#!/usr/bin/python3

# Usage:  dbscan -f /var/lib/dirsrv/slapd-supplier2/db/userroot/entryrdn.db | python3 check_entryrdn
# (using the proper instance and backend names )
# Walk all entries in entryrdn index and follow their parent links until hitting the suffix
# detecting loops and missing entries

import sys
import re
from dataclasses import dataclass

@dataclass
class RDNElem:
    id: int
    rdn: str
    nrdn: str

    def __init__(self, s):
        res = re.match(r'^ID: (\d+); RDN: "([^"]*)"; NRDN: "([^"]*)"', s)
        self.id = res.group(1)
        self.rdn = res.group(2)
        self.nrdn = res.group(3)

    def __str__(self):
        return f"ID: {self.id}; RDN: {self.rdn}; NRDN:  {self.nrdn}"


db = {}
key = None

for line in sys.stdin:
    if key is None:
        key = line.strip()
    else:
        db[key] = RDNElem(line.strip())
        key = None

# print(db)

kp = { k[1:]:v for k,v in db.items() if k.startswith('P') }
kc = { k[1:]:v for k,v in db.items() if k.startswith('C') }
ke = { k:v for k,v in db.items() if k.isdigit() }
ks = { k:v for k,v in db.items() if '=' in k }

nbentries = 0
nberr = 0

for k,v in ke.items():
    dn = []
    ids = []
    v1 = v
    while True:
        dn.append(v1.rdn)
        ids.append(v1.id)
        if len(ids) > 50:
            print(f'Loop detected: ids={ids} dn: {dn}')
            nberr += 1
            break
        if v1.id in kp:
            v1 = kp[v1.id]
        elif v1.nrdn in ks:
            sdn = ",".join(dn)
            # print(f'DN: {sdn}')
            nbentries += 1
            break
        else:
            print(f'Missing parent entry detected: ids={ids} dn: {dn}')
            nberr += 1
            break

print(f'check_entryrdn processed {nbentries} entries with {nberr} errors')