-- ClickHouse queries for file inventory analysis

-- Create database (done automatically by scanner)
CREATE DATABASE IF NOT EXISTS files;

-- Schema (done automatically by scanner)
CREATE TABLE IF NOT EXISTS files.inventory (
    scan_id    String,
    scan_time  DateTime64(3),
    server     LowCardinality(String),
    source     LowCardinality(String),
    folder     String,
    filename   String,
    ext        LowCardinality(String),
    size       UInt64,
    created    DateTime64(3),
    modified   DateTime64(3),
    hash       String DEFAULT ''
) ENGINE = MergeTree
ORDER BY (server, folder, filename);

-- Summary by server
SELECT server, count(*) as files, formatReadableSize(sum(size)) as total_size
FROM files.inventory
GROUP BY server
ORDER BY sum(size) DESC;

-- Find exact duplicates (after running hashupdate)
SELECT 
    hash, 
    count(*) as cnt, 
    formatReadableSize(any(size)) as size,
    groupArray(concat(server, ':', folder, '/', filename)) as files
FROM files.inventory 
WHERE hash != '' 
GROUP BY hash 
HAVING cnt > 1 
ORDER BY any(size) DESC
LIMIT 100;

-- Find files by extension
SELECT ext, count(*) as cnt, formatReadableSize(sum(size)) as total
FROM files.inventory
GROUP BY ext
ORDER BY sum(size) DESC
LIMIT 20;

-- Largest files
SELECT server, folder, filename, formatReadableSize(size) as size
FROM files.inventory
ORDER BY size DESC
LIMIT 50;

-- Find files by name pattern
SELECT server, folder, filename, formatReadableSize(size) as size
FROM files.inventory
WHERE filename ILIKE '%pattern%'
ORDER BY size DESC;

-- Compare two servers - files only on server A
SELECT a.folder, a.filename, formatReadableSize(a.size) as size
FROM files.inventory a
LEFT JOIN files.inventory b ON a.filename = b.filename AND a.size = b.size AND b.server = 'serverB'
WHERE a.server = 'serverA' AND b.filename IS NULL
ORDER BY a.size DESC;

-- Files modified in last 30 days
SELECT server, folder, filename, modified, formatReadableSize(size) as size
FROM files.inventory
WHERE modified > now() - INTERVAL 30 DAY
ORDER BY modified DESC
LIMIT 100;

-- Potential duplicate sizes (before hashing)
SELECT size, count(*) as cnt
FROM files.inventory
WHERE hash = '' AND size > 1000000  -- > 1MB
GROUP BY size
HAVING cnt > 1
ORDER BY size DESC
LIMIT 100;

-- Storage by folder (top level)
SELECT 
    server,
    arrayElement(splitByChar('/', folder), 2) as top_folder,
    count(*) as files,
    formatReadableSize(sum(size)) as total
FROM files.inventory
GROUP BY server, top_folder
ORDER BY sum(size) DESC
LIMIT 50;

-- Clear old scans (keep latest per server)
-- DELETE FROM files.inventory 
-- WHERE (server, scan_time) NOT IN (
--     SELECT server, max(scan_time) FROM files.inventory GROUP BY server
-- );