veloren/voxygen/src/i18n.rs

599 lines
22 KiB
Rust

use common::{
assets,
assets::{load_expect, load_glob, Asset},
};
use deunicode::deunicode;
use ron::de::from_reader;
use serde_derive::*;
use std::{
collections::{HashMap, HashSet},
fs::File,
io::BufReader,
};
use tracing::warn;
/// The reference language, aka the more up-to-date localization data.
/// Also the default language at first startup.
pub const REFERENCE_LANG: &str = "en";
/// How a language can be described
#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize)]
pub struct LanguageMetadata {
/// A human friendly language name (e.g. "English (US)")
pub language_name: String,
/// A short text identifier for this language (e.g. "en_US")
///
/// On the opposite of `language_name` that can change freely,
/// `language_identifier` value shall be stable in time as it
/// is used by setting components to store the language
/// selected by the user.
pub language_identifier: String,
}
/// Store font metadata
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct Font {
/// Key to retrieve the font in the asset system
pub asset_key: String,
/// Scale ratio to resize the UI text dynamicly
pub scale_ratio: f32,
}
impl Font {
/// Scale input size to final UI size
pub fn scale(&self, value: u32) -> u32 { (value as f32 * self.scale_ratio).round() as u32 }
}
/// Store font metadata
pub type VoxygenFonts = HashMap<String, Font>;
/// Store internationalization data
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct VoxygenLocalization {
/// A map storing the localized texts
///
/// Localized content can be accessed using a String key.
pub string_map: HashMap<String, String>,
/// A map for storing variations of localized texts, for example multiple
/// ways of saying "Help, I'm under attack". Used primarily for npc
/// dialogue.
pub vector_map: HashMap<String, Vec<String>>,
/// Whether to convert the input text encoded in UTF-8
/// into a ASCII version by using the `deunicode` crate.
pub convert_utf8_to_ascii: bool,
/// Font configuration is stored here
pub fonts: VoxygenFonts,
pub metadata: LanguageMetadata,
}
impl VoxygenLocalization {
/// Get a localized text from the given key
///
/// If the key is not present in the localization object
/// then the key is returned.
pub fn get<'a>(&'a self, key: &'a str) -> &str {
match self.string_map.get(key) {
Some(localized_text) => localized_text,
None => key,
}
}
/// Get a variation of localized text from the given key
///
/// `index` should be a random number from `0` to `u16::max()`
///
/// If the key is not present in the localization object
/// then the key is returned.
pub fn get_variation<'a>(&'a self, key: &'a str, index: u16) -> &str {
match self.vector_map.get(key) {
Some(v) if !v.is_empty() => &v[index as usize % v.len()],
_ => key,
}
}
/// Return the missing keys compared to the reference language
pub fn list_missing_entries(&self) -> (HashSet<String>, HashSet<String>) {
let reference_localization =
load_expect::<VoxygenLocalization>(i18n_asset_key(REFERENCE_LANG).as_ref());
let reference_string_keys: HashSet<_> =
reference_localization.string_map.keys().cloned().collect();
let string_keys: HashSet<_> = self.string_map.keys().cloned().collect();
let strings = reference_string_keys
.difference(&string_keys)
.cloned()
.collect();
let reference_vector_keys: HashSet<_> =
reference_localization.vector_map.keys().cloned().collect();
let vector_keys: HashSet<_> = self.vector_map.keys().cloned().collect();
let vectors = reference_vector_keys
.difference(&vector_keys)
.cloned()
.collect();
(strings, vectors)
}
/// Log missing entries (compared to the reference language) as warnings
pub fn log_missing_entries(&self) {
let (missing_strings, missing_vectors) = self.list_missing_entries();
for missing_key in missing_strings {
warn!(
"[{:?}] Missing string key {:?}",
self.metadata.language_identifier, missing_key
);
}
for missing_key in missing_vectors {
warn!(
"[{:?}] Missing vector key {:?}",
self.metadata.language_identifier, missing_key
);
}
}
}
impl Asset for VoxygenLocalization {
const ENDINGS: &'static [&'static str] = &["ron"];
/// Load the translations located in the input buffer and convert them
/// into a `VoxygenLocalization` object.
#[allow(clippy::into_iter_on_ref)] // TODO: Pending review in #587
fn parse(buf_reader: BufReader<File>) -> Result<Self, assets::Error> {
let mut asked_localization: VoxygenLocalization =
from_reader(buf_reader).map_err(assets::Error::parse_error)?;
// Update the text if UTF-8 to ASCII conversion is enabled
if asked_localization.convert_utf8_to_ascii {
for value in asked_localization.string_map.values_mut() {
*value = deunicode(value);
}
for value in asked_localization.vector_map.values_mut() {
*value = value.into_iter().map(|s| deunicode(s)).collect();
}
}
asked_localization.metadata.language_name =
deunicode(&asked_localization.metadata.language_name);
Ok(asked_localization)
}
}
/// Load all the available languages located in the Voxygen asset directory
pub fn list_localizations() -> Vec<LanguageMetadata> {
let voxygen_locales_assets = "voxygen.i18n.*";
let lang_list = load_glob::<VoxygenLocalization>(voxygen_locales_assets).unwrap();
lang_list.iter().map(|e| (*e).metadata.clone()).collect()
}
/// Return the asset associated with the language_id
pub fn i18n_asset_key(language_id: &str) -> String { "voxygen.i18n.".to_string() + language_id }
#[cfg(test)]
mod tests {
use super::VoxygenLocalization;
use git2::Repository;
use ron::de::{from_bytes, from_reader};
use std::{
collections::{HashMap, HashSet},
fs,
path::{Path, PathBuf},
};
/// List localization files as a PathBuf vector
fn i18n_files(i18n_dir: &Path) -> Vec<PathBuf> {
fs::read_dir(i18n_dir)
.unwrap()
.map(|res| res.map(|e| e.path()).unwrap())
.filter(|e| match e.extension() {
Some(ext) => ext == "ron",
None => false,
})
.collect()
}
#[derive(Debug, PartialEq)]
enum LocalizationState {
UpToDate,
NotFound,
Outdated,
Unknown,
Unused,
}
#[derive(Debug)]
struct LocalizationEntryState {
pub key_line: Option<usize>,
pub chuck_line_range: Option<(usize, usize)>,
pub commit_id: Option<git2::Oid>,
pub state: LocalizationState,
}
impl LocalizationEntryState {
pub fn new() -> LocalizationEntryState {
LocalizationEntryState {
key_line: None,
chuck_line_range: None,
commit_id: None,
state: LocalizationState::Unknown,
}
}
}
/// Returns the Git blob associated with the given reference and path
#[allow(clippy::expect_fun_call)] // TODO: Pending review in #587
fn read_file_from_path<'a>(
repo: &'a git2::Repository,
reference: &git2::Reference,
path: &std::path::Path,
) -> git2::Blob<'a> {
let tree = reference
.peel_to_tree()
.expect("Impossible to peel HEAD to a tree object");
tree.get_path(path)
.expect(&format!(
"Impossible to find the file {:?} in reference {:?}",
path,
reference.name()
))
.to_object(&repo)
.unwrap()
.peel_to_blob()
.expect("Impossible to fetch the Git object")
}
fn generate_key_version<'a>(
repo: &'a git2::Repository,
localization: &VoxygenLocalization,
path: &std::path::Path,
file_blob: &git2::Blob,
) -> HashMap<String, LocalizationEntryState> {
let mut keys: HashMap<String, LocalizationEntryState> = localization
.string_map
.keys()
.map(|k| (k.to_owned(), LocalizationEntryState::new()))
.collect();
let mut to_process: HashSet<&String> = localization.string_map.keys().collect();
// Find key start lines
let file_content = std::str::from_utf8(file_blob.content()).expect("Got non UTF-8 file");
for (line_nb, line) in file_content.lines().enumerate() {
let mut found_key = None;
for key in to_process.iter() {
if line.contains(key.as_str()) {
found_key = Some(key.to_owned());
break;
}
}
if let Some(key) = found_key {
keys.get_mut(key).unwrap().key_line = Some(line_nb);
to_process.remove(&key);
};
}
let mut error_check_set: Vec<String> = vec![];
// Find commit for each keys
repo.blame_file(path, None)
.expect("Impossible to generate the Git blame")
.iter()
.for_each(|e: git2::BlameHunk| {
for (key, state) in keys.iter_mut() {
let line = match state.key_line {
Some(l) => l,
None => {
if !error_check_set.contains(key) {
eprintln!(
"Key {} does not have a git line in it's state! Skipping key.",
key
);
error_check_set.push(key.clone());
}
continue;
},
};
if line >= e.final_start_line()
&& line < e.final_start_line() + e.lines_in_hunk()
{
state.chuck_line_range = Some((
e.final_start_line(),
e.final_start_line() + e.lines_in_hunk(),
));
state.commit_id = match state.commit_id {
Some(existing_commit) => {
match repo.graph_descendant_of(e.final_commit_id(), existing_commit)
{
Ok(true) => Some(e.final_commit_id()),
Ok(false) => Some(existing_commit),
Err(err) => panic!(err),
}
},
None => Some(e.final_commit_id()),
};
}
}
});
keys
}
// Test to verify all languages that they are VALID and loadable, without
// need of git just on the local assets folder
#[test]
fn verify_all_localizations() {
// Generate paths
let i18n_asset_path = Path::new("assets/voxygen/i18n/");
let en_i18n_path = i18n_asset_path.join("en.ron");
let root_dir = std::env::current_dir()
.map(|p| p.parent().expect("").to_owned())
.unwrap();
assert!(
root_dir.join(&en_i18n_path).is_file(),
"en reference files doesn't exist, something is wrong!"
);
let i18n_files = i18n_files(&root_dir.join(i18n_asset_path));
// This simple check ONLY guarantees that an arbitrary minimum of translation
// files exists. It's just to notice unintentional deletion of all
// files, or modifying the paths. In case you want to delete all
// language you have to adjust this number:
assert!(
i18n_files.len() > 5,
"have less than 5 translation files, arbitrary minimum check failed. Maybe the i18n \
folder is empty?"
);
for path in i18n_files {
let f = fs::File::open(&path).expect("Failed opening file");
let _: VoxygenLocalization = match from_reader(f) {
Ok(v) => v,
Err(e) => {
panic!(
"Could not parse {} RON file, error: {}",
path.to_string_lossy(),
e
);
},
};
}
}
// Test to verify all languages and print missing and faulty localisation
#[test]
#[ignore]
#[allow(clippy::expect_fun_call)]
fn test_all_localizations() {
// Generate paths
let i18n_asset_path = Path::new("assets/voxygen/i18n/");
let en_i18n_path = i18n_asset_path.join("en.ron");
let root_dir = std::env::current_dir()
.map(|p| p.parent().expect("").to_owned())
.unwrap();
let i18n_path = root_dir.join(i18n_asset_path);
if !root_dir.join(&en_i18n_path).is_file() {
panic!("Reference language file not found {:?}", &en_i18n_path)
}
// Initialize Git objects
let repo = Repository::discover(&root_dir).expect(&format!(
"Failed to open the Git repository at {:?}",
&root_dir
));
let head_ref = repo.head().expect("Impossible to get the HEAD reference");
// Read HEAD for the reference language file
let i18n_en_blob = read_file_from_path(&repo, &head_ref, &en_i18n_path);
let loc: VoxygenLocalization = from_bytes(i18n_en_blob.content())
.expect("Expect to parse reference i18n RON file, can't proceed without it");
let i18n_references: HashMap<String, LocalizationEntryState> =
generate_key_version(&repo, &loc, &en_i18n_path, &i18n_en_blob);
// Compare to other reference files
let i18n_files = i18n_files(&i18n_path);
let mut i18n_entry_counts: HashMap<PathBuf, (usize, usize, usize, usize)> = HashMap::new();
for file in &i18n_files {
let relfile = file.strip_prefix(&root_dir).unwrap();
if relfile == en_i18n_path {
continue;
}
println!("\n-----------------------------------");
println!("{:?}", relfile);
println!("-----------------------------------");
// Find the localization entry state
let current_blob = read_file_from_path(&repo, &head_ref, &relfile);
let current_loc: VoxygenLocalization = match from_bytes(current_blob.content()) {
Ok(v) => v,
Err(e) => {
eprintln!(
"Could not parse {} RON file, skipping: {}",
relfile.to_string_lossy(),
e
);
continue;
},
};
let mut current_i18n =
generate_key_version(&repo, &current_loc, &relfile, &current_blob);
for (ref_key, ref_state) in i18n_references.iter() {
match current_i18n.get_mut(ref_key) {
Some(state) => {
let commit_id = match state.commit_id {
Some(c) => c,
None => {
eprintln!(
"Commit ID of key {} in i18n file {} is missing! Skipping key.",
ref_key,
relfile.to_string_lossy()
);
continue;
},
};
let ref_commit_id = match ref_state.commit_id {
Some(c) => c,
None => {
eprintln!(
"Commit ID of key {} in reference i18n file is missing! \
Skipping key.",
ref_key
);
continue;
},
};
if commit_id != ref_commit_id
&& !repo
.graph_descendant_of(commit_id, ref_commit_id)
.unwrap_or(false)
{
state.state = LocalizationState::Outdated;
} else {
state.state = LocalizationState::UpToDate;
}
},
None => {
current_i18n.insert(ref_key.to_owned(), LocalizationEntryState {
key_line: None,
chuck_line_range: None,
commit_id: None,
state: LocalizationState::NotFound,
});
},
}
}
let ref_keys: HashSet<&String> = i18n_references.keys().collect();
for (_, state) in current_i18n
.iter_mut()
.filter(|&(k, _)| !ref_keys.contains(k))
{
state.state = LocalizationState::Unused;
}
// Display
println!(
"\n{:10} | {:60}| {:40} | {:40}\n",
"State",
"Key name",
relfile.to_str().unwrap(),
en_i18n_path.to_str().unwrap()
);
let mut sorted_keys: Vec<&String> = current_i18n.keys().collect();
sorted_keys.sort();
let current_i18n_entry_count = current_i18n.len();
let mut uptodate_entries = 0;
let mut outdated_entries = 0;
let mut unused_entries = 0;
let mut notfound_entries = 0;
let mut unknown_entries = 0;
for key in sorted_keys {
let state = current_i18n.get(key).unwrap();
if state.state != LocalizationState::UpToDate {
match state.state {
LocalizationState::Outdated => outdated_entries += 1,
LocalizationState::NotFound => notfound_entries += 1,
LocalizationState::Unknown => unknown_entries += 1,
LocalizationState::Unused => unused_entries += 1,
LocalizationState::UpToDate => unreachable!(),
};
println!(
"[{:9}] | {:60}| {:40} | {:40}",
format!("{:?}", state.state),
key,
state
.commit_id
.map(|s| format!("{}", s))
.unwrap_or_else(|| "None".to_string()),
i18n_references
.get(key)
.map(|s| s.commit_id)
.flatten()
.map(|s| format!("{}", s))
.unwrap_or_else(|| "None".to_string()),
);
} else {
uptodate_entries += 1;
}
}
println!(
"\n{} up-to-date, {} outdated, {} unused, {} not found, {} unknown entries",
uptodate_entries,
outdated_entries,
unused_entries,
notfound_entries,
unknown_entries
);
// Calculate key count that actually matter for the status of the translation
// Unused entries don't break the game
let real_entry_count = current_i18n_entry_count - unused_entries;
let uptodate_percent = (uptodate_entries as f32 / real_entry_count as f32) * 100_f32;
let outdated_percent = (outdated_entries as f32 / real_entry_count as f32) * 100_f32;
let untranslated_percent =
((notfound_entries + unknown_entries) as f32 / real_entry_count as f32) * 100_f32;
println!(
"{:.2}% up-to-date, {:.2}% outdated, {:.2}% untranslated\n",
uptodate_percent, outdated_percent, untranslated_percent,
);
i18n_entry_counts.insert(
file.clone(),
(
uptodate_entries,
outdated_entries,
notfound_entries + unknown_entries,
real_entry_count,
),
);
}
let mut overall_uptodate_entry_count = 0;
let mut overall_outdated_entry_count = 0;
let mut overall_untranslated_entry_count = 0;
let mut overall_real_entry_count = 0;
println!("-----------------------------------------------------------------------------");
println!("Overall Translation Status");
println!("-----------------------------------------------------------------------------");
println!(
"{:12}| {:8} | {:8} | {:8}",
"", "up-to-date", "outdated", "untranslated"
);
for (path, (uptodate, outdated, untranslated, real)) in i18n_entry_counts {
overall_uptodate_entry_count += uptodate;
overall_outdated_entry_count += outdated;
overall_untranslated_entry_count += untranslated;
overall_real_entry_count += real;
println!(
"{:12}|{:8} |{:6} |{:8}",
path.file_name().unwrap().to_string_lossy(),
uptodate,
outdated,
untranslated
);
}
println!(
"\n{:.2}% up-to-date, {:.2}% outdated, {:.2}% untranslated",
(overall_uptodate_entry_count as f32 / overall_real_entry_count as f32) * 100_f32,
(overall_outdated_entry_count as f32 / overall_real_entry_count as f32) * 100_f32,
(overall_untranslated_entry_count as f32 / overall_real_entry_count as f32) * 100_f32,
);
println!("-----------------------------------------------------------------------------\n");
}
}