CP1.2: build.rs keys the data tests on a whole data folder, handles non-UTF-8 paths; CI pins craft-i18n

Review of craft-libs#482:
- craft_proof_data is set only when every Language data file is embedded, and the new
  craft_proof_no_data only when none is, so a partial folder (en-us but no pl) runs neither
  the pattern tests nor the no-data tests.
- A CRAFT_I18N_DIR without language-data/ is watched (rerun-if-changed), so the build
  notices when the checkout gains it.
- A CRAFT_I18N_DIR that isn't valid UTF-8 is reported and embeds nothing, instead of a lossy
  path in include_str!.
- proof.yml checks out craft-i18n at 062a6375c322c88b96691a17f7b5fed71c51f7a8.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01FypgCX8LsYYE8fsj4byBm2

.github/workflows/proof.yml

@@ -25,10 +25,10 @@ jobs:
         with:
           persist-credentials: false
       # The Language data (craft-i18n `language-data/`), embedded by crates/proof/build.rs.
-      # TODO(CP1.2): pin `ref` to the craft-i18n commit that adds language-data/ once it merges.
       - uses: actions/checkout@v4
         with:
           repository: storytold/craft-i18n
+          ref: 062a6375c322c88b96691a17f7b5fed71c51f7a8
           path: craft-i18n
           persist-credentials: false
       - uses: dtolnay/rust-toolchain@b725eda525362e7cf34af9a7243cac1e3f01a7ec # stable

crates/proof/build.rs

@@ -4,12 +4,16 @@
 //! one constant per file in [`FILES`], each an `include_str!`/`include_bytes!` of
 //! `$CRAFT_I18N_DIR/language-data/<path>`. Unset, or not a checkout, or a file missing (a warning):
 //! that constant is empty, and the crate has no embedded Language data for it (English spelling
-//! falls back to an installed Hunspell dictionary or the crate's own core word list, and
+//! falls back to an installed Hunspell dictionary or the crate's own copy of the Moby word list, and
 //! hyphenation that needs the data is off). With `CRAFT_I18N_REQUIRED=1` (releases) any of those
 //! is an error instead.
 //!
-//! When the English word list is embedded, the `craft_proof_data` cfg is set (the tests that pin
-//! results of the data use it).
+//! When every file in [`FILES`] is embedded, the `craft_proof_data` cfg is set (the tests that pin
+//! results of the data use it); when none is, `craft_proof_no_data` (the tests of a build without
+//! the data). A partial data folder sets neither, so neither set of tests runs against it.
+//!
+//! `include_str!` takes a string, so a `CRAFT_I18N_DIR` that isn't valid UTF-8 can't be embedded:
+//! it is treated like a folder without the data.
 
 use std::fmt::Write as _;
 use std::path::{Path, PathBuf};
@@ -32,6 +36,7 @@ fn main() {
     println!("cargo::rerun-if-env-changed=CRAFT_I18N_DIR");
     println!("cargo::rerun-if-env-changed=CRAFT_I18N_REQUIRED");
     println!("cargo::rustc-check-cfg=cfg(craft_proof_data)");
+    println!("cargo::rustc-check-cfg=cfg(craft_proof_no_data)");
     let required = std::env::var("CRAFT_I18N_REQUIRED").is_ok_and(|v| v.trim() == "1");
     let level = if required { "error" } else { "warning" };
     let dir = match std::env::var_os("CRAFT_I18N_DIR").filter(|d| !d.is_empty()) {
@@ -47,35 +52,36 @@ fn main() {
                 println!("cargo::error=CRAFT_I18N_REQUIRED=1 but CRAFT_I18N_DIR is unset: a release must embed the Language data");
             } else {
                 println!(
-                    "cargo::warning=CRAFT_I18N_DIR is unset: craft-proof embeds no Language data (English spelling falls back to system Hunspell or the core word list; no hyphenation)"
+                    "cargo::warning=CRAFT_I18N_DIR is unset: craft-proof embeds no Language data (English spelling falls back to system Hunspell or the built-in Moby word list; no hyphenation)"
                 );
             }
             None
         }
     };
     let mut o = String::from("// Generated by craft-proof's build.rs from CRAFT_I18N_DIR/language-data. Do not edit.\n");
-    let mut english = false;
+    let mut embedded = 0;
     for &(name, rel, bytes) in FILES {
         let file = dir.as_ref().map(|d| d.join(rel)).filter(|f| f.is_file());
         if let (Some(d), None) = (&dir, &file) {
             println!("cargo::{level}=CRAFT_I18N_DIR: {} is missing; craft-proof embeds none", d.join(rel).display());
         }
         let (ty, empty, mac) = if bytes { ("&[u8]", "b\"\"", "include_bytes") } else { ("&str", "\"\"", "include_str") };
-        let value = match &file {
+        // `language_data` only hands back a UTF-8 folder, so this is the path as it is.
+        let value = match file.as_ref().and_then(|f| f.to_str()) {
             Some(f) => {
-                println!("cargo::rerun-if-changed={}", f.display());
-                if name == "EN_US_DIC" {
-                    english = true;
-                }
+                println!("cargo::rerun-if-changed={f}");
+                embedded += 1;
                 // `{:?}` escapes the path (Windows backslashes) as a Rust string literal.
-                format!("{mac}!({:?})", f.display().to_string())
+                format!("{mac}!({f:?})")
             }
             None => empty.to_string(),
         };
         let _ = writeln!(o, "/// `language-data/{rel}` (empty when it isn't embedded).\npub(crate) static {name}: {ty} = {value};");
     }
-    if english {
+    if embedded == FILES.len() {
         println!("cargo::rustc-cfg=craft_proof_data");
+    } else if embedded == 0 {
+        println!("cargo::rustc-cfg=craft_proof_no_data");
     }
     let out = PathBuf::from(std::env::var_os("OUT_DIR").unwrap_or_default()).join("language_data.rs");
     if let Err(e) = std::fs::write(&out, o) {
@@ -91,8 +97,13 @@ fn language_data(dir: &Path) -> Result<PathBuf, String> {
     if !dir.is_absolute() {
         return Err(format!("{} is not an absolute path", dir.display()));
     }
+    if dir.to_str().is_none() {
+        return Err(format!("{} is not valid UTF-8, which include_str! needs", dir.display()));
+    }
     let data = dir.join("language-data");
     if !data.is_dir() {
+        // Notice when the checkout gains its Language data.
+        println!("cargo::rerun-if-changed={}", dir.display());
         return Err(format!("{} has no language-data/ (a craft-i18n checkout older than its Language data?)", dir.display()));
     }
     println!("cargo::rerun-if-changed={}", data.display());

crates/proof/src/lib.rs

@@ -1115,9 +1115,8 @@ fn starts_consonant_sound(w: &str) -> bool {
 #[cfg(test)]
 pub(crate) static DICTIONARY_TESTS: std::sync::Mutex<()> = std::sync::Mutex::new(());
 
-/// The no-Language-data build (`build.rs` sets `craft_proof_data` when the English word list is
-/// embedded).
-#[cfg(all(test, not(craft_proof_data)))]
+/// The no-Language-data build (`build.rs` sets `craft_proof_no_data` when nothing is embedded).
+#[cfg(all(test, craft_proof_no_data))]
 mod no_data;
 
 #[cfg(test)]