Auto merge of #147918 - yotamofek:pr/stringdex-fork, r=GuillaumeGomez

Upgrade `stringdex` to 0.0.3

Includes a bunch of optimizations for a nice perf win

Author: bors

Cargo.lock

@@ -1289,7 +1289,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "778e2ac28f6c47af28e4907f13ffd1e1ddbd400980a9abd7c8df189bf578a5ad"
 dependencies = [
  "libc",
- "windows-sys 0.52.0",
+ "windows-sys 0.60.2",
 ]
 
 [[package]]
@@ -2155,7 +2155,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "07033963ba89ebaf1584d767badaa2e8fcec21aedea6b8c0346d487d49c28667"
 dependencies = [
  "cfg-if",
- "windows-targets 0.52.6",
+ "windows-targets 0.53.3",
 ]
 
 [[package]]
@@ -4907,7 +4907,7 @@ dependencies = [
  "errno",
  "libc",
  "linux-raw-sys",
- "windows-sys 0.52.0",
+ "windows-sys 0.61.2",
 ]
 
 [[package]]
@@ -5275,9 +5275,9 @@ dependencies = [
 
 [[package]]
 name = "stringdex"
-version = "0.0.2"
+version = "0.0.3"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "18b3bd4f10d15ef859c40291769f0d85209de6b0f1c30713ff9cdf45ac43ea36"
+checksum = "556a6126952cb2f5150057c98a77cc6c771027dea2825bf7fa03d3d638b0a4f8"
 dependencies = [
  "stacker",
 ]

src/librustdoc/Cargo.toml

@@ -21,7 +21,7 @@ rustdoc-json-types = { path = "../rustdoc-json-types" }
 serde = { version = "1.0", features = ["derive"] }
 serde_json = "1.0"
 smallvec = "1.8.1"
-stringdex = "=0.0.2"
+stringdex = "=0.0.3"
 tempfile = "3"
 threadpool = "1.8.1"
 tikv-jemalloc-sys = { version = "0.6.1", optional = true, features = ['override_allocator_on_supported_platforms'] }

src/librustdoc/html/render/search_index.rs

@@ -3,7 +3,9 @@ mod serde;
 
 use std::collections::BTreeSet;
 use std::collections::hash_map::Entry;
+use std::io;
 use std::path::Path;
+use std::string::FromUtf8Error;
 
 use ::serde::de::{self, Deserializer, Error as _};
 use ::serde::ser::{SerializeSeq, Serializer};
@@ -95,21 +97,22 @@ impl SerializedSearchIndex {
         ) -> Result<(), Error> {
             let root_path = doc_root.join(format!("search.index/root{resource_suffix}.js"));
             let column_path = doc_root.join(format!("search.index/{column_name}/"));
+
+            let mut consume = |_, cell: &[u8]| {
+                column.push(String::from_utf8(cell.to_vec())?);
+                Ok::<_, FromUtf8Error>(())
+            };
+
             stringdex_internals::read_data_from_disk_column(
                 root_path,
                 column_name.as_bytes(),
                 column_path.clone(),
-                &mut |_id, item| {
-                    column.push(String::from_utf8(item.to_vec())?);
-                    Ok(())
-                },
-            )
-            .map_err(
-                |error: stringdex_internals::ReadDataError<Box<dyn std::error::Error>>| Error {
-                    file: column_path,
-                    error: format!("failed to read column from disk: {error}"),
-                },
+                &mut consume,
             )
+            .map_err(|error| Error {
+                file: column_path,
+                error: format!("failed to read column from disk: {error}"),
+            })
         }
         fn perform_read_serde(
             resource_suffix: &str,
@@ -119,25 +122,26 @@ impl SerializedSearchIndex {
         ) -> Result<(), Error> {
             let root_path = doc_root.join(format!("search.index/root{resource_suffix}.js"));
             let column_path = doc_root.join(format!("search.index/{column_name}/"));
+
+            let mut consume = |_, cell: &[u8]| {
+                if cell.is_empty() {
+                    column.push(None);
+                } else {
+                    column.push(Some(serde_json::from_slice(cell)?));
+                }
+                Ok::<_, serde_json::Error>(())
+            };
+
             stringdex_internals::read_data_from_disk_column(
                 root_path,
                 column_name.as_bytes(),
                 column_path.clone(),
-                &mut |_id, item| {
-                    if item.is_empty() {
-                        column.push(None);
-                    } else {
-                        column.push(Some(serde_json::from_slice(item)?));
-                    }
-                    Ok(())
-                },
-            )
-            .map_err(
-                |error: stringdex_internals::ReadDataError<Box<dyn std::error::Error>>| Error {
-                    file: column_path,
-                    error: format!("failed to read column from disk: {error}"),
-                },
+                &mut consume,
             )
+            .map_err(|error| Error {
+                file: column_path,
+                error: format!("failed to read column from disk: {error}"),
+            })
         }
         fn perform_read_postings(
             resource_suffix: &str,
@@ -147,23 +151,28 @@ impl SerializedSearchIndex {
         ) -> Result<(), Error> {
             let root_path = doc_root.join(format!("search.index/root{resource_suffix}.js"));
             let column_path = doc_root.join(format!("search.index/{column_name}/"));
+
+            fn consumer(
+                column: &mut Vec<Vec<Vec<u32>>>,
+            ) -> impl FnMut(u32, &[u8]) -> io::Result<()> {
+                |_, cell| {
+                    let mut postings = Vec::new();
+                    encode::read_postings_from_string(&mut postings, cell);
+                    column.push(postings);
+                    Ok(())
+                }
+            }
+
             stringdex_internals::read_data_from_disk_column(
                 root_path,
                 column_name.as_bytes(),
                 column_path.clone(),
-                &mut |_id, buf| {
-                    let mut postings = Vec::new();
-                    encode::read_postings_from_string(&mut postings, buf);
-                    column.push(postings);
-                    Ok(())
-                },
-            )
-            .map_err(
-                |error: stringdex_internals::ReadDataError<Box<dyn std::error::Error>>| Error {
-                    file: column_path,
-                    error: format!("failed to read column from disk: {error}"),
-                },
+                &mut consumer(column),
             )
+            .map_err(|error| Error {
+                file: column_path,
+                error: format!("failed to read column from disk: {error}"),
+            })
         }
 
         assert_eq!(names.len(), path_data.len());
@@ -1055,12 +1064,12 @@ impl Serialize for TypeData {
         let mut buf = Vec::new();
         encode::write_postings_to_string(&self.inverted_function_inputs_index, &mut buf);
         let mut serialized_result = Vec::new();
-        stringdex_internals::encode::write_base64_to_bytes(&buf, &mut serialized_result);
+        stringdex_internals::encode::write_base64_to_bytes(&buf, &mut serialized_result).unwrap();
         seq.serialize_element(&str::from_utf8(&serialized_result).unwrap())?;
         buf.clear();
         serialized_result.clear();
         encode::write_postings_to_string(&self.inverted_function_output_index, &mut buf);
-        stringdex_internals::encode::write_base64_to_bytes(&buf, &mut serialized_result);
+        stringdex_internals::encode::write_base64_to_bytes(&buf, &mut serialized_result).unwrap();
         seq.serialize_element(&str::from_utf8(&serialized_result).unwrap())?;
         if self.search_unbox {
             seq.serialize_element(&1)?;