SerializedSyntaxSet: Load cache data early; only parse lazily

Enselic · Enselic · commit 4197558a351e · 2021-07-25T23:08:55.000+02:00
The slow part is parsing, so we can load the data from the file during startup
without a big performance hit. The benefit of loading the data during startup is
that we can better detect when the cache is incomplete, and fallback to data
from the binary.

In particular, this change makes the following scenario fallback to syntaxes.bin
from the binary during startup (same behavior as current bat), rather than
panicing when we try to lazily load syntaxes.bin that does not exist:

    bat cache --build --source assets
    rm ~/.cache/bat/syntaxes.bin
    bat any-file &gt; /tmp/out.txt
diff --git a/src/assets.rs b/src/assets.rs
@@ -2,7 +2,7 @@ use std::collections::BTreeMap;
 use std::ffi::OsStr;
 use std::fs::{self, File};
 use std::io::BufReader;
-use std::path::{Path, PathBuf};
+use std::path::Path;
 
 use lazycell::LazyCell;
 
@@ -122,9 +122,9 @@ impl HighlightingAssets {
     pub fn from_cache(cache_path: &Path) -> Result<Self> {
         Ok(HighlightingAssets::new(
             None,
-            Some(SerializedSyntaxSet::FromFile(
+            Some(SerializedSyntaxSet::FromFile(std::fs::read(
                 cache_path.join("syntaxes.bin"),
-            )),
+            )?)),
             asset_from_cache(&cache_path.join("themes.bin"), "theme set")?,
         ))
     }
@@ -308,7 +308,7 @@ impl HighlightingAssets {
 #[derive(Debug)]
 enum SerializedSyntaxSet {
     /// The data comes from a user-generated cache file.
-    FromFile(PathBuf),
+    FromFile(Vec<u8>),
 
     /// The data to use is embedded into the bat binary.
     FromBinary(&'static [u8]),
@@ -320,8 +320,8 @@ impl SerializedSyntaxSet {
             SerializedSyntaxSet::FromBinary(data) => {
                 from_binary(data)
             },
-            SerializedSyntaxSet::FromFile(ref path) => {
-                asset_from_cache(&path, "syntax set").expect("cache corrupt, consider rebuilding or clearing, see https://github.com/sharkdp/bat#adding-new-syntaxes--language-definitions on how")
+            SerializedSyntaxSet::FromFile(ref data) => {
+                from_reader(&data[..]).expect("cache corrupt, consider rebuilding or clearing, see https://github.com/sharkdp/bat#adding-new-syntaxes--language-definitions on how")
             },
         }
     }
