EDGAR Release 26.3 preview

.gitignore

@@ -3,3 +3,6 @@
 /.settings
 /.pydevproject
 /.vscode
+/view/rfileviewer/node_modules
+/view/rfileviewer/package-lock.json
+

__builtins__.pyi

@@ -0,0 +1,4 @@
+# __builtins__.pyi
+from typing import Callable
+
+_ = Callable[[str], str]

__init__.py

@@ -1,9 +1,9 @@
 __pluginInfo__ = {
     'name': 'EDGAR',
-    'version': '3.26.1',
-    'description': "This plug-in implements U.S. SEC Edgar Renderer. Arelle version at SEC: 2.39.8 ",
+    'version': '3.26.3',
+    'description': "This plug-in implements U.S. SEC Edgar Renderer. Arelle version at SEC: 2.44.2 ",
     'license': 'Apache-2',
     'author': 'U.S. SEC Employees and Mark V Systems Limited',
     'copyright': '(c) Portions by SEC Employees not subject to domestic copyright, otherwise (c) Copyright 2015 Mark V Systems Limited, All rights reserved.',
-    'import': ('EDGAR/render', )
+    'import': ('EDGAR/render', 'EDGAR/validate')
 }

readme.md

@@ -4,7 +4,7 @@
 The EDGAR plugin, developed and maintained by the staff of the U.S. Securities and Exchange Commission (SEC), is designed to provide traditional and inline XBRL viewers for SEC filings. It also integrates with and extends the EFM Validation plugin, offering EFM validation for SEC filings. For end-user support, please contact the SEC directly at: StructuredData@sec.gov.
 
 ## Arelle Version
-The current version of Arelle in use at the SEC is: **2.39.8**
+The current version of Arelle in use at the SEC is: **2.44.2**
 
 ## Installation
 The EDGAR plugin requires the xule plugin to be present under the Arelle plugin directory. You can clone the xule repository into the plugin directory. 

render/Embedding.py

@@ -353,15 +353,15 @@ def processOrFilterFacts(self):
                     for typedMember in sorted((getMemberOnAxisForFactDict[pseudoAxis]
                                                for fact, getMemberOnAxisForFactDict, periodStartEndLabel in self.cube.factMemberships
                                                if pseudoAxis in getMemberOnAxisForFactDict),
-                                              key=lambda member: member.typedMemberSortKey):
+                                              key=lambda member: member.typedMemberSortKey if isinstance(member,Filing.Member) else str(member)):
                         if typedMember not in giveMemGetPositionDict:
                             giveMemGetPositionDict[typedMember] = len(giveMemGetPositionDict)
 
                 except TypeError:  # if unsortable members, try as string (but will be inconsistent on numbers)
                     for typedMember in sorted((getMemberOnAxisForFactDict[pseudoAxis]
                                                for fact, getMemberOnAxisForFactDict, periodStartEndLabel in self.cube.factMemberships
                                                if pseudoAxis in getMemberOnAxisForFactDict),
-                                              key=lambda member: str(member.typedMemberSortKey)):
+                                              key=lambda member: str(member.typedMemberSortKey) if isinstance(member,Filing.Member) else str(member)):
                         if typedMember not in giveMemGetPositionDict:
                             giveMemGetPositionDict[typedMember] = len(giveMemGetPositionDict)
 

render/Filing.py

@@ -376,6 +376,12 @@ def populateAndLinkClasses(self, uncategorizedCube=None):
             facts = list(self.unusedFactSet)
 
         else:
+            # clear fact.inCubes from any prior Filing.mainFun call on this modelXbrl.
+            # Not clearing them could cause RemoveStuntedCashFlowColumns to treat facts as appearing in other reports.
+            for fact in self.modelXbrl.facts:
+                if hasattr(fact, 'inCubes'):
+                    fact.inCubes.clear()
+
             # build cubes
             for linkroleUri in self.modelXbrl.relationshipSet(arelle.XbrlConst.parentChild).linkRoleUris:
                 cube = Cube.Cube(self, linkroleUri)

render/IoManager.py

@@ -51,21 +51,21 @@ def absPathOnPythonPath(controller, filename):  # if filename is relative, find
     controller.logDebug("No such location {} found in sys path dirs {}.".format(filename, pathdirs))
     return None
 
+def writeTextDoc(filing, text, reportZip, reportFolder, filename, zipDir="", encoding=None):
+    if reportZip:
+        reportZip.writestr(zipDir + filename, text)
+    elif reportFolder is not None:
+        filing.writeFile(os.path.join(reportFolder, filename), text, encoding=encoding)
+
 
 def writeXmlDoc(filing, etree, reportZip, reportFolder, filename, zipDir=""):
     xmlText = treeToString(etree.getroottree(), method='xml', with_tail=False, pretty_print=True, encoding='utf-8', xml_declaration=True)
-    if reportZip:
-        reportZip.writestr(zipDir + filename, xmlText)
-    elif reportFolder is not None:
-        filing.writeFile(os.path.join(reportFolder, filename), xmlText)
+    writeTextDoc(filing, xmlText, reportZip, reportFolder, filename, zipDir=zipDir)
 
 
 def writeHtmlDoc(filing, root, reportZip, reportFolder, filename, zipDir=""):
     htmlText = treeToString(root, method='html', with_tail=False, pretty_print=True, encoding='utf-8')
-    if reportZip:
-        reportZip.writestr(zipDir + filename, htmlText)
-    elif reportFolder is not None:
-        filing.writeFile(os.path.join(reportFolder, filename), htmlText)
+    writeTextDoc(filing, htmlText, reportZip, reportFolder, filename, zipDir=zipDir)
 
 
 def writeJsonDoc(lines, pathOrStream, sort_keys=True):

render/PresentationGroup.py

@@ -253,13 +253,13 @@ def doPreorderTraversal(self, node, giveMemGetPositionDictPrimary, giveMemGetPos
         if depth <= len(visited):  # you can't go deeper than the number of unique relationships you've already visited.
             for childNode in node.childrenList:
                 if parentAxis is not None:  # collect more of this axis' descendants
-                    self.doPreorderTraversal(childNode, giveMemGetPositionDictPrimary, giveMemGetPositionDictAxis, parentAxis, setOfConcepts, visited, visitCounter, depth)
+                    self.doPreorderTraversal(childNode, giveMemGetPositionDictPrimary, giveMemGetPositionDictAxis, parentAxis, setOfConcepts, visited, visitCounter, depth + 1)
                 else:
-                    self.doPreorderTraversal(childNode, giveMemGetPositionDictPrimary, {}, parentAxis, setOfConcepts, visited, visitCounter, depth)
+                    self.doPreorderTraversal(childNode, giveMemGetPositionDictPrimary, {}, parentAxis, setOfConcepts, visited, visitCounter, depth + 1)
         else:
             self.filing.modelXbrl.debug("info",
-                                              ("Presentation group '%{linkRoleName} a an invalid directed cycle"),
-                                              linkrole=self.cube.linkroleUri)
+                                              ("Presentation group \"%(linkRoleName)s\" has an invalid directed cycle"),
+                                              linkRoleName=self.cube.linkroleUri)
 
         if nodeIsAnAxis:
             if concept.isTypedDimension:  # designate this as a typed dimension axis

render/README.md

@@ -10,7 +10,7 @@ domestic copyright protection. 17 U.S.C. 105.
 End user support is by e-mail direct to SEC at: [StructuredData@sec.gov]
 (mailto:StructuredData@sec.gov).
 
-This is EDGAR release 26.1, planned for production March, 2026.
+This is EDGAR release 26.3, planned for production September, 2026.
 
 Developer issue management is by the Jira Edgar Renderer project: https://arelle.atlassian.net/projects/ER
 

render/RFileViewer.py

@@ -0,0 +1,183 @@
+# -*- coding: utf-8 -*-
+"""
+:mod:`EdgarRenderer.Summary`
+Edgar(tm) Renderer was created by staff of the U.S. Securities and Exchange Commission.
+Data and content created by government employees within the scope of their employment
+are not subject to domestic copyright protection. 17 U.S.C. 105.
+"""
+
+# provide choice of python Javascript libraries
+JS_LIB = "pythonmonkey" # 27 Mb extra distribution size, robust implementation for node.js 
+#JS_LIB = "quickjs" # 3 Mb extra distribution size, customary quickjs (dormant project, no recent activity) 
+#JS_LIB = "quickjs-ng" # active quickjs project under maintenance, need PR 8 (https://github.com/genotrance/quickjs-ng/pull/8) to handle long strings for FilingSummary.xml
+if JS_LIB == "pythonmonkey":
+    from pythonmonkey import eval as js_eval 
+elif JS_LIB == "quickjs":
+    import quickjs
+elif JS_LIB == "quickjs-ng": # unsure, has to be built from PR to quickns-ng
+    import quickjs
+
+from lxml import etree
+import json, os
+
+JS_PATH = os.path.join(os.path.dirname(os.path.dirname(__file__)), "view", "rfileviewer", "dist", "r-file-viewer.transform.iife.js")
+
+
+# this section implements an lxml.etree to fast-xml-parser object for python
+
+ATTR_PREFIX = ""
+ATTRS_GRP_NAME = "_attributes"
+TEXT_KEY = "_text"
+
+def _coerce_scalar(s: str):
+    """Coerce strings to bool/int/float when appropriate, else return as str."""
+    if s is None:
+        return ""
+    s = s.strip()
+    # boolean
+    if s.lower() == "true":
+        return True
+    if s.lower() == "false":
+        return False
+    # integer
+    if s.isdigit() or (s.startswith("-") and s[1:].isdigit()):
+        try:
+            return int(s)
+        except ValueError:
+            pass
+    # float
+    try:
+        # allow numbers like 1.23 or -0.5
+        if any(ch in s for ch in ".eE"):
+            return float(s)
+    except ValueError:
+        pass
+    return s
+
+def _merge_child(d: dict, tag: str, value):
+    """Place child value under tag, making a list only if tag repeats."""
+    if tag in d:
+        # Promote to list if repeated
+        if isinstance(d[tag], list):
+            d[tag].append(value)
+        else:
+            d[tag] = [d[tag], value]
+    else:
+        d[tag] = value
+
+def lxml_element_to_fxp_json(el: etree._Element):
+    """
+    Convert any lxml element to a dict shaped like fast-xml-parser JSON output:
+      - attributes in child object ATTRS_GRP_NAME if not null, prefixed with ATTR_PREFIX
+      - text under TEXT_KEY only when attributes or children exist
+      - scalars for simple elements (no attrs/children)
+      - lists created only for repeated sibling tag names
+    """
+    # Collect attributes (prefixed)
+    out = {}
+    for k, v in el.attrib.items():
+        if ATTRS_GRP_NAME:
+            if ATTRS_GRP_NAME not in out: out[ATTRS_GRP_NAME] = {}
+            out[ATTRS_GRP_NAME][f"{ATTR_PREFIX}{k}"] = _coerce_scalar(v)
+        else:
+            out[f"{ATTR_PREFIX}{k}"] = _coerce_scalar(v)
+
+    # Recurse over children, grouping by tag name
+    child_tag_counts = {}
+    for child in el:
+        child_tag_counts[child.tag] = child_tag_counts.get(child.tag, 0) + 1
+
+    for child in el:
+        child_val = lxml_element_to_fxp_json(child)
+        # If child is simple (returns scalar), keep as scalar
+        # Otherwise it's an object (possibly containing attributes/TEXT_KEY)
+        _merge_child(out, child.tag, child_val)
+
+    # Handle element text
+    text = (el.text or "").strip()
+    has_children = len(el) > 0
+    has_attrs = bool(el.attrib)
+
+    if has_children or has_attrs:
+        # If there's meaningful text alongside attrs/children, store under TEXT_KEY
+        if text:
+            out[TEXT_KEY] = _coerce_scalar(text)
+        # If no text and no children produced anything, ensure at least {} is returned
+        if not out:
+            out = {}
+        return out
+    else:
+        # Simple element → scalar
+        return _coerce_scalar(text)
+
+def lxml_etree_to_fxp_json(root: etree._Element):
+    """Top-level wrapper returning {root.tag: ...}"""
+    return {root.tag: lxml_element_to_fxp_json(root)}
+
+# entry point from render/__init__.py to transform etree into html
+
+def transformToHtml(filing_summary_etree, accession_number, title=None, timeout_sec=60, logDebugToConsole=False, secws=False):
+
+    # transform FilingSummary lxml etree to fast-xml-parser JSON object
+    fxp_json_obj = lxml_etree_to_fxp_json(filing_summary_etree)
+
+    # Load the transform-only bundle
+    with open(JS_PATH, "r", encoding="utf-8") as f:
+        js_code = f.read()
+
+    if JS_LIB == "pythonmonkey":
+        js_eval(js_code)
+        js_transform = js_eval("RFileViewer.transform")
+
+        try:
+            html = js_transform(
+                fxp_json_obj,
+                accession_number,
+                title,
+                logDebugToConsole,
+                secws
+            )
+            return html
+        except Exception as exc:
+            print("PythonMonkey JS error:", exc)
+            raise
+
+    elif JS_LIB in ("quickjs", "quickjs-ng"):
+        # quickjs needs stringified object parameter
+        fxp_json_str = json.dumps(fxp_json_obj)
+
+        # add sourceURL for better stack traces
+        js_code = "//@ sourceURL=r-file-viewer.transform.iife.js\n" + js_code
+
+        ctx = quickjs.Context()
+        try:
+            # Evaluate IIFE bundle
+            ctx.eval(js_code)
+            # Create a wrapper for easier calling
+            # Wrap the transform with JSON.parse inside JS
+            ctx.eval("""
+            function __rfv_transform(fxp_json_str, accession_number, title, logDebugToConsole, secws) {
+                const fxp_json_obj = JSON.parse(fxp_json_str);
+                return RFileViewer.transform(fxp_json_obj, accession_number, title, logDebugToConsole, secws);
+            }
+            """)
+
+            # Retrieve JS function reference (Python callable)
+            ctx.set_time_limit(int(timeout_sec * 1000))
+            transform_fn = ctx.get("__rfv_transform")
+
+            # Call JS function
+            html = transform_fn(
+                fxp_json_str,
+                accession_number,
+                title,
+                logDebugToConsole,
+                secws
+            )
+            return html
+        except quickjs.JSException as exc:
+            # Log JS exception with stack trace
+            print("JS exception in RFV transform:", exc)
+            print("JS traceback:\n", str(exc))
+            # Optionally re-raise or wrap as needed
+            raise