lemire

// on success: returns a non-negative integer indicating the size of the
// binary produced, it most be no larger than 2147483647 bytes.
// In case of error, a negativ value is returned:
// * -2 indicates an invalid character,
// * -1 indicates a single character remained,
// * -3 indicates a possible overflow (i.e., more than 2 GB output).

lemire

echo "base64 decode"
node benchmark/buffers/buffer-base64-decode.js
./out/Release/node benchmark/buffers/buffer-base64-decode.js

#echo "base64 encode"
#node benchmark/buffers/buffer-base64-encode.js
#./out/Release/node  benchmark/buffers/buffer-base64-encode.js

echo "base64url decode"
node benchmark/buffers/buffer-base64url-decode.js

lemire

  public unsafe static byte* GetPointerToFirstInvalidByteSse(byte* pInputBuffer, int inputLength)
        {

            int processedLength = 0;

            if (pInputBuffer == null || inputLength <= 0)
            {
                return pInputBuffer;
            }
            if (inputLength > 128)

lemire

                    Vector128<byte> shuf1 = Vector128.Create(TOO_LONG, TOO_LONG, TOO_LONG, TOO_LONG,
                            TOO_LONG, TOO_LONG, TOO_LONG, TOO_LONG,
                            TWO_CONTS, TWO_CONTS, TWO_CONTS, TWO_CONTS,
                            TOO_SHORT | OVERLONG_2,
                            TOO_SHORT,
                            TOO_SHORT | OVERLONG_3 | SURROGATE,
                            TOO_SHORT | TOO_LARGE | TOO_LARGE_1000 | OVERLONG_4);

                    Vector128<byte> shuf2 = Vector128.Create(CARRY | OVERLONG_3 | OVERLONG_2 | OVERLONG_4,

lemire

// See https://twitter.com/pdimov2/status/1462802234761170949
#include <array>
#include <string_view>
#include <string>
#include <iostream>

// For now experimental/reflect is not available generally, but
// it should be standardized for C++ 26 ???? Still, we have
// access to it with the very latest llvm.
#include <experimental/reflect>

lemire

#################
# This starts a web server listening on port 8001, with debugging turned n.
# This should not be be used to run the chatbot on a public website: it is meant
# for testing purposes only.
#################

from flask import Flask, request, jsonify
from flask import Flask, render_template, request, url_for
from langchain.chat_models import ChatOpenAI
from langchain.docstore.document import Document

lemire

###############
# You should basically never use this program. It is only for generating the embeddings for your ChatBot.
# If you want to run the ChatBot, see web_app.py
###############
import os
from openai.embeddings_utils import get_embedding, cosine_similarity
import pandas
import openai
import glob
import time

lemire

// cc test.c -lcurl -o testcurl && ./testcurl
#include <curl/curl.h>
#include <stdio.h>
int main() {
  CURLU *url = curl_url();
  CURLUcode rc = curl_url_set(
      url, CURLUPART_URL, "https://www.7‑Eleven.com/Home/Privacy/Montréal", 0);
  // Returns a CURLUcode error value, which is (0) if everything went fine.
  if (rc == 0) {
    char *buffer;

lemire

#include "simdjson.h"
#include <fstream>
#include <iostream>
#include <sstream>
#include <string>
using namespace simdjson;

std::string read_file(const std::string &filename) {
  std::ifstream file(filename);

lemire

"use strict";

import { bench, run } from "mitata";
import { existsSync, createWriteStream, readFileSync, mkdirSync } from "node:fs";
import path from "node:path";
import axios from "axios";

const fixturesFolderPath = new URL('fixtures', import.meta.url).pathname;
const urls = [
  "https://github.com/ada-url/url-various-datasets/blob/main/files/isaacs_files.txt",