#!/usr/bin/env python3

import json
import sys
import os
import argparse
import time
from datetime import datetime
from rsc import RSCAuth, RSCGraphQL

def find_database_by_name_or_id(identifier):
    """Find database by name or ID and return its details"""
    auth = RSCAuth()
    gql = RSCGraphQL(auth)
    
    # Check if identifier looks like a UUID (contains hyphens)
    if '-' in identifier:
        # It's likely a database ID
        query = """
        query OracleDatabase($fid: UUID!) {
            oracleDatabase(fid: $fid) {
                dbUniqueName
                id
                cluster {
                    id
                    name
                }
                logicalPath {
                    fid
                    name
                    objectType
                }
            }
        }
        """
        variables = {"fid": identifier}
    else:
        # It's a database name
        query = """
        query OracleDatabases($filter: [Filter!]) {
            oracleDatabases(filter: $filter) {
                nodes {
                    dbUniqueName
                    id
                    cluster {
                        id
                        name
                    }
                    logicalPath {
                        fid
                        name
                        objectType
                    }
                }
            }
        }
        """
        variables = {
            "filter": [
                {"texts": [identifier], "field": "NAME_EXACT_MATCH"},
                {"texts": ["false"], "field": "IS_RELIC"},
                {"texts": ["false"], "field": "IS_REPLICATED"}
            ]
        }
    
    response = gql.query(query, variables)
    
    if '-' in identifier:
        # Direct ID lookup
        db = response['data']['oracleDatabase']
        if not db:
            raise ValueError(f"Database with ID '{identifier}' not found")
        return db
    else:
        # Name lookup
        databases = response['data']['oracleDatabases']['nodes']
        if not databases:
            raise ValueError(f"No databases found with name '{identifier}'")
        if len(databases) > 1:
            print(f"Multiple databases found with name '{identifier}':")
            for db in databases:
                host_name = db['logicalPath'][0]['name'] if db['logicalPath'] else 'Unknown'
                print(f"  - {db['dbUniqueName']} (ID: {db['id']}, Host: {host_name})")
            raise ValueError("Please specify the database ID instead")
        return databases[0]

def get_oracle_host_id(host_name, cluster_id):
    """Get Oracle host ID by name and cluster"""
    auth = RSCAuth()
    gql = RSCGraphQL(auth)

    query = """
    query OracleHosts($filter: [Filter!]) {
        oracleTopLevelDescendants(filter: $filter) {
            nodes {
                name
                id
            }
        }
    }
    """

    variables = {
        "filter": [
            {"texts": [host_name], "field": "NAME"},
            {"texts": [cluster_id], "field": "CLUSTER_ID"}
        ]
    }

    response = gql.query(query, variables)
    hosts = response['data']['oracleTopLevelDescendants']['nodes']

    if not hosts:
        raise ValueError(f"Host '{host_name}' not found in cluster")

    if len(hosts) > 1:
        print(f"WARN: Multiple hosts found for '{host_name}':", file=sys.stderr)
        for host in hosts:
            print(f"  - {host['name']} (ID: {host['id']})", file=sys.stderr)
        # Use the first one
        print(f"WARN: Using first match: {hosts[0]['name']}", file=sys.stderr)

    return hosts[0]['id']

def get_latest_pit(db_id):
    """Get the latest Point in Time from recoverable ranges"""
    auth = RSCAuth()
    gql = RSCGraphQL(auth)

    query = """
    query OracleDatabaseRecoverableRangesQuery($fid: String!) {
        oracleRecoverableRanges(
            input: {id: $fid, shouldIncludeDbSnapshotSummaries: false}
        ) {
            data {
                beginTime
                endTime
                __typename
            }
            __typename
        }
        oracleMissedRecoverableRanges(input: {id: $fid}) {
            data {
                beginTime
                endTime
                __typename
            }
            __typename
        }
    }
    """

    variables = {"fid": db_id}
    response = gql.query(query, variables)

    # Get latest endTime from recoverable ranges
    ranges = response['data']['oracleRecoverableRanges']['data']
    if ranges:
        latest_endtime = max(range_item['endTime'] for range_item in ranges)
        print(f"INFO: Latest PIT (ISO8601): {latest_endtime}")

        # Convert to datetime and then to milliseconds since epoch
        dt = datetime.fromisoformat(latest_endtime.replace('Z', '+00:00'))
        unixtime_ms = int(dt.timestamp() * 1000)
        print(f"INFO: Latest PIT unixtime (ms): {unixtime_ms}")

        return unixtime_ms
    else:
        raise ValueError("No recoverable ranges found for database")
    """Get Oracle host ID by name and cluster"""
    auth = RSCAuth()
    gql = RSCGraphQL(auth)

    query = """
    query OracleHosts($filter: [Filter!]) {
        oracleTopLevelDescendants(filter: $filter) {
            nodes {
                name
                id
            }
        }
    }
    """

    variables = {
        "filter": [
            {"texts": [host_name], "field": "NAME"},
            {"texts": [cluster_id], "field": "CLUSTER_ID"}
        ]
    }

    response = gql.query(query, variables)
    hosts = response['data']['oracleTopLevelDescendants']['nodes']

    if not hosts:
        raise ValueError(f"Host '{host_name}' not found in cluster")

    if len(hosts) > 1:
        print(f"WARN: Multiple hosts found for '{host_name}':", file=sys.stderr)
        for host in hosts:
            print(f"  - {host['name']} (ID: {host['id']})", file=sys.stderr)
        # Use the first one
        print(f"WARN: Using first match: {hosts[0]['name']}", file=sys.stderr)

    return hosts[0]['id']

def mount_files_only(db_id, target_host_id, recovery_timestamp_ms, target_mount_path):
    """Execute files-only mount operation"""
    auth = RSCAuth()
    gql = RSCGraphQL(auth)

    variables = {
        "input": {
            "request": {
                "config": {
                    "targetOracleHostOrRacId": target_host_id,
                    "shouldMountFilesOnly": True,
                    "recoveryPoint": {
                        "timestampMs": recovery_timestamp_ms,
                        "scn": None
                    },
                    "targetMountPath": target_mount_path,
                    "shouldAllowRenameToSource": True,
                    "shouldSkipDropDbInUndo": False
                },
                "id": db_id
            },
            "advancedRecoveryConfigMap": []
        }
    }

    query = """
    mutation OracleDatabaseMountMutation($input: MountOracleDatabaseInput!) {
        mountOracleDatabase(input: $input) {
            id
            links {
                href
                rel
                __typename
            }
            __typename
        }
    }
    """

    response = gql.query(query, variables)
    return response['data']['mountOracleDatabase']['id']

def monitor_job_status(job_id, cluster_id):
    """Monitor the mount job status until completion"""
    auth = RSCAuth()
    gql = RSCGraphQL(auth)

    query = """
    query OracleDatabaseAsyncRequestDetails($input: GetOracleAsyncRequestStatusInput!) {
        oracleDatabaseAsyncRequestDetails(input: $input) {
            id
            nodeId
            status
            startTime
            endTime
            progress
            error {
                message
            }
        }
    }
    """

    variables = {
        "input": {
            "id": job_id,
            "clusterUuid": cluster_id
        }
    }

    while True:
        response = gql.query(query, variables)
        details = response['data']['oracleDatabaseAsyncRequestDetails']

        status = details['status']
        progress = details.get('progress', 0)

        print(f"INFO: Job status: {status} ({progress}%)")

        if status == "FAILED":
            error_msg = details.get('error', {}).get('message', 'Unknown error')
            print(f"ERROR: Files-only mount FAILED: {error_msg}", file=sys.stderr)
            print(json.dumps(response, indent=2))
            sys.exit(2)
        elif status == "CANCELLED":
            print("WARN: Files-only mount CANCELLED")
            sys.exit(3)
        elif status == "SUCCEEDED":
            print("INFO: Files-only mount SUCCEEDED")
            print(json.dumps(response, indent=2))
            return

        time.sleep(15)

def main():
    parser = argparse.ArgumentParser(
        description="Mount Oracle database files-only using Rubrik Security Cloud",
        formatter_class=argparse.RawDescriptionHelpFormatter,
        epilog="""
Examples:
  python mount_oracle_filesonly.py --targethost target-host --mountpath /tmp/mount SHED
  python mount_oracle_filesonly.py --targethost target-host --mountpath /tmp/mount --timestamp "2025-11-25 12:00:00" SHED
        """
    )

    parser.add_argument("--targethost", required=True,
                       help="Target host where the files will be mounted")
    parser.add_argument("--mountpath", required=True,
                       help="Target mount path for the files")
    parser.add_argument("--timestamp",
                       help="Optional timestamp for the recovery point in format 'YYYY-MM-DD HH:MM:SS'")
    parser.add_argument("srcdb",
                       help="Source database name or RSC database ID")

    args = parser.parse_args()

    try:
        # Find the source database
        print(f"INFO: Finding source database: {args.srcdb}")
        db = find_database_by_name_or_id(args.srcdb)
        print(f"INFO: Found database: {db['dbUniqueName']} (ID: {db['id']})")
        print(f"INFO: Cluster: {db['cluster']['name']} (ID: {db['cluster']['id']})")

        # Get recovery timestamp
        if args.timestamp:
            print(f"INFO: Using specified timestamp: {args.timestamp}")
            try:
                dt = datetime.strptime(args.timestamp, '%Y-%m-%d %H:%M:%S')
                recovery_timestamp_ms = int(dt.timestamp() * 1000)
                print(f"INFO: Recovery timestamp: {recovery_timestamp_ms} ms")
            except ValueError as e:
                print(f"ERROR: Invalid timestamp format. Use 'YYYY-MM-DD HH:MM:SS': {e}", file=sys.stderr)
                sys.exit(1)
        else:
            print("INFO: No timestamp specified, using latest PIT")
            recovery_timestamp_ms = get_latest_pit(db['id'])

        # Get target host ID
        print(f"INFO: Resolving target host: {args.targethost}")
        target_host_id = get_oracle_host_id(args.targethost, db['cluster']['id'])
        print(f"INFO: Target host ID: {target_host_id}")

        # Execute the files-only mount
        print(f"INFO: Starting files-only mount to path '{args.mountpath}'")
        job_id = mount_files_only(
            db['id'],
            target_host_id,
            recovery_timestamp_ms,
            args.mountpath
        )

        print(f"INFO: Mount job started with ID: {job_id}")

        # Monitor the job
        monitor_job_status(job_id, db['cluster']['id'])

    except Exception as e:
        print(f"ERROR: {e}", file=sys.stderr)
        sys.exit(1)

if __name__ == "__main__":
    main()