parallelize markdown parsing

Signed-off-by: Christine Dodrill <me@christine.website>

4 files changed, 88 insertions, 43 deletions

diff --git a/Cargo.lock b/Cargo.lock
index 0c75a0c..2d9795d 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -574,6 +574,7 @@ checksum = "c70be434c505aee38639abccb918163b63158a4b4bb791b45b7023044bdc3c9c"
 dependencies = [
  "futures-channel",
  "futures-core",
+ "futures-executor",
  "futures-io",
  "futures-sink",
  "futures-task",
@@ -597,12 +598,35 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "db8d3b0917ff63a2a96173133c02818fac4a746b0a57569d3baca9ec0e945e08"
 
 [[package]]
+name = "futures-executor"
+version = "0.3.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9ee9ca2f7eb4475772cf39dd1cd06208dce2670ad38f4d9c7262b3e15f127068"
+dependencies = [
+ "futures-core",
+ "futures-task",
+ "futures-util",
+]
+
+[[package]]
 name = "futures-io"
 version = "0.3.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "e37c1a51b037b80922864b8eed90692c5cd8abd4c71ce49b77146caa47f3253b"
 
 [[package]]
+name = "futures-macro"
+version = "0.3.9"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "0f8719ca0e1f3c5e34f3efe4570ef2c0610ca6da85ae7990d472e9cbfba13664"
+dependencies = [
+ "proc-macro-hack",
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
 name = "futures-sink"
 version = "0.3.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -623,13 +647,17 @@ version = "0.3.9"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "036a2107cdeb57f6d7322f1b6c363dad67cd63ca3b7d1b925bdf75bd5d96cda9"
 dependencies = [
+ "futures-channel",
  "futures-core",
  "futures-io",
+ "futures-macro",
  "futures-sink",
  "futures-task",
  "memchr",
  "pin-project-lite 0.2.3",
  "pin-utils",
+ "proc-macro-hack",
+ "proc-macro-nested",
  "slab",
 ]
 
@@ -1583,6 +1611,12 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "dbf0c48bc1d91375ae5c3cd81e3722dff1abcf81a30960240640d223f59fe0e5"
 
 [[package]]
+name = "proc-macro-nested"
+version = "0.1.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "bc881b2c22681370c6a780e47af9840ef841837bc98118431d4e1868bd0c1086"
+
+[[package]]
 name = "proc-macro2"
 version = "1.0.24"
 source = "registry+https://github.com/rust-lang/crates.io-index"
@@ -2790,6 +2824,7 @@ dependencies = [
  "comrak",
  "envy",
  "eyre",
+ "futures",
  "glob",
  "go_vanity",
  "hyper 0.14.2",
diff --git a/Cargo.toml b/Cargo.toml
index 0219e37..144c0e5 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -13,6 +13,7 @@ color-eyre = "0.5"
 chrono = "0.4"
 comrak = "0.9"
 envy = "0.4"
+futures = "0.3"
 glob = "0.3"
 hyper = "0.14"
 kankyo = "0.3"
diff --git a/src/app/mod.rs b/src/app/mod.rs
index 6c01b2f..c109f3d 100644
--- a/src/app/mod.rs
+++ b/src/app/mod.rs
@@ -70,9 +70,9 @@ pub async fn init(cfg: PathBuf) -> Result<State> {
     let resume = fs::read_to_string(cfg.resume_fname.clone())?;
     let resume: String = markdown::render(&resume)?;
     let mi = mi::Client::new(cfg.mi_token.clone(), crate::APPLICATION_NAME.to_string())?;
-    let blog = crate::post::load("blog", Some(&mi)).await?;
-    let gallery = crate::post::load("gallery", None).await?;
-    let talks = crate::post::load("talks", None).await?;
+    let blog = crate::post::load("blog").await?;
+    let gallery = crate::post::load("gallery").await?;
+    let talks = crate::post::load("talks").await?;
     let mut everything: Vec<Post> = vec![];
 
     {
diff --git a/src/post/mod.rs b/src/post/mod.rs
index eb8ee54..599d9a8 100644
--- a/src/post/mod.rs
+++ b/src/post/mod.rs
@@ -1,7 +1,8 @@
 use chrono::prelude::*;
 use color_eyre::eyre::{eyre, Result, WrapErr};
 use glob::glob;
-use std::{cmp::Ordering, fs};
+use std::{cmp::Ordering, path::PathBuf};
+use tokio::fs;
 
 pub mod frontmatter;
 
@@ -70,45 +71,53 @@ impl Post {
     }
 }
 
-pub async fn load(dir: &str, mi: Option<&mi::Client>) -> Result<Vec<Post>> {
-    let mut result: Vec<Post> = vec![];
-
-    for path in glob(&format!("{}/*.markdown", dir))?.filter_map(Result::ok) {
-        log::debug!("loading {:?}", path);
-        let body =
-            fs::read_to_string(path.clone()).wrap_err_with(|| format!("can't read {:?}", path))?;
-        let (fm, content_offset) = frontmatter::Data::parse(body.clone().as_str())
-            .wrap_err_with(|| format!("can't parse frontmatter of {:?}", path))?;
-        let markup = &body[content_offset..];
-        let date = NaiveDate::parse_from_str(&fm.clone().date, "%Y-%m-%d")
-            .map_err(|why| eyre!("error parsing date in {:?}: {}", path, why))?;
-        let link = format!("{}/{}", dir, path.file_stem().unwrap().to_str().unwrap());
-        let mentions: Vec<mi::WebMention> = match mi {
-            None => vec![],
-            Some(mi) => mi
-                .mentioners(format!("https://christine.website/{}", link))
-                .await
-                .map_err(|why| tracing::error!("error: can't load mentions for {}: {}", link, why))
-                .unwrap_or(vec![]),
-        };
-
-        result.push(Post {
-            front_matter: fm,
-            link: link,
-            body: markup.to_string(),
-            body_html: crate::app::markdown::render(&markup)
-                .wrap_err_with(|| format!("can't parse markdown for {:?}", path))?,
-            date: {
-                DateTime::<Utc>::from_utc(
-                    NaiveDateTime::new(date, NaiveTime::from_hms(0, 0, 0)),
-                    Utc,
-                )
-                .with_timezone(&Utc)
-                .into()
-            },
-            mentions: mentions,
-        })
-    }
+async fn read_post(dir: &str, fname: PathBuf) -> Result<Post> {
+    let body = fs::read_to_string(fname.clone())
+        .await
+        .wrap_err_with(|| format!("can't read {:?}", fname))?;
+    let (front_matter, content_offset) = frontmatter::Data::parse(body.clone().as_str())
+        .wrap_err_with(|| format!("can't parse frontmatter of {:?}", fname))?;
+    let body = &body[content_offset..];
+    let date = NaiveDate::parse_from_str(&front_matter.clone().date, "%Y-%m-%d")
+        .map_err(|why| eyre!("error parsing date in {:?}: {}", fname, why))?;
+    let link = format!("{}/{}", dir, fname.file_stem().unwrap().to_str().unwrap());
+    let body_html = crate::app::markdown::render(&body)
+        .wrap_err_with(|| format!("can't parse markdown for {:?}", fname))?;
+    let body = body.to_string();
+    let date: DateTime<FixedOffset> =
+        DateTime::<Utc>::from_utc(NaiveDateTime::new(date, NaiveTime::from_hms(0, 0, 0)), Utc)
+            .with_timezone(&Utc)
+            .into();
+
+    let mentions: Vec<mi::WebMention> = match std::env::var("MI_TOKEN") {
+        Ok(token) => mi::Client::new(token.to_string(), crate::APPLICATION_NAME.to_string())?
+            .mentioners(format!("https://christine.website/{}", link))
+            .await
+            .map_err(|why| tracing::error!("error: can't load mentions for {}: {}", link, why))
+            .unwrap_or(vec![]),
+        Err(_) => vec![],
+    };
+
+    Ok(Post {
+        front_matter,
+        link,
+        body,
+        body_html,
+        date,
+        mentions,
+    })
+}
+
+pub async fn load(dir: &str) -> Result<Vec<Post>> {
+    let futs = glob(&format!("{}/*.markdown", dir))?
+        .filter_map(Result::ok)
+        .map(|fname| read_post(dir, fname));
+
+    let mut result: Vec<Post> = futures::future::join_all(futs)
+        .await
+        .into_iter()
+        .map(Result::unwrap)
+        .collect();
 
     if result.len() == 0 {
         Err(eyre!("no posts loaded"))