modmapper/src/commands/update.rs

285 lines
12 KiB
Rust
Raw Normal View History

use anyhow::Result;
use chrono::{NaiveDateTime, NaiveTime};
use humansize::{file_size_opts, FileSize};
use reqwest::StatusCode;
use std::collections::HashSet;
use std::io::SeekFrom;
use std::time::Duration;
use tokio::io::{AsyncReadExt, AsyncSeekExt};
use tokio::time::sleep;
use tracing::{debug, info, info_span, warn};
use crate::extractors::{self, extract_with_7zip, extract_with_compress_tools, extract_with_unrar};
use crate::models::file;
use crate::models::game;
use crate::models::{game_mod, game_mod::UnsavedMod};
use crate::nexus_api::{self, GAME_ID, GAME_NAME};
use crate::nexus_scraper;
const REQUEST_TIMEOUT: Duration = Duration::from_secs(7200); // 2 hours
const CONNECT_TIMEOUT: Duration = Duration::from_secs(30);
pub async fn update(pool: &sqlx::Pool<sqlx::Postgres>, start_page: usize) -> Result<()> {
let mut page = start_page;
let mut has_next_page = true;
let game = game::insert(&pool, GAME_NAME, GAME_ID as i32).await?;
let client = reqwest::Client::builder()
.timeout(REQUEST_TIMEOUT)
.connect_timeout(CONNECT_TIMEOUT)
.build()?;
while has_next_page {
let page_span = info_span!("page", page);
let _page_span = page_span.enter();
let mod_list_resp = nexus_scraper::get_mod_list_page(&client, page).await?;
let scraped = mod_list_resp.scrape_mods()?;
has_next_page = scraped.has_next_page;
let processed_mods = game_mod::bulk_get_last_updated_by_nexus_mod_ids(
&pool,
&scraped
.mods
.iter()
.map(|scraped_mod| scraped_mod.nexus_mod_id)
.collect::<Vec<i32>>(),
)
.await?;
let mods_to_create_or_update: Vec<UnsavedMod> = scraped
.mods
.iter()
.filter(|scraped_mod| {
if let Some(processed_mod) = processed_mods
.iter()
.find(|processed_mod| processed_mod.nexus_mod_id == scraped_mod.nexus_mod_id)
{
if processed_mod.last_updated_files_at
> NaiveDateTime::new(
scraped_mod.last_update_at,
NaiveTime::from_hms(0, 0, 0),
)
{
return false;
}
}
true
})
.map(|scraped_mod| UnsavedMod {
name: scraped_mod.name,
nexus_mod_id: scraped_mod.nexus_mod_id,
author_name: scraped_mod.author_name,
author_id: scraped_mod.author_id,
category_name: scraped_mod.category_name,
category_id: scraped_mod.category_id,
description: scraped_mod.desc,
thumbnail_link: scraped_mod.thumbnail_link,
game_id: game.id,
last_update_at: NaiveDateTime::new(
scraped_mod.last_update_at,
NaiveTime::from_hms(0, 0, 0),
),
first_upload_at: NaiveDateTime::new(
scraped_mod.first_upload_at,
NaiveTime::from_hms(0, 0, 0),
),
})
.collect();
let mods = game_mod::batched_insert(&pool, &mods_to_create_or_update).await?;
for db_mod in mods {
let mod_span = info_span!("mod", name = ?&db_mod.name, id = &db_mod.nexus_mod_id);
let _mod_span = mod_span.enter();
let files_resp = nexus_api::files::get(&client, db_mod.nexus_mod_id).await?;
debug!(duration = ?files_resp.wait, "sleeping");
sleep(files_resp.wait).await;
// Filter out replaced/deleted files (indicated by null category) and archived files
let files = files_resp
.files()?
.into_iter()
.filter(|file| match file.category {
None => {
info!(
name = file.file_name,
id = file.file_id,
"skipping file with no category"
);
false
}
Some(category) if category == "ARCHIVED" => false,
Some(_) => true,
});
let processed_file_ids: HashSet<i32> =
file::get_processed_nexus_file_ids_by_mod_id(&pool, db_mod.id)
.await?
.into_iter()
.collect();
for api_file in files {
let file_span =
info_span!("file", name = &api_file.file_name, id = &api_file.file_id,);
let _file_span = file_span.enter();
if processed_file_ids.contains(&(api_file.file_id as i32)) {
info!("skipping file already present and processed in database");
continue;
}
let db_file = file::insert(
&pool,
&file::UnsavedFile {
name: api_file.name,
file_name: api_file.file_name,
nexus_file_id: api_file.file_id as i32,
mod_id: db_mod.id,
category: api_file.category,
version: api_file.version,
mod_version: api_file.mod_version,
size: api_file.size,
uploaded_at: api_file.uploaded_at,
},
)
.await?;
let mut checked_metadata = false;
match nexus_api::metadata::contains_plugin(&client, &api_file).await {
Ok(contains_plugin) => {
if let Some(contains_plugin) = contains_plugin {
checked_metadata = true;
if !contains_plugin {
info!("file metadata does not contain a plugin, skip downloading");
file::update_has_plugin(&pool, db_file.id, false).await?;
continue;
}
} else {
warn!("file has no metadata link, continuing with download");
}
}
Err(err) => {
warn!(error = %err, "error retreiving metadata for file, continuing with download");
}
};
let humanized_size = api_file
.size
.file_size(file_size_opts::CONVENTIONAL)
.expect("unable to create human-readable file size");
info!(size = %humanized_size, "decided to download file");
let download_link_resp =
nexus_api::download_link::get(&client, db_mod.nexus_mod_id, api_file.file_id)
.await;
if let Err(err) = &download_link_resp {
if let Some(reqwest_err) = err.downcast_ref::<reqwest::Error>() {
if reqwest_err.status() == Some(StatusCode::NOT_FOUND) {
warn!(
status = ?reqwest_err.status(),
"failed to get download link for file, skipping file"
);
file::update_has_download_link(&pool, db_file.id, false).await?;
continue;
}
}
}
let download_link_resp = download_link_resp?;
let mut tokio_file = match download_link_resp.download_file(&client).await {
Ok(file) => {
info!(bytes = api_file.size, "download finished");
file::update_downloaded_at(&pool, db_file.id).await?;
file
}
Err(err) => {
warn!(error = %err, "failed all attempts at downloading file, skipping file");
continue;
}
};
let mut initial_bytes = [0; 8];
tokio_file.seek(SeekFrom::Start(0)).await?;
if let Err(err) = tokio_file.read_exact(&mut initial_bytes).await {
warn!(error = %err, "failed to read initial bytes, skipping file");
file::update_unable_to_extract_plugins(&pool, db_file.id, true).await?;
continue;
}
let kind = match infer::get(&initial_bytes) {
Some(kind) => kind,
None => {
warn!(initial_bytes = ?initial_bytes, "unable to determine file type of archive, skipping file");
file::update_unable_to_extract_plugins(&pool, db_file.id, true).await?;
continue;
}
};
info!(
mime_type = kind.mime_type(),
"inferred mime_type of downloaded archive"
);
match kind.mime_type() {
"application/vnd.rar" => {
info!("downloaded archive is RAR archive, attempt to uncompress entire archive");
// Use unrar to uncompress the entire .rar file to avoid bugs with compress_tools uncompressing certain .rar files:
// https://github.com/libarchive/libarchive/issues/373, https://github.com/libarchive/libarchive/issues/1426
tokio_file.seek(SeekFrom::Start(0)).await?;
let mut file = tokio_file.try_clone().await?.into_std().await;
match extract_with_unrar(
&mut file,
&pool,
&db_file,
&db_mod,
checked_metadata,
)
.await
{
Ok(_) => Ok(()),
Err(err) => {
// unrar failed to extract rar file (e.g. archive has unicode filenames)
// Attempt to uncompress the archive using `7z` unix command instead
warn!(error = %err, "failed to extract file with unrar, extracting whole archive with 7z instead");
extract_with_7zip(&mut file, &pool, &db_file, &db_mod).await
}
}?;
}
_ => {
tokio_file.seek(SeekFrom::Start(0)).await?;
let mut file = tokio_file.try_clone().await?.into_std().await;
match extract_with_compress_tools(&mut file, &pool, &db_file, &db_mod).await
{
Ok(_) => Ok(()),
Err(err) => {
if err
.downcast_ref::<extractors::compress_tools::ExtractorError>()
.is_some()
&& (kind.mime_type() == "application/zip"
|| kind.mime_type() == "application/x-7z-compressed")
{
// compress_tools or libarchive failed to extract zip/7z file (e.g. archive is deflate64 compressed)
// Attempt to uncompress the archive using `7z` unix command instead
warn!(error = %err, "failed to extract file with compress_tools, extracting whole archive with 7z instead");
extract_with_7zip(&mut file, &pool, &db_file, &db_mod).await
} else {
Err(err)
}
}
}?;
}
}
debug!(duration = ?download_link_resp.wait, "sleeping");
sleep(download_link_resp.wait).await;
}
game_mod::update_last_updated_files_at(&pool, db_mod.id).await?;
}
page += 1;
debug!(?page, ?has_next_page, "sleeping 1 second");
sleep(Duration::from_secs(1)).await;
}
Ok(())
}