datui_cli/
lib.rs

1//! Shared CLI definitions for datui.
2//!
3//! Used by the main application and by the build script (manpage) and
4//! gen_docs binary (command-line-options markdown).
5
6use clap::{CommandFactory, Parser, ValueEnum};
7use std::path::Path;
8
9/// File format for data files (used to bypass extension-based detection).
10/// When `--format` is not specified, format is auto-detected from the file extension.
11#[derive(Debug, Clone, Copy, ValueEnum, PartialEq, Eq)]
12pub enum FileFormat {
13    /// Parquet columnar format
14    Parquet,
15    /// Comma-separated values
16    Csv,
17    /// Tab-separated values
18    Tsv,
19    /// Pipe-separated values
20    Psv,
21    /// JSON array format
22    Json,
23    /// JSON Lines / NDJSON (one JSON object per line)
24    Jsonl,
25    /// Arrow IPC / Feather
26    Arrow,
27    /// Avro row format
28    Avro,
29    /// ORC columnar format
30    Orc,
31    /// Excel (.xls, .xlsx, .xlsm, .xlsb)
32    Excel,
33}
34
35impl FileFormat {
36    /// Detect file format from path extension. Returns None when extension is missing or unknown.
37    pub fn from_path(path: &Path) -> Option<Self> {
38        path.extension()
39            .and_then(|e| e.to_str())
40            .and_then(Self::from_extension)
41    }
42
43    /// Parse format from extension string (e.g. "parquet", "csv").
44    pub fn from_extension(ext: &str) -> Option<Self> {
45        match ext.to_lowercase().as_str() {
46            "parquet" => Some(Self::Parquet),
47            "csv" => Some(Self::Csv),
48            "tsv" => Some(Self::Tsv),
49            "psv" => Some(Self::Psv),
50            "json" => Some(Self::Json),
51            "jsonl" | "ndjson" => Some(Self::Jsonl),
52            "arrow" | "ipc" | "feather" => Some(Self::Arrow),
53            "avro" => Some(Self::Avro),
54            "orc" => Some(Self::Orc),
55            "xls" | "xlsx" | "xlsm" | "xlsb" => Some(Self::Excel),
56            _ => None,
57        }
58    }
59}
60
61/// Compression format for data files
62#[derive(Debug, Clone, Copy, ValueEnum, PartialEq, Eq)]
63pub enum CompressionFormat {
64    /// Gzip compression (.gz) - Most common, good balance of speed and compression
65    Gzip,
66    /// Zstandard compression (.zst) - Modern, fast compression with good ratios
67    Zstd,
68    /// Bzip2 compression (.bz2) - Good compression ratio, slower than gzip
69    Bzip2,
70    /// XZ compression (.xz) - Excellent compression ratio, slower than bzip2
71    Xz,
72}
73
74impl CompressionFormat {
75    /// Detect compression format from file extension
76    pub fn from_extension(path: &Path) -> Option<Self> {
77        if let Some(ext) = path.extension().and_then(|e| e.to_str()) {
78            match ext.to_lowercase().as_str() {
79                "gz" => Some(Self::Gzip),
80                "zst" | "zstd" => Some(Self::Zstd),
81                "bz2" | "bz" => Some(Self::Bzip2),
82                "xz" => Some(Self::Xz),
83                _ => None,
84            }
85        } else {
86            None
87        }
88    }
89
90    /// Get file extension for this compression format
91    pub fn extension(&self) -> &'static str {
92        match self {
93            Self::Gzip => "gz",
94            Self::Zstd => "zst",
95            Self::Bzip2 => "bz2",
96            Self::Xz => "xz",
97        }
98    }
99}
100
101/// Accepted values for `--number-format`.
102///
103/// This crate cannot depend on datui-lib (the dependency runs the other way),
104/// so the list is duplicated here to give clap proper `--help` output and shell
105/// completion. `number_format_values_match_presets` in datui-lib asserts the two
106/// lists stay in sync.
107pub const NUMBER_FORMAT_VALUES: &[&str] = &[
108    "none",
109    "thousands",
110    "european",
111    "si",
112    "swiss",
113    "indian",
114    "underscore",
115    "system",
116];
117
118/// Command-line arguments for datui
119#[derive(Clone, Parser, Debug)]
120#[command(
121    name = "datui",
122    version,
123    about = "Data Exploration in the Terminal",
124    long_about = include_str!("../long_about.txt")
125)]
126pub struct Args {
127    /// Path(s) to the data file(s) to open.
128    /// Multiple files of the same format are concatenated into one table (not required with --generate-config, --clear-cache, or --remove-templates)
129    #[arg(required_unless_present_any = ["generate_config", "clear_cache", "remove_templates"], num_args = 1.., value_name = "PATH")]
130    pub paths: Vec<std::path::PathBuf>,
131
132    /// Skip this many lines when reading a file
133    #[arg(long = "skip-lines")]
134    pub skip_lines: Option<usize>,
135
136    /// Skip this many rows when reading a file
137    #[arg(long = "skip-rows")]
138    pub skip_rows: Option<usize>,
139
140    /// Skip this many rows at the end of the file (e.g. to ignore vendor footer or trailing garbage)
141    #[arg(long = "skip-tail-rows", value_name = "N")]
142    pub skip_tail_rows: Option<usize>,
143
144    /// Specify that the file has no header
145    #[arg(long = "no-header")]
146    pub no_header: Option<bool>,
147
148    /// Specify the delimiter to use when reading a delimited text file
149    #[arg(long = "delimiter")]
150    pub delimiter: Option<u8>,
151
152    /// Number of rows to use when inferring CSV schema (default: 1000). Larger values reduce risk of wrong type (e.g. int then N/A).
153    #[arg(long = "infer-schema-length", value_name = "N")]
154    pub infer_schema_length: Option<usize>,
155
156    /// When reading CSV, ignore parse errors and continue with the next batch (default: false)
157    #[arg(long = "ignore-errors", value_name = "BOOL", value_parser = clap::value_parser!(bool))]
158    pub ignore_errors: Option<bool>,
159
160    /// Treat these values as null when reading CSV. Use once per value; no "=" means all columns, COL=VAL means column COL only (first "=" separates column from value). Example: --null-value NA --null-value amount=
161    #[arg(long = "null-value", value_name = "VAL")]
162    pub null_value: Vec<String>,
163
164    /// Specify the compression format explicitly (gzip, zstd, bzip2, xz)
165    /// If not specified, compression is auto-detected from file extension.
166    #[arg(long = "compression", value_enum)]
167    pub compression: Option<CompressionFormat>,
168
169    /// Force file format (parquet, csv, tsv, psv, json, jsonl, arrow, avro, orc, excel).
170    /// By default format is auto-detected from the file extension. Use this for URLs or paths without an extension.
171    #[arg(long = "format", value_enum)]
172    pub format: Option<FileFormat>,
173
174    /// Enable debug mode to show operational information
175    #[arg(long = "debug", action)]
176    pub debug: bool,
177
178    /// Enable Hive-style partitioning for directory or glob paths; ignored for a single file
179    #[arg(long = "hive", action)]
180    pub hive: bool,
181
182    /// Infer Hive/partitioned Parquet schema from one file for faster load (default: true). Set to false to use full schema scan.
183    #[arg(long = "single-spine-schema", value_name = "BOOL", value_parser = clap::value_parser!(bool))]
184    pub single_spine_schema: Option<bool>,
185
186    /// Try to parse CSV string columns as dates (e.g. YYYY-MM-DD, ISO datetime). Default: true
187    #[arg(long = "parse-dates", value_name = "BOOL", value_parser = clap::value_parser!(bool))]
188    pub parse_dates: Option<bool>,
189
190    /// Trim whitespace and parse CSV string columns as date, datetime, time, duration, int, or float. Default: applied to all string columns. Use --parse-strings=COL (repeatable) to limit to specific columns, or --no-parse-strings to disable.
191    #[arg(long = "parse-strings", value_name = "COL", num_args = 0.., default_missing_value = "")]
192    pub parse_strings: Vec<String>,
193
194    /// Disable parse-strings for CSV (trim and type inference). Overrides config and default.
195    #[arg(long = "no-parse-strings", action)]
196    pub no_parse_strings: bool,
197
198    /// Decompress into memory. Default: decompress to temp file and use lazy scan
199    #[arg(long = "decompress-in-memory", default_missing_value = "true", num_args = 0..=1, value_parser = clap::value_parser!(bool))]
200    pub decompress_in_memory: Option<bool>,
201
202    /// Directory for decompression temp files (default: system temp, e.g. TMPDIR)
203    #[arg(long = "temp-dir", value_name = "DIR")]
204    pub temp_dir: Option<std::path::PathBuf>,
205
206    /// Excel sheet to load: 0-based index (e.g. 0) or sheet name (e.g. "Sales")
207    #[arg(long = "sheet", value_name = "SHEET")]
208    pub excel_sheet: Option<String>,
209
210    /// Clear all cache data and exit
211    #[arg(long = "clear-cache", action)]
212    pub clear_cache: bool,
213
214    /// Apply a template by name when starting the application
215    #[arg(long = "template")]
216    pub template: Option<String>,
217
218    /// Remove all templates and exit
219    #[arg(long = "remove-templates", action)]
220    pub remove_templates: bool,
221
222    /// When set, datasets with this many or more rows are sampled for analysis (faster, less memory).
223    /// Overrides config [performance] sampling_threshold. Use 0 to disable sampling (full dataset) for this run.
224    /// When omitted, config or full-dataset mode is used.
225    #[arg(long = "sampling-threshold", value_name = "N")]
226    pub sampling_threshold: Option<usize>,
227
228    /// Use Polars streaming engine for LazyFrame collect when available (default: true). Set to false to disable.
229    #[arg(long = "polars-streaming", value_name = "BOOL", value_parser = clap::value_parser!(bool))]
230    pub polars_streaming: Option<bool>,
231
232    /// Apply workaround for Polars 0.52 pivot with Date/Datetime index (default: true). Set to false to test without it.
233    #[arg(long = "workaround-pivot-date-index", value_name = "BOOL", value_parser = clap::value_parser!(bool))]
234    pub workaround_pivot_date_index: Option<bool>,
235
236    /// Number of pages to buffer ahead of the visible area (default: 3)
237    /// Larger values provide smoother scrolling but use more memory
238    #[arg(long = "pages-lookahead")]
239    pub pages_lookahead: Option<usize>,
240
241    /// Number of pages to buffer behind the visible area (default: 3)
242    /// Larger values provide smoother scrolling but use more memory
243    #[arg(long = "pages-lookback")]
244    pub pages_lookback: Option<usize>,
245
246    /// Display row numbers on the left side of the table
247    #[arg(long = "row-numbers", action)]
248    pub row_numbers: bool,
249
250    /// Starting index for row numbers (default: 1)
251    #[arg(long = "row-start-index")]
252    pub row_start_index: Option<usize>,
253
254    /// Colorize main table cells by column type (default: true). Set to false to disable.
255    #[arg(long = "column-colors", value_name = "BOOL", value_parser = clap::value_parser!(bool))]
256    pub column_colors: Option<bool>,
257
258    /// Digit grouping for numbers in the data table (default: none). Press F to toggle while running.
259    /// "system" reads LC_ALL/LC_NUMERIC/LANG and picks a matching style.
260    #[arg(long = "number-format", value_name = "FORMAT", value_parser = clap::builder::PossibleValuesParser::new(NUMBER_FORMAT_VALUES))]
261    pub number_format: Option<String>,
262
263    /// Right-align numeric columns and their headers (default: true). Set to false to left-align.
264    #[arg(long = "align-numeric-right", value_name = "BOOL", value_parser = clap::value_parser!(bool))]
265    pub align_numeric_right: Option<bool>,
266
267    /// Generate default configuration file at ~/.config/datui/config.toml
268    #[arg(long = "generate-config", action)]
269    pub generate_config: bool,
270
271    /// Force overwrite existing config file when using --generate-config
272    #[arg(long = "force", requires = "generate_config", action)]
273    pub force: bool,
274
275    /// S3-compatible endpoint URL (overrides config and AWS_ENDPOINT_URL). Example: http://localhost:9000
276    #[arg(long = "s3-endpoint-url", value_name = "URL")]
277    pub s3_endpoint_url: Option<String>,
278
279    /// S3 access key (overrides config and AWS_ACCESS_KEY_ID)
280    #[arg(long = "s3-access-key-id", value_name = "KEY")]
281    pub s3_access_key_id: Option<String>,
282
283    /// S3 secret key (overrides config and AWS_SECRET_ACCESS_KEY)
284    #[arg(long = "s3-secret-access-key", value_name = "SECRET")]
285    pub s3_secret_access_key: Option<String>,
286
287    /// S3 region (overrides config and AWS_REGION). Example: us-east-1
288    #[arg(long = "s3-region", value_name = "REGION")]
289    pub s3_region: Option<String>,
290}
291
292/// Escape `|` and newlines for use in markdown table cells.
293fn escape_table_cell(s: &str) -> String {
294    s.replace('|', "\\|").replace(['\n', '\r'], " ")
295}
296
297/// Render command-line options as markdown.
298///
299/// Used by the gen_docs binary; output is written to stdout and then
300/// to `docs/reference/command-line-options.md` by the docs build process.
301pub fn render_options_markdown() -> String {
302    let mut cmd = Args::command();
303    cmd.build();
304
305    let mut out = String::from("# Command Line Options\n\n");
306
307    out.push_str("## Usage\n\n```\n");
308    let usage = cmd.render_usage();
309    out.push_str(&usage.to_string());
310    out.push_str("\n```\n\n");
311
312    out.push_str("## Options\n\n");
313    out.push_str("| Option | Description |\n");
314    out.push_str("|--------|-------------|\n");
315
316    for arg in cmd.get_arguments() {
317        let id = arg.get_id().as_ref().to_string();
318        if id == "help" || id == "version" {
319            continue;
320        }
321
322        let option_str = if arg.is_positional() {
323            let placeholder: String = arg
324                .get_value_names()
325                .map(|names| {
326                    names
327                        .iter()
328                        .map(|n: &clap::builder::Str| format!("<{}>", n.as_ref() as &str))
329                        .collect::<Vec<_>>()
330                        .join(" ")
331                })
332                .unwrap_or_default();
333            if arg.is_required_set() {
334                placeholder
335            } else {
336                format!("[{placeholder}]")
337            }
338        } else {
339            let mut parts = Vec::new();
340            if let Some(s) = arg.get_short() {
341                parts.push(format!("-{s}"));
342            }
343            if let Some(l) = arg.get_long() {
344                parts.push(format!("--{l}"));
345            }
346            let op = parts.join(", ");
347            let takes_val = arg.get_action().takes_values();
348            let placeholder: String = if takes_val {
349                arg.get_value_names()
350                    .map(|names| {
351                        names
352                            .iter()
353                            .map(|n: &clap::builder::Str| format!("<{}>", n.as_ref() as &str))
354                            .collect::<Vec<_>>()
355                            .join(" ")
356                    })
357                    .unwrap_or_default()
358            } else {
359                String::new()
360            };
361            if placeholder.is_empty() {
362                op
363            } else {
364                format!("{op} {placeholder}")
365            }
366        };
367
368        let help = arg
369            .get_help()
370            .map(|h| escape_table_cell(&h.to_string()))
371            .unwrap_or_else(|| "-".to_string());
372
373        out.push_str(&format!("| `{option_str}` | {help} |\n"));
374    }
375
376    out
377}
378
379#[cfg(test)]
380mod tests {
381    use super::*;
382
383    #[test]
384    fn test_compression_detection() {
385        assert_eq!(
386            CompressionFormat::from_extension(Path::new("file.csv.gz")),
387            Some(CompressionFormat::Gzip)
388        );
389        assert_eq!(
390            CompressionFormat::from_extension(Path::new("file.csv.zst")),
391            Some(CompressionFormat::Zstd)
392        );
393        assert_eq!(
394            CompressionFormat::from_extension(Path::new("file.csv.bz2")),
395            Some(CompressionFormat::Bzip2)
396        );
397        assert_eq!(
398            CompressionFormat::from_extension(Path::new("file.csv.xz")),
399            Some(CompressionFormat::Xz)
400        );
401        assert_eq!(
402            CompressionFormat::from_extension(Path::new("file.csv")),
403            None
404        );
405        assert_eq!(CompressionFormat::from_extension(Path::new("file")), None);
406    }
407
408    #[test]
409    fn test_compression_extension() {
410        assert_eq!(CompressionFormat::Gzip.extension(), "gz");
411        assert_eq!(CompressionFormat::Zstd.extension(), "zst");
412        assert_eq!(CompressionFormat::Bzip2.extension(), "bz2");
413        assert_eq!(CompressionFormat::Xz.extension(), "xz");
414    }
415
416    #[test]
417    fn test_file_format_from_path() {
418        assert_eq!(
419            FileFormat::from_path(Path::new("data.parquet")),
420            Some(FileFormat::Parquet)
421        );
422        assert_eq!(
423            FileFormat::from_path(Path::new("data.csv")),
424            Some(FileFormat::Csv)
425        );
426        assert_eq!(
427            FileFormat::from_path(Path::new("file.jsonl")),
428            Some(FileFormat::Jsonl)
429        );
430        assert_eq!(FileFormat::from_path(Path::new("noext")), None);
431        assert_eq!(
432            FileFormat::from_path(Path::new("file.NDJSON")),
433            Some(FileFormat::Jsonl)
434        );
435    }
436}
datui_cli/lib.rs

datui_cli/
lib.rs
