id
stringlengths
25
25
kind
stringclasses
1 value
title
stringlengths
17
17
provisional
bool
1 class
output_code
stringlengths
60
4.12k
input_data_sample
stringlengths
3
637
output_data_sample
unknown
transformation_instruction
stringlengths
54
1.91k
cmsx8ire8004kkup2f9xsd0kf
contributor_item
Submission XSD0KF
false
import csv, io, json def infer_type(values): def is_int(v): try: int(v) return True except ValueError: return False def is_float(v): try: float(v) return True except ValueError: return False if all(is_in...
id,price,label 1,9.99,shoe 2,14.50,hat 3,3.00,sock
{ "id": "int", "price": "float", "label": "str" }
Given a plain CSV, output a JSON object reporting each column's name and its inferred type ('int', 'float', or 'str') based on scanning all rows.
cmsx8ire9004nkup287dy24d5
contributor_item
Submission DY24D5
false
import json, re def transform(text): pattern = r'[\w.+-]+@[\w-]+\.[\w.-]+' found = re.findall(pattern, text) seen = [] for f in found: if f not in seen: seen.append(f) return json.dumps(seen)
Contact alice@example.com or bob@example.org for details. Cc: alice@example.com always.
[ "alice@example.com", "bob@example.org" ]
Given a block of text containing email addresses scattered among other words, extract all valid email addresses into a JSON array, preserving order and removing duplicates.
cmsx8ire9004okup2atgwmt0c
contributor_item
Submission GWMT0C
false
import json def transform(text): data = json.loads(text) subtotal = sum(item['price'] * item['qty'] for item in data['items']) tax = round(subtotal * 0.08, 2) subtotal = round(subtotal, 2) total = round(subtotal + tax, 2) return json.dumps({"subtotal": subtotal, "tax": tax, "total": total})
{"items": [{"name": "Book", "price": 12.99, "qty": 2}, {"name": "Pen", "price": 1.50, "qty": 3}]}
{ "subtotal": 30.48, "tax": 2.44, "total": 32.92 }
Given a JSON object representing a shopping cart ({items: [{name, price, qty}]}), compute the subtotal, a flat 8% tax, and the total, each rounded to 2 decimals, and return as JSON.
cmsx8ire9004mkup2k37b7use
contributor_item
Submission 7B7USE
false
import json from collections import defaultdict def transform(text): data = json.loads(text) totals = defaultdict(int) for d in data: totals[d['sku']] += d['quantity'] return json.dumps(dict(totals))
[{"sku": "A1", "warehouse": "east", "quantity": 10}, {"sku": "A1", "warehouse": "west", "quantity": 5}, {"sku": "B2", "warehouse": "east", "quantity": 7}]
{ "A1": 15, "B2": 7 }
Given a JSON array of {sku, warehouse, quantity} records, produce a JSON object mapping sku to total quantity summed across all warehouses.
cmsx8ire8004ekup2zwiyc1dk
contributor_item
Submission IYC1DK
false
import json def transform(text): data = json.loads(text) pairs = sorted(data.items(), key=lambda kv: kv[1]) return json.dumps([list(p) for p in pairs])
{"Widget": 19.99, "Gadget": 9.99, "Gizmo": 29.50}
[ [ "Gadget", 9.99 ], [ "Widget", 19.99 ], [ "Gizmo", 29.5 ] ]
Given a JSON object mapping product names to prices, return a JSON array of [name, price] pairs sorted by price ascending.
cmsx8ire8003nkup2c8k1gz1e
contributor_item
Submission K1GZ1E
false
import json def flatten(d, prefix=''): out = {} for k, v in d.items(): key = f"{prefix}.{k}" if prefix else k if isinstance(v, dict): out.update(flatten(v, key)) else: out[key] = v return out def transform(text): data = json.loads(text) return json.d...
{"user": {"name": "Dan", "address": {"city": "Austin", "zip": "78701"}}, "active": true}
{ "user.name": "Dan", "user.address.city": "Austin", "user.address.zip": "78701", "active": true }
Flatten a nested JSON object into a single-level dict with dot-separated keys.
cmsx8ire8003rkup248o1zk5t
contributor_item
Submission O1ZK5T
false
import csv, io, json from collections import defaultdict def transform(text): reader = csv.reader(io.StringIO(text.strip())) rows = list(reader)[1:] sums = defaultdict(float) for cat, amt in rows: sums[cat] += float(amt) return json.dumps({k: round(v, 2) for k, v in sums.items()})
category,amount food,12.50 travel,45.00 food,7.25 utilities,60.00 travel,15.75
{ "food": 19.75, "travel": 60.75, "utilities": 60 }
Group a CSV of (category,amount) rows by category and output JSON mapping category to the sum of amounts.
cmsx8ire8003vkup2m8q4yqmo
contributor_item
Submission Q4YQMO
false
import csv, io def transform(text): reader = csv.reader(io.StringIO(text.strip())) rows = list(reader) header = rows[0] age_idx = header.index('age') filtered = [r for r in rows[1:] if int(r[age_idx]) >= 18] out = io.StringIO() writer = csv.writer(out, lineterminator='\n') writer.writer...
name,age Tom,15 Jerry,22 Spike,17 Tyke,19
"name,age\nJerry,22\nTyke,19"
Filter rows of a CSV to only those where the 'age' column is 18 or older, keeping the header, and output as CSV.
cmsx8ire8003ukup2inh4mljl
contributor_item
Submission H4MLJL
false
import json, csv, io from collections import defaultdict def transform(text): data = json.loads(text) regions = sorted(set(d['region'] for d in data)) products = sorted(set(d['product'] for d in data)) table = defaultdict(lambda: defaultdict(int)) for d in data: table[d['region']][d['produc...
[{"region": "East", "product": "Widget", "sales": 100}, {"region": "East", "product": "Gadget", "sales": 50}, {"region": "West", "product": "Widget", "sales": 75}]
"region,Gadget,Widget\nEast,50,100\nWest,0,75"
Convert a JSON array of objects into a pivoted CSV: rows are unique 'region' values, columns are unique 'product' values, cells are 'sales' totals.
cmsx8ire80041kup2odbl1uzu
contributor_item
Submission BL1UZU
false
import json from urllib.parse import urlparse, parse_qs def transform(text): p = urlparse(text.strip()) query = {k: v[0] if len(v) == 1 else v for k, v in parse_qs(p.query).items()} return json.dumps({ "scheme": p.scheme, "netloc": p.netloc, "path": p.path, "query": query ...
https://shop.example.com/products/42?color=red&size=M
{ "scheme": "https", "netloc": "shop.example.com", "path": "/products/42", "query": { "color": "red", "size": "M" } }
Parse a full URL into its components (scheme, netloc, path, query params as a JSON object) and return as JSON.
cmsx8ire80045kup28o59hlu0
contributor_item
Submission 59HLU0
false
import json def transform(text): lines = [l for l in text.strip().split('\n') if l] rows = [] for l in lines: fields = {} for pair in l.split(';'): k, v = pair.split(':', 1) fields[k.strip()] = v.strip() rows.append(fields) return json.dumps(rows)
name:Alice;age:30;city:Reno name:Bob;age:41;city:Provo
[ { "name": "Alice", "age": "30", "city": "Reno" }, { "name": "Bob", "age": "41", "city": "Provo" } ]
Convert semicolon-separated key:value pairs on each line into a JSON array of objects.
cmsx8ire80048kup2dvselwaj
contributor_item
Submission SELWAJ
false
import json def transform(text): nums = json.loads(text) return json.dumps({ "min": min(nums), "max": max(nums), "mean": round(sum(nums) / len(nums), 2), "count": len(nums) })
[4, 8, 15, 16, 23, 42]
{ "min": 4, "max": 42, "mean": 18, "count": 6 }
Given a JSON array of numbers, return a JSON object with min, max, mean (rounded to 2 decimals), and count.
cmsx8ire8003pkup2dd530e6r
contributor_item
Submission 530E6R
false
import json, re from collections import Counter def transform(text): lines = [l for l in text.strip().split('\n') if l] levels = [] for l in lines: m = re.match(r'\[(\w+)\]', l) if m: levels.append(m.group(1)) return json.dumps(dict(Counter(levels)))
[INFO] server started [ERROR] connection failed [INFO] retrying [WARN] slow response [ERROR] timeout
{ "INFO": 2, "ERROR": 2, "WARN": 1 }
Parse a block of log lines like '[LEVEL] message' and return a JSON object counting occurrences of each level.
cmsx8ire8003skup287b1cifu
contributor_item
Submission B1CIFU
false
import json from collections import Counter def transform(text): words = json.loads(text) counts = Counter(w.lower() for w in words) return json.dumps(dict(counts))
["Apple", "banana", "apple", "Cherry", "banana", "apple"]
{ "apple": 3, "banana": 2, "cherry": 1 }
Convert a JSON array of word strings into a JSON object mapping each unique word (lowercased) to its frequency count.
cmsx8ire8003qkup2y46w1tql
contributor_item
Submission 6W1TQL
false
import json def transform(text): lines = [l for l in text.split('\n') if l.strip()] rows = [] for l in lines: name = l[0:10].strip() age = l[10:15].strip() city = l[15:25].strip() rows.append({"name": name, "age": int(age), "city": city}) return json.dumps(rows)
Alice 30 Chicago Bob 45 Denver Cara 27 Miami
[ { "name": "Alice", "age": 30, "city": "Chicago" }, { "name": "Bob", "age": 45, "city": "Denver" }, { "name": "Cara", "age": 27, "city": "Miami" } ]
Convert a fixed-width text table (columns: name 10 chars, age 5 chars, city 10 chars) into JSON array of objects.
cmsx8ire8003xkup2nuorsv6f
contributor_item
Submission ORSV6F
false
import csv, io def transform(text): reader = csv.reader(io.StringIO(text.strip())) rows = list(reader) header, body = rows[0], rows[1:] seen = set() unique = [] for r in body: key = tuple(r) if key not in seen: seen.add(key) unique.append(r) out = io....
sku,qty A1,5 A2,3 A1,5 A3,7 A2,3
"sku,qty\nA1,5\nA2,3\nA3,7"
Remove duplicate rows from a CSV (matching on all columns), keeping only the first occurrence, output as CSV.
cmsx8ire8003ykup24iww1cge
contributor_item
Submission WW1CGE
false
import json, re def to_snake(name): return re.sub(r'(?<!^)(?=[A-Z])', '_', name).lower() def transform(text): data = json.loads(text) return json.dumps({to_snake(k): v for k, v in data.items()})
{"firstName": "Ann", "lastLoginTime": "2026-01-01T00:00:00Z", "isActive": true}
{ "first_name": "Ann", "last_login_time": "2026-01-01T00:00:00Z", "is_active": true }
Convert JSON object keys from camelCase to snake_case, preserving values, for a flat JSON object.
cmsx8ire80046kup212siekac
contributor_item
Submission SIEKAC
false
import json, re def transform(text): lines = [l.strip() for l in text.strip().split('\n') if l.strip()] name, street, last = lines[0], lines[1], lines[2] m = re.match(r'(.+),\s*(\w{2})\s+(\d{5})', last) city, state, zip_code = m.group(1), m.group(2), m.group(3) return json.dumps({ "name": n...
Jane Doe 123 Maple St Springfield, IL 62704
{ "name": "Jane Doe", "street": "123 Maple St", "city": "Springfield", "state": "IL", "zip": "62704" }
Parse a multi-line US-style address block (name, street, 'city, state zip') into a structured JSON object.
cmsx8ire80043kup2ix0cl37b
contributor_item
Submission 0CL37B
false
import json from datetime import datetime def transform(text): lines = [l for l in text.strip().split('\n') if l] result = [] for l in lines: dt = datetime.fromisoformat(l.replace('Z', '+00:00')) if 9 <= dt.hour < 17: result.append(l) return json.dumps(result)
2026-04-01T08:00:00Z 2026-04-01T10:15:00Z 2026-04-01T16:59:00Z 2026-04-01T18:00:00Z
[ "2026-04-01T10:15:00Z", "2026-04-01T16:59:00Z" ]
Given a block of ISO 8601 timestamps (one per line), filter to only those between 09:00 and 17:00 UTC and return as a JSON array of the original strings.
cmsx8ire8004fkup2ol72svnh
contributor_item
Submission 72SVNH
false
import csv, io def transform(text): reader = csv.DictReader(io.StringIO(text.strip())) rows = [] for row in reader: first, last = row['full_name'].split(' ', 1) rows.append({'first_name': first, 'last_name': last, 'email': row['email']}) out = io.StringIO() writer = csv.DictWriter(o...
full_name,email John Smith,john@example.com Mary Ann Lee,mary@example.com
"first_name,last_name,email\nJohn,Smith,john@example.com\nMary,Ann Lee,mary@example.com"
Given a CSV with a 'full_name' column, split it into 'first_name' and 'last_name' columns and output the modified CSV (dropping full_name).
cmsx8ire8004gkup2rxx3t7jm
contributor_item
Submission X3T7JM
false
import json def transform(text): data = json.loads(text) leaves = [] def walk(node): children = node.get('children') or [] if not children: leaves.append(node['name']) else: for c in children: walk(c) for root in data: walk(root) ...
[{"name": "Electronics", "children": [{"name": "Phones", "children": []}, {"name": "Laptops", "children": [{"name": "Gaming", "children": []}]}]}]
[ "Phones", "Gaming" ]
Given a JSON array of nested category trees ({name, children:[...]}), return a JSON array of all leaf node names (nodes with no children).
cmsx8ire8004jkup2qlmbnnoj
contributor_item
Submission MBNNOJ
false
import json from datetime import datetime def transform(text): data = json.loads(text) data.sort(key=lambda d: datetime.fromisoformat(d['timestamp'].replace('Z', '+00:00'))) return json.dumps(data)
[{"timestamp": "2026-02-01T12:00:00Z", "event": "logout"}, {"timestamp": "2026-02-01T08:00:00Z", "event": "login"}, {"timestamp": "2026-02-01T09:30:00Z", "event": "click"}]
[ { "timestamp": "2026-02-01T08:00:00Z", "event": "login" }, { "timestamp": "2026-02-01T09:30:00Z", "event": "click" }, { "timestamp": "2026-02-01T12:00:00Z", "event": "logout" } ]
Given a JSON array of log event objects with 'timestamp' (ISO 8601) and 'event', sort them chronologically and return the sorted JSON array.
cmsx8ire9004pkup233nj9h88
contributor_item
Submission NJ9H88
false
import json from collections import Counter def transform(text): lines = [l for l in text.strip().split('\n') if l] statuses = [] for l in lines: parts = l.split() statuses.append(parts[-1]) return json.dumps(dict(Counter(statuses)))
10.0.0.1 - GET /home 200 10.0.0.2 - POST /login 401 10.0.0.1 - GET /about 200 10.0.0.3 - GET /missing 404
{ "200": 2, "401": 1, "404": 1 }
Given lines of 'IP - method path status' access-log entries, return a JSON object mapping each HTTP status code (as string) to the number of occurrences.
cmsx8ire8003tkup2z6xdaheb
contributor_item
Submission XDAHEB
false
import csv, io def transform(text): reader = csv.reader(io.StringIO(text.strip())) rows = list(reader) header = rows[0] lines = ['| ' + ' | '.join(header) + ' |'] lines.append('| ' + ' | '.join(['---'] * len(header)) + ' |') for row in rows[1:]: lines.append('| ' + ' | '.join(row) + ' |...
name,score Alice,88 Bob,74
"| name | score |\n| --- | --- |\n| Alice | 88 |\n| Bob | 74 |"
Convert a CSV table into a Markdown table string with a header separator row.
cmsxm07z400mvkup2b1p6hw5r
contributor_item
Submission P6HW5R
false
import json def transform(input): lines = input.strip().splitlines() import re stats = {} for line in lines: m = re.search(r'service=(\w+) path=(\S+) status=(\d+) latency_ms=(\d+)', line) if not m or m.group(2) == '/health': continue service, status, latency = m.grou...
2026-08-18T10:00:01Z service=auth path=/login status=200 latency_ms=84 2026-08-18T10:00:02Z service=auth path=/login status=503 latency_ms=420 2026-08-18T10:00:03Z service=payments path=/charge status=201 latency_ms=190 2026-08-18T10:00:04Z service=payments path=/charge status=500 latency_ms=610 2026-08-18T10:00:05Z se...
{ "auth": { "error_rate_pct": 50, "avg_latency_ms": 252 }, "payments": { "error_rate_pct": 33.3, "avg_latency_ms": 318.3 } }
Parse the supplied raw log text and compute error rate and latency by service, excluding health checks. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400n3kup2o8xf0rfy
contributor_item
Submission XF0RFY
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: [0, 0]) for line in lines: f = dict(x.split('=', 1) for x in line.split()) key = (f['service'], f['phase']); d[key][0] += 1; d[key][1] += int(f['status']) >= ...
phase=before status=200 service=api phase=before status=200 service=api phase=before status=500 service=api phase=after status=200 service=api phase=after status=500 service=api phase=after status=503 service=api phase=before status=200 service=worker phase=after status=200 service=worker
{ "api": { "before_error_pct": 33.3, "after_error_pct": 66.7, "delta_points": 33.3 }, "worker": { "before_error_pct": 0, "after_error_pct": 0, "delta_points": 0 } }
Parse the supplied raw log text and compute deployment error regression compared with pre-deploy traffic. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400n5kup2b25pfpdj
contributor_item
Submission 5PFPDJ
false
import json def transform(input): lines = input.strip().splitlines() from datetime import datetime open_at = {}; totals = {}; episodes = {} for line in lines: stamp, cfield, sfield = line.split(); c = cfield.split('=')[1]; state = sfield.split('=')[1] t = datetime.fromisoformat(stamp.re...
2026-08-18T13:00:00Z circuit=payments state=CLOSED 2026-08-18T13:01:10Z circuit=payments state=OPEN 2026-08-18T13:02:00Z circuit=payments state=HALF_OPEN 2026-08-18T13:02:20Z circuit=payments state=CLOSED 2026-08-18T13:05:00Z circuit=search state=OPEN 2026-08-18T13:07:30Z circuit=search state=CLOSED
{ "payments": { "open_episodes": 1, "total_open_seconds": 70 }, "search": { "open_episodes": 1, "total_open_seconds": 150 } }
Parse the supplied raw log text and compute circuit breaker open intervals and total outage seconds. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nkkup2l3u3zovm
contributor_item
Submission U3ZOVM
false
import json def transform(input): lines = input.strip().splitlines() import os from collections import defaultdict groups = {'.jpg':'image','.png':'image','.js':'script','.html':'document'} d = defaultdict(lambda: {'served': 0, 'client_error': 0}) for line in lines: f = dict(x.split('='...
path=/img/a.jpg status=200 bytes=4000 path=/img/b.png status=404 bytes=220 path=/app/main.js status=200 bytes=9000 path=/app/old.js status=404 bytes=310 path=/docs/readme.html status=200 bytes=1500 path=/img/c.jpg status=206 bytes=1800
{ "document": { "served": 1500, "client_error": 0 }, "image": { "served": 5800, "client_error": 220 }, "script": { "served": 9000, "client_error": 310 } }
Parse the supplied raw log text and compute cDN bandwidth by content type and 4xx waste. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400n4kup2w7aer95n
contributor_item
Submission AER95N
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict seen = defaultdict(lambda: {'active': set(), 'buyers': set(), 'purchases': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); row = seen[f['campaign']] row['acti...
campaign=spring user=u1 event=view campaign=spring user=u1 event=buy campaign=spring user=u2 event=view campaign=spring user=u3 event=buy campaign=summer user=u4 event=view campaign=summer user=u5 event=view campaign=summer user=u5 event=buy campaign=summer user=u5 event=buy
{ "spring": { "unique_users": 3, "buyer_conversion_pct": 66.7, "purchases": 2 }, "summer": { "unique_users": 2, "buyer_conversion_pct": 50, "purchases": 2 } }
Parse the supplied raw log text and compute unique active users and conversion rate by campaign. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400mykup28ied32fd
contributor_item
Submission ED32FD
false
import json def transform(input): lines = input.strip().splitlines() import math from collections import defaultdict lat = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()) if int(f['status']) < 500: lat[f['endpoint']].append(int(f['lat...
endpoint=/search latency_ms=90 status=200 endpoint=/search latency_ms=120 status=200 endpoint=/search latency_ms=410 status=200 endpoint=/search latency_ms=230 status=504 endpoint=/export latency_ms=800 status=200 endpoint=/export latency_ms=1200 status=200 endpoint=/export latency_ms=950 status=500 endpoint=/export la...
{ "/export": { "p95_ms": 1200, "over_500ms": 3 }, "/search": { "p95_ms": 410, "over_500ms": 0 } }
Parse the supplied raw log text and compute per-endpoint p95 and slow-request count. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400mzkup2kcctcgkf
contributor_item
Submission CTCGKF
false
import json def transform(input): lines = input.strip().splitlines() from datetime import datetime sessions = {} for line in lines: stamp, sfield, efield = line.split() sid = sfield.split('=', 1)[1]; event = efield.split('=', 1)[1] row = sessions.setdefault(sid, {'events': 0}) ...
2026-08-18T09:00:00Z session=s1 event=start 2026-08-18T09:00:12Z session=s1 event=click 2026-08-18T09:01:05Z session=s1 event=end 2026-08-18T09:02:00Z session=s2 event=start 2026-08-18T09:04:30Z session=s2 event=end 2026-08-18T09:05:00Z session=s3 event=start 2026-08-18T09:05:20Z session=s3 event=click
{ "s1": { "duration_seconds": 65, "events": 3 }, "s2": { "duration_seconds": 150, "events": 2 } }
Parse the supplied raw log text and compute session duration and event count, excluding incomplete sessions. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500ndkup25y1hquzm
contributor_item
Submission 1HQUZM
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: [0, 0]) for line in lines: f = dict(x.split('=', 1) for x in line.split()) if f['maintenance'] == 'true': continue d[f['zone']][0] += 1; d[f['zone']][...
zone=a result=ok maintenance=false zone=a result=fail maintenance=false zone=a result=fail maintenance=true zone=a result=ok maintenance=false zone=b result=ok maintenance=false zone=b result=ok maintenance=false zone=b result=fail maintenance=false
{ "a": { "eligible_checks": 3, "availability_pct": 66.67 }, "b": { "eligible_checks": 3, "availability_pct": 66.67 } }
Parse the supplied raw log text and compute availability by zone with maintenance excluded. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z700o5kup2evavef7b
contributor_item
Submission AVEF7B
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(dict) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['token']][f['event']] = int(f['ts']) lifetimes = []; revoked = [] for events in d.values...
token=a event=issued ts=100 token=a event=revoked ts=460 token=b event=issued ts=200 token=b event=expired ts=800 token=c event=issued ts=300 token=c event=revoked ts=320
{ "tokens_completed": 3, "median_lifetime_s": 360, "max_revocation_lag_s": 360 }
Parse the supplied raw log text and compute token issuance lifetime and revocation lag. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400n2kup2of5d25u9
contributor_item
Submission 5D25U9
false
import json def transform(input): lines = input.strip().splitlines() import re from collections import defaultdict d = defaultdict(list) for line in lines: m = re.match(r'duration=(\d+)ms rows=(\d+) sql="(.+)"', line) duration, sql = int(m.group(1)), m.group(3) fingerprint =...
duration=42ms rows=1 sql="SELECT * FROM users WHERE id=17" duration=380ms rows=1 sql="SELECT * FROM users WHERE id=22" duration=510ms rows=80 sql="SELECT * FROM orders WHERE account_id=9" duration=620ms rows=65 sql="SELECT * FROM orders WHERE account_id=14" duration=75ms rows=1 sql="SELECT * FROM users WHERE id=31"
{ "SELECT * FROM orders WHERE account_id=?": { "calls": 2, "slow_calls": 2, "max_ms": 620 }, "SELECT * FROM users WHERE id=?": { "calls": 3, "slow_calls": 1, "max_ms": 380 } }
Parse the supplied raw log text and compute database slow-query summary by normalized statement fingerprint. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400n8kup27r5fle0t
contributor_item
Submission 5FLE0T
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['request_id']].append((int(f['status']), f['host'])) dupes = {rid: rows for rid, rows in d.i...
request_id=r1 status=200 host=a request_id=r2 status=500 host=a request_id=r1 status=200 host=b request_id=r3 status=201 host=b request_id=r2 status=200 host=c request_id=r4 status=404 host=a request_id=r4 status=404 host=b
{ "duplicate_ids": [ "r1", "r2", "r4" ], "conflicting_ids": [ "r2" ], "duplicate_log_lines": 3 }
Parse the supplied raw log text and compute detect duplicate request IDs and conflicting outcomes. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nckup2kd7t2n81
contributor_item
Submission 7T2N81
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[int(f['partition'])].append(int(f['lag'])) result = {str(p): {'peak_lag': max(v), 'net_change'...
time=1 partition=0 lag=120 time=2 partition=0 lag=90 time=3 partition=0 lag=40 time=1 partition=1 lag=20 time=2 partition=1 lag=85 time=3 partition=1 lag=60 time=1 partition=2 lag=0 time=2 partition=2 lag=0
{ "0": { "peak_lag": 120, "net_change": -80, "recovered_pct": 66.7 }, "1": { "peak_lag": 85, "net_change": 40, "recovered_pct": 0 }, "2": { "peak_lag": 0, "net_change": 0, "recovered_pct": 0 } }
Parse the supplied raw log text and compute consumer lag recovery and peak by partition. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500ngkup2o6kdvoo1
contributor_item
Submission KDVOO1
false
import json def transform(input): lines = input.strip().splitlines() from collections import Counter first = {} for line in lines: f = dict(x.split('=', 1) for x in line.split()) if f['payment'] not in first or int(f['attempt']) < int(first[f['payment']]['attempt']): first[f['payment']]...
payment=p1 attempt=1 outcome=decline reason=insufficient_funds payment=p1 attempt=2 outcome=approved reason=none payment=p2 attempt=1 outcome=decline reason=expired_card payment=p3 attempt=1 outcome=approved reason=none payment=p4 attempt=1 outcome=decline reason=insufficient_funds payment=p4 attempt=2 outcome=decline ...
{ "first_attempt_declines": 3, "reason_share_pct": { "expired_card": 33.3, "insufficient_funds": 66.7 } }
Parse the supplied raw log text and compute payment authorization decline reason share, excluding retries. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nhkup2ge0yz0r5
contributor_item
Submission 0YZ0R5
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'total': 0, 'malformed': 0, 'valid_bytes': 0}) for line in lines: parts = [x.split('=', 1) for x in line.split() if '=' in x]; f = dict(parts); row = d.get(f.get('so...
source=app bytes=120 level=INFO source=app bytes=oops level=ERROR source=worker bytes=80 level=INFO source=worker level=WARN source=app bytes=220 level=ERROR source=worker bytes=140 level=INFO
{}
Parse the supplied raw log text and compute log ingestion malformed-rate and valid byte totals by source. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nlkup2laghshh3
contributor_item
Submission GHSHH3
false
import json def transform(input): lines = input.strip().splitlines() import statistics from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['priority']].append(int(f['delivered']) - int(f['promised'])) result =...
priority=high promised=100 delivered=95 priority=high promised=110 delivered=125 priority=high promised=120 delivered=145 priority=low promised=200 delivered=230 priority=low promised=220 delivered=210 priority=low promised=240 delivered=260
{ "high": { "median_delay": 15, "late_pct": 66.7, "worst_delay": 25 }, "low": { "median_delay": 20, "late_pct": 66.7, "worst_delay": 30 } }
Parse the supplied raw log text and compute median delivery delay and late share by priority. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500njkup2cdemxpwp
contributor_item
Submission EMXPWP
false
import json def transform(input): lines = input.strip().splitlines() from datetime import datetime breach = {}; delays = {} for line in lines: stamp, s, cpu, action = line.split(); service = s.split('=')[1]; value = int(cpu.split('=')[1]); act = action.split('=')[1] t = datetime.fromiso...
2026-08-18T14:00:00Z service=api cpu=72 action=none 2026-08-18T14:00:30Z service=api cpu=86 action=none 2026-08-18T14:01:10Z service=api cpu=91 action=scale_out 2026-08-18T14:02:00Z service=worker cpu=83 action=none 2026-08-18T14:03:45Z service=worker cpu=88 action=scale_out
{ "delay_seconds": { "api": 40, "worker": 105 }, "slowest_service": "worker" }
Parse the supplied raw log text and compute autoscaling response delay from threshold breach to scale action. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z600ntkup2atga7all
contributor_item
Submission GA7ALL
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['queue']].append((f['id'], int(f['now']) - int(f['enqueued']))) result = {} for q, rows ...
now=1000 queue=orders id=a enqueued=990 now=1000 queue=orders id=b enqueued=920 now=1000 queue=orders id=c enqueued=600 now=1000 queue=users id=d enqueued=970 now=1000 queue=users id=e enqueued=850
{ "orders": { "age_buckets": { "under_60s": 1, "60_to_299s": 1, "300s_plus": 1 }, "oldest_id": "c", "oldest_age_s": 400 }, "users": { "age_buckets": { "under_60s": 1, "60_to_299s": 1, "300s_plus": 0 }, "oldest_id": "e", "oldest_age_s": 150 } }
Parse the supplied raw log text and compute dead-letter queue age buckets and oldest message. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z600nukup2lok9dvi6
contributor_item
Submission K9DVI6
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: [0, 0, 0]) for line in lines: f = dict(x.split('=', 1) for x in line.split()) if int(f['status']) != 200: continue row = d[f['encoding']]; row[0] += i...
encoding=gzip raw=10000 sent=3200 status=200 encoding=gzip raw=8000 sent=2800 status=200 encoding=br raw=12000 sent=3000 status=200 encoding=br raw=5000 sent=0 status=304 encoding=identity raw=2000 sent=2000 status=200
{ "br": { "responses": 1, "bytes_saved": 9000, "reduction_pct": 75 }, "gzip": { "responses": 2, "bytes_saved": 12000, "reduction_pct": 66.7 }, "identity": { "responses": 1, "bytes_saved": 0, "reduction_pct": 0 } }
Parse the supplied raw log text and compute compression savings by encoding. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z600nxkup2hkpi61xu
contributor_item
Submission PI61XU
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['pool']].append((int(f['used']), int(f['max']))) result = {} for pool, rows in sorted(d....
pool=main used=7 max=10 pool=main used=9 max=10 pool=main used=10 max=10 pool=main used=8 max=10 pool=analytics used=4 max=5 pool=analytics used=5 max=5 pool=analytics used=5 max=5
{ "analytics": { "peak_utilization_pct": 100, "saturation_samples": 2, "episodes": 1 }, "main": { "peak_utilization_pct": 100, "saturation_samples": 1, "episodes": 1 } }
Parse the supplied raw log text and compute connection pool saturation episodes. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z600nykup2vq5n2qcx
contributor_item
Submission 5N2QCX
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict weights = {'express': 2, 'standard': 1}; d = defaultdict(lambda: {'shipments': 0, 'breaches': 0, 'weighted_delay': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); del...
carrier=x tier=express promised=2 actual=3 carrier=x tier=standard promised=5 actual=5 carrier=x tier=express promised=2 actual=6 carrier=y tier=express promised=2 actual=2 carrier=y tier=standard promised=5 actual=7
{ "x": { "breach_pct": 66.7, "weighted_delay_days": 10 }, "y": { "breach_pct": 50, "weighted_delay_days": 2 } }
Parse the supplied raw log text and compute shipping SLA breach by carrier and weighted delay. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z600nwkup2z0k3jyuh
contributor_item
Submission K3JYUH
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'lags': [], 'missing': 0, 'total': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); r = d[f['region']]; r['total'] += 1 if f['replica'...
tx=t1 primary=100 replica=108 region=east tx=t2 primary=120 replica=155 region=east tx=t3 primary=140 replica=- region=east tx=t4 primary=200 replica=212 region=west tx=t5 primary=220 replica=225 region=west
{ "east": { "max_lag": 35, "avg_lag": 21.5, "missing_ack_pct": 33.3 }, "west": { "max_lag": 12, "avg_lag": 8.5, "missing_ack_pct": 0 } }
Parse the supplied raw log text and compute replication latency with missing acknowledgements. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z700o8kup2j3ywf7v4
contributor_item
Submission YWF7V4
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict minutes = defaultdict(dict) for line in lines: f = dict(x.split('=', 1) for x in line.split()); minutes[int(f['minute'])][f['region']] = (int(f['requests']), int(f['errors'])) result = {...
minute=1 region=primary requests=900 errors=9 minute=1 region=secondary requests=100 errors=1 minute=2 region=primary requests=500 errors=20 minute=2 region=secondary requests=500 errors=5 minute=3 region=primary requests=100 errors=8 minute=3 region=secondary requests=900 errors=9
{ "1": { "secondary_traffic_pct": 10, "secondary_error_pct": 1 }, "2": { "secondary_traffic_pct": 50, "secondary_error_pct": 1 }, "3": { "secondary_traffic_pct": 90, "secondary_error_pct": 1 }, "shift_points": 80 }
Parse the supplied raw log text and compute regional failover traffic shift. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z700o4kup2vyallc2z
contributor_item
Submission ALLC2Z
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'tp': 0, 'fp': 0, 'fn': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); r = d[f['rule']] r['tp'] += f['decision'] == 'block' and f['r...
rule=velocity decision=block review=fraud rule=velocity decision=block review=legit rule=velocity decision=allow review=fraud rule=geo decision=block review=fraud rule=geo decision=block review=fraud rule=geo decision=allow review=legit
{ "geo": { "precision_pct": 100, "recall_pct": 100 }, "velocity": { "precision_pct": 50, "recall_pct": 50 } }
Parse the supplied raw log text and compute fraud rule precision on reviewed decisions. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nnkup2f5l9m7gz
contributor_item
Submission L9M7GZ
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); event = (int(f['term']), f['leader']) if not d[f['cluster']] or d[f['cluster']][-1] != event...
cluster=c1 term=7 leader=n1 cluster=c1 term=8 leader=n2 cluster=c1 term=9 leader=n1 cluster=c2 term=3 leader=n4 cluster=c2 term=3 leader=n4 cluster=c2 term=4 leader=n5
{ "c1": { "elections": 2, "unique_leaders": 2, "latest_term": 9 }, "c2": { "elections": 1, "unique_leaders": 2, "latest_term": 4 } }
Parse the supplied raw log text and compute leader election instability by cluster. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z600nvkup22tsge1s1
contributor_item
Submission SGE1S1
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['user']].append(f['event']) locked = [u for u, e in d.items() if 'lock' in e] post_lock_...
user=a event=fail ip=1.1.1.1 user=a event=fail ip=1.1.1.1 user=a event=lock ip=1.1.1.1 user=a event=fail ip=1.1.1.1 user=b event=fail ip=2.2.2.2 user=b event=success ip=2.2.2.2 user=c event=lock ip=3.3.3.3 user=c event=success ip=3.3.3.3
{ "locked_users": [ "a", "c" ], "post_lock_attempts": 2, "post_lock_successes": 1 }
Parse the supplied raw log text and compute authentication lockout effectiveness. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z700o1kup2sip5vv6y
contributor_item
Submission P5VV6Y
false
import json def transform(input): lines = input.strip().splitlines() import re from collections import defaultdict d = defaultdict(lambda: {'total': 0, 'zero': 0, 'lat': []}) for line in lines: m = re.match(r'query="([^"]+)" results=(\d+) latency=(\d+)', line); words = len(m.group(1).split(...
query="red shoes" results=12 latency=50 query="red hat" results=0 latency=40 query="wireless noise cancelling headphones" results=4 latency=120 query="very specific antique brass fixture" results=0 latency=150 query="pen" results=0 latency=15
{ "long": { "zero_result_pct": 50, "avg_latency_ms": 135 }, "short": { "zero_result_pct": 66.7, "avg_latency_ms": 35 } }
Parse the supplied raw log text and compute search zero-result rate by normalized query length bucket. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z700o6kup2rj42yb73
contributor_item
Submission 42YB73
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'play': 0, 'stall': 0, 'sessions': 0, 'bad': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); r = d[f['device']]; play = int(f['play_ms']); st...
device=mobile session=s1 play_ms=60000 stall_ms=3000 device=mobile session=s2 play_ms=30000 stall_ms=0 device=tv session=s3 play_ms=120000 stall_ms=12000 device=tv session=s4 play_ms=90000 stall_ms=4500
{ "mobile": { "stall_ratio_pct": 3.33, "bad_sessions": 0, "sessions": 2 }, "tv": { "stall_ratio_pct": 7.86, "bad_sessions": 1, "sessions": 2 } }
Parse the supplied raw log text and compute video playback stall ratio by device. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z700o2kup26vklncgg
contributor_item
Submission KLNCGG
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict order = {'compile': 0, 'test': 1, 'deploy': 2}; d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['build']].append(f) result = {} for buil...
build=b1 stage=compile duration=40 outcome=ok build=b1 stage=test duration=80 outcome=fail build=b1 stage=deploy duration=0 outcome=skipped build=b2 stage=compile duration=35 outcome=ok build=b2 stage=test duration=70 outcome=ok build=b2 stage=deploy duration=25 outcome=ok
{ "b1": { "completed_duration": 120, "first_failed_stage": "test", "success": false }, "b2": { "completed_duration": 130, "first_failed_stage": null, "success": true } }
Parse the supplied raw log text and compute build pipeline critical failure stage. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400mwkup2sdgtiern
contributor_item
Submission GTIERN
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'hits': 0, 'total': 0, 'saved': 0}) for line in lines: fields = dict(token.split('=', 1) for token in line.split()) r = fields['region']; hit = fields['cache...
ts=10:01 region=us-east cache=HIT bytes=8400 origin_bytes=8400 ts=10:02 region=us-east cache=MISS bytes=1200 origin_bytes=1200 ts=10:03 region=eu-west cache=HIT bytes=5100 origin_bytes=5100 ts=10:04 region=eu-west cache=HIT bytes=3200 origin_bytes=3200 ts=10:05 region=eu-west cache=MISS bytes=700 origin_bytes=700 ts=10...
{ "eu-west": { "hit_ratio": 0.667, "origin_bytes_saved": 8300 }, "us-east": { "hit_ratio": 0.667, "origin_bytes_saved": 10900 } }
Parse the supplied raw log text and compute cache hit ratio and bytes saved per region. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400mxkup24dozcxmu
contributor_item
Submission OZCXMU
false
import json def transform(input): lines = input.strip().splitlines() requests = {} for line in lines: f = dict(x.split('=', 1) for x in line.split()) row = requests.setdefault(f['rid'], {'op': f['op'], 'attempts': 0, 'done': False}) row['attempts'] += 1 row['done'] = row['do...
rid=a1 op=read attempt=1 outcome=retry rid=a1 op=read attempt=2 outcome=ok rid=b2 op=write attempt=1 outcome=retry rid=b2 op=write attempt=2 outcome=retry rid=b2 op=write attempt=3 outcome=ok rid=c3 op=read attempt=1 outcome=ok rid=d4 op=write attempt=1 outcome=retry
{ "read": { "completed": 2, "retry_amplification": 1.5 }, "write": { "completed": 1, "retry_amplification": 3 } }
Parse the supplied raw log text and compute retry amplification by operation, counting only completed request ids. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400n0kup2e6rziee5
contributor_item
Submission RZIEE5
false
import json def transform(input): lines = input.strip().splitlines() from collections import Counter counts = Counter(line[:16] for line in lines) peak = max(counts.values()) result = {'busiest_minute': min(k for k, v in counts.items() if v == peak), 'requests': peak, 'minutes_observed': len(counts...
2026-08-18T11:00:01Z GET /a 200 2026-08-18T11:00:30Z GET /b 200 2026-08-18T11:01:02Z POST /c 201 2026-08-18T11:01:17Z GET /a 500 2026-08-18T11:01:45Z GET /a 200 2026-08-18T11:02:03Z GET /b 200 2026-08-18T11:02:44Z GET /b 200
{ "busiest_minute": "2026-08-18T11:01", "requests": 3, "minutes_observed": 3 }
Parse the supplied raw log text and compute peak requests/minute and busiest minute with tie broken chronologically. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400n1kup2zv161s98
contributor_item
Submission 161S98
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict series = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()) series[f['queue']].append((int(f['depth']), int(f['limit']))) result = {} for q,...
time=12:00 queue=email depth=40 limit=100 time=12:01 queue=email depth=115 limit=100 time=12:02 queue=email depth=130 limit=100 time=12:03 queue=email depth=90 limit=100 time=12:00 queue=video depth=210 limit=200 time=12:01 queue=video depth=260 limit=200 time=12:02 queue=video depth=310 limit=200
{ "email": { "net_growth": 50, "longest_breach_streak": 2 }, "video": { "net_growth": 100, "longest_breach_streak": 3 } }
Parse the supplied raw log text and compute queue backlog growth and breach streak. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400n6kup2rpn5fvnn
contributor_item
Submission N5FVNN
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list); max_uptime = 0 for line in lines: f = dict(x.split('=', 1) for x in line.split()); max_uptime = max(max_uptime, int(f['uptime_ms'])) d[f['collector']].append(i...
uptime_ms=10000 collector=young pause_ms=12 uptime_ms=20000 collector=young pause_ms=18 uptime_ms=30000 collector=full pause_ms=240 uptime_ms=40000 collector=young pause_ms=15 uptime_ms=50000 collector=full pause_ms=310
{ "full": { "events": 2, "max_pause_ms": 310, "uptime_share_pct": 1.1 }, "young": { "events": 3, "max_pause_ms": 18, "uptime_share_pct": 0.09 } }
Parse the supplied raw log text and compute gC pause share and maximum pause by collector. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z400n7kup250g5ggyf
contributor_item
Submission G5GGYF
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'total': 0, 'classes': defaultdict(int), 'success_bytes': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); row = d[f['method']]; status = int(...
method=GET status=200 bytes=400 method=GET status=304 bytes=0 method=GET status=404 bytes=120 method=POST status=201 bytes=80 method=POST status=400 bytes=90 method=POST status=503 bytes=40
{ "GET": { "class_counts": { "2xx": 1, "3xx": 1, "4xx": 1 }, "success_bytes": 400 }, "POST": { "class_counts": { "2xx": 1, "4xx": 1, "5xx": 1 }, "success_bytes": 80 } }
Parse the supplied raw log text and compute status-class distribution and success-weighted bytes. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nakup2k9nypzgt
contributor_item
Submission NYPZGT
false
import json def transform(input): lines = input.strip().splitlines() import re from collections import defaultdict d = defaultdict(lambda: {'count': 0, 'hosts': set()}) for line in lines: if 'level=ERROR' not in line: continue host = re.search(r'host=(\w+)', line).group(1); message ...
host=web1 level=ERROR error="Timeout after 120ms" trace=t1 host=web2 level=ERROR error="Timeout after 450ms" trace=t2 host=web1 level=INFO msg="ready" host=web3 level=ERROR error="KeyError user_184" trace=t3 host=web2 level=ERROR error="KeyError user_992" trace=t4
{ "KeyError user_#": { "count": 2, "hosts": [ "web2", "web3" ] }, "Timeout after #ms": { "count": 2, "hosts": [ "web1", "web2" ] } }
Parse the supplied raw log text and compute exception fingerprints with affected hosts. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500n9kup25tzog9n3
contributor_item
Submission ZOG9N3
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['pid']].append(int(f['rss_mb'])) result = {} for pid, values in sorted(d.items()): ...
t=1 pid=api rss_mb=220 t=2 pid=worker rss_mb=310 t=3 pid=api rss_mb=245 t=4 pid=worker rss_mb=295 t=5 pid=api rss_mb=330 t=6 pid=worker rss_mb=410
{ "api": { "high_water_mb": 330, "largest_increase_mb": 85, "net_change_mb": 110 }, "worker": { "high_water_mb": 410, "largest_increase_mb": 115, "net_change_mb": 100 } }
Parse the supplied raw log text and compute memory high-water mark and largest positive delta by process. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nbkup2xwvx605p
contributor_item
Submission VX605P
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['tenant']].append((int(f['minute']), int(f['used']), int(f['limit']))) result = {} for t...
minute=0 tenant=alpha used=100 limit=1000 minute=10 tenant=alpha used=240 limit=1000 minute=20 tenant=alpha used=410 limit=1000 minute=0 tenant=beta used=50 limit=500 minute=10 tenant=beta used=90 limit=500 minute=20 tenant=beta used=150 limit=500
{ "alpha": { "current_pct": 41, "projected_exhaustion_minute": 58.1 }, "beta": { "current_pct": 30, "projected_exhaustion_minute": 90 } }
Parse the supplied raw log text and compute aPI quota consumption and projected exhaustion. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nekup2cfcn8bm5
contributor_item
Submission CN8BM5
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['stream']].append((int(f['seq']), int(f['ts']))) result = {} for stream, rows in sorted(...
stream=orders ts=100 seq=1 stream=orders ts=108 seq=2 stream=orders ts=105 seq=3 stream=orders ts=120 seq=4 stream=users ts=50 seq=1 stream=users ts=45 seq=2 stream=users ts=41 seq=3
{ "orders": { "out_of_order_events": 1, "largest_regression": 3 }, "users": { "out_of_order_events": 2, "largest_regression": 5 } }
Parse the supplied raw log text and compute out-of-order event count and largest timestamp regression per stream. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nfkup23lz2ae0a
contributor_item
Submission Z2AE0A
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'exposed': set(), 'converted': set()}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); key = 'exposed' if f['event'] == 'expose' else 'converted'...
variant=control user=u1 event=expose variant=control user=u1 event=convert variant=control user=u2 event=expose variant=test user=u3 event=expose variant=test user=u4 event=expose variant=test user=u4 event=convert variant=test user=u5 event=expose variant=test user=u5 event=convert
{ "exposures": { "control": 2, "test": 3 }, "conversion_pct": { "control": 50, "test": 66.7 }, "test_lift_points": 16.7 }
Parse the supplied raw log text and compute feature flag exposure imbalance and conversion lift. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nikup2ossrxx91
contributor_item
Submission SRXX91
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'elapsed': 0, 'processed': 0, 'failed': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); row = d[f['job']] for key in ('elapsed_s', 'p...
job=import batch=1 elapsed_s=10 processed=500 failed=5 job=import batch=2 elapsed_s=15 processed=750 failed=15 job=export batch=1 elapsed_s=20 processed=600 failed=0 job=export batch=2 elapsed_s=25 processed=900 failed=9
{ "export": { "throughput_per_s": 33.33, "failed_pct": 0.6000000000000001 }, "import": { "throughput_per_s": 50, "failed_pct": 1.6 } }
Parse the supplied raw log text and compute batch job throughput and failed-record rate. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nmkup2f1qovtq7
contributor_item
Submission QOVTQ7
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['sensor']].append(float(f['value'])) result = {} for sensor, values in sorted(d.items())...
sensor=a value=20.0 sensor=a value=21.0 sensor=a value=20.5 sensor=a value=29.0 sensor=a value=21.5 sensor=b value=10.0 sensor=b value=10.5 sensor=b value=11.0 sensor=b value=12.0
{ "a": { "anomaly_samples": [ 4 ], "range": 9 }, "b": { "anomaly_samples": [], "range": 2 } }
Parse the supplied raw log text and compute rolling three-sample temperature anomaly count. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nqkup2v87icgge
contributor_item
Submission 7ICGGE
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: defaultdict(set)) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['provider']][f['event']].add(f['id']) result = {} for provider, e in...
id=m1 event=queued provider=a id=m1 event=delivered provider=a id=m1 event=open provider=a id=m2 event=queued provider=a id=m2 event=bounced provider=a id=m3 event=queued provider=b id=m3 event=delivered provider=b id=m3 event=open provider=b id=m3 event=open provider=b
{ "a": { "delivery_pct": 50, "unique_open_pct_of_delivered": 100, "bounces": 1 }, "b": { "delivery_pct": 100, "unique_open_pct_of_delivered": 100, "bounces": 0 } }
Parse the supplied raw log text and compute email delivery funnel with unique message ids. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500npkup2ygx5o2hs
contributor_item
Submission X5O2HS
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['trace']].append(f) result = {} for trace, spans in sorted(d.items()): roots = [...
trace=t1 span=root parent=- start=0 end=120 trace=t1 span=db parent=root start=10 end=80 trace=t1 span=cache parent=root start=85 end=100 trace=t2 span=db parent=root start=5 end=55 trace=t2 span=render parent=root start=60 end=90
{ "t1": { "complete_root": true, "missing_parents": [], "observed_duration_ms": 120 }, "t2": { "complete_root": false, "missing_parents": [ "root" ], "observed_duration_ms": 85 } }
Parse the supplied raw log text and compute trace completeness and critical path duration. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nokup2dhgopmw1
contributor_item
Submission GOPMW1
false
import json def transform(input): lines = input.strip().splitlines() windows = {} target = None for line in lines: f = dict(x.split('=', 1) for x in line.split()) if 'window' in f: windows[f['window']] = (int(f['good']), int(f['total'])) else: target = float(f['target']) bud...
window=5m good=940 total=1000 window=1h good=11900 total=12000 service=checkout target=99.9
{ "burn_rate": { "1h": 8.33, "5m": 60 }, "page": true }
Parse the supplied raw log text and compute sLO burn rate across short and long windows. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nrkup2pg95f25i
contributor_item
Submission 95F25I
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'reserve': 0, 'release': 0, 'commit': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['sku']][f['action']] += int(f['qty']) result = ...
sku=A order=o1 action=reserve qty=3 sku=A order=o1 action=release qty=1 sku=A order=o1 action=commit qty=2 sku=A order=o2 action=reserve qty=5 sku=B order=o3 action=reserve qty=4 sku=B order=o3 action=commit qty=3 sku=B order=o3 action=release qty=1
{ "A": { "reserved": 8, "accounted": 3, "leaked": 5 }, "B": { "reserved": 4, "accounted": 4, "leaked": 0 } }
Parse the supplied raw log text and compute inventory reservation leakage. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z500nskup20attr1jk
contributor_item
Submission TTR1JK
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'total': 0, 'failed': 0, 'versions': defaultdict(int), 'ok_ms': []}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); r = d[f['host']]; r['total']...
host=api tls=1.3 outcome=ok ms=35 host=api tls=1.2 outcome=ok ms=55 host=api tls=1.2 outcome=fail ms=80 host=cdn tls=1.3 outcome=ok ms=20 host=cdn tls=1.3 outcome=ok ms=22 host=cdn tls=1.2 outcome=fail ms=100
{ "api": { "failure_pct": 33.3, "version_mix": { "1.2": 2, "1.3": 1 }, "avg_success_ms": 45 }, "cdn": { "failure_pct": 33.3, "version_mix": { "1.2": 1, "1.3": 2 }, "avg_success_ms": 21 } }
Parse the supplied raw log text and compute tLS handshake version mix and failure rate. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z700nzkup2ingw1lze
contributor_item
Submission GW1LZE
false
import json def transform(input): lines = input.strip().splitlines() from datetime import datetime from collections import defaultdict d = defaultdict(list) for line in lines: stamp, w = line.split(); d[w.split('=')[1]].append(datetime.fromisoformat(stamp.replace('Z', '+00:00'))) result...
2026-08-18T15:00:00Z worker=w1 2026-08-18T15:00:30Z worker=w1 2026-08-18T15:02:10Z worker=w1 2026-08-18T15:00:05Z worker=w2 2026-08-18T15:00:50Z worker=w2 2026-08-18T15:01:35Z worker=w2
{ "w1": { "max_gap_seconds": 100, "gaps_over_60s": 1 }, "w2": { "max_gap_seconds": 45, "gaps_over_60s": 0 } }
Parse the supplied raw log text and compute worker heartbeat gaps. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z700o0kup2rk246m5p
contributor_item
Submission 246M5P
false
import json def transform(input): lines = input.strip().splitlines() from decimal import Decimal from collections import defaultdict d = defaultdict(lambda: {'gross': Decimal('0'), 'fees': Decimal('0'), 'count': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()) ...
currency=USD refund=100.00 fee=2.50 outcome=settled currency=USD refund=40.00 fee=1.00 outcome=failed currency=USD refund=25.00 fee=0.75 outcome=settled currency=EUR refund=80.00 fee=2.00 outcome=settled currency=EUR refund=20.00 fee=0.50 outcome=settled
{ "EUR": { "settled": 2, "net": "97.50", "fee_pct": 2.5 }, "USD": { "settled": 2, "net": "121.75", "fee_pct": 2.6 } }
Parse the supplied raw log text and compute refund net amount by currency after fees. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z700o3kup2nlgfiifl
contributor_item
Submission GFIIFL
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'hits': 0, 'total': 0, 'lat': []}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); r = d[f['type']]; r['total'] += 1; r['hits'] += f['cache'] == ...
type=A cache=hit latency_ms=2 type=A cache=miss latency_ms=40 type=A cache=hit latency_ms=3 type=AAAA cache=miss latency_ms=55 type=AAAA cache=miss latency_ms=60 type=AAAA cache=hit latency_ms=4
{ "A": { "hit_pct": 66.7, "mean_latency_ms": 15 }, "AAAA": { "hit_pct": 33.3, "mean_latency_ms": 39.7 } }
Parse the supplied raw log text and compute dNS resolver cache effectiveness by record type. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm07z700o7kup2jpubjj8b
contributor_item
Submission UBJJ8B
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['phase']].append(int(f['wait_ms'])) result = {phase: {'avg_wait_ms': round(sum(v) / len(v), ...
phase=before query=q1 wait_ms=5 phase=before query=q2 wait_ms=8 phase=during query=q3 wait_ms=450 phase=during query=q4 wait_ms=700 phase=during query=q5 wait_ms=20 phase=after query=q6 wait_ms=12 phase=after query=q7 wait_ms=9
{ "after": { "avg_wait_ms": 10.5, "blocked_over_100ms": 0 }, "before": { "avg_wait_ms": 6.5, "blocked_over_100ms": 0 }, "during": { "avg_wait_ms": 390, "blocked_over_100ms": 2 }, "migration_added_avg_ms": 383.5 }
Parse the supplied raw log text and compute schema migration lock impact. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00ogkup2v7y510ku
contributor_item
Submission Y510KU
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict weights = {'low':1,'medium':2,'high':3}; d = defaultdict(lambda: {'drifted': [], 'score': 0, 'total': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); r = d[f['env']];...
env=prod key=timeout expected=30 actual=60 severity=high env=prod key=retries expected=3 actual=3 severity=medium env=stage key=timeout expected=30 actual=25 severity=high env=stage key=region expected=us actual=us severity=low env=stage key=debug expected=false actual=true severity=medium
{ "prod": { "drift_pct": 50, "weighted_score": 3, "keys": [ "timeout" ] }, "stage": { "drift_pct": 66.7, "weighted_score": 5, "keys": [ "debug", "timeout" ] } }
Parse the supplied raw log text and compute configuration drift by environment and key severity. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00oskup237ei9pun
contributor_item
Submission EI9PUN
false
import json def transform(input): lines = input.strip().splitlines() import sqlite3 db=sqlite3.connect(':memory:'); db.execute('create table logs(region text, amount integer, outcome text)') for line in lines: f=dict(x.split('=',1) for x in line.split()); db.execute('insert into logs values(?,?...
region=us amount=120 outcome=commit region=us amount=50 outcome=rollback region=us amount=80 outcome=commit region=eu amount=200 outcome=commit region=eu amount=100 outcome=rollback
{ "eu": { "transactions": 2, "committed": 1, "committed_amount": 200 }, "us": { "transactions": 3, "committed": 2, "committed_amount": 200 } }
Parse the supplied raw log text and compute sQL conditional aggregate for transaction outcomes. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00o9kup2qj1v52um
contributor_item
Submission 1V52UM
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'attempted': 0, 'protected': 0, 'ok_time': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); r = d[f['target']]; size = int(f['bytes']); r['att...
target=db outcome=ok bytes=500 duration_s=50 target=db outcome=fail bytes=200 duration_s=40 target=files outcome=ok bytes=900 duration_s=120 target=files outcome=ok bytes=600 duration_s=90 target=files outcome=fail bytes=300 duration_s=60
{ "db": { "byte_success_pct": 71.4, "successful_throughput": 10 }, "files": { "byte_success_pct": 83.3, "successful_throughput": 7.14 } }
Parse the supplied raw log text and compute backup reliability weighted by protected bytes. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00owkup2n4laskp0
contributor_item
Submission LASKP0
false
import json def transform(input): lines = input.strip().splitlines() import sqlite3 db=sqlite3.connect(':memory:');db.execute('create table logs(source text,severity text)') for line in lines: f=dict(x.split('=',1) for x in line.split());db.execute('insert into logs values(?,?)',(f['source'],f[...
source=cpu severity=warn source=cpu severity=critical source=cpu severity=critical source=disk severity=warn source=disk severity=info source=network severity=critical
[ { "source": "cpu", "total": 3, "critical": 2, "critical_pct": 66.7 }, { "source": "network", "total": 1, "critical": 1, "critical_pct": 100 } ]
Parse the supplied raw log text and compute sQL group/HAVING for noisy alert sources. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00obkup2o6qpra12
contributor_item
Submission QPRA12
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['device']].append((int(f['minute']), int(f['battery']))) result = {} for device, rows in...
device=d1 minute=0 battery=90 device=d1 minute=30 battery=75 device=d1 minute=60 battery=48 device=d2 minute=0 battery=50 device=d2 minute=20 battery=19 device=d2 minute=40 battery=15
{ "d1": { "drain_pct_per_hour": 42, "first_below_20_minute": null }, "d2": { "drain_pct_per_hour": 52.5, "first_below_20_minute": 20 } }
Parse the supplied raw log text and compute battery drain and low-battery crossings by device. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00ohkup2cxvefzy1
contributor_item
Submission VEFZY1
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'clients': set(), 'reconnects': 0}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); r = d[f['version']]; r['clients'].add(f['client']); r['reconn...
version=1.0 client=a event=connect version=1.0 client=a event=reconnect version=1.0 client=b event=connect version=1.0 client=b event=reconnect version=2.0 client=c event=connect version=2.0 client=d event=connect version=2.0 client=d event=reconnect version=2.0 client=d event=reconnect
{ "1.0": { "clients": 2, "reconnects_per_client": 1 }, "2.0": { "clients": 2, "reconnects_per_client": 1 } }
Parse the supplied raw log text and compute webSocket reconnect burden by client version. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00ofkup252igcsu3
contributor_item
Submission IGCSU3
false
import json def transform(input): lines = input.strip().splitlines() result = {} for line in lines: f = dict(x.split('=', 1) for x in line.split()); read = int(f['read_mb']); written = int(f['written_mb']) result[f['shard']] = {'reclaimed_mb': int(f['before_mb']) - int(f['after_mb']), 'writ...
shard=s1 read_mb=500 written_mb=650 before_mb=900 after_mb=600 shard=s2 read_mb=400 written_mb=300 before_mb=700 after_mb=550 shard=s3 read_mb=0 written_mb=0 before_mb=200 after_mb=200
{ "s1": { "reclaimed_mb": 300, "write_amplification": 1.3 }, "s2": { "reclaimed_mb": 150, "write_amplification": 0.75 }, "s3": { "reclaimed_mb": 0, "write_amplification": null } }
Parse the supplied raw log text and compute storage compaction amplification and reclaimed space. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00omkup2ilr2thfg
contributor_item
Submission R2THFG
false
import json def transform(input): lines = input.strip().splitlines() result={} for line in lines: f=dict(x.split('=',1) for x in line.split()); space=int(f['used_gb'])/int(f['total_gb']); inode=int(f['used_inodes'])/int(f['total_inodes']); limiting='space' if space>=inode else 'inodes' resu...
mount=/data used_gb=850 total_gb=1000 used_inodes=40 total_inodes=100 mount=/tmp used_gb=20 total_gb=100 used_inodes=95 total_inodes=100 mount=/logs used_gb=450 total_gb=500 used_inodes=88 total_inodes=100
{ "/data": { "space_pct": 85, "inode_pct": 40, "limiting_resource": "space", "alert": false }, "/logs": { "space_pct": 90, "inode_pct": 88, "limiting_resource": "space", "alert": true }, "/tmp": { "space_pct": 20, "inode_pct": 95, "limiting_resource": "inodes", ...
Parse the supplied raw log text and compute filesystem capacity risk using both bytes and inodes. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00oqkup21u82s5g3
contributor_item
Submission 82S5G3
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d=defaultdict(lambda:defaultdict(int)) for line in lines: f=dict(x.split('=',1) for x in line.split()); d[f['version']][f['session']]+=f['event']=='crash' result={v:{'sessions':len(s),'c...
version=3.1 session=a event=start version=3.1 session=a event=crash version=3.1 session=b event=start version=3.2 session=c event=start version=3.2 session=d event=start version=3.2 session=d event=crash version=3.2 session=d event=crash
{ "3.1": { "sessions": 2, "crash_free_pct": 50, "crash_events": 1 }, "3.2": { "sessions": 2, "crash_free_pct": 50, "crash_events": 2 } }
Parse the supplied raw log text and compute crash-free sessions by application version. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00oukup2suc75nnz
contributor_item
Submission C75NNZ
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d=defaultdict(lambda:{'total':0,'fallback':0,'keys':[]}) for line in lines: f=dict(x.split('=',1) for x in line.split()); r=d[f['locale']]; r['total']+=1 if f['source']!=f['locale']:...
locale=fr key=home.title source=fr locale=fr key=home.cta source=en locale=de key=home.title source=de locale=de key=home.cta source=en locale=de key=help source=en locale=es key=home.title source=es
{ "de": { "fallback_pct": 66.7, "fallback_keys": [ "help", "home.cta" ] }, "es": { "fallback_pct": 0, "fallback_keys": [] }, "fr": { "fallback_pct": 50, "fallback_keys": [ "home.cta" ] } }
Parse the supplied raw log text and compute localization fallback rate by locale. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00ovkup2tdr6bkyj
contributor_item
Submission R6BKYJ
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d=defaultdict(lambda:{'requests':0,'errors':0,'clients':set()}) for line in lines: f=dict(x.split('=',1) for x in line.split()); r=d[f['version']]; status=int(f['status']); r['requests']+=1;...
version=v1 status=200 client=a version=v1 status=500 client=b version=v1 status=200 client=a version=v2 status=200 client=c version=v2 status=201 client=d version=v2 status=400 client=e
{ "v1": { "traffic_share_pct": 50, "error_pct": 33.3, "unique_clients": 2 }, "v2": { "traffic_share_pct": 50, "error_pct": 33.3, "unique_clients": 3 } }
Parse the supplied raw log text and compute aPI-version adoption and error rate. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00ookup2epikg5vo
contributor_item
Submission IKG5VO
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d=defaultdict(lambda:{'delays':[],'missed':0}) for line in lines: f=dict(x.split('=',1) for x in line.split()); r=d[f['job']] if f['actual']=='-':r['missed']+=1 else:r['delay...
job=hourly scheduled=100 actual=105 job=hourly scheduled=200 actual=260 job=hourly scheduled=300 actual=- job=daily scheduled=1000 actual=1010 job=daily scheduled=2000 actual=2005
{ "daily": { "missed": 0, "max_lateness": 10, "late_over_30": 0 }, "hourly": { "missed": 1, "max_lateness": 60, "late_over_30": 1 } }
Parse the supplied raw log text and compute scheduled-job lateness and missed-run detection. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00oakup2dv87z6qt
contributor_item
Submission 87Z6QT
false
import json def transform(input): lines = input.strip().splitlines() rates = {} for line in lines: f = dict(x.split('=', 1) for x in line.split()); allowed = int(f['allowed']); limited = int(f['limited']) rates[f['client']] = allowed / (allowed + limited) values = list(rates.values()); ...
client=a allowed=90 limited=10 client=b allowed=45 limited=5 client=c allowed=40 limited=40 client=d allowed=18 limited=2
{ "allow_pct": { "a": 90, "b": 90, "c": 50, "d": 90 }, "jain_fairness": 0.9552, "worst_client": "c" }
Parse the supplied raw log text and compute rate-limit fairness across clients. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00otkup2mkwfcv3p
contributor_item
Submission WFCV3P
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d=defaultdict(lambda:{'bytes':0,'parts':0,'terminal':None}) for line in lines: f=dict(x.split('=',1) for x in line.split()); r=d[f['upload']] if f['state']=='stored':r['bytes']+=int(...
upload=u1 part=1 bytes=100 state=stored upload=u1 part=2 bytes=120 state=stored upload=u1 part=0 bytes=0 state=complete upload=u2 part=1 bytes=200 state=stored upload=u2 part=2 bytes=180 state=stored upload=u3 part=1 bytes=90 state=stored upload=u3 part=0 bytes=0 state=abort
{ "completed_bytes": 220, "orphaned_bytes": 380, "aborted_uploads": 1 }
Parse the supplied raw log text and compute multipart upload completion and orphaned-byte totals. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00ockup276m7zfts
contributor_item
Submission M7ZFTS
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(dict) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['request']][f['stage']] = int(f['ms']) result = {} for req, stages in sorted(d.items()):...
request=r1 stage=queue ms=20 request=r1 stage=compute ms=80 request=r1 stage=network ms=40 request=r2 stage=queue ms=60 request=r2 stage=compute ms=70 request=r2 stage=network ms=20
{ "r1": { "total_ms": 140, "bottleneck": "compute", "bottleneck_share_pct": 57.1 }, "r2": { "total_ms": 150, "bottleneck": "compute", "bottleneck_share_pct": 46.7 } }
Parse the supplied raw log text and compute request stage contribution to end-to-end latency. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00odkup22yfz5jqh
contributor_item
Submission FZ5JQH
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['hook']].append((int(f['attempt']), int(f['status']))) result = {} for hook, rows in sor...
hook=h1 attempt=1 status=500 hook=h1 attempt=2 status=200 hook=h2 attempt=1 status=429 hook=h2 attempt=2 status=503 hook=h2 attempt=3 status=204 hook=h3 attempt=1 status=400 hook=h3 attempt=2 status=400
{ "h1": { "delivered": true, "attempts_to_delivery": 2, "retryable_failures": 1 }, "h2": { "delivered": true, "attempts_to_delivery": 3, "retryable_failures": 2 }, "h3": { "delivered": false, "attempts_to_delivery": null, "retryable_failures": 0 } }
Parse the supplied raw log text and compute webhook eventual success and attempts-to-delivery. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00oekup26pez9e6v
contributor_item
Submission EZ9E6V
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(list) for line in lines: f = dict(x.split('=', 1) for x in line.split()); d[f['pod']].append((int(f['throttled_ms']), int(f['latency_ms']))) result = {} for pod, rows...
pod=a throttled_ms=0 latency_ms=80 pod=a throttled_ms=20 latency_ms=120 pod=a throttled_ms=40 latency_ms=180 pod=b throttled_ms=0 latency_ms=70 pod=b throttled_ms=10 latency_ms=75
{ "a": { "throttled_samples": 2, "latency_increase_ms": 70 }, "b": { "throttled_samples": 1, "latency_increase_ms": 5 } }
Parse the supplied raw log text and compute cPU throttling impact on latency. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00oikup20h99qukz
contributor_item
Submission 99QUKZ
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d = defaultdict(lambda: {'attempts': 0, 'denied': 0, 'actors': set()}) for line in lines: f = dict(x.split('=', 1) for x in line.split()); kind = f['resource'].split('://')[0]; r=d[kind]; r[...
actor=u1 resource=s3://a action=read outcome=denied actor=u1 resource=s3://b action=write outcome=denied actor=u2 resource=db://orders action=read outcome=allowed actor=u2 resource=db://users action=write outcome=denied actor=u3 resource=s3://c action=read outcome=allowed
{ "db": { "denial_pct": 50, "affected_actors": [ "u2" ] }, "s3": { "denial_pct": 66.7, "affected_actors": [ "u1" ] } }
Parse the supplied raw log text and compute permission-denied concentration by resource class. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00ojkup2imj8tf6s
contributor_item
Submission J8TF6S
false
import json def transform(input): lines = input.strip().splitlines() from datetime import date rows=[] for line in lines: f=dict(x.split('=',1) for x in line.split()); days=(date.fromisoformat(f['expires'])-date.fromisoformat(f['observed'])).days rows.append({'host':f['host'],'days_rema...
observed=2026-08-18 host=api expires=2026-08-25 issuer=A observed=2026-08-18 host=cdn expires=2026-10-01 issuer=B observed=2026-08-18 host=old expires=2026-08-17 issuer=A
{ "certificates": [ { "host": "old", "days_remaining": -1, "risk": "expired" }, { "host": "api", "days_remaining": 7, "risk": "urgent" }, { "host": "cdn", "days_remaining": 44, "risk": "normal" } ], "urgent_or_expired": 2 }
Parse the supplied raw log text and compute certificate expiry risk relative to observation time. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00olkup2wpjum25a
contributor_item
Submission JUM25A
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict,Counter order=['cart','shipping','payment','complete']; d=defaultdict(set) for line in lines: f=dict(x.split('=',1) for x in line.split()); d[f['session']].add(f['event']) abandon=Counte...
session=s1 event=cart session=s1 event=shipping session=s1 event=payment session=s1 event=complete session=s2 event=cart session=s2 event=shipping session=s3 event=cart session=s3 event=shipping session=s3 event=payment
{ "sessions": 3, "completed": 1, "abandoned_after": { "shipping": 1, "payment": 1 } }
Parse the supplied raw log text and compute checkout funnel abandonment stage. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00okkup2fs4ni4yc
contributor_item
Submission 4NI4YC
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d=defaultdict(lambda:{'correct':0,'total':0,'lat':[]}) for line in lines: f=dict(x.split('=',1) for x in line.split()); r=d[f['model']]; r['total']+=1; r['correct']+=f['prediction']==f['trut...
model=v1 prediction=cat truth=cat latency=40 model=v1 prediction=dog truth=cat latency=45 model=v1 prediction=dog truth=dog latency=50 model=v2 prediction=cat truth=cat latency=55 model=v2 prediction=dog truth=dog latency=60 model=v2 prediction=bird truth=bird latency=70
{ "v1": { "accuracy_pct": 66.7, "median_latency_ms": 45 }, "v2": { "accuracy_pct": 100, "median_latency_ms": 60 } }
Parse the supplied raw log text and compute inference accuracy proxy and latency by model version. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00opkup2n21qnh1a
contributor_item
Submission 1QNH1A
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d=defaultdict(lambda:[0,0,0,0.0]) for line in lines: f=dict(x.split('=',1) for x in line.split()); r=d[f['campaign']]; vals=[int(f['impressions']),int(f['clicks']),int(f['conversions']),floa...
campaign=a impressions=1000 clicks=50 conversions=5 spend=100 campaign=a impressions=500 clicks=20 conversions=2 spend=40 campaign=b impressions=800 clicks=80 conversions=4 spend=160 campaign=b impressions=200 clicks=10 conversions=1 spend=30
{ "a": { "ctr_pct": 4.67, "conversion_pct_of_clicks": 10, "cost_per_conversion": 20 }, "b": { "ctr_pct": 9, "conversion_pct_of_clicks": 5.5600000000000005, "cost_per_conversion": 38 } }
Parse the supplied raw log text and compute advertising click-through and spend per conversion. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00onkup21n8t2vy3
contributor_item
Submission 8T2VY3
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict d=defaultdict(list); offenders=[] for line in lines: f=dict(x.split('=',1) for x in line.split()); offset=int(f['offset_ms']); d[f['source']].append(abs(offset)) if abs(offset)>50: o...
node=n1 offset_ms=12 source=ntp-a node=n2 offset_ms=-85 source=ntp-a node=n3 offset_ms=140 source=ntp-b node=n4 offset_ms=-20 source=ntp-b
{ "by_source": { "ntp-a": { "mean_abs_skew_ms": 48.5, "max_abs_skew_ms": 85 }, "ntp-b": { "mean_abs_skew_ms": 80, "max_abs_skew_ms": 140 } }, "outside_50ms": [ "n2", "n3" ] }
Parse the supplied raw log text and compute clock-skew summary and nodes outside tolerance. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00orkup2s7pf8kul
contributor_item
Submission PF8KUL
false
import json def transform(input): lines = input.strip().splitlines() from collections import defaultdict starts={}; durations=defaultdict(list) for line in lines: f=dict(x.split('=',1) for x in line.split()); g=f['group']; t=int(f['ts']) if f['event']=='rebalance_start':starts[g]=t ...
group=g1 event=rebalance_start ts=100 group=g1 event=rebalance_end ts=112 group=g1 event=rebalance_start ts=200 group=g1 event=rebalance_end ts=245 group=g2 event=rebalance_start ts=300 group=g2 event=rebalance_end ts=308
{ "g1": { "rebalances": 2, "total_unavailable_s": 57, "max_rebalance_s": 45 }, "g2": { "rebalances": 1, "total_unavailable_s": 8, "max_rebalance_s": 8 } }
Parse the supplied raw log text and compute broker consumer-group rebalance stability. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00oxkup2yrbxqlql
contributor_item
Submission BXQLQL
false
import json def transform(input): lines = input.strip().splitlines() import sqlite3 db=sqlite3.connect(':memory:');db.execute('create table logs(product text,type text,amount integer)') for line in lines: f=dict(x.split('=',1) for x in line.split());db.execute('insert into logs values(?,?,?)',(...
product=a type=sale amount=100 product=a type=sale amount=80 product=a type=refund amount=30 product=b type=sale amount=200 product=b type=refund amount=50
{ "a": { "gross": 180, "refunds": 30, "net": 150, "refund_pct": 16.7 }, "b": { "gross": 200, "refunds": 50, "net": 150, "refund_pct": 25 } }
Parse the supplied raw log text and compute sQL revenue and refund net by product. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.
cmsxm0nwe00oykup2j0my35xk
contributor_item
Submission MY35XK
false
import json def transform(input): lines = input.strip().splitlines() import sqlite3 db=sqlite3.connect(':memory:');db.execute('create table logs(sensor text,ts integer,value integer)') for line in lines: f=dict(x.split('=',1) for x in line.split());db.execute('insert into logs values(?,?,?)',(f...
sensor=a ts=1 value=10 sensor=a ts=2 value=12 sensor=a ts=3 value=30 sensor=b ts=1 value=50 sensor=b ts=2 value=40 sensor=b ts=3 value=43
{ "sensor": "a", "timestamp": 3, "delta": 18 }
Parse the supplied raw log text and compute sQL window function for largest reading jump. Ignore or specially handle the boundary cases encoded in the fixture, preserve deterministic ordering, and return the resulting metrics as a JSON string.