ambers 0.3.8

Pure Rust reader for SPSS .sav and .zsav files
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187

//! SAV/ZSAV file writer.
//!
//! Writes Arrow RecordBatch + SpssMetadata to SPSS .sav/.zsav binary format.
//! Supports uncompressed, bytecode-compressed (.sav), and zlib-compressed (.zsav) output.

mod data;
mod layout;
mod records;

#[cfg(test)]
mod tests;

use std::fs::File;
use std::io::{BufWriter, Seek, Write};
use std::path::Path;

use crate::constants::*;
use crate::error::Result;
use crate::metadata::SpssMetadata;
use arrow::datatypes::{DataType, TimeUnit};
use arrow::record_batch::RecordBatch;

// ---------------------------------------------------------------------------
// Public API
// ---------------------------------------------------------------------------

/// Write an Arrow RecordBatch to an SPSS .sav file.
///
/// If `metadata` is provided, it controls variable labels, value labels,
/// formats, and other SPSS metadata. If default metadata is used, types
/// are inferred from the Arrow schema.
///
/// # Compression
/// - `Compression::None` — uncompressed .sav
/// - `Compression::Bytecode` — row-compressed .sav
/// - `Compression::Zlib` — block-compressed .zsav
///
/// # Compression Level (zsav only)
/// Controls zlib compression intensity. Ignored for non-zsav files.
/// - `None` — default (level 6, compact)
/// - `Some(1)` — "fast": fastest compression, largest files
/// - `Some(3)` — "balanced": moderate speed, moderate size
/// - `Some(6)` — "compact": slower compression, smallest files (default)
pub fn write_sav(
    path: impl AsRef<Path>,
    batch: &RecordBatch,
    metadata: &SpssMetadata,
    compression: Compression,
    compression_level: Option<u32>,
) -> Result<()> {
    let file = File::create(path)?;
    let writer = BufWriter::with_capacity(4 * 1024 * 1024, file);
    write_sav_to_writer(writer, batch, metadata, compression, compression_level)
}

/// Fill missing metadata fields from the Arrow schema with sensible defaults.
///
/// For each column in the schema, if the metadata is missing a field
/// (format, measure, alignment, display_width, role, storage_width),
/// it is filled using the same type-based logic as `from_arrow_schema()`.
/// User-set fields are never overwritten.
fn fill_defaults_from_schema(meta: &mut SpssMetadata, schema: &arrow::datatypes::Schema) {
    for field in schema.fields() {
        let name = field.name();

        // Infer defaults from Arrow type (same logic as SpssMetadata::from_arrow_schema)
        let (default_fmt, default_measure, default_alignment) = match field.data_type() {
            DataType::Float64 => ("F8.2", Measure::Scale, Alignment::Right),
            DataType::Int64 | DataType::Int32 | DataType::Int16 | DataType::Int8 => {
                ("F8.0", Measure::Scale, Alignment::Right)
            }
            DataType::Boolean => ("F1.0", Measure::Nominal, Alignment::Right),
            DataType::Date32 => ("DATE11", Measure::Scale, Alignment::Right),
            DataType::Timestamp(TimeUnit::Microsecond, _) => {
                ("DATETIME23.2", Measure::Scale, Alignment::Right)
            }
            DataType::Duration(TimeUnit::Microsecond) => {
                ("TIME11.2", Measure::Scale, Alignment::Right)
            }
            DataType::Utf8 | DataType::Utf8View | DataType::LargeUtf8 => {
                ("A255", Measure::Nominal, Alignment::Left)
            }
            _ => ("F8.2", Measure::Scale, Alignment::Right),
        };

        if !meta.variable_formats.contains_key(name.as_str()) {
            meta.variable_formats
                .insert(name.clone(), default_fmt.to_string());
        }

        if !meta.variable_measures.contains_key(name.as_str()) {
            meta.variable_measures.insert(name.clone(), default_measure);
        }

        if !meta.variable_alignments.contains_key(name.as_str()) {
            meta.variable_alignments
                .insert(name.clone(), default_alignment);
        }

        if !meta.variable_display_widths.contains_key(name.as_str()) {
            // Use the format width (already set above or by user)
            let fmt_str = meta.variable_formats.get(name.as_str()).unwrap();
            let w = SpssFormat::from_string(fmt_str)
                .map(|f| f.width as u32)
                .unwrap_or(8);
            meta.variable_display_widths.insert(name.clone(), w);
        }

        if !meta.variable_roles.contains_key(name.as_str()) {
            meta.variable_roles.insert(name.clone(), Role::Input);
        }

        if !meta.variable_storage_widths.contains_key(name.as_str()) {
            let fmt_str = meta.variable_formats.get(name.as_str()).unwrap();
            let sw = if let Some(rest) = fmt_str.strip_prefix('A') {
                rest.split('.')
                    .next()
                    .and_then(|w| w.parse::<usize>().ok())
                    .unwrap_or(255)
            } else {
                8
            };
            meta.variable_storage_widths.insert(name.clone(), sw);
        }
    }
}

/// Write to any writer that implements Write + Seek.
///
/// Seek is required for zsav (zlib) compression to backpatch the zheader.
/// For non-zsav files, `std::io::Cursor<Vec<u8>>` is a convenient seekable wrapper.
///
/// # Compression Level (zsav only)
/// Controls zlib compression intensity. Ignored for non-zsav files.
/// - `None` — default (level 6, compact)
/// - `Some(1)` — "fast": fastest compression, largest files
/// - `Some(3)` — "balanced": moderate speed, moderate size
/// - `Some(6)` — "compact": slower compression, smallest files (default)
pub fn write_sav_to_writer<W: Write + Seek>(
    mut writer: W,
    batch: &RecordBatch,
    metadata: &SpssMetadata,
    compression: Compression,
    compression_level: Option<u32>,
) -> Result<()> {
    // Fill any missing metadata fields from the schema
    let mut meta = metadata.clone();
    fill_defaults_from_schema(&mut meta, batch.schema().as_ref());
    let metadata = &meta;

    // Validate inputs before heavy computation
    layout::validate_write_inputs(batch.schema().as_ref(), metadata)?;

    let layout = layout::compute_layout(batch, metadata)?;
    let nrows = batch.num_rows() as i32;

    // Write dictionary
    records::write_header(&mut writer, &layout, metadata, compression, nrows)?;
    records::write_variable_records(&mut writer, &layout)?;
    records::write_value_label_records(&mut writer, &layout, metadata)?;
    records::write_document_record(&mut writer, metadata)?;
    records::write_info_integer(&mut writer, compression)?;
    records::write_info_float(&mut writer)?;
    records::write_info_var_display(&mut writer, &layout)?;
    records::write_info_long_names(&mut writer, &layout)?;
    records::write_info_very_long_strings(&mut writer, &layout)?;
    records::write_info_encoding(&mut writer)?;
    records::write_info_long_string_labels(&mut writer, &layout, metadata)?;
    records::write_info_long_string_missing(&mut writer, &layout, metadata)?;
    records::write_info_mr_sets(&mut writer, metadata, &layout)?;
    records::write_info_mr_sets_v2(&mut writer, metadata)?;
    records::write_info_var_attributes(&mut writer, metadata)?;
    records::write_dict_termination(&mut writer)?;

    // Write data
    match compression {
        Compression::None => data::write_data_uncompressed(&mut writer, batch, &layout)?,
        Compression::Bytecode => data::write_data_bytecode(&mut writer, batch, &layout)?,
        Compression::Zlib => {
            let level = flate2::Compression::new(compression_level.unwrap_or(6).min(9));
            data::write_data_zsav(&mut writer, batch, &layout, level)?;
        }
    }

    writer.flush()?;
    Ok(())
}