X-Git-Url: https://pintos-os.org/cgi-bin/gitweb.cgi?a=blobdiff_plain;f=rust%2Fsrc%2Fmain.rs;h=404e96d57d07becaf09c4688319a0ac3739ceee4;hb=e0cbdf0daefcca81be9572aab0deedf945687f5a;hp=319d7aea7c0f32a020db403a21fe60fad96cac67;hpb=13d0badc005f46d33eb5a7ec4ef5d794b4512450;p=pspp

diff --git a/rust/src/main.rs b/rust/src/main.rs
index 319d7aea7c..404e96d57d 100644
--- a/rust/src/main.rs
+++ b/rust/src/main.rs
@@ -15,14 +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},
-};
+use clap::{Parser, ValueEnum};
+use encoding_rs::Encoding;
+use pspp::cooked::decode;
+use pspp::raw::{Reader, Record, Magic};
 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)]
@@ -35,36 +36,84 @@ 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 {
+    Identify,
+    Raw,
+    #[default]
+    Cooked,
 }
 
 fn main() -> Result<()> {
-    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()?;
 
-    let mut n_cases = 0;
-    for record in records {
-        println!("{record:?}");
-        match record {
-            Record::EndOfHeaders(_) if max_cases == 0 => break,
-            Record::Case(_) => {
-                n_cases += 1;
-                if n_cases >= max_cases {
-                    break;
-                }
+    match mode {
+        Mode::Identify => {
+            let Record::Header(header) = reader.next().unwrap()? else { unreachable!() };
+            match header.magic {
+                Magic::Sav => println!("SPSS System File"),
+                Magic::Zsav => println!("SPSS System File with Zlib compression"),
+                Magic::Ebcdic => println!("EBCDIC-encoded SPSS System File"),
+            }
+            return Ok(())
+        }
+        Mode::Raw => {
+            let headers: Vec<Record> = reader.collect_headers()?;
+            for header in headers {
+                println!("{header:?}");
             }
-            _ => (),
         }
+        Mode::Cooked => {
+            let headers: Vec<Record> = reader.collect_headers()?;
+            let headers = decode(headers, encoding, &|e| panic!("{e}"))?;
+            for header in headers {
+                println!("{header:?}");
+            }
+        }
+    }
+
+    for _ in 0..max_cases {
+        let Some(Ok(record)) = reader.next() else {
+            break;
+        };
+        println!("{:?}", record);
     }
     Ok(())
 }