refactor: make everything a little more modular

This commit is contained in:
VC
2022-04-22 13:36:02 +02:00
parent de375b9f28
commit 080218f385
11 changed files with 478 additions and 323 deletions

51
src/config.rs Normal file
View File

@@ -0,0 +1,51 @@
// std
use std::fs::read_to_string;
// toml
use serde::Deserialize;
/// General configuration Struct
#[derive(Debug, Deserialize)]
pub struct Config {
pub twitter: TwitterConfig,
pub mastodon: MastodonConfig,
pub scootaloo: ScootalooConfig,
}
#[derive(Debug, Deserialize)]
pub struct TwitterConfig {
pub username: String,
pub consumer_key: String,
pub consumer_secret: String,
pub access_key: String,
pub access_secret: String,
}
#[derive(Debug, Deserialize)]
pub struct MastodonConfig {
pub base: String,
pub client_id: String,
pub client_secret: String,
pub redirect: String,
pub token: String,
}
#[derive(Debug, Deserialize)]
pub struct ScootalooConfig {
pub db_path: String,
pub cache_path: String,
}
/// Parses the TOML file into a Config Struct
pub fn parse_toml(toml_file: &str) -> Config {
let toml_config = read_to_string(toml_file).unwrap_or_else(|e|
panic!("Cannot open config file {}: {}", toml_file, e)
);
let config: Config = toml::from_str(&toml_config).unwrap_or_else(|e|
panic!("Cannot parse TOML file {}: {}", toml_file, e)
);
config
}

27
src/error.rs Normal file
View File

@@ -0,0 +1,27 @@
use std::fmt;
#[derive(Debug)]
pub struct ScootalooError {
details: String,
}
impl ScootalooError {
pub fn new(msg: &str) -> ScootalooError {
ScootalooError {
details: String::from(msg),
}
}
}
impl fmt::Display for ScootalooError {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(f, "{}", self.details)
}
}
impl std::error::Error for ScootalooError {
fn description(&self) -> &str {
&self.details
}
}

View File

@@ -1,334 +1,53 @@
// auto-imports
mod error;
use error::ScootalooError;
mod config;
use config::Config;
pub use config::parse_toml;
mod mastodon;
use mastodon::{get_mastodon_token, build_basic_status};
pub use mastodon::register;
mod twitter;
use twitter::*;
mod util;
mod state;
use state::{read_state, write_state};
pub use state::init_db;
// std
use std::{
borrow::Cow,
collections::HashMap,
io::stdin,
fmt,
fs::{read_to_string, write},
error::Error,
};
use std::borrow::Cow;
// toml
use serde::Deserialize;
// egg-mode
use egg_mode::{
Token,
KeyPair,
entities::{UrlEntity, MediaEntity, MentionEntity, MediaType},
user::UserID,
tweet::{
Tweet,
user_timeline,
},
};
// tokio
use tokio::fs::remove_file;
// elefren
use elefren::{
prelude::*,
apps::App,
status_builder::StatusBuilder,
scopes::Scopes,
};
// reqwest
use reqwest::Url;
// tokio
use tokio::{
io::copy,
fs::{File, create_dir_all, remove_file},
};
// htmlescape
use htmlescape::decode_html;
// log
use log::{info, warn, error, debug};
/**********
* Generic usage functions
***********/
/*
* Those functions are related to the Twitter side of things
*/
/// Reads last tweet id from a file
fn read_state(s: &str) -> Option<u64> {
let state = read_to_string(s);
if let Ok(s) = state {
debug!("Last Tweet ID (from file): {}", &s);
return s.parse::<u64>().ok();
}
None
}
/// Writes last treated tweet id to a file
fn write_state(f: &str, s: u64) -> Result<(), std::io::Error> {
write(f, format!("{}", s))
}
/// Gets Twitter oauth2 token
fn get_oauth2_token(config: &Config) -> Token {
let con_token = KeyPair::new(String::from(&config.twitter.consumer_key), String::from(&config.twitter.consumer_secret));
let access_token = KeyPair::new(String::from(&config.twitter.access_key), String::from(&config.twitter.access_secret));
Token::Access {
consumer: con_token,
access: access_token,
}
}
/// Gets Twitter user timeline
async fn get_user_timeline(config: &Config, token: Token, lid: Option<u64>) -> Result<Vec<Tweet>, Box<dyn Error>> {
// fix the page size to 200 as it is the maximum Twitter authorizes
let (_, feed) = user_timeline(UserID::from(String::from(&config.twitter.username)), true, false, &token)
.with_page_size(200)
.older(lid)
.await?;
Ok(feed.to_vec())
}
/// Decodes urls from UrlEntities
fn decode_urls(urls: &Vec<UrlEntity>) -> HashMap<String, String> {
let mut decoded_urls = HashMap::new();
for url in urls {
if url.expanded_url.is_some() {
// unwrap is safe here as we just verified that there is something inside expanded_url
decoded_urls.insert(String::from(&url.url), String::from(url.expanded_url.as_deref().unwrap()));
}
}
decoded_urls
}
/// Decodes the Twitter mention to something that will make sense once Twitter has joined the
/// Fediverse
fn twitter_mentions(ums: &Vec<MentionEntity>) -> HashMap<String, String> {
let mut decoded_mentions = HashMap::new();
for um in ums {
decoded_mentions.insert(format!("@{}", um.screen_name), format!("@{}@twitter.com", um.screen_name));
}
decoded_mentions
}
/// Retrieves a single media from a tweet and store it in a temporary file
async fn get_tweet_media(m: &MediaEntity, t: &str) -> Result<String, Box<dyn Error>> {
match m.media_type {
MediaType::Photo => {
return cache_media(&m.media_url_https, t).await;
},
_ => {
match &m.video_info {
Some(v) => {
for variant in &v.variants {
if variant.content_type == "video/mp4" {
return cache_media(&variant.url, t).await;
}
}
return Err(ScootalooError::new(&format!("Media Type for {} is video but no mp4 file URL is available", &m.url)).into());
},
None => {
return Err(ScootalooError::new(&format!("Media Type for {} is video but does not contain any video_info", &m.url)).into());
},
}
},
};
}
/*
* Those functions are related to the Mastodon side of things
*/
/// Gets Mastodon Data
fn get_mastodon_token(masto: &MastodonConfig) -> Mastodon {
let data = Data {
base: Cow::from(String::from(&masto.base)),
client_id: Cow::from(String::from(&masto.client_id)),
client_secret: Cow::from(String::from(&masto.client_secret)),
redirect: Cow::from(String::from(&masto.redirect)),
token: Cow::from(String::from(&masto.token)),
};
Mastodon::from(data)
}
/// Builds toot text from tweet
fn build_basic_status(tweet: &Tweet) -> Result<String, Box<dyn Error>> {
let mut toot = String::from(&tweet.text);
let decoded_urls = decode_urls(&tweet.entities.urls);
for decoded_url in decoded_urls {
toot = toot.replace(&decoded_url.0, &decoded_url.1);
}
let decoded_mentions = twitter_mentions(&tweet.entities.user_mentions);
for decoded_mention in decoded_mentions {
toot = toot.replace(&decoded_mention.0, &decoded_mention.1);
}
if let Ok(t) = decode_html(&toot) {
toot = t;
}
Ok(toot)
}
/*
* Generic private functions
*/
/// Gets and caches Twitter Media inside the determined temp dir
async fn cache_media(u: &str, t: &str) -> Result<String, Box<dyn Error>> {
// create dir
create_dir_all(t).await?;
// get file
let mut response = reqwest::get(u).await?;
// create local file
let url = Url::parse(u)?;
let dest_filename = url.path_segments().ok_or_else(|| ScootalooError::new(&format!("Cannot determine the destination filename for {}", u)))?
.last().ok_or_else(|| ScootalooError::new(&format!("Cannot determine the destination filename for {}", u)))?;
let dest_filepath = format!("{}/{}", t, dest_filename);
let mut dest_file = File::create(&dest_filepath).await?;
while let Some(chunk) = response.chunk().await? {
copy(&mut &*chunk, &mut dest_file).await?;
}
Ok(dest_filepath)
}
/**********
* local error handler
**********/
#[derive(Debug)]
struct ScootalooError {
details: String,
}
impl ScootalooError {
fn new(msg: &str) -> ScootalooError {
ScootalooError {
details: String::from(msg),
}
}
}
impl fmt::Display for ScootalooError {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(f, "{}", self.details)
}
}
impl std::error::Error for ScootalooError {
fn description(&self) -> &str {
&self.details
}
}
/**********
* Config structure
***********/
/// General configuration Struct
#[derive(Debug, Deserialize)]
pub struct Config {
twitter: TwitterConfig,
mastodon: MastodonConfig,
scootaloo: ScootalooConfig,
}
#[derive(Debug, Deserialize)]
struct TwitterConfig {
username: String,
consumer_key: String,
consumer_secret: String,
access_key: String,
access_secret: String,
}
#[derive(Debug, Deserialize)]
struct MastodonConfig {
base: String,
client_id: String,
client_secret: String,
redirect: String,
token: String,
}
#[derive(Debug, Deserialize)]
struct ScootalooConfig {
last_tweet_path: String,
cache_path: String,
}
/*********
* Main functions
*********/
/// Parses the TOML file into a Config Struct
pub fn parse_toml(toml_file: &str) -> Config {
let toml_config = read_to_string(toml_file).unwrap_or_else(|e|
panic!("Cannot open config file {}: {}", toml_file, e)
);
let config: Config = toml::from_str(&toml_config).unwrap_or_else(|e|
panic!("Cannot parse TOML file {}: {}", toml_file, e)
);
config
}
/// Generic register function
/// As this function is supposed to be run only once, it will panic for every error it encounters
/// Most of this function is a direct copy/paste of the official `elefren` crate
pub fn register(host: &str) {
let mut builder = App::builder();
builder.client_name(Cow::from(String::from(env!("CARGO_PKG_NAME"))))
.redirect_uris(Cow::from(String::from("urn:ietf:wg:oauth:2.0:oob")))
.scopes(Scopes::write_all())
.website(Cow::from(String::from("https://framagit.org/veretcle/scootaloo")));
let app = builder.build().expect("Cannot build the app");
let registration = Registration::new(host).register(app).expect("Cannot build registration object");
let url = registration.authorize_url().expect("Cannot generate registration URI!");
println!("Click this link to authorize on Mastodon: {}", url);
println!("Paste the returned authorization code: ");
let mut input = String::new();
stdin().read_line(&mut input).expect("Unable to read back registration code!");
let code = input.trim();
let mastodon = registration.complete(code).expect("Unable to create access token!");
let toml = toml::to_string(&*mastodon).unwrap();
println!("Please insert the following block at the end of your configuration file:\n[mastodon]\n{}", toml);
}
/// This is where the magic happens
#[tokio::main]
pub async fn run(config: Config) {
// retrieve the last tweet ID for the username
let last_tweet_id = read_state(&config.scootaloo.last_tweet_path);
let last_tweet_id = read_state(&config.scootaloo.db_path);
// get OAuth2 token
let token = get_oauth2_token(&config);
let token = get_oauth2_token(&config.twitter);
// get Mastodon instance
let mastodon = get_mastodon_token(&config.mastodon);
// get user timeline feed (Vec<tweet>)
let mut feed = get_user_timeline(&config, token, last_tweet_id)
let mut feed = get_user_timeline(&config.twitter, token, last_tweet_id)
.await
.unwrap_or_else(|e|
panic!("Something went wrong when trying to retrieve {}s timeline: {}", &config.twitter.username, e)
@@ -411,7 +130,7 @@ pub async fn run(config: Config) {
// last_tweet gathered not to be written
// write the current state (tweet ID) to avoid copying it another time
write_state(&config.scootaloo.last_tweet_path, tweet.id).unwrap_or_else(|e|
write_state(&config.scootaloo.db_path, tweet.id).unwrap_or_else(|e|
panic!("Cant write the last tweet retrieved: {}", e)
);
}

View File

@@ -11,6 +11,8 @@ use simple_logger::SimpleLogger;
// std
use std::str::FromStr;
const DEFAULT_CONFIG_PATH: &'static str = "/usr/local/etc/scootaloo.toml";
fn main() {
let matches = App::new(env!("CARGO_PKG_NAME"))
.version(env!("CARGO_PKG_VERSION"))
@@ -40,10 +42,29 @@ fn main() {
.takes_value(true)
.required(true)
.display_order(1)))
.subcommand(SubCommand::with_name("init")
.version(env!("CARGO_PKG_VERSION"))
.about("Command to init Scootaloo DB")
.arg(Arg::with_name("config")
.short("c")
.long("config")
.value_name("CONFIG_FILE")
.help("TOML config file for scootaloo (default /usr/local/etc/scootaloo.toml")
.takes_value(true)
.display_order(1)))
.get_matches();
if let Some(matches) = matches.subcommand_matches("register") {
register(matches.value_of("host").unwrap());
return;
match matches.subcommand() {
("register", Some(sub_m)) => {
register(sub_m.value_of("host").unwrap());
return;
},
("init", Some(sub_m)) => {
let config = parse_toml(sub_m.value_of("config").unwrap_or(DEFAULT_CONFIG_PATH));
init_db(&config).unwrap();
return;
},
_ => (),
}
if matches.is_present("log_level") {
@@ -56,7 +77,7 @@ fn main() {
};
}
let config = parse_toml(matches.value_of("config").unwrap_or("/usr/local/etc/scootaloo.toml"));
let config = parse_toml(matches.value_of("config").unwrap_or(DEFAULT_CONFIG_PATH));
run(config);
}

106
src/mastodon.rs Normal file
View File

@@ -0,0 +1,106 @@
// auto imports
use crate::config::MastodonConfig;
use crate::twitter::decode_urls;
// std
use std::{
borrow::Cow,
error::Error,
collections::HashMap,
io::stdin,
};
// htmlescape
use htmlescape::decode_html;
// egg-mode
use egg_mode::{
tweet::Tweet,
entities::MentionEntity,
};
// elefren
use elefren::{
prelude::*,
apps::App,
scopes::Scopes,
};
/// Decodes the Twitter mention to something that will make sense once Twitter has joined the
/// Fediverse
fn twitter_mentions(ums: &Vec<MentionEntity>) -> HashMap<String, String> {
let mut decoded_mentions = HashMap::new();
for um in ums {
decoded_mentions.insert(format!("@{}", um.screen_name), format!("@{}@twitter.com", um.screen_name));
}
decoded_mentions
}
/// Gets Mastodon Data
pub fn get_mastodon_token(masto: &MastodonConfig) -> Mastodon {
let data = Data {
base: Cow::from(String::from(&masto.base)),
client_id: Cow::from(String::from(&masto.client_id)),
client_secret: Cow::from(String::from(&masto.client_secret)),
redirect: Cow::from(String::from(&masto.redirect)),
token: Cow::from(String::from(&masto.token)),
};
Mastodon::from(data)
}
/// Builds toot text from tweet
pub fn build_basic_status(tweet: &Tweet) -> Result<String, Box<dyn Error>> {
let mut toot = String::from(&tweet.text);
let decoded_urls = decode_urls(&tweet.entities.urls);
for decoded_url in decoded_urls {
toot = toot.replace(&decoded_url.0, &decoded_url.1);
}
let decoded_mentions = twitter_mentions(&tweet.entities.user_mentions);
for decoded_mention in decoded_mentions {
toot = toot.replace(&decoded_mention.0, &decoded_mention.1);
}
if let Ok(t) = decode_html(&toot) {
toot = t;
}
Ok(toot)
}
/// Generic register function
/// As this function is supposed to be run only once, it will panic for every error it encounters
/// Most of this function is a direct copy/paste of the official `elefren` crate
pub fn register(host: &str) {
let mut builder = App::builder();
builder.client_name(Cow::from(String::from(env!("CARGO_PKG_NAME"))))
.redirect_uris(Cow::from(String::from("urn:ietf:wg:oauth:2.0:oob")))
.scopes(Scopes::write_all())
.website(Cow::from(String::from("https://framagit.org/veretcle/scootaloo")));
let app = builder.build().expect("Cannot build the app");
let registration = Registration::new(host).register(app).expect("Cannot build registration object");
let url = registration.authorize_url().expect("Cannot generate registration URI!");
println!("Click this link to authorize on Mastodon: {}", url);
println!("Paste the returned authorization code: ");
let mut input = String::new();
stdin().read_line(&mut input).expect("Unable to read back registration code!");
let code = input.trim();
let mastodon = registration.complete(code).expect("Unable to create access token!");
let toml = toml::to_string(&*mastodon).unwrap();
println!("Please insert the following block at the end of your configuration file:\n[mastodon]\n{}", toml);
}

38
src/state.rs Normal file
View File

@@ -0,0 +1,38 @@
// auto-imports
use crate::config::Config;
// std
use std::{
fs::{read_to_string, write},
error::Error,
};
// log
use log::debug;
/// Reads last tweet id from a file
pub fn read_state(s: &str) -> Option<u64> {
let state = read_to_string(s);
if let Ok(s) = state {
debug!("Last Tweet ID (from file): {}", &s);
return s.parse::<u64>().ok();
}
None
}
/// Writes last treated tweet id to a file
pub fn write_state(f: &str, s: u64) -> Result<(), std::io::Error> {
write(f, format!("{}", s))
}
/*********
* Main functions
*********/
/// Initiates the DB from path
pub fn init_db(config: &Config) -> Result<(), Box<dyn Error>> {
println!("config.scootaloo.db_path: {}", config.scootaloo.db_path);
Ok(())
}

83
src/twitter.rs Normal file
View File

@@ -0,0 +1,83 @@
// auto-imports
use crate::ScootalooError;
use crate::config::TwitterConfig;
use crate::util::cache_media;
// std
use std::{
error::Error,
collections::HashMap,
};
// egg-mode
use egg_mode::{
Token,
KeyPair,
entities::{UrlEntity, MediaEntity, MediaType},
user::UserID,
tweet::{
Tweet,
user_timeline,
},
};
/// Gets Twitter oauth2 token
pub fn get_oauth2_token(config: &TwitterConfig) -> Token {
let con_token = KeyPair::new(String::from(&config.consumer_key), String::from(&config.consumer_secret));
let access_token = KeyPair::new(String::from(&config.access_key), String::from(&config.access_secret));
Token::Access {
consumer: con_token,
access: access_token,
}
}
/// Gets Twitter user timeline
pub async fn get_user_timeline(config: &TwitterConfig, token: Token, lid: Option<u64>) -> Result<Vec<Tweet>, Box<dyn Error>> {
// fix the page size to 200 as it is the maximum Twitter authorizes
let (_, feed) = user_timeline(UserID::from(String::from(&config.username)), true, false, &token)
.with_page_size(200)
.older(lid)
.await?;
Ok(feed.to_vec())
}
/// Decodes urls from UrlEntities
pub fn decode_urls(urls: &Vec<UrlEntity>) -> HashMap<String, String> {
let mut decoded_urls = HashMap::new();
for url in urls {
if url.expanded_url.is_some() {
// unwrap is safe here as we just verified that there is something inside expanded_url
decoded_urls.insert(String::from(&url.url), String::from(url.expanded_url.as_deref().unwrap()));
}
}
decoded_urls
}
/// Retrieves a single media from a tweet and store it in a temporary file
pub async fn get_tweet_media(m: &MediaEntity, t: &str) -> Result<String, Box<dyn Error>> {
match m.media_type {
MediaType::Photo => {
return cache_media(&m.media_url_https, t).await;
},
_ => {
match &m.video_info {
Some(v) => {
for variant in &v.variants {
if variant.content_type == "video/mp4" {
return cache_media(&variant.url, t).await;
}
}
return Err(ScootalooError::new(&format!("Media Type for {} is video but no mp4 file URL is available", &m.url)).into());
},
None => {
return Err(ScootalooError::new(&format!("Media Type for {} is video but does not contain any video_info", &m.url)).into());
},
}
},
};
}

37
src/util.rs Normal file
View File

@@ -0,0 +1,37 @@
// std
use std::error::Error;
use crate::ScootalooError;
// reqwest
use reqwest::Url;
// tokio
use tokio::{
io::copy,
fs::{File, create_dir_all},
};
/// Gets and caches Twitter Media inside the determined temp dir
pub async fn cache_media(u: &str, t: &str) -> Result<String, Box<dyn Error>> {
// create dir
create_dir_all(t).await?;
// get file
let mut response = reqwest::get(u).await?;
// create local file
let url = Url::parse(u)?;
let dest_filename = url.path_segments().ok_or_else(|| ScootalooError::new(&format!("Cannot determine the destination filename for {}", u)))?
.last().ok_or_else(|| ScootalooError::new(&format!("Cannot determine the destination filename for {}", u)))?;
let dest_filepath = format!("{}/{}", t, dest_filename);
let mut dest_file = File::create(&dest_filepath).await?;
while let Some(chunk) = response.chunk().await? {
copy(&mut &*chunk, &mut dest_file).await?;
}
Ok(dest_filepath)
}