectoryEntryNesting" type="xs:integer"/>
       <xs:attribute name="logLevelFilter" />