Here's a list of examples of basic operations to do with twitter-archive-reader package.
This complete the documentation associated with each part, this isn't meant to be a comprehensive list of what the package can do !
This import statement will be printed every time a new thing need to be imported.
import TwitterArchive from "twitter-archive-reader";
const archive = new TwitterArchive('filename.zip');
// Wait for file read
await archive.ready();import TwitterArchive from "twitter-archive-reader";
const file_input = document.querySelector('input[type="file"]') as HTMLInputElement;
const archive = new TwitterArchive(file_input.files[0]);
// Wait for file read
await archive.ready();const tweets = archive.tweets;
// Take the 30ths first tweets and get their text
const texts = tweets.all.slice(0, 30).map(t => t.text);
// Extract real text (without the starting(s) @)
texts.map(t => t.replace(/^(@\w+ ?)*/g, ''));Warning:
To extract tweet text without leading mentions, do not use .display_text_range: In GDPR archives, this property is unproperly set and .display_text_range[0] is always "0".
archive.tweets.all.find(t => t.display_text_range[0] !== "0") // => undefinedarchive.tweets.all.reduce((acc, val) => {
if (val.retweeted_status) {
// This a retweet, the property exists
return acc + 1;
}
return acc;
}, 0);import TwitterArchive, { PartialTweet } from 'twitter-archive-reader';
function getVideoUrlOfTweet(tweet: PartialTweet) {
// This is for GDPR archives — classic archives does not
// contains .extended_entities, only .entities
// (that can't contain videos).
if (tweet.extended_entities && tweet.extended_entities.media) {
const video = tweet.extended_entities.media.find(v => v.video_info);
if (video) {
// This contains a video.
// A video can contains multiple links:
// multiple formats/bitrate
// We need to find the best bitrate for MP4 format
// Then sort per bitrate, the higher the first
const best_video = video.video_info.variants
.filter(variant => variant.content_type === "video/mp4")
.filter(variant => variant.bitrate)
.sort((a, b) => Number(b.bitrate) - Number(a.bitrate));
if (best_video.length) {
const video_url = best_video[0].url;
// Final step, trim the possible query string
// at the end of the url...
return video_url.split('?')[0];
}
}
}
return undefined;
}
// Execute the fn !
getVideoUrlOfTweet(archive.tweets.single('tweet_id'));const messages = archive.messages;
// We have only access to user IDs in DMs
const conversations = messages.all.map(c =>
`Conversation #${c.id} with users #${[...c.participants].join(', #')},
containing ${c.length} messages.`
);
console.log(...conversations);archive.messages.all.reduce((acc, val) =>
acc + val.all.reduce((acc, val) =>
acc + (val.senderId === archive.user.id ? 1 : 0)
, 0)
, 0);
// or, in a less compact way
let count = 0;
for (const conversation of archive.messages) {
for (const msg of conversation) {
if (msg.senderId === archive.user.id) {
count++;
}
}
}if (archive.messages.groups.length) {
console.log("This archive contains group conversations !");
}const user = archive.user;
console.log(`
Screen name: ${user.screen_name},
Name: ${user.name},
Bio: ${user.bio},
Registered location: ${user.location},
User ID: ${user.id},
You are${user.verified ? "" : " not"} verified.
`);console.log(`
Creation date: ${archive.user.created_at},
Creation IP: ${archive.user.account_creation_ip}
`);const history = [
...archive.user.screen_name_history.map(s => s.changedFrom),
archive.user.screen_name
];
console.log(`You used the following names: @${history.join(', @')}`);const impressions = archive.ads.impressions_by_advertiser;
const five_most_common = Object.entries(impressions)
.sort((a, b) => b[1].length - a[1].length)
.slice(0, 5)
.map(e => e[0].slice(1)); // Get the screen name and trim the @
console.log("Most common advertisers: ", five_most_common.join(', '));import TwitterArchive, { AdImpression, TwitterHelpers } from 'twitter-archive-reader';
const days_to_impressions: { [day: string]: AdImpression[] } = {};
for (const impression of archive.ads.impressions) {
const date = TwitterHelpers.parseAdDate(impression.impressionTime)
const [day, month, year] = [
String(date.getDate()).padStart(2, "0"),
String(date.getMonth() + 1).padStart(2, "0"),
date.getFullYear(),
];
const date_string = `${year}-${month}-${day}`;
if (date_string in days_to_impressions) {
days_to_impressions[date_string].push(impression);
}
else {
days_to_impressions[date_string] = [impression];
}
}const my_dm = archive.messages.single('dm_id');
const medias_of_dm = await archive.medias.ofDm(my_dm);const my_dm = archive.messages.single('dm_id');
if (my_dm.mediaUrls.length) {
const first_media = await archive.medias.fromDmMediaUrl(
my_dm.mediaUrls[0],
false /* Is DM in a group conversation ? */
);
}// Find a tweet with a media defined
const tweet = archive.tweets.all.find(t => t.extended_entities && t.extended_entities.media);
if (tweet) {
// All medias of the tweet
const medias = await archive.medias.ofTweet(tweet);
// From a specific media
const media_1 = tweet.extended_entities.media[0];
const media_1_bin = await archive.medias.fromTweetMediaEntity(media_1);
}const [profile, header] = await Promise.all([
archive.medias.getProfilePictureOf(archive.user),
archive.medias.getProfileBannerOf(archive.user)
]);import { promises as FsPromise } from 'fs';
// Find the tweets
const tweets = archive.tweets.between("2019-01-01T12:01:00Z", "2019-06-06T16:32:45Z");
const destination = "/path/to/dir/";
for (const tweet of tweets) {
// Get the medias of the tweets
const medias: ArrayBuffer[] = await archive.medias.ofTweet(tweet);
let i = 0;
for (const media of medias) {
// Write a media on the disk
await FsPromise.writeFile(destination + `${tweet.id_str}-${i}`, Buffer.from(media));
i++;
}
}