-#![allow(unused_variables)]
-#![allow(dead_code)]
/* PSPP - a program for statistical analysis.
* Copyright (C) 2023 Free Software Foundation, Inc.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>. */
-use anyhow::{anyhow, Result};
-use clap::Parser;
-use hexplay::HexView;
-use hexplay::HexViewBuilder;
-use num::{Float, Num};
-use std::cmp::Ordering;
+use anyhow::Result;
+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::prelude::*;
use std::io::BufReader;
use std::path::{Path, PathBuf};
use std::str;
-use std::{fmt, num::FpCategory};
+use thiserror::Error as ThisError;
/// A utility to dissect SPSS system files.
#[derive(Parser, Debug)]
struct Args {
/// Maximum number of cases to print.
#[arg(long = "data", default_value_t = 0)]
- max_cases: usize,
+ max_cases: u64,
/// Files to dissect.
#[arg(required = true)]
files: Vec<PathBuf>,
-}
-
-fn main() -> Result<()> {
- let Args { max_cases, files } = Args::parse();
-
- for file in files {
- Dissector::new(file)?;
- }
- Ok(())
-}
-
-#[derive(Copy, Clone, Debug)]
-enum Compression {
- Simple,
- ZLib,
-}
-
-#[derive(Copy, Clone, Debug)]
-enum Endianness {
- BigEndian,
- LittleEndian,
-}
-use Endianness::*;
-
-trait Parse<T, const N: usize> {
- fn parse(self, bytes: [u8; N]) -> T;
-}
-impl Parse<u64, 8> for Endianness {
- fn parse(self, bytes: [u8; 8]) -> u64 {
- match self {
- BigEndian => u64::from_be_bytes(bytes),
- LittleEndian => u64::from_le_bytes(bytes),
- }
- }
-}
-impl Parse<u32, 4> for Endianness {
- fn parse(self, bytes: [u8; 4]) -> u32 {
- match self {
- BigEndian => u32::from_be_bytes(bytes),
- LittleEndian => u32::from_le_bytes(bytes),
- }
- }
-}
-impl Parse<u16, 2> for Endianness {
- fn parse(self, bytes: [u8; 2]) -> u16 {
- match self {
- BigEndian => u16::from_be_bytes(bytes),
- LittleEndian => u16::from_le_bytes(bytes),
- }
- }
-}
-impl Parse<u8, 1> for Endianness {
- fn parse(self, bytes: [u8; 1]) -> u8 {
- match self {
- BigEndian => u8::from_be_bytes(bytes),
- LittleEndian => u8::from_le_bytes(bytes),
- }
- }
-}
-impl Parse<i64, 8> for Endianness {
- fn parse(self, bytes: [u8; 8]) -> i64 {
- match self {
- BigEndian => i64::from_be_bytes(bytes),
- LittleEndian => i64::from_le_bytes(bytes),
- }
- }
-}
-impl Parse<i32, 4> for Endianness {
- fn parse(self, bytes: [u8; 4]) -> i32 {
- match self {
- BigEndian => i32::from_be_bytes(bytes),
- LittleEndian => i32::from_le_bytes(bytes),
- }
- }
-}
-impl Parse<i16, 2> for Endianness {
- fn parse(self, bytes: [u8; 2]) -> i16 {
- match self {
- BigEndian => i16::from_be_bytes(bytes),
- LittleEndian => i16::from_le_bytes(bytes),
- }
- }
-}
-impl Parse<i8, 1> for Endianness {
- fn parse(self, bytes: [u8; 1]) -> i8 {
- match self {
- BigEndian => i8::from_be_bytes(bytes),
- LittleEndian => i8::from_le_bytes(bytes),
- }
- }
-}
-impl Parse<f64, 8> for Endianness {
- fn parse(self, bytes: [u8; 8]) -> f64 {
- match self {
- BigEndian => f64::from_be_bytes(bytes),
- LittleEndian => f64::from_le_bytes(bytes),
- }
- }
-}
-
-fn read_bytes<const N: usize>(r: &mut BufReader<File>) -> Result<[u8; N]> {
- let mut buf = [0; N];
- r.read_exact(&mut buf)?;
- Ok(buf)
-}
-
-fn read_vec(r: &mut BufReader<File>, n: usize) -> Result<Vec<u8>> {
- let mut vec = vec![0; n];
- r.read_exact(&mut vec)?;
- Ok(vec)
-}
-
-trait ReadSwap<T> {
- fn read_swap(&mut self) -> Result<T>;
-}
-
-impl ReadSwap<u32> for Dissector {
- fn read_swap(&mut self) -> Result<u32> {
- Ok(self.endianness.parse(read_bytes(&mut self.r)?))
- }
-}
-impl ReadSwap<u8> for Dissector {
- fn read_swap(&mut self) -> Result<u8> {
- Ok(self.endianness.parse(read_bytes(&mut self.r)?))
- }
-}
-
-impl ReadSwap<i32> for Dissector {
- fn read_swap(&mut self) -> Result<i32> {
- Ok(self.endianness.parse(read_bytes(&mut self.r)?))
- }
-}
-impl ReadSwap<f64> for Dissector {
- fn read_swap(&mut self) -> Result<f64> {
- Ok(self.endianness.parse(read_bytes(&mut self.r)?))
- }
-}
+ /// How to dissect the file.
+ #[arg(short, long, value_enum, default_value_t)]
+ mode: Mode,
-struct Dissector {
- filename: String,
- r: BufReader<File>,
- compression: Option<Compression>,
- endianness: Endianness,
- fp_format: Endianness,
- bias: f64,
- n_variable_records: usize,
- n_variables: usize,
- var_widths: Vec<i32>,
-}
-
-fn detect_endianness(layout_code: [u8; 4]) -> Option<Endianness> {
- for endianness in [BigEndian, LittleEndian] {
- match endianness.parse(layout_code) {
- 2 | 3 => return Some(endianness),
- _ => (),
- }
- }
- None
+ /// The encoding to use.
+ #[arg(long, value_parser = parse_encoding)]
+ encoding: Option<&'static Encoding>,
}
-fn detect_fp_format(bias: [u8; 8]) -> Option<Endianness> {
- for endianness in [BigEndian, LittleEndian] {
- let value: f64 = endianness.parse(bias);
- if value == 100.0 {
- return Some(endianness);
- }
- }
- None
-}
+#[derive(ThisError, Debug)]
+#[error("{0}: unknown encoding")]
+struct UnknownEncodingError(String);
-fn trim_end(mut s: Vec<u8>, c: u8) -> Vec<u8> {
- while s.last() == Some(&c) {
- s.pop();
+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())),
}
- s
}
-fn slice_trim_end(mut s: &[u8], c: u8) -> &[u8] {
- while s.last() == Some(&c) {
- s = s.split_last().unwrap().1;
- }
- s
+#[derive(Clone, Copy, Debug, Default, ValueEnum)]
+enum Mode {
+ Identify,
+ Raw,
+ #[default]
+ Cooked,
}
-fn format_name(type_: u32) -> &'static str {
- match type_ {
- 1 => "A",
- 2 => "AHEX",
- 3 => "COMMA",
- 4 => "DOLLAR",
- 5 => "F",
- 6 => "IB",
- 7 => "PIBHEX",
- 8 => "P",
- 9 => "PIB",
- 10 => "PK",
- 11 => "RB",
- 12 => "RBHEX",
- 15 => "Z",
- 16 => "N",
- 17 => "E",
- 20 => "DATE",
- 21 => "TIME",
- 22 => "DATETIME",
- 23 => "ADATE",
- 24 => "JDATE",
- 25 => "DTIME",
- 26 => "WKDAY",
- 27 => "MONTH",
- 28 => "MOYR",
- 29 => "QYR",
- 30 => "WKYR",
- 31 => "PCT",
- 32 => "DOT",
- 33 => "CCA",
- 34 => "CCB",
- 35 => "CCC",
- 36 => "CCD",
- 37 => "CCE",
- 38 => "EDATE",
- 39 => "SDATE",
- 40 => "MTIME",
- 41 => "YMDHMS",
- _ => "invalid",
- }
-}
-
-fn round_up<T: Num + Copy>(x: T, y: T) -> T {
- (x + (y - T::one())) / y * y
-}
-
-struct UntypedValue {
- raw: [u8; 8],
- endianness: Endianness,
-}
-
-impl UntypedValue {
- fn new(raw: [u8; 8], endianness: Endianness) -> UntypedValue {
- UntypedValue { raw, endianness }
- }
-}
-
-impl fmt::Display for UntypedValue {
- fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
- let numeric: f64 = self.endianness.parse(self.raw);
- let n_printable = self
- .raw
- .iter()
- .take_while(|&&x| x == b' ' || x.is_ascii_graphic())
- .count();
- let printable_prefix = std::str::from_utf8(&self.raw[0..n_printable]).unwrap();
- write!(f, "{numeric}/\"{printable_prefix}\"")
- }
-}
-
-struct HexFloat<T: Float>(T);
-
-impl<T: Float> fmt::Display for HexFloat<T> {
- fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
- let sign = if self.0.is_sign_negative() { "-" } else { "" };
- match self.0.classify() {
- FpCategory::Nan => return write!(f, "NaN"),
- FpCategory::Infinite => return write!(f, "{sign}Infinity"),
- FpCategory::Zero => return write!(f, "{sign}0.0"),
- _ => (),
- };
- let (significand, mut exponent, _) = self.0.integer_decode();
- let mut hex_sig = format!("{:x}", significand);
- while hex_sig.ends_with('0') {
- hex_sig.pop();
- exponent += 4;
- }
- match hex_sig.len() {
- 0 => write!(f, "{sign}0.0"),
- 1 => write!(f, "{sign}0x{hex_sig}.0p{exponent}"),
- len => write!(
- f,
- "{sign}0x{}.{}p{}",
- hex_sig.chars().next().unwrap(),
- &hex_sig[1..],
- exponent + 4 * (len as i16 - 1)
- ),
- }
- }
-}
-
-#[cfg(test)]
-mod hex_float_tests {
- use crate::HexFloat;
- use num::Float;
+fn main() -> Result<()> {
+ let Args {
+ max_cases,
+ files,
+ mode,
+ encoding,
+ } = Args::parse();
- #[test]
- fn test() {
- assert_eq!(format!("{}", HexFloat(1.0)), "0x1.0p0");
- assert_eq!(format!("{}", HexFloat(123.0)), "0x1.ecp6");
- assert_eq!(format!("{}", HexFloat(1.0 / 16.0)), "0x1.0p-4");
- assert_eq!(format!("{}", HexFloat(f64::infinity())), "Infinity");
- assert_eq!(format!("{}", HexFloat(f64::neg_infinity())), "-Infinity");
- assert_eq!(format!("{}", HexFloat(f64::nan())), "NaN");
- assert_eq!(format!("{}", HexFloat(0.0)), "0.0");
- assert_eq!(format!("{}", HexFloat(f64::neg_zero())), "-0.0");
+ for file in files {
+ dissect(&file, max_cases, mode, encoding)?;
}
+ Ok(())
}
-impl Dissector {
- fn new<P: AsRef<Path>>(filename: P) -> Result<Dissector> {
- let mut r = BufReader::new(File::open(&filename)?);
- let filename = filename.as_ref().to_string_lossy().into_owned();
- let rec_type: [u8; 4] = read_bytes(&mut r)?;
- let zmagic = match &rec_type {
- b"$FL2" => false,
- b"$FL3" => true,
- _ => Err(anyhow!("This is not an SPSS system file."))?,
- };
-
- let eye_catcher: [u8; 60] = read_bytes(&mut r)?;
- let layout_code: [u8; 4] = read_bytes(&mut r)?;
- let endianness = detect_endianness(layout_code)
- .ok_or_else(|| anyhow!("This is not an SPSS system file."))?;
- let layout_code: u32 = endianness.parse(layout_code);
- let _nominal_case_size: [u8; 4] = read_bytes(&mut r)?;
- let compressed: u32 = endianness.parse(read_bytes(&mut r)?);
- let compression = match (zmagic, compressed) {
- (false, 0) => None,
- (false, 1) => Some(Compression::Simple),
- (true, 2) => Some(Compression::ZLib),
- _ => Err(anyhow!(
- "{} file header has invalid compression value {compressed}.",
- if zmagic { "ZSAV" } else { "SAV" }
- ))?,
- };
-
- let weight_index: u32 = endianness.parse(read_bytes(&mut r)?);
- let n_cases: u32 = endianness.parse(read_bytes(&mut r)?);
-
- let bias: [u8; 8] = read_bytes(&mut r)?;
- let fp_format = detect_fp_format(bias)
- .unwrap_or_else(|| { eprintln!("Compression bias is not the usual value of 100, or system file uses unrecognized floating-point format."); endianness });
- let bias: f64 = fp_format.parse(bias);
-
- let mut d = Dissector {
- filename,
- r,
- compression,
- endianness,
- fp_format,
- bias,
- n_variable_records: 0,
- n_variables: 0,
- var_widths: Vec::new(),
- };
-
- let creation_date: [u8; 9] = read_bytes(&mut d.r)?;
- let creation_time: [u8; 8] = read_bytes(&mut d.r)?;
- let file_label: [u8; 64] = read_bytes(&mut d.r)?;
- let file_label = trim_end(Vec::from(file_label), b' ');
- d.skip_bytes(3)?;
-
- println!("File header record:");
- println!(
- "{:>17}: {}",
- "Product name",
- String::from_utf8_lossy(&eye_catcher)
- );
- println!("{:>17}: {}", "Layout code", layout_code);
- println!(
- "{:>17}: {} ({})",
- "Compressed",
- compressed,
- match compression {
- None => "no compression",
- Some(Compression::Simple) => "simple compression",
- Some(Compression::ZLib) => "ZLIB compression",
- }
- );
- println!("{:>17}: {}", "Weight index", weight_index);
- println!("{:>17}: {}", "Number of cases", n_cases);
- println!("{:>17}: {}", "Compression bias", bias);
- println!(
- "{:>17}: {}",
- "Creation date",
- String::from_utf8_lossy(&creation_date)
- );
- println!(
- "{:>17}: {}",
- "Creation time",
- String::from_utf8_lossy(&creation_time)
- );
- println!(
- "{:>17}: \"{}\"",
- "File label",
- String::from_utf8_lossy(&file_label)
- );
-
- loop {
- let rec_type: u32 = d.read_swap()?;
- match rec_type {
- 2 => d.read_variable_record()?,
- 3 => d.read_value_label_record()?,
- 4 => Err(anyhow!("Misplaced type 4 record."))?,
- 6 => d.read_document_record()?,
- 7 => d.read_extension_record()?,
- 999 => break,
- _ => Err(anyhow!("Unrecognized record type {rec_type}."))?,
- }
- }
-
- let pos = d.r.stream_position()?;
- println!(
- "{:08x}: end-of-dictionary record (first byte of data at {:0x})",
- pos,
- pos + 4
- );
-
- Ok(d)
- }
-
- fn read_extension_record(&mut self) -> Result<()> {
- let offset = self.r.stream_position()?;
- let subtype: u32 = self.read_swap()?;
- let size: u32 = self.read_swap()?;
- let count: u32 = self.read_swap()?;
- println!("{offset:08x}: Record 7, subtype {subtype}, size={size}, count={count}");
- match subtype {
- 3 => self.read_machine_integer_info(size, count),
- 4 => self.read_machine_float_info(size, count),
- 5 => self.read_variable_sets(size, count),
- 6 => {
- // DATE variable information. We don't use it yet, but we should.
- Ok(())
- }
- 7 | 19 => self.read_mrsets(size, count),
- 10 => self.read_extra_product_info(size, count),
- 11 => self.read_display_parameters(size, count),
- _ => self.read_unknown_extension(subtype, size, count),
- }
- }
-
- fn warn(&mut self, s: String) -> Result<()> {
- println!(
- "\"{}\" near offset 0x{:08x}: {s}",
- self.filename,
- self.r.stream_position()?
- );
- Ok(())
- }
-
- fn skip_bytes(&mut self, mut n: u64) -> Result<()> {
- let mut buf = [0; 1024];
- while n > 0 {
- let chunk = u64::min(n, buf.len() as u64);
- self.r.read_exact(&mut buf[0..chunk as usize])?;
- n -= chunk;
- }
- Ok(())
- }
-
- fn read_unknown_extension(&mut self, subtype: u32, size: u32, count: u32) -> Result<()> {
- self.warn(format!("Unrecognized record type 7, subtype {subtype}."))?;
- if size == 0 || count > 65536 / size {
- self.skip_bytes(size as u64 * count as u64)?;
- } else if size != 1 {
- let mut offset = 0;
- for _ in 0..count {
- let vec = read_vec(&mut self.r, size as usize)?;
- println!(
- "{}",
- HexViewBuilder::new(&vec).address_offset(offset).finish()
- );
- offset += size as usize;
- }
- }
- Ok(())
- }
-
- fn read_variable_record(&mut self) -> Result<()> {
- self.n_variable_records += 1;
- println!(
- "{:08x}: variable record {}",
- self.r.stream_position()?,
- self.n_variable_records
- );
- let width: i32 = self.read_swap()?;
- let has_variable_label: u32 = self.read_swap()?;
- let missing_value_code: i32 = self.read_swap()?;
- let print_format: u32 = self.read_swap()?;
- let write_format: u32 = self.read_swap()?;
- let name: [u8; 8] = read_bytes(&mut self.r)?;
- let name: Vec<u8> = trim_end(Vec::from(name), b'\0');
-
- if width >= 0 {
- self.n_variables += 1;
- }
- self.var_widths.push(width);
-
- println!(
- "\tWidth: {width} ({})",
- match width {
- _ if width > 0 => "string",
- _ if width == 0 => "numeric",
- _ => "long string continuation record",
- }
- );
-
- println!("\tVariable label: {has_variable_label}");
- println!(
- "\tMissing values code: {missing_value_code} ({})",
- match missing_value_code {
- 0 => "no missing values",
- 1 => "one missing value",
- 2 => "two missing values",
- 3 => "three missing values",
- -2 => "one missing value range",
- -3 => "one missing value, one range",
- _ => "bad value",
- }
- );
- for (which, format) in [("Print", print_format), ("Worite", write_format)] {
- let type_ = format_name(format >> 16);
- let w = (format >> 8) & 0xff;
- let d = format & 0xff;
- println!("\t{which} format: {format:06x} ({type_}{w}.{d})");
- }
- println!("\tName: {}", String::from_utf8_lossy(&name));
-
- // Read variable label.
- match has_variable_label {
- 0 => (),
- 1 => {
- let offset = self.r.stream_position()?;
- let len: u32 = self.read_swap()?;
- let read_len = len.min(65535) as usize;
- let label = read_vec(&mut self.r, read_len)?;
- println!(
- "\t{offset:08x} Variable label: \"{}\"",
- String::from_utf8_lossy(&label)
- );
-
- self.skip_bytes((round_up(len, 4) - len).into())?;
- }
- _ => Err(anyhow!("Variable label indicator field is not 0 or 1."))?,
- };
-
- // Read missing values.
- if missing_value_code != 0 {
- print!("\t{:08x} Missing values:", self.r.stream_position()?);
- match width.cmp(&0) {
- Ordering::Equal => {
- let (has_range, n_individual) = match missing_value_code {
- -3 => (true, 1),
- -2 => (true, 0),
- 1 | 2 | 3 => (false, missing_value_code),
- _ => Err(anyhow!(
- "Numeric missing value indicator field is not -3, -2, 0, 1, 2, or 3."
- ))?,
- };
- if has_range {
- let low: f64 = self.read_swap()?;
- let high: f64 = self.read_swap()?;
- print!(" {low}...{high}");
- }
- for _ in 0..n_individual {
- let value: f64 = self.read_swap()?;
- print!(" {value}");
- }
- }
- Ordering::Greater => {
- if !(0..=3).contains(&missing_value_code) {
- Err(anyhow!(
- "String missing value indicator field is not 0, 1, 2, or 3."
- ))?;
- }
- for _ in 0..missing_value_code {
- let string: [u8; 8] = read_bytes(&mut self.r)?;
- let string: Vec<u8> = trim_end(Vec::from(string), b'\0');
- println!(" {}", String::from_utf8_lossy(&string));
- }
- }
- Ordering::Less => (),
- }
- println!();
- }
-
- Ok(())
- }
-
- fn read_value_label_record(&mut self) -> Result<()> {
- println!("{:08x}: value labels record", self.r.stream_position()?);
-
- // Read the labels.
- let n_labels: u32 = self.read_swap()?;
- for _ in 0..n_labels {
- let raw: [u8; 8] = read_bytes(&mut self.r)?;
- let value = UntypedValue::new(raw, self.fp_format);
- let label_len: u8 = self.read_swap()?;
- let padded_len = round_up(label_len as usize + 1, 8);
-
- let mut label = read_vec(&mut self.r, padded_len)?;
- label.truncate(label_len as usize);
- let label = String::from_utf8_lossy(&label);
-
- println!("\t{value}: {label}");
- }
-
- // Read the type-4 record with the corresponding variable indexes.
- let rec_type: u32 = self.read_swap()?;
- if rec_type != 4 {
- Err(anyhow!(
- "Variable index record (type 4) does not immediately \
- follow value label record (type 3) as it should."
- ))?;
- }
-
- println!("\t{:08x}: apply to variables", self.r.stream_position()?);
- let n_vars: u32 = self.read_swap()?;
- for _ in 0..n_vars {
- let index: u32 = self.read_swap()?;
- print!(" {index}");
- }
- println!();
-
- Ok(())
- }
-
- fn read_document_record(&mut self) -> Result<()> {
- println!("{:08x}: document record", self.r.stream_position()?);
- let n_lines: u32 = self.read_swap()?;
- println!("\t{n_lines} lines of documents");
-
- for i in 0..n_lines {
- print!("\t{:08x}: ", self.r.stream_position()?);
- let line: [u8; 64] = read_bytes(&mut self.r)?;
- let line = trim_end(Vec::from(line), b' ');
- println!("line {i}: \"{}\"", String::from_utf8_lossy(&line));
- }
- Ok(())
- }
-
- fn read_machine_integer_info(&mut self, size: u32, count: u32) -> Result<()> {
- let offset = self.r.stream_position()?;
- let version_major: u32 = self.read_swap()?;
- let version_minor: u32 = self.read_swap()?;
- let version_revision: u32 = self.read_swap()?;
- let machine_code: u32 = self.read_swap()?;
- let float_representation: u32 = self.read_swap()?;
- let compression_code: u32 = self.read_swap()?;
- let integer_representation: u32 = self.read_swap()?;
- let character_code: u32 = self.read_swap()?;
-
- println!("{offset:08x}: machine integer info");
- if size != 4 || count != 8 {
- Err(anyhow!(
- "Bad size ({size}) or count ({count}) field on record type 7, subtype 3"
- ))?;
- }
- println!("\tVersion: {version_major}.{version_minor}.{version_revision}");
- println!("\tMachine code: {machine_code}");
- println!(
- "\tFloating point representation: {float_representation} ({})",
- match float_representation {
- 1 => "IEEE 754",
- 2 => "IBM 370",
- 3 => "DEC VAX",
- _ => "unknown",
- }
- );
- println!("\tCompression code: {compression_code}");
- println!(
- "\tEndianness: {integer_representation} ({})",
- match integer_representation {
- 1 => "big",
- 2 => "little",
- _ => "unknown",
- }
- );
- println!("\tCharacter code: {character_code}");
- Ok(())
- }
-
- fn read_machine_float_info(&mut self, size: u32, count: u32) -> Result<()> {
- let offset = self.r.stream_position()?;
- let sysmis: f64 = self.read_swap()?;
- let highest: f64 = self.read_swap()?;
- let lowest: f64 = self.read_swap()?;
-
- println!("{offset:08x}: machine float info");
- if size != 4 || count != 8 {
- Err(anyhow!(
- "Bad size ({size}) or count ({count}) field on extension 4."
- ))?;
- }
-
- println!("\tsysmis: {sysmis} ({})", HexFloat(sysmis));
- println!("\thighest: {highest} ({})", HexFloat(highest));
- println!("\tlowest: {lowest} ({})", HexFloat(lowest));
- Ok(())
- }
+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)?;
- fn read_variable_sets(&mut self, size: u32, count: u32) -> Result<()> {
- println!("{:08x}: variable sets", self.r.stream_position()?);
- let mut text = self.open_text_record(size, count)?;
- loop {
- while text.match_byte(b'\n') {
- continue;
+ 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"),
}
- let set = match text.tokenize(b'=') {
- Some(set) => String::from_utf8_lossy(set).into_owned(),
- None => break,
- };
-
- // Always present even for an empty set.
- text.match_byte(b' ');
-
- match text.tokenize(b'\n') {
- None => println!("\tset \"{set}\" is empty"),
- Some(variables) => {
- println!(
- "\tset \"{set}\" contains \"{}\"",
- String::from_utf8_lossy(variables).trim_end_matches('\r')
- );
- }
- };
+ return Ok(())
}
- Ok(())
- }
-
- // Read record type 7, subtype 7.
- fn read_mrsets(&mut self, size: u32, count: u32) -> Result<()> {
- print!("{:08x}: multiple response sets", self.r.stream_position()?);
- let mut text = self.open_text_record(size, count)?;
- loop {
- #[derive(PartialEq, Eq)]
- enum MrSet {
- MC,
- MD,
+ Mode::Raw => {
+ let headers: Vec<Record> = reader.collect_headers()?;
+ for header in headers {
+ println!("{header:?}");
}
-
- while text.match_byte(b'\n') {}
- let Some(name) = text.tokenize(b'=') else {
- break;
- };
-
- let (mrset, cat_label_from_counted_values, label_from_var_label) = if text
- .match_byte(b'C')
- {
- if !text.match_byte(b' ') {
- Err(anyhow!(
- "missing space following 'C' at offset {} in mrsets record",
- text.pos
- ))?;
- }
- (MrSet::MC, false, false)
- } else if text.match_byte(b'D') {
- (MrSet::MD, false, false)
- } else if text.match_byte(b'E') {
- if !text.match_byte(b' ') {
- Err(anyhow!(
- "missing space following 'E' at offset {} in mrsets record",
- text.pos
- ))?;
- }
-
- let pos = text.pos;
- let Some(number) = text.tokenize(b' ') else {
- Err(anyhow!(
- "Missing label source value following `E' at offset {}u in MRSETS record",
- text.pos
- ))?
- };
-
- let label_from_var_label = if number == b"11" {
- true
- } else if number == b"1" {
- false
- } else {
- Err(anyhow!("Unexpected label source value `{}' following `E' at offset {pos} in MRSETS record", String::from_utf8_lossy(number)))?
- };
- (MrSet::MD, true, label_from_var_label)
- } else {
- Err(anyhow!(
- "missing `C', `D', or `E' at offset {} in mrsets record",
- text.pos
- ))?
- };
-
- let counted_value = if mrset == MrSet::MD {
- Some(text.parse_counted_string()?)
- } else { None };
-
- let label = text.parse_counted_string()?;
-
- let variables = text.tokenize(b'\n');
-
- print!("\t\"{}\": multiple {} set",
- String::from_utf8_lossy(name),
- if mrset == MrSet::MC { "category" } else { "dichotomy" });
-
}
- Ok(())
- }
-
- fn read_extra_product_info(&mut self, size: u32, count: u32) -> Result<()> {
- print!("{:08x}: extra product info", self.r.stream_position()?);
- let text = self.open_text_record(size, count)?;
- print_string(&text.buffer);
- Ok(())
- }
-
- fn read_display_parameters(&mut self, size: u32, count: u32) -> Result<()> {
- println!(
- "{:08x}: variable display parameters",
- self.r.stream_position()?
- );
- if size != 4 {
- Err(anyhow!("Bad size ({size}) on extension 11."))?;
- }
- let n_vars = self.n_variables;
- let includes_width = if count as usize == 3 * n_vars {
- true
- } else if count as usize == 2 * n_vars {
- false
- } else {
- Err(anyhow!(
- "Extension 11 has bad count {count} (for {n_vars} variables)."
- ))?
- };
-
- for i in 0..n_vars {
- let measure: u32 = self.read_swap()?;
- print!(
- "\tVar #{i}: measure={measure} ({})",
- match measure {
- 1 => "nominal",
- 2 => "ordinal",
- 3 => "scale",
- _ => "invalid",
- }
- );
-
- if includes_width {
- let width: u32 = self.read_swap()?;
- print!(", width={width}");
+ Mode::Cooked => {
+ let headers: Vec<Record> = reader.collect_headers()?;
+ let headers = decode(headers, encoding, &|e| panic!("{e}"))?;
+ for header in headers {
+ println!("{header:?}");
}
-
- let align: u32 = self.read_swap()?;
- println!(
- ", align={align} ({})",
- match align {
- 0 => "left",
- 1 => "right",
- 2 => "centre",
- _ => "invalid",
- }
- );
}
- Ok(())
}
- fn open_text_record(&mut self, size: u32, count: u32) -> Result<TextRecord> {
- let n_bytes = match u32::checked_mul(size, count) {
- Some(n) => n,
- None => Err(anyhow!("Extension record too large."))?,
+ for _ in 0..max_cases {
+ let Some(Ok(record)) = reader.next() else {
+ break;
};
- Ok(TextRecord::new(read_vec(&mut self.r, n_bytes as usize)?))
- }
-}
-
-fn print_string(s: &[u8]) {
- if s.contains(&b'\0') {
- println!("{}", HexView::new(s));
- } else {
- for &c in s {
- match c {
- b'\\' => print!("\\\\"),
- b'\n' => println!(),
- c if (b' '..=b'~').contains(&c) => print!("{}", c as char),
- c => print!("\\{:2x}", c),
- }
- }
- }
-}
-
-struct TextRecord {
- buffer: Vec<u8>,
- pos: usize,
-}
-
-impl TextRecord {
- fn new(buffer: Vec<u8>) -> TextRecord {
- TextRecord { buffer, pos: 0 }
- }
-
- fn tokenize(&mut self, delimiter: u8) -> Option<&[u8]> {
- let start = self.pos;
- while self.pos < self.buffer.len()
- && self.buffer[self.pos] != delimiter
- && self.buffer[self.pos] != 0
- {
- self.pos += 1
- }
- if start == self.pos {
- None
- } else {
- Some(&self.buffer[start..self.pos])
- }
- }
-
- fn match_byte(&mut self, c: u8) -> bool {
- if self.pos < self.buffer.len() && self.buffer[self.pos] == c {
- self.pos += 1;
- true
- } else {
- false
- }
- }
-
- fn parse_usize(&mut self) -> Result<usize> {
- let n_digits = self.buffer[self.pos..]
- .iter()
- .take_while(|c| c.is_ascii_digit())
- .count();
- if n_digits == 0 {
- Err(anyhow!("expecting digit at offset {} in record", self.pos))?;
- }
- let start = self.pos;
- self.pos += n_digits;
- let end = self.pos;
- let digits = str::from_utf8(&self.buffer[start..end]).unwrap();
- let Ok(number) = digits.parse::<usize>() else {
- Err(anyhow!(
- "expecting number in [0,{}] at offset {} in record",
- usize::MAX,
- self.pos
- ))?
- };
- self.pos = end;
- Ok(number)
- }
-
- fn get_n_bytes(&mut self, n: usize) -> Option<(usize, usize)> {
- let start = self.pos;
- let Some(end) = start.checked_add(n) else {
- return None;
- };
- self.pos = end;
- Some((start, end))
- }
-
- fn parse_counted_string(&mut self) -> Result<&[u8]> {
- let length = self.parse_usize()?;
- if !self.match_byte(b' ') {
- Err(anyhow!("expecting space at offset {} in record", self.pos))?;
- }
-
- let Some((start, end)) = self.get_n_bytes(length) else {
- Err(anyhow!("{length}-byte string starting at offset {} exceeds record length {}",
- self.pos, self.buffer.len()))?
- };
- if !self.match_byte(b' ') {
- Err(anyhow!(
- "expecting space at offset {} following {}-byte string",
- self.pos,
- end - start
- ))?;
- }
- Ok(&self.buffer[start..end])
+ println!("{:?}", record);
}
+ Ok(())
}
projects / pspp / blobdiff