Initial commit.

This commit is contained in:
2020-08-03 14:21:50 +02:00
commit 94de59d768
8 changed files with 45779 additions and 0 deletions

2
.gitignore vendored Normal file
View File

@@ -0,0 +1,2 @@
/target
Cargo.lock

5
Cargo.toml Normal file
View File

@@ -0,0 +1,5 @@
[workspace]
members = [
"txtlang",
"txtlang-gen",
]

11
txtlang-gen/Cargo.toml Normal file
View File

@@ -0,0 +1,11 @@
[package]
name = "txtlang-gen"
version = "0.1.0"
authors = ["Anders Olsson <anders.e.olsson@gmail.com>"]
edition = "2018"
[dependencies]
anyhow = "1.0"
isolang = "1.0"
quick-xml = "0.18"
thiserror = "1.0"

157
txtlang-gen/src/main.rs Normal file
View File

@@ -0,0 +1,157 @@
use std::env;
use std::fs::File;
use std::io::BufReader;
use std::io::{self, Write};
use anyhow::Result;
use isolang::Language;
use quick_xml::{events::Event, Reader};
use thiserror::Error;
#[derive(Error, Debug)]
enum Error {
#[error("xml error")]
Xml(#[from] quick_xml::Error),
}
#[derive(PartialEq, Debug)]
struct LocalDisplayName {
name: String,
lang: Language,
}
fn main() -> Result<()> {
let inputs = env::args().skip(1);
let mut results = Vec::new();
for input in inputs {
let mut reader = File::open(&input)
.map(BufReader::new)
.map(Reader::from_reader)?;
reader.trim_text(true);
let mut buf = Vec::new();
loop {
match reader.read_event(&mut buf) {
Ok(Event::Start(ref e)) if e.name() == b"localeDisplayNames" => {
results.extend(parse::locale_display_names(&mut reader, &mut buf));
}
Err(e) => panic!("Error at position {}: {:?}", reader.buffer_position(), e),
Ok(Event::Eof) => break,
_ => (),
}
buf.clear();
}
}
let before = results.len();
results.sort_unstable_by(|a, b| a.name.cmp(&b.name));
results.dedup();
let dedup = results.len();
let mut remove = Vec::new();
results.dedup_by(|a, b| {
if a.name == b.name {
remove.push(a.name.clone());
}
a.name == b.name
});
results.retain(|x| !remove.contains(&x.name));
let clean = results.len();
eprintln!("before={}", before);
eprintln!("dedup={}", dedup);
eprintln!("clean={}", clean);
let stdout = io::stdout();
let mut lock = stdout.lock();
for result in results {
assert!(!result.name.contains('\n'));
assert!(!result.name.contains('\t'));
writeln!(lock, "{}\t{}", result.name, result.lang.to_639_3())
.expect("failed to write to stdout");
}
lock.flush().expect("failed to flush to stdout");
Ok(())
}
mod parse {
use std::fs::File;
use std::io::BufReader;
use anyhow::Result;
use isolang::Language;
use quick_xml::{events::Event, Reader};
use crate::LocalDisplayName;
pub(crate) fn locale_display_names(
reader: &mut Reader<BufReader<File>>,
buf: &mut Vec<u8>,
) -> Vec<LocalDisplayName> {
let mut result = Vec::new();
let mut parse = false;
let mut language = None;
loop {
match reader.read_event(buf) {
Ok(Event::End(ref e)) if e.name() == b"localeDisplayNames" => {
break;
}
Ok(Event::Start(ref e)) if e.name() == b"languages" => {
parse = true;
}
Ok(Event::End(ref e)) if e.name() == b"languages" => {
parse = false;
}
Ok(Event::Start(ref e)) if e.name() == b"language" && parse => {
let a = e
.attributes()
.filter_map(Result::ok)
.find(|a| a.key == b"type");
if let Some(a) = a {
if let Ok(lang) = reader.decode_without_bom(&a.value) {
language = Language::from_639_1(lang)
.or_else(|| Language::from_639_3(lang))
.or_else(|| Language::from_locale(lang));
}
}
}
Ok(Event::End(ref e)) if e.name() == b"language" && parse => {
language = None;
}
Ok(Event::Text(e)) if language.is_some() => {
if let Ok(text) = reader.decode_without_bom(&e) {
result.push(LocalDisplayName {
name: text.to_lowercase(),
lang: language.take().unwrap(),
});
}
}
Err(e) => panic!("Error at position {}: {:?}", reader.buffer_position(), e),
Ok(Event::Eof) => break,
_ => (),
}
buf.clear();
}
result
}
}

13
txtlang/Cargo.toml Normal file
View File

@@ -0,0 +1,13 @@
[package]
name = "txtlang"
version = "0.1.0"
authors = ["Anders Olsson <anders.e.olsson@gmail.com>"]
edition = "2018"
build = "build.rs"
[dependencies]
isolang = "1.0"
[build-dependencies]
isolang = "1.0"
phf_codegen = "0.8"

44
txtlang/build.rs Normal file
View File

@@ -0,0 +1,44 @@
use std::collections::HashMap;
use std::env;
use std::error::Error;
use std::fs::{self, File};
use std::io::{self, BufRead, BufReader, BufWriter, Write};
use std::path::Path;
fn title(s: &str) -> String {
let mut c = s.chars();
match c.next() {
None => String::new(),
Some(f) => f.to_uppercase().chain(c).collect(),
}
}
fn main() -> Result<(), Box<dyn Error>> {
let mut map = phf_codegen::Map::new();
let lines = File::open("languages.txt")
.map(BufReader::new)?
.lines()
.filter_map(Result::ok)
.collect::<Vec<_>>();
for line in &lines {
let parts = line.splitn(2, '\t').collect::<Vec<_>>();
let name = parts[0];
let language = parts[1];
map.entry(name, &format!("Language::{}", title(language)));
}
let path = Path::new(&env::var("OUT_DIR")?).join("languages.rs");
let mut file = BufWriter::new(File::create(&path)?);
write!(
&mut file,
"static LANGUAGES: phf::Map<&'static str, Language> = {};\n",
map.build()
)?;
Ok(())
}

45534
txtlang/languages.txt Normal file

File diff suppressed because it is too large Load Diff

13
txtlang/src/lib.rs Normal file
View File

@@ -0,0 +1,13 @@
use isolang::Language;
include!(concat!(env!("OUT_DIR"), "/languages.rs"));
pub trait LanguageExt {
fn from_name(name: &str) -> Option<Language>;
}
impl LanguageExt for Language {
fn from_name(name: &str) -> Option<Language> {
LANGUAGES.get(name).cloned()
}
}