open-webui/src/lib/utils/index.ts
2024-09-22 11:49:18 +02:00

863 lines
24 KiB
TypeScript

import { v4 as uuidv4 } from 'uuid';
import sha256 from 'js-sha256';
import { WEBUI_BASE_URL } from '$lib/constants';
import { TTS_RESPONSE_SPLIT } from '$lib/types';
//////////////////////////
// Helper functions
//////////////////////////
const convertLatexToSingleLine = (content) => {
// Patterns to match multiline LaTeX blocks
const patterns = [
/(\$\$\s[\s\S]*?\s\$\$)/g, // Match $$ ... $$
/(\\\[[\s\S]*?\\\])/g, // Match \[ ... \]
/(\\begin\{[a-z]+\}[\s\S]*?\\end\{[a-z]+\})/g // Match \begin{...} ... \end{...}
];
patterns.forEach((pattern) => {
content = content.replace(pattern, (match) => {
return match.replace(/\s*\n\s*/g, ' ').trim();
});
});
return content;
};
export const replaceTokens = (content, char, user) => {
const charToken = /{{char}}/gi;
const userToken = /{{user}}/gi;
const videoIdToken = /{{VIDEO_FILE_ID_([a-f0-9-]+)}}/gi; // Regex to capture the video ID
const htmlIdToken = /{{HTML_FILE_ID_([a-f0-9-]+)}}/gi; // Regex to capture the HTML ID
// Replace {{char}} if char is provided
if (char !== undefined && char !== null) {
content = content.replace(charToken, char);
}
// Replace {{user}} if user is provided
if (user !== undefined && user !== null) {
content = content.replace(userToken, user);
}
// Replace video ID tags with corresponding <video> elements
content = content.replace(videoIdToken, (match, fileId) => {
const videoUrl = `${WEBUI_BASE_URL}/api/v1/files/${fileId}/content`;
return `<video src="${videoUrl}" controls></video>`;
});
// Replace HTML ID tags with corresponding HTML content
content = content.replace(htmlIdToken, (match, fileId) => {
const htmlUrl = `${WEBUI_BASE_URL}/api/v1/files/${fileId}/content`;
return `<iframe src="${htmlUrl}" width="100%" frameborder="0" onload="this.style.height=(this.contentWindow.document.body.scrollHeight+20)+'px';"></iframe>`;
});
return content;
};
export const sanitizeResponseContent = (content: string) => {
return content
.replace(/<\|[a-z]*$/, '')
.replace(/<\|[a-z]+\|$/, '')
.replace(/<$/, '')
.replaceAll(/<\|[a-z]+\|>/g, ' ')
.replaceAll('<', '&lt;')
.replaceAll('>', '&gt;')
.trim();
};
export const processResponseContent = (content: string) => {
content = convertLatexToSingleLine(content);
return content.trim();
};
export const revertSanitizedResponseContent = (content: string) => {
return content.replaceAll('&lt;', '<').replaceAll('&gt;', '>');
};
export function unescapeHtml(html: string) {
const doc = new DOMParser().parseFromString(html, 'text/html');
return doc.documentElement.textContent;
}
export const capitalizeFirstLetter = (string) => {
return string.charAt(0).toUpperCase() + string.slice(1);
};
export const splitStream = (splitOn) => {
let buffer = '';
return new TransformStream({
transform(chunk, controller) {
buffer += chunk;
const parts = buffer.split(splitOn);
parts.slice(0, -1).forEach((part) => controller.enqueue(part));
buffer = parts[parts.length - 1];
},
flush(controller) {
if (buffer) controller.enqueue(buffer);
}
});
};
export const convertMessagesToHistory = (messages) => {
const history = {
messages: {},
currentId: null
};
let parentMessageId = null;
let messageId = null;
for (const message of messages) {
messageId = uuidv4();
if (parentMessageId !== null) {
history.messages[parentMessageId].childrenIds = [
...history.messages[parentMessageId].childrenIds,
messageId
];
}
history.messages[messageId] = {
...message,
id: messageId,
parentId: parentMessageId,
childrenIds: []
};
parentMessageId = messageId;
}
history.currentId = messageId;
return history;
};
export const getGravatarURL = (email) => {
// Trim leading and trailing whitespace from
// an email address and force all characters
// to lower case
const address = String(email).trim().toLowerCase();
// Create a SHA256 hash of the final string
const hash = sha256(address);
// Grab the actual image URL
return `https://www.gravatar.com/avatar/${hash}`;
};
export const canvasPixelTest = () => {
// Test a 1x1 pixel to potentially identify browser/plugin fingerprint blocking or spoofing
// Inspiration: https://github.com/kkapsner/CanvasBlocker/blob/master/test/detectionTest.js
const canvas = document.createElement('canvas');
const ctx = canvas.getContext('2d');
canvas.height = 1;
canvas.width = 1;
const imageData = new ImageData(canvas.width, canvas.height);
const pixelValues = imageData.data;
// Generate RGB test data
for (let i = 0; i < imageData.data.length; i += 1) {
if (i % 4 !== 3) {
pixelValues[i] = Math.floor(256 * Math.random());
} else {
pixelValues[i] = 255;
}
}
ctx.putImageData(imageData, 0, 0);
const p = ctx.getImageData(0, 0, canvas.width, canvas.height).data;
// Read RGB data and fail if unmatched
for (let i = 0; i < p.length; i += 1) {
if (p[i] !== pixelValues[i]) {
console.log(
'canvasPixelTest: Wrong canvas pixel RGB value detected:',
p[i],
'at:',
i,
'expected:',
pixelValues[i]
);
console.log('canvasPixelTest: Canvas blocking or spoofing is likely');
return false;
}
}
return true;
};
export const generateInitialsImage = (name) => {
const canvas = document.createElement('canvas');
const ctx = canvas.getContext('2d');
canvas.width = 100;
canvas.height = 100;
if (!canvasPixelTest()) {
console.log(
'generateInitialsImage: failed pixel test, fingerprint evasion is likely. Using default image.'
);
return '/user.png';
}
ctx.fillStyle = '#F39C12';
ctx.fillRect(0, 0, canvas.width, canvas.height);
ctx.fillStyle = '#FFFFFF';
ctx.font = '40px Helvetica';
ctx.textAlign = 'center';
ctx.textBaseline = 'middle';
const sanitizedName = name.trim();
const initials =
sanitizedName.length > 0
? sanitizedName[0] +
(sanitizedName.split(' ').length > 1
? sanitizedName[sanitizedName.lastIndexOf(' ') + 1]
: '')
: '';
ctx.fillText(initials.toUpperCase(), canvas.width / 2, canvas.height / 2);
return canvas.toDataURL();
};
export const copyToClipboard = async (text) => {
let result = false;
if (!navigator.clipboard) {
const textArea = document.createElement('textarea');
textArea.value = text;
// Avoid scrolling to bottom
textArea.style.top = '0';
textArea.style.left = '0';
textArea.style.position = 'fixed';
document.body.appendChild(textArea);
textArea.focus();
textArea.select();
try {
const successful = document.execCommand('copy');
const msg = successful ? 'successful' : 'unsuccessful';
console.log('Fallback: Copying text command was ' + msg);
result = true;
} catch (err) {
console.error('Fallback: Oops, unable to copy', err);
}
document.body.removeChild(textArea);
return result;
}
result = await navigator.clipboard
.writeText(text)
.then(() => {
console.log('Async: Copying to clipboard was successful!');
return true;
})
.catch((error) => {
console.error('Async: Could not copy text: ', error);
return false;
});
return result;
};
export const compareVersion = (latest, current) => {
return current === '0.0.0'
? false
: current.localeCompare(latest, undefined, {
numeric: true,
sensitivity: 'case',
caseFirst: 'upper'
}) < 0;
};
export const findWordIndices = (text) => {
const regex = /\[([^\]]+)\]/g;
const matches = [];
let match;
while ((match = regex.exec(text)) !== null) {
matches.push({
word: match[1],
startIndex: match.index,
endIndex: regex.lastIndex - 1
});
}
return matches;
};
export const removeLastWordFromString = (inputString, wordString) => {
// Split the string into an array of words
const words = inputString.split(' ');
if (words.at(-1) === wordString) {
words.pop();
}
// Join the remaining words back into a string
let resultString = words.join(' ');
if (resultString !== '') {
resultString += ' ';
}
return resultString;
};
export const removeFirstHashWord = (inputString) => {
// Split the string into an array of words
const words = inputString.split(' ');
// Find the index of the first word that starts with #
const index = words.findIndex((word) => word.startsWith('#'));
// Remove the first word with #
if (index !== -1) {
words.splice(index, 1);
}
// Join the remaining words back into a string
const resultString = words.join(' ');
return resultString;
};
export const transformFileName = (fileName) => {
// Convert to lowercase
const lowerCaseFileName = fileName.toLowerCase();
// Remove special characters using regular expression
const sanitizedFileName = lowerCaseFileName.replace(/[^\w\s]/g, '');
// Replace spaces with dashes
const finalFileName = sanitizedFileName.replace(/\s+/g, '-');
return finalFileName;
};
export const calculateSHA256 = async (file) => {
// Create a FileReader to read the file asynchronously
const reader = new FileReader();
// Define a promise to handle the file reading
const readFile = new Promise((resolve, reject) => {
reader.onload = () => resolve(reader.result);
reader.onerror = reject;
});
// Read the file as an ArrayBuffer
reader.readAsArrayBuffer(file);
try {
// Wait for the FileReader to finish reading the file
const buffer = await readFile;
// Convert the ArrayBuffer to a Uint8Array
const uint8Array = new Uint8Array(buffer);
// Calculate the SHA-256 hash using Web Crypto API
const hashBuffer = await crypto.subtle.digest('SHA-256', uint8Array);
// Convert the hash to a hexadecimal string
const hashArray = Array.from(new Uint8Array(hashBuffer));
const hashHex = hashArray.map((byte) => byte.toString(16).padStart(2, '0')).join('');
return `${hashHex}`;
} catch (error) {
console.error('Error calculating SHA-256 hash:', error);
throw error;
}
};
export const getImportOrigin = (_chats) => {
// Check what external service chat imports are from
if ('mapping' in _chats[0]) {
return 'openai';
}
return 'webui';
};
export const getUserPosition = async (raw = false) => {
// Get the user's location using the Geolocation API
const position = await new Promise((resolve, reject) => {
navigator.geolocation.getCurrentPosition(resolve, reject);
}).catch((error) => {
console.error('Error getting user location:', error);
throw error;
});
if (!position) {
return 'Location not available';
}
// Extract the latitude and longitude from the position
const { latitude, longitude } = position.coords;
if (raw) {
return { latitude, longitude };
} else {
return `${latitude.toFixed(3)}, ${longitude.toFixed(3)} (lat, long)`;
}
};
const convertOpenAIMessages = (convo) => {
// Parse OpenAI chat messages and create chat dictionary for creating new chats
const mapping = convo['mapping'];
const messages = [];
let currentId = '';
let lastId = null;
for (const message_id in mapping) {
const message = mapping[message_id];
currentId = message_id;
try {
if (
messages.length == 0 &&
(message['message'] == null ||
(message['message']['content']['parts']?.[0] == '' &&
message['message']['content']['text'] == null))
) {
// Skip chat messages with no content
continue;
} else {
const new_chat = {
id: message_id,
parentId: lastId,
childrenIds: message['children'] || [],
role: message['message']?.['author']?.['role'] !== 'user' ? 'assistant' : 'user',
content:
message['message']?.['content']?.['parts']?.[0] ||
message['message']?.['content']?.['text'] ||
'',
model: 'gpt-3.5-turbo',
done: true,
context: null
};
messages.push(new_chat);
lastId = currentId;
}
} catch (error) {
console.log('Error with', message, '\nError:', error);
}
}
const history: Record<PropertyKey, (typeof messages)[number]> = {};
messages.forEach((obj) => (history[obj.id] = obj));
const chat = {
history: {
currentId: currentId,
messages: history // Need to convert this to not a list and instead a json object
},
models: ['gpt-3.5-turbo'],
messages: messages,
options: {},
timestamp: convo['create_time'],
title: convo['title'] ?? 'New Chat'
};
return chat;
};
const validateChat = (chat) => {
// Because ChatGPT sometimes has features we can't use like DALL-E or migh have corrupted messages, need to validate
const messages = chat.messages;
// Check if messages array is empty
if (messages.length === 0) {
return false;
}
// Last message's children should be an empty array
const lastMessage = messages[messages.length - 1];
if (lastMessage.childrenIds.length !== 0) {
return false;
}
// First message's parent should be null
const firstMessage = messages[0];
if (firstMessage.parentId !== null) {
return false;
}
// Every message's content should be a string
for (const message of messages) {
if (typeof message.content !== 'string') {
return false;
}
}
return true;
};
export const convertOpenAIChats = (_chats) => {
// Create a list of dictionaries with each conversation from import
const chats = [];
let failed = 0;
for (const convo of _chats) {
const chat = convertOpenAIMessages(convo);
if (validateChat(chat)) {
chats.push({
id: convo['id'],
user_id: '',
title: convo['title'],
chat: chat,
timestamp: convo['timestamp']
});
} else {
failed++;
}
}
console.log(failed, 'Conversations could not be imported');
return chats;
};
export const isValidHttpUrl = (string: string) => {
let url;
try {
url = new URL(string);
} catch (_) {
return false;
}
return url.protocol === 'http:' || url.protocol === 'https:';
};
export const removeEmojis = (str: string) => {
// Regular expression to match emojis
const emojiRegex = /[\uD800-\uDBFF][\uDC00-\uDFFF]|\uD83C[\uDC00-\uDFFF]|\uD83D[\uDC00-\uDE4F]/g;
// Replace emojis with an empty string
return str.replace(emojiRegex, '');
};
export const removeFormattings = (str: string) => {
return str.replace(/(\*)(.*?)\1/g, '').replace(/(```)(.*?)\1/gs, '');
};
export const cleanText = (content: string) => {
return removeFormattings(removeEmojis(content.trim()));
};
// This regular expression matches code blocks marked by triple backticks
const codeBlockRegex = /```[\s\S]*?```/g;
export const extractSentences = (text: string) => {
const codeBlocks: string[] = [];
let index = 0;
// Temporarily replace code blocks with placeholders and store the blocks separately
text = text.replace(codeBlockRegex, (match) => {
const placeholder = `\u0000${index}\u0000`; // Use a unique placeholder
codeBlocks[index++] = match;
return placeholder;
});
// Split the modified text into sentences based on common punctuation marks, avoiding these blocks
let sentences = text.split(/(?<=[.!?])\s+/);
// Restore code blocks and process sentences
sentences = sentences.map((sentence) => {
// Check if the sentence includes a placeholder for a code block
return sentence.replace(/\u0000(\d+)\u0000/g, (_, idx) => codeBlocks[idx]);
});
return sentences.map(cleanText).filter(Boolean);
};
export const extractParagraphsForAudio = (text: string) => {
const codeBlocks: string[] = [];
let index = 0;
// Temporarily replace code blocks with placeholders and store the blocks separately
text = text.replace(codeBlockRegex, (match) => {
const placeholder = `\u0000${index}\u0000`; // Use a unique placeholder
codeBlocks[index++] = match;
return placeholder;
});
// Split the modified text into paragraphs based on newlines, avoiding these blocks
let paragraphs = text.split(/\n+/);
// Restore code blocks and process paragraphs
paragraphs = paragraphs.map((paragraph) => {
// Check if the paragraph includes a placeholder for a code block
return paragraph.replace(/\u0000(\d+)\u0000/g, (_, idx) => codeBlocks[idx]);
});
return paragraphs.map(cleanText).filter(Boolean);
};
export const extractSentencesForAudio = (text: string) => {
return extractSentences(text).reduce((mergedTexts, currentText) => {
const lastIndex = mergedTexts.length - 1;
if (lastIndex >= 0) {
const previousText = mergedTexts[lastIndex];
const wordCount = previousText.split(/\s+/).length;
const charCount = previousText.length;
if (wordCount < 4 || charCount < 50) {
mergedTexts[lastIndex] = previousText + ' ' + currentText;
} else {
mergedTexts.push(currentText);
}
} else {
mergedTexts.push(currentText);
}
return mergedTexts;
}, [] as string[]);
};
export const getMessageContentParts = (content: string, split_on: string = 'punctuation') => {
const messageContentParts: string[] = [];
switch (split_on) {
default:
case TTS_RESPONSE_SPLIT.PUNCTUATION:
messageContentParts.push(...extractSentencesForAudio(content));
break;
case TTS_RESPONSE_SPLIT.PARAGRAPHS:
messageContentParts.push(...extractParagraphsForAudio(content));
break;
case TTS_RESPONSE_SPLIT.NONE:
messageContentParts.push(cleanText(content));
break;
}
return messageContentParts;
};
export const blobToFile = (blob, fileName) => {
// Create a new File object from the Blob
const file = new File([blob], fileName, { type: blob.type });
return file;
};
/**
* @param {string} template - The template string containing placeholders.
* @returns {string} The template string with the placeholders replaced by the prompt.
*/
export const promptTemplate = (
template: string,
user_name?: string,
user_location?: string
): string => {
// Get the current date
const currentDate = new Date();
// Format the date to YYYY-MM-DD
const formattedDate =
currentDate.getFullYear() +
'-' +
String(currentDate.getMonth() + 1).padStart(2, '0') +
'-' +
String(currentDate.getDate()).padStart(2, '0');
// Format the time to HH:MM:SS AM/PM
const currentTime = currentDate.toLocaleTimeString('en-US', {
hour: 'numeric',
minute: 'numeric',
second: 'numeric',
hour12: true
});
// Replace {{CURRENT_DATETIME}} in the template with the formatted datetime
template = template.replace('{{CURRENT_DATETIME}}', `${formattedDate} ${currentTime}`);
// Replace {{CURRENT_DATE}} in the template with the formatted date
template = template.replace('{{CURRENT_DATE}}', formattedDate);
// Replace {{CURRENT_TIME}} in the template with the formatted time
template = template.replace('{{CURRENT_TIME}}', currentTime);
if (user_name) {
// Replace {{USER_NAME}} in the template with the user's name
template = template.replace('{{USER_NAME}}', user_name);
}
if (user_location) {
// Replace {{USER_LOCATION}} in the template with the current location
template = template.replace('{{USER_LOCATION}}', user_location);
}
return template;
};
/**
* This function is used to replace placeholders in a template string with the provided prompt.
* The placeholders can be in the following formats:
* - `{{prompt}}`: This will be replaced with the entire prompt.
* - `{{prompt:start:<length>}}`: This will be replaced with the first <length> characters of the prompt.
* - `{{prompt:end:<length>}}`: This will be replaced with the last <length> characters of the prompt.
* - `{{prompt:middletruncate:<length>}}`: This will be replaced with the prompt truncated to <length> characters, with '...' in the middle.
*
* @param {string} template - The template string containing placeholders.
* @param {string} prompt - The string to replace the placeholders with.
* @returns {string} The template string with the placeholders replaced by the prompt.
*/
export const titleGenerationTemplate = (template: string, prompt: string): string => {
template = template.replace(
/{{prompt}}|{{prompt:start:(\d+)}}|{{prompt:end:(\d+)}}|{{prompt:middletruncate:(\d+)}}/g,
(match, startLength, endLength, middleLength) => {
if (match === '{{prompt}}') {
return prompt;
} else if (match.startsWith('{{prompt:start:')) {
return prompt.substring(0, startLength);
} else if (match.startsWith('{{prompt:end:')) {
return prompt.slice(-endLength);
} else if (match.startsWith('{{prompt:middletruncate:')) {
if (prompt.length <= middleLength) {
return prompt;
}
const start = prompt.slice(0, Math.ceil(middleLength / 2));
const end = prompt.slice(-Math.floor(middleLength / 2));
return `${start}...${end}`;
}
return '';
}
);
template = promptTemplate(template);
return template;
};
export const approximateToHumanReadable = (nanoseconds: number) => {
const seconds = Math.floor((nanoseconds / 1e9) % 60);
const minutes = Math.floor((nanoseconds / 6e10) % 60);
const hours = Math.floor((nanoseconds / 3.6e12) % 24);
const results: string[] = [];
if (seconds >= 0) {
results.push(`${seconds}s`);
}
if (minutes > 0) {
results.push(`${minutes}m`);
}
if (hours > 0) {
results.push(`${hours}h`);
}
return results.reverse().join(' ');
};
export const getTimeRange = (timestamp) => {
const now = new Date();
const date = new Date(timestamp * 1000); // Convert Unix timestamp to milliseconds
// Calculate the difference in milliseconds
const diffTime = now.getTime() - date.getTime();
const diffDays = diffTime / (1000 * 3600 * 24);
const nowDate = now.getDate();
const nowMonth = now.getMonth();
const nowYear = now.getFullYear();
const dateDate = date.getDate();
const dateMonth = date.getMonth();
const dateYear = date.getFullYear();
if (nowYear === dateYear && nowMonth === dateMonth && nowDate === dateDate) {
return 'Today';
} else if (nowYear === dateYear && nowMonth === dateMonth && nowDate - dateDate === 1) {
return 'Yesterday';
} else if (diffDays <= 7) {
return 'Previous 7 days';
} else if (diffDays <= 30) {
return 'Previous 30 days';
} else if (nowYear === dateYear) {
return date.toLocaleString('default', { month: 'long' });
} else {
return date.getFullYear().toString();
}
};
/**
* Extract frontmatter as a dictionary from the specified content string.
* @param content {string} - The content string with potential frontmatter.
* @returns {Object} - The extracted frontmatter as a dictionary.
*/
export const extractFrontmatter = (content) => {
const frontmatter = {};
let frontmatterStarted = false;
let frontmatterEnded = false;
const frontmatterPattern = /^\s*([a-z_]+):\s*(.*)\s*$/i;
// Split content into lines
const lines = content.split('\n');
// Check if the content starts with triple quotes
if (lines[0].trim() !== '"""') {
return {};
}
frontmatterStarted = true;
for (let i = 1; i < lines.length; i++) {
const line = lines[i];
if (line.includes('"""')) {
if (frontmatterStarted) {
frontmatterEnded = true;
break;
}
}
if (frontmatterStarted && !frontmatterEnded) {
const match = frontmatterPattern.exec(line);
if (match) {
const [, key, value] = match;
frontmatter[key.trim()] = value.trim();
}
}
}
return frontmatter;
};
// Function to determine the best matching language
export const bestMatchingLanguage = (supportedLanguages, preferredLanguages, defaultLocale) => {
const languages = supportedLanguages.map((lang) => lang.code);
const match = preferredLanguages
.map((prefLang) => languages.find((lang) => lang.startsWith(prefLang)))
.find(Boolean);
return match || defaultLocale;
};
// Get the date in the format YYYY-MM-DD
export const getFormattedDate = () => {
const date = new Date();
return date.toISOString().split('T')[0];
};
// Get the time in the format HH:MM:SS
export const getFormattedTime = () => {
const date = new Date();
return date.toTimeString().split(' ')[0];
};
// Get the current date and time in the format YYYY-MM-DD HH:MM:SS
export const getCurrentDateTime = () => {
return `${getFormattedDate()} ${getFormattedTime()}`;
};
// Get the user's timezone
export const getUserTimezone = () => {
return Intl.DateTimeFormat().resolvedOptions().timeZone;
};
// Get the weekday
export const getWeekday = () => {
const date = new Date();
const weekdays = ['Sunday', 'Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday', 'Saturday'];
return weekdays[date.getDay()];
};