Works for at least one test file now

[pspp] / rust / src / main.rs

diff --git a/rust/src/main.rs b/rust/src/main.rs
index 8085957daad73d8b57df657ce3779855438d8b0a..2251b760fe7dab58b5ae78d60b9cc7258853b906 100644 (file)
--- a/rust/src/main.rs
+++ b/rust/src/main.rs
@@ -15,12 +15,15 @@
  * along with this program.  If not, see <http://www.gnu.org/licenses/>. */
 
 use anyhow::Result;
-use clap::Parser;
-use pspp::{raw::{Reader, Record}, locale_charset::locale_charset};
+use clap::{Parser, ValueEnum};
+use encoding_rs::Encoding;
+use pspp::cooked::decode;
+use pspp::raw::{Reader, Record};
 use std::fs::File;
 use std::io::BufReader;
 use std::path::{Path, PathBuf};
 use std::str;
+use thiserror::Error as ThisError;
 
 /// A utility to dissect SPSS system files.
 #[derive(Parser, Debug)]
@@ -33,29 +36,66 @@ struct Args {
     /// Files to dissect.
     #[arg(required = true)]
     files: Vec<PathBuf>,
+
+    /// How to dissect the file.
+    #[arg(short, long, value_enum, default_value_t)]
+    mode: Mode,
+
+    /// The encoding to use.
+    #[arg(long, value_parser = parse_encoding)]
+    encoding: Option<&'static Encoding>,
+}
+
+#[derive(ThisError, Debug)]
+#[error("{0}: unknown encoding")]
+struct UnknownEncodingError(String);
+
+fn parse_encoding(arg: &str) -> Result<&'static Encoding, UnknownEncodingError> {
+    match Encoding::for_label_no_replacement(arg.as_bytes()) {
+        Some(encoding) => Ok(encoding),
+        None => Err(UnknownEncodingError(arg.to_string())),
+    }
+}
+
+#[derive(Clone, Copy, Debug, Default, ValueEnum)]
+enum Mode {
+    Raw,
+    #[default]
+    Cooked,
 }
 
 fn main() -> Result<()> {
-    println!("locale_charset={}", locale_charset());
-    let Args { max_cases, files } = Args::parse();
+    let Args {
+        max_cases,
+        files,
+        mode,
+        encoding,
+    } = Args::parse();
 
     for file in files {
-        dissect(&file, max_cases)?;
+        dissect(&file, max_cases, mode, encoding)?;
     }
     Ok(())
 }
 
-fn dissect(file_name: &Path, max_cases: u64) -> Result<()> {
+fn dissect(file_name: &Path, max_cases: u64, mode: Mode, encoding: Option<&'static Encoding>) -> Result<()> {
     let reader = File::open(file_name)?;
     let reader = BufReader::new(reader);
     let mut reader = Reader::new(reader)?;
-    let records: Vec<Record> = reader.collect_headers()?;
 
-    for record in records {
-        println!("{record:?}");
-        if let Record::EndOfHeaders(_) = record {
-            break;
-        };
+    let headers: Vec<Record> = reader.collect_headers()?;
+    match mode {
+        Mode::Raw => {
+            for header in headers {
+                println!("{header:?}");
+            }
+        }
+        Mode::Cooked => {
+            let headers = decode(headers, encoding, &|e| panic!("{e}"))?;
+            for header in headers {
+                println!("{header:?}");
+            }
+        }
     }
 
     for _ in 0..max_cases {