Filter current NeTEx_publication for elements that do not exists in t…

scripts/xsd_selector_clear.py

@@ -0,0 +1,147 @@
+from pathlib import Path
+from lxml import etree
+import json
+import sys
+
+# --- CONFIG ---
+NS = {"xsd": "http://www.w3.org/2001/XMLSchema"}
+BASE_DIR = Path(sys.argv[1])  # map met alle XSD bestanden
+INPUT_XSD = OUTPUT_XSD = BASE_DIR / "NeTEx_publication.xsd"     # bestand dat je wilt herschrijven
+REPORT_JSON = "report.json"
+
+# --- 1. Verzamel alle element-namen uit alle XSD's (recursief, inclusief nested, negeer abstract) ---
+all_elements = set()
+for xsd_file in BASE_DIR.rglob("*.xsd"):
+    tree = etree.parse(str(xsd_file))
+    root = tree.getroot()
+    for elem in root.findall(".//xsd:element", namespaces=NS):
+        if elem.get("abstract") == "true":
+            continue  # negeer abstract elements
+        name = elem.get("name")
+        if name:
+            all_elements.add(name)
+
+print(f"Totale elementen gevonden (excl. abstract): {len(all_elements)}")
+
+# --- 2. Open het doelbestand ---
+tree = etree.parse(str(INPUT_XSD))
+root = tree.getroot()
+
+# --- 3. Prepare report dictionary ---
+report = {"removed_selectors": [], "removed_elements": [], "removed_comments": []}
+
+# --- 4. Functie om selector xpath te filteren ---
+def filter_selector(node):
+    selector = node.find("xsd:selector", namespaces=NS)
+    if selector is None:
+        return False  # geen selector
+    xpath_expr = selector.get("xpath")
+    if not xpath_expr:
+        return False
+    exprs = [e.strip() for e in xpath_expr.split("|")]
+    kept_exprs = []
+    removed_exprs = []
+    for e in exprs:
+        # strip leading .// en namespace prefix
+        e_clean = e.replace(".//", "").split(":")[-1]
+        if e_clean in all_elements:
+            kept_exprs.append(e)
+        else:
+            removed_exprs.append(e)
+    if removed_exprs:
+        report["removed_selectors"].append({
+            "parent_name": node.get("name"),
+            "removed_expressions": removed_exprs
+        })
+    selector.set("xpath", " | ".join(kept_exprs))  # altijd bijwerken
+    print(f"[DEBUG] Node: {node.tag}, name: {node.get('name')}, kept expressions: {kept_exprs}")
+    return bool(kept_exprs)  # True als minstens één expression overblijft
+
+# --- 5. Helper: verwijder alle comments direct boven een node ---
+def remove_comments_above(node):
+    parent = node.getparent()
+    prev = node.getprevious()
+    removed_comments = []
+    while prev is not None and isinstance(prev, etree._Comment):
+        removed_comments.append(prev.text)
+        parent.remove(prev)
+        prev = node.getprevious()  # update na verwijderen
+    if removed_comments:
+        report["removed_comments"].append({
+            "parent_name": node.get("name"),
+            "removed_comments": removed_comments
+        })
+        print(f"[DEBUG] Removed comment(s) above '{node.get('name')}': {removed_comments}")
+
+# --- 6. Eerste pass: filter key, unique, keyref selectors ---
+for tag in ["key", "unique"]:
+    for node in root.findall(f".//xsd:{tag}", namespaces=NS):
+        has_expr = filter_selector(node)
+        if not has_expr:
+            remove_comments_above(node)
+            report["removed_elements"].append({
+                "type": tag,
+                "name": node.get("name"),
+                "reason": "all selector expressions removed"
+            })
+            print(f"[DEBUG] Removing {tag} '{node.get('name')}' (no expressions left)")
+            node.getparent().remove(node)
+
+for kr in root.findall(".//xsd:keyref", namespaces=NS):
+    has_expr = filter_selector(kr)
+    refer_name = kr.get("refer")
+    if not has_expr or refer_name is None:
+        remove_comments_above(kr)
+        report["removed_elements"].append({
+            "type": "keyref",
+            "name": kr.get("name"),
+            "reason": "all selector expressions removed or no refer"
+        })
+        print(f"[DEBUG] Removing keyref '{kr.get('name')}' (no expressions or no refer)")
+        kr.getparent().remove(kr)
+
+# Tweede pass: verwijder keys die nergens meer door keyrefs gebruikt worden
+for k in root.findall(".//xsd:key", namespaces=NS):
+    kname = k.get("name")  # gebruik exact dezelfde string als in refer
+    still_used = False
+    for kr in root.findall(".//xsd:keyref", namespaces=NS):
+        if kr.get("refer") == kname or kr.get("refer") == f"netex:{kname}":
+            still_used = True
+            break
+    if not still_used:
+        remove_comments_above(k)
+        report["removed_elements"].append({
+            "type": "key",
+            "name": kname,
+            "reason": "no keyrefs reference this key anymore"
+        })
+        print(f"[DEBUG] Removing key '{kname}' (no keyrefs reference it)")
+        k.getparent().remove(k)
+
+# --- 3e pass: verwijder keyrefs waarvan de refererende key is verwijderd ---
+existing_keys = {k.get("name") for k in root.findall(".//xsd:key", namespaces=NS)}
+
+for kr in root.findall(".//xsd:keyref", namespaces=NS):
+    refer = kr.get("refer")
+    if refer:
+        local_refer = refer.split(":")[-1]  # simpele prefix strip
+        if local_refer not in existing_keys:
+            remove_comments_above(kr)
+            report["removed_elements"].append({
+                "type": "keyref",
+                "name": kr.get("name"),
+                "reason": f"referenced key '{refer}' was removed"
+            })
+            print(f"[DEBUG] Removing keyref '{kr.get('name')}' (referenced key '{refer}' removed)")
+            kr.getparent().remove(kr)
+
+# --- 8. Schrijf het XSD-bestand opnieuw ---
+tree.write(str(OUTPUT_XSD), encoding="utf-8", xml_declaration=True, pretty_print=True)
+
+# --- 9. Schrijf JSON rapport ---
+# with open(REPORT_JSON, "w", encoding="utf-8") as jf:
+#    json.dump(report, jf, indent=2, ensure_ascii=False, default=str)
+
+# print(f"Verwerkt XSD opgeslagen als {OUTPUT_XSD}")
+# print(f"JSON rapport opgeslagen als {REPORT_JSON}")
+