Mokkari

A python wrapper for the Metron Comic Book Database API.

Installation

pip install mokkari

Authentication

Authenticate with an API token, which you can generate from your metron.cloud account page:

import mokkari

m = mokkari.api(api_token="your-api-token")

Username/password (Basic Auth) with mokkari.api(username, password) is deprecated and raises a DeprecationWarning. It will be removed in mokkari 5.0, so switch to a token.

Example Usage

import mokkari

# Your own config file to keep your credentials secret
from config import api_token

m = mokkari.api(api_token=api_token)

# Get all Marvel comics for the week of 2021-06-07
this_week = m.issues_list(
    {
        "store_date_range_after": "2021-06-07",
        "store_date_range_before": "2021-06-13",
        "publisher_name": "marvel",
    }
)

# Print the results
for i in this_week:
    print(f"{i.id} {i.issue_name}")

    # Retrieve the detail for an individual issue
    asm_68 = m.issue(31660)

# Print the issue Description
print(asm_68.desc)

Rate Limiting

The API allows at least 20 requests per minute (the server may allow more when load is low), plus a daily limit that starts at 5,000 requests and is raised for OpenCollective donors (up to 25,000/day). Because the daily limit varies per user, mokkari doesn't hardcode it — it reads the X-RateLimit-* headers Metron returns with every response and pre-empts a request once those headers show a window is exhausted, avoiding an HTTP call that would fail anyway. When a rate limit is exceeded, a RateLimitError is raised.

The most recently observed state is available via session.rate_limit_status:

status = m.rate_limit_status
print(f"Sustained remaining: {status.sustained.remaining}/{status.sustained.limit}")

Handling Rate Limits

The RateLimitError includes a retry_after attribute that tells you exactly how many seconds to wait before making another request:

import mokkari
from mokkari.exceptions import RateLimitError
import time

m = mokkari.api(api_token=api_token)

try:
    issue = m.issue(31660)
except RateLimitError as e:
    # Display user-friendly message
    print(f"Rate limited: {e}")

    # Programmatically wait for the exact time needed
    print(f"Waiting {e.retry_after} seconds...")
    time.sleep(e.retry_after)

    # Retry the request
    issue = m.issue(31660)

Thread Safety

A Session can be shared across threads, but the rate-limit check above is advisory rather than a hard gate: it only blocks once the last known response headers show a window is exhausted, and that check isn't synchronized with sending the request. Concurrent threads can therefore each pass the check and send their requests before either response updates rate_limit_status, letting a burst of threads momentarily exceed the per-minute limit (Metron's server-side limit still applies and will reject the excess requests).

If you're calling a shared Session from multiple threads, cap your own concurrency instead of relying on Session to do it for you, e.g. keep a ThreadPoolExecutor at or below the burst limit:

from concurrent.futures import ThreadPoolExecutor

m = mokkari.api(api_token=api_token)

# Keep worker count at or below the burst limit (20/min at minimum) to avoid
# racing past the local rate-limit check.
with ThreadPoolExecutor(max_workers=20) as executor:
    issues = list(executor.map(m.issue, issue_ids))

Pacing (opt-in)

Passing rate_limiter closes the gap described above: instead of racing past an advisory check, every HTTP send is dispatched through the rate limiter first, which can block a caller until capacity actually frees rather than letting it send anyway. mokkari.rate_limit.HeaderPacedRateLimiter is a ready-to-use implementation. It sizes the per-minute window from the X-RateLimit-* headers but paces it from its own monotonic log of send times, so a local clock that has drifted from Metron's doesn't matter, and it spaces sends evenly across the window (at the 20/min floor, one every 3 seconds). If Metron still answers with a 429, it backs every caller off by the Retry-After value the server sent:

from concurrent.futures import ThreadPoolExecutor

import mokkari
from mokkari.rate_limit import HeaderPacedRateLimiter

m = mokkari.api(api_token=api_token, rate_limiter=HeaderPacedRateLimiter())

with ThreadPoolExecutor(max_workers=20) as executor:
    issues = list(executor.map(m.issue, issue_ids))

The limiter only blocks for the per-minute window, whose waits are seconds long. When the daily limit is exhausted it raises RateLimitError instead of blocking for what could be hours, with retry_after set to the time until the daily window resets. That leaves it to your application to tell the user and either wait or quit:

import time

from mokkari.exceptions import RateLimitError
from mokkari.utils import format_time

try:
    issue = m.issue(31660)
except RateLimitError as e:
    if input(f"Daily limit reached. Wait {format_time(e.retry_after)}? (y/n): ") == "y":
        time.sleep(e.retry_after)
        issue = m.issue(31660)

Paginated list calls follow the same rules: if a page is rejected with a 429 they retry it through the limiter, which blocks until it's safe to send, and an exhausted daily limit raises RateLimitError from the list call rather than being waited out.

A rate limiter object is scoped to the Session it's passed to — construct one per Session rather than sharing an instance across sessions using different credentials. Passing your own object works too, as long as it implements the acquire/on_rate_limited/release methods described in mokkari.rate_limit.RateLimiter. Leaving rate_limiter unset (the default) keeps the raise-immediately behavior described above.

Pacing across processes with Redis

HeaderPacedRateLimiter only sees the requests its own process sends, so several workers using one account can overrun the per-minute window together until Metron answers with 429s. mokkari.redis_rate_limit.RedisRateLimiter paces the same way but keeps its state in Redis, so every process and host using the same account shares one per-minute window, one daily estimate and one 429 backoff. Install the redis extra and pass it a client:

pip install mokkari[redis]

import redis

import mokkari
from mokkari.redis_rate_limit import RedisRateLimiter

client = redis.Redis(host="localhost", port=6379)
# Any stable name for your Metron account, such as its username. Don't use the
# token itself: it becomes part of key names anyone with Redis access can read.
m = mokkari.api(
    api_token=api_token,
    rate_limiter=RedisRateLimiter(client, account="your-username"),
)

It raises RateLimitError on an exhausted daily limit just like HeaderPacedRateLimiter. Times come from the Redis server's clock, and every key expires on its own, so a worker that crashes mid-request can't leave the account blocked. If Redis is unreachable, acquire raises the client's connection error and the request isn't sent.

Connection Reuse

A Session keeps its connections to Metron open between requests, so repeated calls don't pay a new TCP/TLS handshake each time. Cookies are never stored or sent back. To release the connections deterministically, use the session as a context manager or call close():

with mokkari.api(api_token="your-token") as m:
    issue = m.issue(1)

Closing is optional, and a closed session can still be used; it just opens new connections. If the server drops an idle connection, the underlying HTTP library notices and discards it when the next request checks it out, so a request opens a fresh connection instead. The only failure is a race: if the server closes a connection in the instant it is reused, that request raises an ApiError. This is rare, and mokkari doesn't retry it automatically. Don't share a Session across forked processes; create one per process.

Documentation

Read the project documentation

Bugs/Requests

Please use the GitHub issue tracker to submit bugs or request features.

Metadata

Release files for mokkari 4.10.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for mokkari 4.10.0
File	Size	Uploaded
mokkari-4.10.0.tar.gz	109.5 kB	Oct 11, 2026	Details

Built distribution (wheel)

Table of built distributions (wheels) for mokkari 4.10.0
File	Interpreter	ABI	Platform	Reset
mokkari-4.10.0-py3-none-any.whl 78.3 kB Oct 11, 2026	Python 3	none	any	Details

Total release size: 187.8 kB

Release files / mokkari-4.10.0.tar.gz

Download URL	mokkari-4.10.0.tar.gz
Size	109.5 kB
Tags	Source
SHA-256 checksum How to use checksums	`5d7678316219b7505879ea5118e0c128e743b437b588f7d2b5fe417a65a54ac1`
BLAKE2b-256 checksum How to use checksums	`b2685868d8282371a07a48bbe8047453b2606d8ad2dbc08418096eca2db6b579`
Upload date	Oct 11, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`uv/0.13.0 {"installer":{"name":"uv","version":"0.13.0","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}`

Release files / mokkari-4.10.0-py3-none-any.whl

Download URL	mokkari-4.10.0-py3-none-any.whl
Size	78.3 kB
Tags	Python 3
SHA-256 checksum How to use checksums	`75b42f99d1f59ad6c7e6036d66898d833d30874ae4be6b29aeb77746256a0b24`
BLAKE2b-256 checksum How to use checksums	`bc01b4040bf17f501d9fad4dd4fcec85a7c79c8c7dbbf5fbee4519f5464649c6`
Upload date	Oct 11, 2026
Uploaded using Trusted Publishing? What is trusted publishing?	No
Uploaded via	`uv/0.13.0 {"installer":{"name":"uv","version":"0.13.0","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}`