Lars/mindnet
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 21 Wiki Activity

scripts/edges_full_check.py aktualisiert
All checks were successful
Deploy mindnet to llm-node / deploy (push) Successful in 3s
Details

Browse Source
This commit is contained in:
Lars 2025-11-17 15:31:58 +01:00
parent e4115e705e
commit 12c600edbe
1 changed files with 13 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
scripts/edges_full_check.py
View File

@ -1,10 +1,12 @@
#!/usr/bin/env python3 #!/usr/bin/env python3
# -*- coding: utf-8 -*- # -*- coding: utf-8 -*-
from __future__ import annotations from __future__ import annotations
import json import json
from collections import Counter from collections import Counter
from app.core.qdrant import QdrantConfig, get_client from app.core.qdrant import QdrantConfig, get_client
def fetch_all(client, col): def fetch_all(client, col):
points = [] points = []
next_offset = None next_offset = None
@ -17,6 +19,14 @@ def fetch_all(client, col):
break break
return points return points
def is_callout_rule(rule_id: str) -> bool:
if not rule_id:
return False
r = rule_id.lower()
return r.startswith("callout:edge:v1") or ("callout" in r)
def main(): def main():
cfg = QdrantConfig.from_env() cfg = QdrantConfig.from_env()
cl = get_client(cfg) cl = get_client(cfg)
@ -35,7 +45,6 @@ def main():
chunks_by_note = Counter([c.payload.get("note_id") for c in chunks]) chunks_by_note = Counter([c.payload.get("note_id") for c in chunks])
e_by_kind = Counter([e.payload.get("kind") or e.payload.get("relation") for e in edges]) e_by_kind = Counter([e.payload.get("kind") or e.payload.get("relation") for e in edges])
# pro Note: belongs_to == #chunks; next == prev == max(chunks-1, 0)
belongs_by_note = Counter() belongs_by_note = Counter()
next_by_note = Counter() next_by_note = Counter()
prev_by_note = Counter() prev_by_note = Counter()
@ -57,14 +66,10 @@ def main():
b = belongs_by_note[nid] b = belongs_by_note[nid]
n = next_by_note[nid] n = next_by_note[nid]
p = prev_by_note[nid] p = prev_by_note[nid]
per_note[nid] = {"chunks": ccount, "belongs_to": b, "next": n, "prev": p, "checks": { per_note[nid] = {"chunks": ccount, "belongs_to": b, "next": n, "prev": p, "checks": {"belongs_to_equals_chunks": (b == ccount), "next_prev_match": (n == p == max(ccount-1, 0))}}
"belongs_to_equals_chunks": (b == ccount),
"next_prev_match": (n == p == max(ccount-1, 0)),
}}
ok_belongs &= (b == ccount) ok_belongs &= (b == ccount)
ok_nextprev &= (n == p == max(ccount-1, 0)) ok_nextprev &= (n == p == max(ccount-1, 0))
# Rule-Statistiken & Dubletten-Prüfung
explicit = defaults = callout = inline = 0 explicit = defaults = callout = inline = 0
multi_callout_detected = False multi_callout_detected = False
callout_key_counts = Counter() callout_key_counts = Counter()
@ -83,7 +88,7 @@ def main():
else: else:
seen.add(key) seen.add(key)
if rule.startswith("callout:edge:v1"): if is_callout_rule(rule):
callout += 1 callout += 1
callout_key_counts[(cid, kind, rule)] += 1 callout_key_counts[(cid, kind, rule)] += 1
if rule.startswith("inline:rel:v1"): if rule.startswith("inline:rel:v1"):
@ -116,5 +121,6 @@ def main():
} }
print(json.dumps(report, ensure_ascii=False, indent=2)) print(json.dumps(report, ensure_ascii=False, indent=2))
if __name__ == "__main__": if __name__ == "__main__":
main() main()