Merge pull request 'Recursive, parallel crawl' (#11) from recursive-crawl into main

Reviewed-on: https://yerbamate.ml/LemmyNet/lemmy-stats-crawler/pulls/11
This commit is contained in:
nutomic 2022-05-13 11:37:52 +00:00
commit 575672cbe3
5 changed files with 326 additions and 172 deletions

173
Cargo.lock generated
View file

@ -205,6 +205,15 @@ dependencies = [
"memchr",
]
[[package]]
name = "ansi_term"
version = "0.12.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "d52a9bb7ec0cf484c551830a7ce27bd20d67eac647e1befb56b0be4ee39a55d2"
dependencies = [
"winapi",
]
[[package]]
name = "anyhow"
version = "1.0.57"
@ -220,6 +229,17 @@ dependencies = [
"event-listener",
]
[[package]]
name = "async-recursion"
version = "1.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "2cda8f4bcc10624c4e85bc66b3f452cca98cfa5ca002dc83a16aad2367641bea"
dependencies = [
"proc-macro2 1.0.37",
"quote 1.0.18",
"syn 1.0.92",
]
[[package]]
name = "async-trait"
version = "0.1.53"
@ -433,26 +453,17 @@ dependencies = [
[[package]]
name = "clap"
version = "3.1.15"
version = "2.34.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "85a35a599b11c089a7f49105658d089b8f2cf0882993c17daf6de15285c2c35d"
checksum = "a0610544180c38b88101fecf2dd634b174a62eef6946f84dfc6a7127512b381c"
dependencies = [
"ansi_term",
"atty",
"bitflags",
"clap_lex",
"indexmap",
"strsim",
"termcolor",
"strsim 0.8.0",
"textwrap",
]
[[package]]
name = "clap_lex"
version = "0.2.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a37c35f1112dad5e6e0b1adaff798507497a18fceeb30cceb3bae7d1427b9213"
dependencies = [
"os_str_bytes",
"unicode-width",
"vec_map",
]
[[package]]
@ -545,7 +556,7 @@ dependencies = [
"ident_case",
"proc-macro2 1.0.37",
"quote 1.0.18",
"strsim",
"strsim 0.10.0",
"syn 1.0.92",
]
@ -560,6 +571,17 @@ dependencies = [
"syn 1.0.92",
]
[[package]]
name = "derive-new"
version = "0.5.9"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "3418329ca0ad70234b9735dc4ceed10af4df60eff9c8e7b06cb5e520d92c3535"
dependencies = [
"proc-macro2 1.0.37",
"quote 1.0.18",
"syn 1.0.92",
]
[[package]]
name = "derive_more"
version = "0.99.17"
@ -1008,6 +1030,15 @@ version = "0.11.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "ab5ef0d4909ef3724cc8cce6ccc8572c5c817592e9285f5464f8e86f8bd3726e"
[[package]]
name = "heck"
version = "0.3.3"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "6d621efb26863f0e9924c6ac577e8275e5e6b77455db64ffa6c65c904e9e132c"
dependencies = [
"unicode-segmentation",
]
[[package]]
name = "heck"
version = "0.4.0"
@ -1295,14 +1326,18 @@ name = "lemmy-stats-crawler"
version = "0.1.0"
dependencies = [
"anyhow",
"clap",
"async-recursion",
"derive-new",
"futures",
"lemmy_api_common",
"log",
"once_cell",
"reqwest",
"semver",
"serde",
"serde_json",
"stderrlog",
"structopt",
"tokio",
]
@ -1806,12 +1841,6 @@ dependencies = [
"vcpkg",
]
[[package]]
name = "os_str_bytes"
version = "6.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8e22443d1643a904602595ba1cd8f7d896afe56d26712531c5ff73a15b2fbf64"
[[package]]
name = "parking_lot"
version = "0.11.2"
@ -2040,6 +2069,30 @@ version = "0.1.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "925383efa346730478fb4838dbe9137d2a47675ad789c546d150a6e1dd4ab31c"
[[package]]
name = "proc-macro-error"
version = "1.0.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "da25490ff9892aab3fcf7c36f08cfb902dd3e71ca0f9f9517bea02a73a5ce38c"
dependencies = [
"proc-macro-error-attr",
"proc-macro2 1.0.37",
"quote 1.0.18",
"syn 1.0.92",
"version_check",
]
[[package]]
name = "proc-macro-error-attr"
version = "1.0.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a1be40180e52ecc98ad80b184934baf3d0d29f979574e439af5a55274b35f869"
dependencies = [
"proc-macro2 1.0.37",
"quote 1.0.18",
"version_check",
]
[[package]]
name = "proc-macro2"
version = "0.4.30"
@ -2587,6 +2640,19 @@ version = "0.5.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "6e63cff320ae2c57904679ba7cb63280a3dc4613885beafb148ee7bf9aa9042d"
[[package]]
name = "stderrlog"
version = "0.5.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "45a53e2eff3e94a019afa6265e8ee04cb05b9d33fe9f5078b14e4e391d155a38"
dependencies = [
"atty",
"chrono",
"log",
"termcolor",
"thread_local",
]
[[package]]
name = "string_cache"
version = "0.8.4"
@ -2613,12 +2679,42 @@ dependencies = [
"quote 1.0.18",
]
[[package]]
name = "strsim"
version = "0.8.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "8ea5119cdb4c55b55d432abb513a0429384878c15dde60cc77b1c99de1a95a6a"
[[package]]
name = "strsim"
version = "0.10.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "73473c0e59e6d5812c5dfe2a064a6444949f089e20eec9a2e5506596494e4623"
[[package]]
name = "structopt"
version = "0.3.26"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "0c6b5c64445ba8094a6ab0c3cd2ad323e07171012d9c98b0b15651daf1787a10"
dependencies = [
"clap",
"lazy_static",
"structopt-derive",
]
[[package]]
name = "structopt-derive"
version = "0.4.18"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "dcb5ae327f9cc13b68763b5749770cb9e048a99bd9dfdfa58d0cf05d5f64afe0"
dependencies = [
"heck 0.3.3",
"proc-macro-error",
"proc-macro2 1.0.37",
"quote 1.0.18",
"syn 1.0.92",
]
[[package]]
name = "strum"
version = "0.24.0"
@ -2631,7 +2727,7 @@ version = "0.24.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "6878079b17446e4d3eba6192bb0a2950d5b14f0ed8424b852310e5a94345d0ef"
dependencies = [
"heck",
"heck 0.4.0",
"proc-macro2 1.0.37",
"quote 1.0.18",
"rustversion",
@ -2705,9 +2801,12 @@ dependencies = [
[[package]]
name = "textwrap"
version = "0.15.0"
version = "0.11.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b1141d4d61095b28419e22cb0bbf02755f5e54e0526f97f1e3d1d160e60885fb"
checksum = "d326610f408c7a4eb6f51c37c330e496b08506c9457c9d34287ecc38809fb060"
dependencies = [
"unicode-width",
]
[[package]]
name = "thiserror"
@ -2731,11 +2830,11 @@ dependencies = [
[[package]]
name = "thread_local"
version = "1.1.4"
version = "1.0.1"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "5516c27b78311c50bf42c071425c560ac799b11c30b31f87e3081965fe5e0180"
checksum = "d40c6d1b69745a6ec6fb1ca717914848da4b44ae29d9b3080cbee91d72a69b14"
dependencies = [
"once_cell",
"lazy_static",
]
[[package]]
@ -2918,9 +3017,9 @@ dependencies = [
[[package]]
name = "tracing-subscriber"
version = "0.3.11"
version = "0.3.6"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "4bc28f93baff38037f64e6f43d34cfa1605f27a49c34e8a04c5e78b0babf2596"
checksum = "77be66445c4eeebb934a7340f227bfe7b338173d3f8c00a60a5a58005c9faecf"
dependencies = [
"sharded-slab",
"thread_local",
@ -2991,6 +3090,12 @@ dependencies = [
"tinyvec",
]
[[package]]
name = "unicode-segmentation"
version = "1.9.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "7e8820f5d777f6224dc4be3632222971ac30164d4a258d595640799554ebfd99"
[[package]]
name = "unicode-width"
version = "0.1.9"
@ -3062,6 +3167,12 @@ version = "0.2.15"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "accd4ea62f7bb7a82fe23066fb0957d48ef677f6eeb8215f372f52e48bb32426"
[[package]]
name = "vec_map"
version = "0.8.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f1bddf1187be692e79c5ffeab891132dfb0f236ed36a43c7ed39f1165ee20191"
[[package]]
name = "version_check"
version = "0.9.4"

View file

@ -11,7 +11,11 @@ anyhow = "1.0.57"
tokio = { version = "1.18.1", features = ["macros", "rt-multi-thread"] }
futures = "0.3.21"
serde_json = "1.0.81"
clap = "3.1.15"
semver = "1.0.9"
once_cell = "1.10.0"
lemmy_api_common = "0.16.0"
async-recursion = "1.0.0"
log = "0.4.17"
derive-new = "0.5.9"
stderrlog = "0.5.1"
structopt = "0.3.26"

View file

@ -1,65 +1,30 @@
use crate::CLIENT;
use crate::REQUEST_TIMEOUT;
use anyhow::anyhow;
use anyhow::Error;
use async_recursion::async_recursion;
use futures::future::join_all;
use lemmy_api_common::site::GetSiteResponse;
use once_cell::sync::Lazy;
use reqwest::Client;
use log::debug;
use semver::Version;
use serde::Serialize;
use std::collections::VecDeque;
use std::collections::HashSet;
use std::ops::Deref;
use std::sync::Arc;
use tokio::sync::Mutex;
static CLIENT: Lazy<Client> = Lazy::new(Client::default);
#[derive(new)]
pub struct CrawlJob {
domain: String,
current_distance: i32,
params: Arc<CrawlParams>,
}
pub async fn crawl(
start_instances: Vec<String>,
exclude: Vec<String>,
#[derive(new)]
pub struct CrawlParams {
min_lemmy_version: Version,
exclude_domains: Vec<String>,
max_depth: i32,
) -> Result<(Vec<InstanceDetails>, i32), Error> {
let mut pending_instances: VecDeque<CrawlInstance> = start_instances
.iter()
.map(|s| CrawlInstance::new(s.to_string(), 0))
.collect();
let min_lemmy_version = min_lemmy_version().await?;
let mut crawled_instances = vec![];
let mut instance_details = vec![];
let mut failed_instances = 0;
while let Some(current_instance) = pending_instances.pop_back() {
crawled_instances.push(current_instance.domain.clone());
if current_instance.depth > max_depth || exclude.contains(&current_instance.domain) {
continue;
}
match fetch_instance_details(&current_instance.domain, &min_lemmy_version).await {
Ok(details) => {
if let Some(federated) = &details.site_info.federated_instances.as_ref() {
for i in &federated.linked {
let is_in_crawled = crawled_instances.contains(i);
let is_in_pending = pending_instances.iter().any(|p| &p.domain == i);
if !is_in_crawled && !is_in_pending {
let ci = CrawlInstance::new(i.clone(), current_instance.depth + 1);
pending_instances.push_back(ci);
}
}
}
instance_details.push(details);
}
Err(e) => {
failed_instances += 1;
eprintln!("Failed to crawl {}: {}", current_instance.domain, e)
}
}
}
// Sort by active monthly users descending
instance_details.sort_by_key(|i| {
i.site_info
.site_view
.as_ref()
.map(|s| s.counts.users_active_month)
.unwrap_or(0)
});
instance_details.reverse();
Ok((instance_details, failed_instances))
crawled_instances: Arc<Mutex<HashSet<String>>>,
}
#[derive(Serialize, Debug)]
@ -68,54 +33,65 @@ pub struct InstanceDetails {
pub site_info: GetSiteResponse,
}
struct CrawlInstance {
domain: String,
depth: i32,
}
impl CrawlInstance {
pub fn new(domain: String, depth: i32) -> CrawlInstance {
CrawlInstance { domain, depth }
impl CrawlJob {
#[async_recursion]
pub async fn crawl(self) -> Vec<Result<InstanceDetails, Error>> {
// need to acquire and release mutex before recursing, otherwise it will deadlock
{
let mut crawled_instances = self.params.crawled_instances.deref().lock().await;
if crawled_instances.contains(&self.domain) {
return vec![];
} else {
crawled_instances.insert(self.domain.clone());
}
}
}
async fn fetch_instance_details(
domain: &str,
min_lemmy_version: &Version,
) -> Result<InstanceDetails, Error> {
let client = Client::default();
if self.current_distance > self.params.max_depth
|| self.params.exclude_domains.contains(&self.domain)
{
return vec![];
}
let site_info_url = format!("https://{}/api/v3/site", domain);
let site_info = client
debug!("Starting crawl for {}, distance {}", &self.domain, &self.current_distance);
let site_info = match self.fetch_instance_details().await {
Ok(o) => o,
Err(e) => return vec![Err(e)],
};
if site_info.1 < self.params.min_lemmy_version {
return vec![];
}
let mut result = vec![];
if let Some(federated) = &site_info.0.federated_instances {
for domain in federated.linked.iter() {
let crawl_job =
CrawlJob::new(domain.clone(), self.current_distance + 1, self.params.clone());
result.push(crawl_job.crawl());
}
}
let mut result2: Vec<Result<InstanceDetails, Error>> =
join_all(result).await.into_iter().flatten().collect();
debug!("Successfully finished crawl for {}", &self.domain);
result2.push(Ok(InstanceDetails {
domain: self.domain,
site_info: site_info.0,
}));
result2
}
async fn fetch_instance_details(&self) -> Result<(GetSiteResponse, Version), Error> {
let site_info_url = format!("https://{}/api/v3/site", &self.domain);
let site_info = CLIENT
.get(&site_info_url)
.timeout(REQUEST_TIMEOUT)
.send()
.await?
.json::<GetSiteResponse>()
.await?;
let version = Version::parse(&site_info.version)?;
if &version < min_lemmy_version {
return Err(anyhow!("lemmy version is too old ({})", version));
Ok((site_info, version))
}
Ok(InstanceDetails {
domain: domain.to_owned(),
site_info,
})
}
/// calculate minimum allowed lemmy version based on current version. in case of current version
/// 0.16.3, the minimum from this function is 0.15.3. this is to avoid rejecting all instances on
/// the previous version when a major lemmy release is published.
async fn min_lemmy_version() -> Result<Version, Error> {
let lemmy_version_url = "https://raw.githubusercontent.com/LemmyNet/lemmy-ansible/main/VERSION";
let req = CLIENT
.get(lemmy_version_url)
.timeout(REQUEST_TIMEOUT)
.send()
.await?;
let mut version = Version::parse(req.text().await?.trim())?;
version.minor -= 1;
Ok(version)
}

View file

@ -1,9 +1,72 @@
#[macro_use]
extern crate derive_new;
use crate::crawl::{CrawlJob, CrawlParams, InstanceDetails};
use anyhow::Error;
use futures::future::join_all;
use once_cell::sync::Lazy;
use reqwest::Client;
use semver::Version;
use std::collections::HashSet;
use std::sync::Arc;
use std::time::Duration;
use tokio::sync::Mutex;
pub mod crawl;
pub const REQUEST_TIMEOUT: Duration = Duration::from_secs(10);
pub const DEFAULT_START_INSTANCES: &str = "lemmy.ml";
pub const DEFAULT_MAX_CRAWL_DEPTH: &str = "20";
pub const EXCLUDE_INSTANCES: &str =
"ds9.lemmy.ml, enterprise.lemmy.ml, voyager.lemmy.ml, test.lemmy.ml";
static CLIENT: Lazy<Client> = Lazy::new(Client::default);
pub async fn start_crawl(
start_instances: Vec<String>,
exclude_domains: Vec<String>,
max_distance: i32,
) -> Result<Vec<InstanceDetails>, Error> {
let params = Arc::new(CrawlParams::new(
min_lemmy_version().await?,
exclude_domains,
max_distance,
Arc::new(Mutex::new(HashSet::new())),
));
let mut jobs = vec![];
for domain in start_instances.into_iter() {
let job = CrawlJob::new(domain, 0, params.clone());
jobs.push(job.crawl());
}
// TODO: log the errors
let mut instance_details: Vec<InstanceDetails> = join_all(jobs)
.await
.into_iter()
.flatten()
.filter_map(|r| r.ok())
.collect();
// Sort by active monthly users descending
instance_details.sort_unstable_by_key(|i| {
i.site_info
.site_view
.as_ref()
.map(|s| s.counts.users_active_month)
.unwrap_or(0)
});
instance_details.reverse();
Ok(instance_details)
}
/// calculate minimum allowed lemmy version based on current version. in case of current version
/// 0.16.3, the minimum from this function is 0.15.3. this is to avoid rejecting all instances on
/// the previous version when a major lemmy release is published.
async fn min_lemmy_version() -> Result<Version, Error> {
let lemmy_version_url = "https://raw.githubusercontent.com/LemmyNet/lemmy-ansible/main/VERSION";
let req = CLIENT
.get(lemmy_version_url)
.timeout(REQUEST_TIMEOUT)
.send()
.await?;
let mut version = Version::parse(req.text().await?.trim())?;
version.minor -= 1;
Ok(version)
}

View file

@ -1,54 +1,55 @@
use anyhow::Error;
use clap::{Arg, Command};
use lemmy_stats_crawler::crawl::{crawl, InstanceDetails};
use lemmy_stats_crawler::{DEFAULT_MAX_CRAWL_DEPTH, DEFAULT_START_INSTANCES, EXCLUDE_INSTANCES};
use lemmy_stats_crawler::crawl::InstanceDetails;
use lemmy_stats_crawler::start_crawl;
use serde::Serialize;
use structopt::StructOpt;
#[derive(StructOpt, Debug)]
#[structopt()]
struct Parameters {
#[structopt(short, long, default_value = "lemmy.ml")]
start_instances: Vec<String>,
#[structopt(short, long, default_value = "ds9.lemmy.ml, enterprise.lemmy.ml, voyager.lemmy.ml, test.lemmy.ml")]
exclude_instances: Vec<String>,
#[structopt(short, long, default_value = "20")]
max_crawl_distance: i32,
/// Silence all output
#[structopt(short, long)]
quiet: bool,
/// Verbose mode (-v, -vv, -vvv, etc)
#[structopt(short = "v", long = "verbose", parse(from_occurrences))]
verbose: usize,
}
#[tokio::main]
pub async fn main() -> Result<(), Error> {
let matches = Command::new("Lemmy Stats Crawler")
.arg(
Arg::new("start-instances")
.long("start-instances")
.takes_value(true),
)
.arg(Arg::new("exclude").long("exclude").takes_value(true))
.arg(
Arg::new("max-crawl-depth")
.long("max-crawl-depth")
.takes_value(true),
)
.get_matches();
let start_instances: Vec<String> = matches
.value_of("start-instances")
.unwrap_or(DEFAULT_START_INSTANCES)
.split(',')
.map(|s| s.trim().to_string())
.collect();
let exclude: Vec<String> = matches
.value_of("exclude")
.unwrap_or(EXCLUDE_INSTANCES)
.split(',')
.map(|s| s.trim().to_string())
.collect();
let max_crawl_depth: i32 = matches
.value_of("max-crawl-depth")
.unwrap_or(DEFAULT_MAX_CRAWL_DEPTH)
.parse()?;
let params = Parameters::from_args();
stderrlog::new()
.module(module_path!())
.quiet(params.quiet)
.verbosity(params.verbose)
.init()?;
eprintln!("Crawling...");
let (instance_details, failed_instances) =
crawl(start_instances, exclude, max_crawl_depth).await?;
let total_stats = aggregate(instance_details, failed_instances);
let instance_details = start_crawl(
params.start_instances,
params.exclude_instances,
params.max_crawl_distance,
)
.await?;
let total_stats = aggregate(instance_details);
println!("{}", serde_json::to_string_pretty(&total_stats)?);
Ok(())
}
// TODO: lemmy stores these numbers in SiteAggregates, would be good to simply use that as a member
// (to avoid many members). but SiteAggregates also has id, site_id fields
#[derive(Serialize)]
struct TotalStats {
crawled_instances: i32,
failed_instances: i32,
online_users: usize,
total_users: i64,
users_active_day: i64,
@ -58,7 +59,7 @@ struct TotalStats {
instance_details: Vec<InstanceDetails>,
}
fn aggregate(instance_details: Vec<InstanceDetails>, failed_instances: i32) -> TotalStats {
fn aggregate(instance_details: Vec<InstanceDetails>) -> TotalStats {
let mut online_users = 0;
let mut total_users = 0;
let mut users_active_day = 0;
@ -79,7 +80,6 @@ fn aggregate(instance_details: Vec<InstanceDetails>, failed_instances: i32) -> T
}
TotalStats {
crawled_instances,
failed_instances,
online_users,
total_users,
users_active_day,