From 2c1c5eaaa7b698ea1d7ba6331b5c5da6c685b0c6 Mon Sep 17 00:00:00 2001
From: Robert Bastian <robertbastian@users.noreply.github.com>
Date: Wed, 27 Apr 2022 13:29:19 +0200
Subject: [PATCH] key convenience fns

---
 provider/datagen/src/bin/datagen.rs | 38 ++++++---------------
 provider/datagen/src/lib.rs         | 53 +++++++++++++++++++++++++++++
 2 files changed, 64 insertions(+), 27 deletions(-)

diff --git a/provider/datagen/src/bin/datagen.rs b/provider/datagen/src/bin/datagen.rs
index 1db9e56fc59..de1c6757d9a 100644
--- a/provider/datagen/src/bin/datagen.rs
+++ b/provider/datagen/src/bin/datagen.rs
@@ -10,11 +10,7 @@ use icu_locid::LanguageIdentifier;
 use icu_provider::hello_world::HelloWorldV1Marker;
 use icu_provider::prelude::*;
 use simple_logger::SimpleLogger;
-use std::borrow::Cow;
-use std::collections::HashSet;
 use std::fs::File;
-use std::io;
-use std::io::BufRead;
 use std::path::PathBuf;
 use std::str::FromStr;
 
@@ -254,32 +250,20 @@ fn main() -> eyre::Result<()> {
         all_keys
     } else if matches.is_present("HELLO_WORLD") {
         vec![HelloWorldV1Marker::KEY]
+    } else if let Some(paths) = matches.values_of("KEYS") {
+        icu_datagen::keys(&paths.collect::<Vec<_>>())
+    } else if let Some(key_file_path) = matches.value_of_os("KEY_FILE") {
+        File::open(key_file_path)
+            .and_then(icu_datagen::keys_from_file)
+            .with_context(|| key_file_path.to_string_lossy().into_owned())?
     } else {
-        let mut keys = HashSet::new();
-
-        if let Some(paths) = matches.values_of("KEYS") {
-            keys.extend(paths.map(Cow::Borrowed));
-        } else if let Some(key_file_path) = matches.value_of_os("KEY_FILE") {
-            let file = File::open(key_file_path)
-                .with_context(|| key_file_path.to_string_lossy().into_owned())?;
-            for line in io::BufReader::new(file).lines() {
-                let path = line.with_context(|| key_file_path.to_string_lossy().into_owned())?;
-                keys.insert(Cow::Owned(path));
-            }
-        }
-
-        let filtered: Vec<_> = all_keys
-            .into_iter()
-            .filter(|k| keys.contains(k.get_path()))
-            .collect();
-
-        if filtered.is_empty() {
-            eyre::bail!("No keys selected");
-        }
-
-        filtered
+        unreachable!();
     };
 
+    if selected_keys.is_empty() {
+        eyre::bail!("No keys selected");
+    }
+
     let mut source_data = SourceData::default();
 
     if matches.is_present("INPUT_FROM_TESTDATA") {
diff --git a/provider/datagen/src/lib.rs b/provider/datagen/src/lib.rs
index 563091dc46a..cfdea2462e7 100644
--- a/provider/datagen/src/lib.rs
+++ b/provider/datagen/src/lib.rs
@@ -61,8 +61,61 @@ use icu_provider::serde::SerializeMarker;
 use icu_provider_adapters::filter::Filterable;
 use icu_provider_fs::export::serializers;
 use rayon::prelude::*;
+use std::collections::HashSet;
+use std::io::{Read, BufRead, BufReader};
 use std::path::PathBuf;
 
+/// Parses a list of human-readable key identifiers and returns a
+/// list of [`ResourceKey`]s. Invalid key names are ignored.
+///
+/// # Example
+/// ```
+/// # use icu_provider::ResourceMarker;
+/// assert_eq!(
+///     icu_datagen::keys(&["list/and@1", "list/or@1"]), 
+///     vec![
+///         icu_list::provider::AndListV1Marker::KEY,
+///         icu_list::provider::OrListV1Marker::KEY,
+///     ],
+/// );
+/// ```
+pub fn keys<S: AsRef<str>>(strings: &[S]) -> Vec<ResourceKey> {
+    let keys = strings.iter().map(AsRef::as_ref).collect::<HashSet<&str>>();
+    get_all_keys()
+        .into_iter()
+        .filter(|k| keys.contains(k.get_path()))
+        .collect()
+}
+
+/// Parses a file of human-readable key identifiers and returns a
+/// list of [`ResourceKey`]s. Invalid key names are ignored.
+///
+/// # Example
+/// 
+/// #### keys.txt
+/// ```text
+/// list/and@1
+/// list/or@1
+/// ```
+/// #### build.rs
+/// ```no_run
+/// # use icu_provider::ResourceMarker;
+/// # use std::fs::File;
+/// # fn main() -> std::io::Result<()> {
+/// assert_eq!(
+///     icu_datagen::keys_from_file(File::open("keys.txt")?)?, 
+///     vec![
+///         icu_list::provider::AndListV1Marker::KEY,
+///         icu_list::provider::OrListV1Marker::KEY,
+///     ],
+/// );
+/// # Ok(())
+/// # }
+/// ```
+pub fn keys_from_file<R: Read>(file: R) -> std::io::Result<Vec<ResourceKey>> {
+    Ok(keys(&BufReader::new(file).lines().collect::<std::io::Result<Vec<String>>>()?))
+}
+
 /// The output format.
 #[non_exhaustive]
 pub enum Out {