add scraper cli using clap

2026-01-19 10:10:38 +01:00 · 2026-01-19 10:10:38 +01:00 · e45daf2971
parent 146faa015b
commit e45daf2971
7 changed files with 228 additions and 36 deletions
--- a/Cargo.lock
+++ b/Cargo.lock
@ -269,6 +269,56 @@ dependencies = [
 "libc",
 ]

+[[package]]
+name = "anstream"
+version = "0.6.21"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "43d5b281e737544384e969a5ccad3f1cdd24b48086a0fc1b2a5262a26b8f4f4a"
+dependencies = [
+ "anstyle",
+ "anstyle-parse",
+ "anstyle-query",
+ "anstyle-wincon",
+ "colorchoice",
+ "is_terminal_polyfill",
+ "utf8parse",
+]
+
+[[package]]
+name = "anstyle"
+version = "1.0.13"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "5192cca8006f1fd4f7237516f40fa183bb07f8fbdfedaa0036de5ea9b0b45e78"
+
+[[package]]
+name = "anstyle-parse"
+version = "0.2.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "4e7644824f0aa2c7b9384579234ef10eb7efb6a0deb83f9630a49594dd9c15c2"
+dependencies = [
+ "utf8parse",
+]
+
+[[package]]
+name = "anstyle-query"
+version = "1.1.5"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "40c48f72fd53cd289104fc64099abca73db4166ad86ea0b4341abe65af83dadc"
+dependencies = [
+ "windows-sys 0.61.2",
+]
+
+[[package]]
+name = "anstyle-wincon"
+version = "3.0.11"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "291e6a250ff86cd4a820112fb8898808a366d8f9f58ce16d1f538353ad55747d"
+dependencies = [
+ "anstyle",
+ "once_cell_polyfill",
+ "windows-sys 0.61.2",
+]
+
 [[package]]
 name = "anyhow"
 version = "1.0.100"
@ -475,6 +525,52 @@ dependencies = [
 "windows-link",
 ]

+[[package]]
+name = "clap"
+version = "4.5.54"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c6e6ff9dcd79cff5cd969a17a545d79e84ab086e444102a591e288a8aa3ce394"
+dependencies = [
+ "clap_builder",
+ "clap_derive",
+]
+
+[[package]]
+name = "clap_builder"
+version = "4.5.54"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "fa42cf4d2b7a41bc8f663a7cab4031ebafa1bf3875705bfaf8466dc60ab52c00"
+dependencies = [
+ "anstream",
+ "anstyle",
+ "clap_lex",
+ "strsim",
+]
+
+[[package]]
+name = "clap_derive"
+version = "4.5.49"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2a0b5487afeab2deb2ff4e03a807ad1a03ac532ff5a2cee5d86884440c7f7671"
+dependencies = [
+ "heck",
+ "proc-macro2",
+ "quote",
+ "syn 2.0.113",
+]
+
+[[package]]
+name = "clap_lex"
+version = "0.7.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "c3e64b0cc0439b12df2fa678eae89a1c56a529fd067a9115f7827f1fffd22b32"
+
+[[package]]
+name = "colorchoice"
+version = "1.0.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b05b61dc5112cbb17e4b6cd61790d9845d13888356391624cbe7e41efeac1e75"
+
 [[package]]
 name = "concurrent-queue"
 version = "2.5.0"
@ -1443,6 +1539,12 @@ dependencies = [
 "serde",
 ]

+[[package]]
+name = "is_terminal_polyfill"
+version = "1.70.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a6cb138bb79a146c1bd460005623e142ef0181e3d0219cb493e02f7d08a35695"
+
 [[package]]
 name = "itertools"
 version = "0.14.0"
@ -1644,6 +1746,7 @@ version = "0.2.1"
 dependencies = [
 "anyhow",
 "chrono",
+ "clap",
 "const_format",
 "dotenvy",
 "futures",
@ -1765,6 +1868,12 @@ version = "1.21.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "42f5e15c9953c5e4ccceeb2e7382a716482c34515315f7b03532b8b4e8393d2d"

+[[package]]
+name = "once_cell_polyfill"
+version = "1.70.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "384b8ab6d37215f3c5301a95a4accb5d64aa607f1fcb26a11b5303878451b4fe"
+
 [[package]]
 name = "parking"
 version = "2.2.1"
@ -2885,6 +2994,12 @@ dependencies = [
 "unicode-properties",
 ]

+[[package]]
+name = "strsim"
+version = "0.11.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "7da8b5736845d9f2fcb837ea5d9e2628564b3b043a70948a3f0b778838c5fb4f"
+
 [[package]]
 name = "strum"
 version = "0.27.2"
@ -3355,6 +3470,12 @@ version = "1.0.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "b6c140620e7ffbb22c2dee59cafe6084a59b5ffc27a8859a5f0d494b5d52b6be"

+[[package]]
+name = "utf8parse"
+version = "0.2.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "06abde3611657adf66d383f00b093d7faecc7fa57071cce2578660c9f1010821"
+
 [[package]]
 name = "utoipa"
 version = "5.4.0"
--- a/4
+++ b/4
@ -32,7 +32,8 @@ RUN cargo chef cook --release --recipe-path recipe.json
 COPY . .
 RUN OFFLINE=true cargo build --release \
    --bin mensa-upb-api \
-    --bin mensa-upb-scraper
+    --bin mensa-upb-scraper \
+    --bin scraper-cli

 # =====================================================
 # Runtime image: scraper (cron-based)
@ -47,6 +48,7 @@ RUN echo "0 0/8 * * * /app/mensa-upb-scraper >> /var/log/cron.log 2>&1" \
    touch /var/log/cron.log

 COPY --from=builder /app/target/release/mensa-upb-scraper /app/mensa-upb-scraper
+COPY --from=builder /app/target/release/scraper-cli /app/scraper-cli

 ENTRYPOINT ["/sbin/tini", "--"]
 CMD sh -c 'env > /etc/environment && crond -l 2 && tail -f /var/log/cron.log'
--- a/scraper/Cargo.toml
+++ b/scraper/Cargo.toml
@ -9,9 +9,14 @@ version = "0.2.1"
 edition = "2024"
 publish = false

+[[bin]]
+name = "scraper-cli"
+path = "src/bin/cli.rs"
+
 [dependencies]
 anyhow = { workspace = true }
 chrono = { workspace = true }
+clap = { version = "4.5.54", features = ["derive", "env"] }
 const_format = "0.2.33"
 dotenvy = { workspace = true }
 futures = { workspace = true }
--- a/scraper/src/bin/cli.rs
+++ b/scraper/src/bin/cli.rs
@ -0,0 +1,60 @@
+use anyhow::Result;
+use clap::Parser;
+use futures::future;
+use mensa_upb_scraper::check_refresh;
+use sqlx::postgres::PgPoolOptions;
+use strum::IntoEnumIterator as _;
+use tracing::level_filters::LevelFilter;
+use tracing_subscriber::EnvFilter;
+
+#[derive(Debug, Clone, clap::Parser)]
+struct Cli {
+    /// Database connection string
+    #[clap(env = "DATABASE_URL")]
+    database: String,
+    /// Canteen to scrape
+    #[clap(short, long = "canteen")]
+    canteens: Vec<shared::Canteen>,
+    /// Date to scrape (YYYY-MM-DD)
+    #[clap(short, long = "date", required = true)]
+    dates: Vec<chrono::NaiveDate>,
+    /// Force refresh even if not needed
+    #[clap(short, long)]
+    force: bool,
+}
+
+#[tokio::main]
+async fn main() -> Result<()> {
+    dotenvy::dotenv().ok();
+
+    let mut cli = Cli::parse();
+
+    if cli.canteens.is_empty() {
+        cli.canteens = shared::Canteen::iter().collect();
+    }
+
+    let db = PgPoolOptions::new().connect_lazy(&cli.database)?;
+
+    let env_filter = EnvFilter::builder()
+        .with_default_directive(LevelFilter::WARN.into())
+        .from_env()
+        .expect("Invalid filter")
+        .add_directive("mensa_upb_scraper=debug".parse().unwrap());
+    tracing_subscriber::fmt().with_env_filter(env_filter).init();
+
+    sqlx::migrate!("../migrations").run(&db).await?;
+
+    tracing::info!("Starting up...");
+
+    let handles = cli.dates.into_iter().map(|date| {
+        let db = db.clone();
+        let canteens = cli.canteens.clone();
+        tokio::spawn(async move { check_refresh(&db, date, &canteens, cli.force).await })
+    });
+
+    future::join_all(handles).await;
+
+    tracing::info!("Finished scraping menu");
+
+    Ok(())
+}
--- a/scraper/src/main.rs
+++ b/scraper/src/main.rs
@ -2,8 +2,8 @@ use std::sync::LazyLock;

 use anyhow::Result;
 use chrono::{Duration, Utc};
-use futures::{future, StreamExt};
-use mensa_upb_scraper::{check_refresh, util, FILTER_CANTEENS};
+use futures::future;
+use mensa_upb_scraper::{FILTER_CANTEENS, check_refresh, util};
 use shared::Canteen;
 use strum::IntoEnumIterator as _;
 use tracing::level_filters::LevelFilter;
@ -36,17 +36,11 @@ async fn main() -> Result<()> {
        .map(|d| (Utc::now() + Duration::days(d)).date_naive())
        .map(|date| {
            let db = db.clone();
-            tokio::spawn(async move { check_refresh(&db, date, &CANTEENS).await })
+            tokio::spawn(async move { check_refresh(&db, date, &CANTEENS, false).await })
        });

    future::join_all(handles).await;

-    futures::stream::iter((0..7).map(|d| (Utc::now() + Duration::days(d)).date_naive()))
-        .for_each_concurrent(None, async |date| {
-            check_refresh(&db, date, &CANTEENS).await;
-        })
-        .await;
-
    tracing::info!("Finished scraping menu");

    Ok(())
--- a/scraper/src/refresh.rs
+++ b/scraper/src/refresh.rs
@ -27,18 +27,26 @@ static NON_FILTERED_CANTEENS: LazyLock<Vec<Canteen>> = LazyLock::new(|| {
 });

 #[tracing::instrument(skip(db))]
-pub async fn check_refresh(db: &sqlx::PgPool, date: NaiveDate, canteens: &[Canteen]) -> bool {
-    if date > Utc::now().date_naive() + chrono::Duration::days(31) {
+pub async fn check_refresh(
+    db: &sqlx::PgPool,
+    date: NaiveDate,
+    canteens: &[Canteen],
+    force: bool,
+) -> bool {
+    if !force && date > Utc::now().date_naive() + chrono::Duration::days(31) {
        tracing::debug!("Not refreshing menu for date {date} as it is too far in the future");
        return false;
    }

-    if date < Utc::now().date_naive() {
+    if !force && date < Utc::now().date_naive() {
        tracing::trace!("Not refreshing menu for date {date} as it is in the past");
        return false;
    }

-    let canteens_needing_refresh = match sqlx::query!(
+    let canteens_needing_refresh = if force {
+        canteens.iter().cloned().collect::<BTreeSet<_>>()
+    } else {
+        match sqlx::query!(
            r#"SELECT canteen, max(scraped_at) AS "scraped_at!" FROM canteens_scraped WHERE canteen = ANY($1) AND scraped_for = $2 GROUP BY canteen"#,
            &canteens
                .iter()
@ -54,13 +62,15 @@ pub async fn check_refresh(db: &sqlx::PgPool, date: NaiveDate, canteens: &[Cante
                .map(|r| (Canteen::from_str(&r.canteen).expect("malformed db entry"), Some(r.scraped_at)))
                .chain(NON_FILTERED_CANTEENS.iter().filter(|c| canteens.contains(c)).map(|c| (*c, None)))
                .unique_by(|(c, _)| *c)
-            .filter(|(_, scraped_at)| scraped_at.is_none_or(|scraped_at| needs_refresh(scraped_at, date)))
+                .filter(|(c, scraped_at)| 
+                    canteens.contains(c) && scraped_at.is_none_or(|scraped_at| needs_refresh(scraped_at, date)))
                .map(|(c, _)| c)
                .collect::<BTreeSet<_>>(),
            Err(err) => {
                tracing::error!("Error checking for existing scrapes: {}", err);
                return false;
            }
+        }
    };

    if canteens_needing_refresh.is_empty() {
--- a/web-api/src/menu.rs
+++ b/web-api/src/menu.rs
@ -28,7 +28,7 @@ impl Menu {
            .collect::<Vec<_>>();

        if allow_refresh {
-            check_refresh(db, date, canteens).await;
+            check_refresh(db, date, canteens, false).await;
        };

        let result = sqlx::query!(r#"SELECT name, array_agg(DISTINCT canteen ORDER BY canteen) AS "canteens!", dish_type AS "dish_type: DishType", image_src, price_students, price_employees, price_guests, vegan, vegetarian