checkpoint

Cargo.toml

@@ -11,8 +11,8 @@ metrics-exporter-prometheus = { version = "0.16.2", features=["http-listener"]}
 # minio = "0.1.0"
 minio = {git="https://github.com/minio/minio-rs.git", rev = "c28f576"}
 reqwest = { version = "0.12", features = ["gzip", "default", "rustls-tls"] }
+rusqlite = { version = "0.34.0", features = ["bundled"] }
 serde = { version = "1.0", features = ["derive"] }
-surrealdb = "2.2"
 tokio = { version="1.41.0", features = ["full"] }
 toml = "0.8.20"
 tracing = "0.1"

Crawler.toml

@@ -3,9 +3,8 @@ surreal_url = "localhost:8000"
 surreal_username = "root"
 surreal_password = "root"
 surreal_ns = "test"
-surreal_db = "v1.19.5"
+surreal_db = "v1.19.2"
 
 # Crawler config
-crawl_filter = "en.wikipedia.org" 
+crawl_filter = "en.wikipedia.com" 
-start_url = "https://en.wikipedia.org"
+budget = 1000
-budget = 100

README.md

@@ -4,12 +4,11 @@ Crawls sites saving all the found links to a surrealdb database. It then proceed
 
 ### TODO
 
-- [x] Domain filtering - prevent the crawler from going on alternate versions of wikipedia.
+- [ ] Domain filtering - prevent the crawler from going on alternate versions of wikipedia.
 - [ ] Conditionally save content - based on filename or file contents
 - [x] GUI / TUI ? - Graphana
 - [x] Better asynchronous getting of the sites. Currently it all happens serially.
-- [x] Allow for storing asynchronously - dropping the "links to" logic fixes this need
+- [ ] Allow for storing asynchronously
-- [x] Control crawler via config file (no recompliation needed)
 
 3/17/25: Took >1hr to crawl 100 pages
 

docker/compose.yml

@@ -66,3 +66,4 @@ volumes:
   grafana_storage:
   alloy_storage:
   surrealdb_storage:
+  minio_storage:

src/db.rs

@@ -1,12 +1,7 @@
 use metrics::counter;
+use rusqlite::Connection;
 use std::fmt::Debug;
 use serde::{Deserialize, Serialize};
-use surrealdb::{
-    engine::remote::ws::{Client, Ws},
-    opt::auth::Root,
-    sql::Thing,
-    Surreal,
-};
 use tracing::{error, instrument, trace};
 use url::Url;
 
@@ -49,11 +44,16 @@ impl Website {
 
     // Insert ever item in the vec into surreal, crawled state will be preserved as TRUE
     // if already in the database as such or incoming data is TRUE.
-    #[instrument(skip(db))]
+    pub async fn store_all(all: Vec<Self>, db: &Connection) {
-    pub async fn store_all(all: Vec<Self>, db: &Surreal<Client>) -> Vec<Thing> {
         counter!(STORE).increment(1);
         let mut things = Vec::with_capacity(all.len());
 
+        rusqlite::ParamsFromIter;
+
+        db.execute("",
+            params![]
+        );
+
         match db
             .query(
                 "INSERT INTO website $array
@@ -90,32 +90,10 @@ pub struct Record {
     pub id: Thing,
 }
 
-#[instrument(skip_all, name = "SurrealDB")]
+#[instrument(skip_all, name = "sqlite_connect")]
-pub async fn connect(config: &Config) -> surrealdb::Result<Surreal<Client>> {
+pub async fn connect(config: &Config) -> Result<Connection, rusqlite::Error> {
-    trace!("Establishing connection to surreal...");
+    trace!("Establishing connection to sqlite...");
     // Connect to the server
-    let db = Surreal::new::<Ws>(&config.surreal_url).await?;
+    Connection::open("./squeelite.db")
-
-    trace!("Logging in...");
-    // Signin as a namespace, database, or root user
-    db.signin(Root {
-        username: &config.surreal_username,
-        password: &config.surreal_password,
-    })
-    .await?;
-
-    // Select a specific namespace / database
-    db.use_ns(&config.surreal_ns)
-        .use_db(&config.surreal_db)
-        .await?;
-
-    let setup = include_bytes!("setup.surql");
-    let file = setup.iter().map(|c| *c as char).collect::<String>();
-
-    db.query(file)
-        .await
-        .expect("Failed to setup surreal tables.");
-
-    Ok(db)
 }
 

src/filesystem.rs

@@ -59,13 +59,6 @@ fn valid_file_extension(take: &&OsStr) -> bool {
                 "pdf" => true,
                 "json" => true,
                 "xml" => true,
-
-                // IGNORE
-                // TODO Should this be a list of all domains?
-                "org" => false,
-                "com" => false,
-                "net" => false,
-
                 _ => {
                     warn!("Might be forgetting a file extension: {s}");
                     false

src/main.rs

@@ -33,14 +33,11 @@ struct Config {
     surreal_password: String,
 
     crawl_filter: String,
-    start_url: String,
     budget: usize,
 }
 
 #[tokio::main]
 async fn main() {
-    println!("Logs and metrics are provided to the Grafana dashboard");
-
     let writer = std::fs::OpenOptions::new()
         .append(true)
         .create(true)
@@ -73,7 +70,8 @@ async fn main() {
         .expect("failed to install recorder/exporter");
 
     info!("Starting...");
-
+    // Would probably take these in as parameters from a cli
+    let starting_url = "https://en.wikipedia.org/";
     // When getting uncrawled pages, name must contain this variable. "" will effectively get ignored.
     // let crawl_filter = "en.wikipedia.org/";
     // let budget = 50;
@@ -84,7 +82,6 @@ async fn main() {
     let _ = file.read_to_string(&mut buf);
 
     let config: Config = toml::from_str(&buf).expect("Failed to parse Crawler.toml");
-    let starting_url = &config.start_url;
 
     let db = connect(&config)
         .await
@@ -141,15 +138,6 @@ async fn main() {
     }
     drop(span);
 
-    if let Ok(mut ok) = db.query("count(select id from website where crawled = true)").await {
-        let res = ok.take::<Option<usize>>(0);
-        if let Ok(i) = res {
-            if let Some(n) = i {
-                info!("Total crawled pages now equals {n}");
-            }
-        }
-    }
-
     info!("Done");
 }