-
Notifications
You must be signed in to change notification settings - Fork 4
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Showing
8 changed files
with
453 additions
and
75 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,56 @@ | ||
use airtable_api::Record; | ||
use anyhow::Result; | ||
use chrono::Utc; | ||
use dotenv::var; | ||
use replit_takeout::{ | ||
airtable::{self, AirtableSyncedUser, ProcessState}, | ||
replit_graphql::ProfileRepls, | ||
}; | ||
|
||
#[tokio::main] | ||
async fn main() -> Result<()> { | ||
env_logger::init(); | ||
dotenv::dotenv().ok(); | ||
|
||
let token = var("REPLIT_TEST_TOKEN")?; | ||
|
||
let fields = AirtableSyncedUser { | ||
id: 29999230, | ||
token, | ||
username: "malted".into(), | ||
status: ProcessState::Registered, | ||
email: "test@malted.dev".into(), | ||
r2_link: "http://example.com".into(), | ||
failed_ids: "none".into(), | ||
started_at: Some(Utc::now()), | ||
finished_at: None, | ||
repl_count: 0, | ||
file_count: 0, | ||
statistics: vec!["recpWEjc0zLoKEtZP".into()], | ||
}; | ||
|
||
let mut user = Record { | ||
id: String::new(), | ||
fields, | ||
created_time: None, | ||
}; | ||
|
||
log::info!("Starting..."); | ||
if let Err(err) = ProfileRepls::download(&user.fields.token, user.clone()).await { | ||
log::error!("Error with `{}`'s download: {err:#?}", user.fields.username); | ||
|
||
user.fields.status = ProcessState::ErroredMain; | ||
//arst airtable::update_records(vec![user.clone()]).await?; | ||
|
||
// user.fields.failed_ids = errored.join(","); | ||
|
||
// send_email( | ||
// &user.fields.email, | ||
// "Your Replitβ export is slightly delayed :/".into(), | ||
// format!("Hey {}, We have run into an issue processing your Replitβ takeout π₯‘.\nWe will manually review and confirm that all your data is included. If you don't hear back again within a few days email malted@hackclub.com. Sorry for the inconvenience.", user.fields.username), | ||
// ) | ||
// .await; | ||
} | ||
|
||
Ok(()) | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,29 @@ | ||
use std::collections::HashMap; | ||
|
||
use anyhow::Result; | ||
use dotenv::var; | ||
use log::error; | ||
use replit_takeout::replit::repls::Repl; | ||
|
||
#[tokio::main] | ||
async fn main() -> Result<()> { | ||
env_logger::init(); | ||
dotenv::dotenv().ok(); | ||
|
||
let token = var("REPLIT_TEST_TOKEN")?; | ||
|
||
let repls = Repl::fetch(&token, None).await.expect("some repls"); | ||
error!("got {} repls", repls.len()); | ||
|
||
let mut map: HashMap<String, Repl> = HashMap::new(); | ||
|
||
for repl in repls { | ||
if map.contains_key(&repl.id) { | ||
log::error!("ALREADY CONTAINS {:?}", repl.clone()); | ||
} | ||
|
||
map.insert(repl.id.clone(), repl); | ||
} | ||
|
||
Ok(()) | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,33 @@ | ||
query ReplList($path: String!, $starred: Boolean, $after: String) { | ||
currentUser { | ||
id | ||
username | ||
replFolderByPath(path: $path) { | ||
ownerId: userId | ||
pathnames | ||
parent { | ||
pathnames | ||
} | ||
folders { | ||
id | ||
name | ||
pathnames | ||
replsCount | ||
folderType | ||
} | ||
repls(starred: $starred, after: $after) { | ||
items { | ||
id | ||
title | ||
isPrivate | ||
slug | ||
url | ||
timeCreated | ||
} | ||
pageInfo { | ||
nextCursor | ||
} | ||
} | ||
} | ||
} | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,46 @@ | ||
use reqwest::{ | ||
cookie::Jar, | ||
header::{self, HeaderMap}, | ||
Client, Url, | ||
}; | ||
use std::sync::Arc; | ||
|
||
pub mod repls; | ||
|
||
pub static REPLIT_GQL_URL: &str = "https://replit.com/graphql"; | ||
|
||
pub fn create_client(token: &String, client: Option<Client>) -> Result<Client, reqwest::Error> { | ||
if let Some(client) = client { | ||
return Ok(client); | ||
} | ||
|
||
Client::builder() | ||
.user_agent(crate::utils::random_user_agent()) | ||
.default_headers(create_client_headers()) | ||
.cookie_provider(create_client_cookie_jar(token)) | ||
.build() | ||
} | ||
|
||
fn create_client_headers() -> HeaderMap { | ||
let mut headers = header::HeaderMap::new(); | ||
headers.insert( | ||
"X-Requested-With", | ||
header::HeaderValue::from_static("XMLHttpRequest"), | ||
); | ||
headers.insert( | ||
reqwest::header::REFERER, | ||
header::HeaderValue::from_static("https://replit.com/~"), | ||
); | ||
|
||
headers | ||
} | ||
|
||
fn create_client_cookie_jar(token: &String) -> Arc<Jar> { | ||
let cookie = &format!("connect.sid={token}; Domain=replit.com"); | ||
let url = REPLIT_GQL_URL.parse::<Url>().unwrap(); | ||
|
||
let jar = Jar::default(); | ||
jar.add_cookie_str(cookie, &url); | ||
|
||
Arc::new(jar) | ||
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,144 @@ | ||
use std::collections::HashSet; | ||
|
||
use super::{create_client, REPLIT_GQL_URL}; | ||
use anyhow::Result; | ||
use graphql_client::{GraphQLQuery, Response}; | ||
use log::{debug, info, trace, warn}; | ||
use reqwest::{Client, StatusCode}; | ||
use tokio::time::{sleep, Duration}; | ||
|
||
type DateTime = String; | ||
#[derive(GraphQLQuery)] | ||
#[graphql( | ||
schema_path = "src/graphql/schema 7.graphql", | ||
query_path = "src/graphql/repls-query.graphql", | ||
response_derives = "Debug" | ||
)] | ||
struct ReplList; | ||
|
||
#[derive(Debug, Clone, PartialEq, Eq, Hash)] | ||
pub struct Repl { | ||
pub id: String, | ||
pub title: String, | ||
pub slug: String, | ||
pub private: bool, | ||
pub url: String, | ||
pub time_created: String, | ||
} | ||
impl Repl { | ||
pub async fn fetch(token: &str, client_opt: Option<Client>) -> Result<HashSet<Repl>> { | ||
let client = create_client(&token.into(), client_opt)?; | ||
let mut all_repls = HashSet::new(); | ||
let mut visited_folder_ids = HashSet::new(); | ||
|
||
Self::fetch_recursive("", "", &client, &mut all_repls, &mut visited_folder_ids).await?; | ||
|
||
info!("got {} repls", all_repls.len()); | ||
|
||
Ok(all_repls) | ||
} | ||
|
||
async fn fetch_recursive( | ||
path: &str, | ||
folder_id: &str, | ||
client: &Client, | ||
all_repls: &mut HashSet<Repl>, | ||
visited_folder_ids: &mut HashSet<String>, | ||
) -> Result<()> { | ||
if !folder_id.is_empty() && visited_folder_ids.contains(folder_id) { | ||
info!("Skipping already visited folder: {} ({})", path, folder_id); | ||
return Ok(()); | ||
} | ||
|
||
if !folder_id.is_empty() { | ||
visited_folder_ids.insert(folder_id.to_string()); | ||
} | ||
|
||
info!("Traversing {} ({folder_id})", path); | ||
|
||
let mut cursor = None; | ||
let mut retry_count = 0; | ||
let max_retries = 5; | ||
|
||
loop { | ||
let folder_query = ReplList::build_query(repl_list::Variables { | ||
path: path.to_string(), | ||
starred: None, | ||
after: cursor.clone(), | ||
}); | ||
|
||
let folder_data = loop { | ||
if retry_count >= max_retries { | ||
return Err(anyhow::anyhow!("Max retries reached for path {path}")); | ||
} | ||
|
||
let response = client.post(REPLIT_GQL_URL).json(&folder_query).send().await; | ||
|
||
match response { | ||
Ok(res) if res.status() == StatusCode::TOO_MANY_REQUESTS => { | ||
let wait_time = Duration::from_secs(2u64.pow(retry_count)); | ||
warn!("Rate-limited - waiting {:?} before retrying", wait_time); | ||
sleep(wait_time).await; | ||
retry_count = (retry_count + 1).min(max_retries); | ||
continue; | ||
} | ||
Ok(res) => break res, | ||
Err(e) => { | ||
warn!("Error fetching data: {:?}", e); | ||
let wait_time = Duration::from_secs(2u64.pow(retry_count)); | ||
warn!("Waiting {:?} before retrying", wait_time); | ||
sleep(wait_time).await; | ||
retry_count = (retry_count + 1).min(max_retries); | ||
|
||
continue; | ||
} | ||
} | ||
}; | ||
|
||
let folder_data = folder_data.text().await?; | ||
|
||
let folder: Response<repl_list::ResponseData> = serde_json::from_str(&folder_data)?; | ||
log::trace!("{path}-{:#?}", folder); | ||
|
||
let folder = folder | ||
.data | ||
.and_then(|data| data.current_user) | ||
.and_then(|user| user.repl_folder_by_path) | ||
.ok_or_else(|| anyhow::anyhow!("Failed to get folder data"))?; | ||
|
||
// Process subfolders | ||
for subfolder in folder.folders { | ||
Box::pin(Self::fetch_recursive( | ||
&subfolder.pathnames.join("/"), | ||
&subfolder.id, | ||
client, | ||
all_repls, | ||
visited_folder_ids, | ||
)) | ||
.await?; | ||
} | ||
|
||
for repl in folder.repls.items { | ||
all_repls.insert(Repl { | ||
id: repl.id, | ||
title: repl.title, | ||
slug: repl.slug, | ||
private: repl.is_private, | ||
url: repl.url, | ||
time_created: repl.time_created, | ||
}); | ||
|
||
} | ||
|
||
sleep(Duration::from_millis(250)).await; | ||
|
||
// Check for next page | ||
match folder.repls.page_info.next_cursor { | ||
Some(next_cursor) => cursor = Some(next_cursor), | ||
None => break, | ||
} | ||
} | ||
|
||
Ok(()) | ||
} | ||
} |
Oops, something went wrong.