kurtbrose

from sqlalchemy import select, text


def reset_sequences(engine, schema="public", gap=10000):
    """
    Reset all sequences in a given schema to 10000, 20000, etc. so that ids won't overlap easily.
    Ensures we don't get "lucky" and crossing ids between tables works because they are both id=1,
    passing a test that should fail.
    """
    with engine.connect() as conn:

justinvanwinkle

# I was looking for a rate limiting library to call rate limited apis as closely 
#   as possible to their enforced limits.  I looked at the first few python libraries
#   that I found, and when I glanced at the source, they were all clearly broken.
# Curious how this could be, I took all the top google and pip search results for: python rate limiting
#   and tried to get them to do the wrong thing and fail to rate limit in situations that could come up 
#   in normal use (though in some cases very specific use)


# https://github.com/tomasbasham/ratelimit
# Where broken:

sebmarkbage

I heard some points of criticism to how React deals with reactivity and it's focus on "purity". It's interesting because there are really two approaches evolving. There's a mutable + change tracking approach and there's an immutability + referential equality testing approach. It's difficult to mix and match them when you build new features on top. So that's why React has been pushing a bit harder on immutability lately to be able to build on top of it. Both have various tradeoffs but others are doing good research in other areas, so we've decided to focus on this direction and see where it leads us.

I did want to address a few points that I didn't see get enough consideration around the tradeoffs. So here's a small brain dump.

"Compiled output results in smaller apps" - E.g. Svelte apps start smaller but the compiler output is 3-4x larger per component than the equivalent VDOM approach. This is mostly due to the code that is usually shared in the VDOM "VM" needs to be inlined into each component. The tr

zzzeek

""" "Writing SQL is just as fast as using an ORM" proof of concept

Below is a simple Python object model, where we represent a database that
stores the names of employees at a company, some of whom are "engineers",
and a list of the jobs they do and the programming languages they use.

We'd like to persist the state represented by this Python object model
in a relational database, using our Python objects as a start.  Then we'd
like to write SQL queries for rows in this database, and we get back instances
of Python objects exactly as they were created.

zzzeek

# note: we are using SQLAlchemy that includes _cache_key
# currently at:
# https://gerrit.sqlalchemy.org/#/c/sqlalchemy/sqlalchemy/+/1204/5/

import typing

from sqlalchemy import bindparam
from sqlalchemy import Column
from sqlalchemy import func
from sqlalchemy import inspection

ngtvspc

# Uses the Python Imaging Library
# `pip install Pillow` works too
from PIL import Image

image_filename = "picture_with_EXIF.jpg"
image_file = open('image_filename)
image = Image.open(image_file)

# next 3 lines strip exif
image_data = list(image.getdata())

jakevdp

foca

#!/usr/bin/env bash

set -e
[ -z "$DEBUG" ] || set -x;

usage() {
  echo "$0 <repo> <tag> [<release name>] [-- <asset>...]" >&2;
}

if [ "$1" = "-h" -o "$1" = "--help" ]; then

zzzeek

from sqlalchemy import event
from sqlalchemy import DDL


def _mysql_cidr_overlap(metadata):
    @event.listens_for(metadata, "after_create")
    def _create_mysql_proc(target, connection, **kw):
        if connection.engine.name != 'mysql':
            return

kurtbrose

'''
Simple distributed, streaming algorithm for keeping counts on
an adaptive sample of data.
'''
import time
import hashlib
import json
import base64