mirror of
https://github.com/uutils/coreutils
synced 2024-12-14 15:22:38 +00:00
Merge pull request #5252 from zhitkoff/split-gnu-test-fail.sh
split: implement remaining -n variants and pass GNU tests/split/fail.sh
This commit is contained in:
Sylvestre Ledru
GitHub
commit
80f8eb6724
3 changed files with 947 additions and 59 deletions
|
|
@ -13,14 +13,16 @@ use crate::filenames::FilenameIterator;
|
|||
use crate::filenames::SuffixType;
|
||||
use clap::{crate_version, parser::ValueSource, Arg, ArgAction, ArgMatches, Command};
|
||||
use std::env;
|
||||
use std::ffi::OsString;
|
||||
use std::fmt;
|
||||
use std::fs::{metadata, File};
|
||||
use std::io;
|
||||
use std::io::{stdin, BufRead, BufReader, BufWriter, ErrorKind, Read, Write};
|
||||
use std::path::Path;
|
||||
use std::u64;
|
||||
use uucore::display::Quotable;
|
||||
use uucore::error::{FromIo, UIoError, UResult, USimpleError, UUsageError};
|
||||
use uucore::parse_size::{parse_size, ParseSizeError};
|
||||
use uucore::parse_size::{parse_size, parse_size_max, ParseSizeError};
|
||||
use uucore::uio_error;
|
||||
use uucore::{format_usage, help_about, help_section, help_usage};
|
||||
|
||||
|
|
@ -52,14 +54,175 @@ const AFTER_HELP: &str = help_section!("after help", "split.md");
|
|||
|
||||
#[uucore::main]
|
||||
pub fn uumain(args: impl uucore::Args) -> UResult<()> {
|
||||
let (args, obs_lines) = handle_obsolete(args);
|
||||
|
||||
let matches = uu_app().try_get_matches_from(args)?;
|
||||
match Settings::from(&matches) {
|
||||
|
||||
match Settings::from(&matches, &obs_lines) {
|
||||
Ok(settings) => split(&settings),
|
||||
Err(e) if e.requires_usage() => Err(UUsageError::new(1, format!("{e}"))),
|
||||
Err(e) => Err(USimpleError::new(1, format!("{e}"))),
|
||||
}
|
||||
}
|
||||
|
||||
/// Extract obsolete shorthand (if any) for specifying lines in following scenarios (and similar)
|
||||
/// `split -22 file` would mean `split -l 22 file`
|
||||
/// `split -2de file` would mean `split -l 2 -d -e file`
|
||||
/// `split -x300e file` would mean `split -x -l 300 -e file`
|
||||
/// `split -x300e -22 file` would mean `split -x -e -l 22 file` (last obsolete lines option wins)
|
||||
/// following GNU `split` behavior
|
||||
fn handle_obsolete(args: impl uucore::Args) -> (Vec<OsString>, Option<String>) {
|
||||
let mut obs_lines = None;
|
||||
let mut preceding_long_opt_req_value = false;
|
||||
let mut preceding_short_opt_req_value = false;
|
||||
|
||||
let filtered_args = args
|
||||
.filter_map(|os_slice| {
|
||||
filter_args(
|
||||
os_slice,
|
||||
&mut obs_lines,
|
||||
&mut preceding_long_opt_req_value,
|
||||
&mut preceding_short_opt_req_value,
|
||||
)
|
||||
})
|
||||
.collect();
|
||||
|
||||
(filtered_args, obs_lines)
|
||||
}
|
||||
|
||||
/// Helper function to [`handle_obsolete`]
|
||||
/// Filters out obsolete lines option from args
|
||||
fn filter_args(
|
||||
os_slice: OsString,
|
||||
obs_lines: &mut Option<String>,
|
||||
preceding_long_opt_req_value: &mut bool,
|
||||
preceding_short_opt_req_value: &mut bool,
|
||||
) -> Option<OsString> {
|
||||
let filter: Option<OsString>;
|
||||
if let Some(slice) = os_slice.to_str() {
|
||||
if should_extract_obs_lines(
|
||||
slice,
|
||||
preceding_long_opt_req_value,
|
||||
preceding_short_opt_req_value,
|
||||
) {
|
||||
// start of the short option string
|
||||
// that can have obsolete lines option value in it
|
||||
filter = handle_extract_obs_lines(slice, obs_lines);
|
||||
} else {
|
||||
// either not a short option
|
||||
// or a short option that cannot have obsolete lines value in it
|
||||
filter = Some(OsString::from(slice));
|
||||
}
|
||||
handle_preceding_options(
|
||||
slice,
|
||||
preceding_long_opt_req_value,
|
||||
preceding_short_opt_req_value,
|
||||
);
|
||||
} else {
|
||||
// Cannot cleanly convert os_slice to UTF-8
|
||||
// Do not process and return as-is
|
||||
// This will cause failure later on, but we should not handle it here
|
||||
// and let clap panic on invalid UTF-8 argument
|
||||
filter = Some(os_slice);
|
||||
}
|
||||
filter
|
||||
}
|
||||
|
||||
/// Helper function to [`filter_args`]
|
||||
/// Checks if the slice is a true short option (and not hyphen prefixed value of an option)
|
||||
/// and if so, a short option that can contain obsolete lines value
|
||||
fn should_extract_obs_lines(
|
||||
slice: &str,
|
||||
preceding_long_opt_req_value: &bool,
|
||||
preceding_short_opt_req_value: &bool,
|
||||
) -> bool {
|
||||
slice.starts_with('-')
|
||||
&& !slice.starts_with("--")
|
||||
&& !preceding_long_opt_req_value
|
||||
&& !preceding_short_opt_req_value
|
||||
&& !slice.starts_with("-a")
|
||||
&& !slice.starts_with("-b")
|
||||
&& !slice.starts_with("-C")
|
||||
&& !slice.starts_with("-l")
|
||||
&& !slice.starts_with("-n")
|
||||
}
|
||||
|
||||
/// Helper function to [`filter_args`]
|
||||
/// Extracts obsolete lines numeric part from argument slice
|
||||
/// and filters it out
|
||||
fn handle_extract_obs_lines(slice: &str, obs_lines: &mut Option<String>) -> Option<OsString> {
|
||||
let mut obs_lines_extracted: Vec<char> = vec![];
|
||||
let mut obs_lines_end_reached = false;
|
||||
let filtered_slice: Vec<char> = slice
|
||||
.chars()
|
||||
.filter(|c| {
|
||||
// To correctly process scenario like '-x200a4'
|
||||
// we need to stop extracting digits once alphabetic character is encountered
|
||||
// after we already have something in obs_lines_extracted
|
||||
if c.is_ascii_digit() && !obs_lines_end_reached {
|
||||
obs_lines_extracted.push(*c);
|
||||
false
|
||||
} else {
|
||||
if !obs_lines_extracted.is_empty() {
|
||||
obs_lines_end_reached = true;
|
||||
}
|
||||
true
|
||||
}
|
||||
})
|
||||
.collect();
|
||||
|
||||
if obs_lines_extracted.is_empty() {
|
||||
// no obsolete lines value found/extracted
|
||||
Some(OsString::from(slice))
|
||||
} else {
|
||||
// obsolete lines value was extracted
|
||||
let extracted: String = obs_lines_extracted.iter().collect();
|
||||
*obs_lines = Some(extracted);
|
||||
if filtered_slice.get(1).is_some() {
|
||||
// there were some short options in front of or after obsolete lines value
|
||||
// i.e. '-xd100' or '-100de' or similar, which after extraction of obsolete lines value
|
||||
// would look like '-xd' or '-de' or similar
|
||||
let filtered_slice: String = filtered_slice.iter().collect();
|
||||
Some(OsString::from(filtered_slice))
|
||||
} else {
|
||||
None
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
/// Helper function to [`handle_extract_obs_lines`]
|
||||
/// Captures if current slice is a preceding option
|
||||
/// that requires value
|
||||
fn handle_preceding_options(
|
||||
slice: &str,
|
||||
preceding_long_opt_req_value: &mut bool,
|
||||
preceding_short_opt_req_value: &mut bool,
|
||||
) {
|
||||
// capture if current slice is a preceding long option that requires value and does not use '=' to assign that value
|
||||
// following slice should be treaded as value for this option
|
||||
// even if it starts with '-' (which would be treated as hyphen prefixed value)
|
||||
if slice.starts_with("--") {
|
||||
*preceding_long_opt_req_value = &slice[2..] == OPT_BYTES
|
||||
|| &slice[2..] == OPT_LINE_BYTES
|
||||
|| &slice[2..] == OPT_LINES
|
||||
|| &slice[2..] == OPT_ADDITIONAL_SUFFIX
|
||||
|| &slice[2..] == OPT_FILTER
|
||||
|| &slice[2..] == OPT_NUMBER
|
||||
|| &slice[2..] == OPT_SUFFIX_LENGTH;
|
||||
}
|
||||
// capture if current slice is a preceding short option that requires value and does not have value in the same slice (value separated by whitespace)
|
||||
// following slice should be treaded as value for this option
|
||||
// even if it starts with '-' (which would be treated as hyphen prefixed value)
|
||||
*preceding_short_opt_req_value =
|
||||
slice == "-b" || slice == "-C" || slice == "-l" || slice == "-n" || slice == "-a";
|
||||
// slice is a value
|
||||
// reset preceding option flags
|
||||
if !slice.starts_with('-') {
|
||||
*preceding_short_opt_req_value = false;
|
||||
*preceding_long_opt_req_value = false;
|
||||
}
|
||||
}
|
||||
|
||||
pub fn uu_app() -> Command {
|
||||
Command::new(uucore::util_name())
|
||||
.version(crate_version!())
|
||||
|
|
@ -72,6 +235,7 @@ pub fn uu_app() -> Command {
|
|||
Arg::new(OPT_BYTES)
|
||||
.short('b')
|
||||
.long(OPT_BYTES)
|
||||
.allow_hyphen_values(true)
|
||||
.value_name("SIZE")
|
||||
.help("put SIZE bytes per output file"),
|
||||
)
|
||||
|
|
@ -79,14 +243,15 @@ pub fn uu_app() -> Command {
|
|||
Arg::new(OPT_LINE_BYTES)
|
||||
.short('C')
|
||||
.long(OPT_LINE_BYTES)
|
||||
.allow_hyphen_values(true)
|
||||
.value_name("SIZE")
|
||||
.default_value("2")
|
||||
.help("put at most SIZE bytes of lines per output file"),
|
||||
)
|
||||
.arg(
|
||||
Arg::new(OPT_LINES)
|
||||
.short('l')
|
||||
.long(OPT_LINES)
|
||||
.allow_hyphen_values(true)
|
||||
.value_name("NUMBER")
|
||||
.default_value("1000")
|
||||
.help("put NUMBER lines/records per output file"),
|
||||
|
|
@ -95,6 +260,7 @@ pub fn uu_app() -> Command {
|
|||
Arg::new(OPT_NUMBER)
|
||||
.short('n')
|
||||
.long(OPT_NUMBER)
|
||||
.allow_hyphen_values(true)
|
||||
.value_name("CHUNKS")
|
||||
.help("generate CHUNKS output files; see explanation below"),
|
||||
)
|
||||
|
|
@ -102,6 +268,7 @@ pub fn uu_app() -> Command {
|
|||
.arg(
|
||||
Arg::new(OPT_ADDITIONAL_SUFFIX)
|
||||
.long(OPT_ADDITIONAL_SUFFIX)
|
||||
.allow_hyphen_values(true)
|
||||
.value_name("SUFFIX")
|
||||
.default_value("")
|
||||
.help("additional SUFFIX to append to output file names"),
|
||||
|
|
@ -109,6 +276,7 @@ pub fn uu_app() -> Command {
|
|||
.arg(
|
||||
Arg::new(OPT_FILTER)
|
||||
.long(OPT_FILTER)
|
||||
.allow_hyphen_values(true)
|
||||
.value_name("COMMAND")
|
||||
.value_hint(clap::ValueHint::CommandName)
|
||||
.help(
|
||||
|
|
@ -178,9 +346,10 @@ pub fn uu_app() -> Command {
|
|||
Arg::new(OPT_SUFFIX_LENGTH)
|
||||
.short('a')
|
||||
.long(OPT_SUFFIX_LENGTH)
|
||||
.allow_hyphen_values(true)
|
||||
.value_name("N")
|
||||
.default_value(OPT_DEFAULT_SUFFIX_LENGTH)
|
||||
.help("use suffixes of fixed length N. 0 implies dynamic length."),
|
||||
.help("use suffixes of fixed length N. 0 implies dynamic length, starting with 2"),
|
||||
)
|
||||
.arg(
|
||||
Arg::new(OPT_VERBOSE)
|
||||
|
|
@ -217,6 +386,10 @@ enum NumberType {
|
|||
/// Split into a specific number of chunks by byte.
|
||||
Bytes(u64),
|
||||
|
||||
/// Split into a specific number of chunks by byte
|
||||
/// but output only the *k*th chunk.
|
||||
KthBytes(u64, u64),
|
||||
|
||||
/// Split into a specific number of chunks by line (approximately).
|
||||
Lines(u64),
|
||||
|
||||
|
|
@ -237,6 +410,7 @@ impl NumberType {
|
|||
fn num_chunks(&self) -> u64 {
|
||||
match self {
|
||||
Self::Bytes(n) => *n,
|
||||
Self::KthBytes(_, n) => *n,
|
||||
Self::Lines(n) => *n,
|
||||
Self::KthLines(_, n) => *n,
|
||||
Self::RoundRobin(n) => *n,
|
||||
|
|
@ -255,6 +429,7 @@ enum NumberTypeError {
|
|||
///
|
||||
/// ```ignore
|
||||
/// -n N
|
||||
/// -n K/N
|
||||
/// -n l/N
|
||||
/// -n l/K/N
|
||||
/// -n r/N
|
||||
|
|
@ -265,9 +440,12 @@ enum NumberTypeError {
|
|||
/// The chunk number was invalid.
|
||||
///
|
||||
/// This can happen if the value of `K` in any of the following
|
||||
/// command-line options is not a positive integer:
|
||||
/// command-line options is not a positive integer
|
||||
/// or if `K` is 0
|
||||
/// or if `K` is greater than `N`:
|
||||
///
|
||||
/// ```ignore
|
||||
/// -n K/N
|
||||
/// -n l/K/N
|
||||
/// -n r/K/N
|
||||
/// ```
|
||||
|
|
@ -281,6 +459,7 @@ impl NumberType {
|
|||
///
|
||||
/// ```ignore
|
||||
/// "N"
|
||||
/// "K/N"
|
||||
/// "l/N"
|
||||
/// "l/K/N"
|
||||
/// "r/N"
|
||||
|
|
@ -292,15 +471,20 @@ impl NumberType {
|
|||
///
|
||||
/// # Errors
|
||||
///
|
||||
/// If the string is not one of the valid number types, if `K` is
|
||||
/// not a nonnegative integer, or if `N` is not a positive
|
||||
/// integer, then this function returns [`NumberTypeError`].
|
||||
/// If the string is not one of the valid number types,
|
||||
/// if `K` is not a nonnegative integer,
|
||||
/// or if `K` is 0,
|
||||
/// or if `N` is not a positive integer,
|
||||
/// or if `K` is greater than `N`
|
||||
/// then this function returns [`NumberTypeError`].
|
||||
fn from(s: &str) -> Result<Self, NumberTypeError> {
|
||||
fn is_invalid_chunk(chunk_number: u64, num_chunks: u64) -> bool {
|
||||
chunk_number > num_chunks || chunk_number == 0
|
||||
}
|
||||
let parts: Vec<&str> = s.split('/').collect();
|
||||
match &parts[..] {
|
||||
[n_str] => {
|
||||
let num_chunks = n_str
|
||||
.parse()
|
||||
let num_chunks = parse_size(n_str)
|
||||
.map_err(|_| NumberTypeError::NumberOfChunks(n_str.to_string()))?;
|
||||
if num_chunks > 0 {
|
||||
Ok(Self::Bytes(num_chunks))
|
||||
|
|
@ -308,34 +492,44 @@ impl NumberType {
|
|||
Err(NumberTypeError::NumberOfChunks(s.to_string()))
|
||||
}
|
||||
}
|
||||
[k_str, n_str] if !k_str.starts_with('l') && !k_str.starts_with('r') => {
|
||||
let num_chunks = parse_size(n_str)
|
||||
.map_err(|_| NumberTypeError::NumberOfChunks(n_str.to_string()))?;
|
||||
let chunk_number = parse_size(k_str)
|
||||
.map_err(|_| NumberTypeError::ChunkNumber(k_str.to_string()))?;
|
||||
if is_invalid_chunk(chunk_number, num_chunks) {
|
||||
return Err(NumberTypeError::ChunkNumber(k_str.to_string()));
|
||||
}
|
||||
Ok(Self::KthBytes(chunk_number, num_chunks))
|
||||
}
|
||||
["l", n_str] => {
|
||||
let num_chunks = n_str
|
||||
.parse()
|
||||
let num_chunks = parse_size(n_str)
|
||||
.map_err(|_| NumberTypeError::NumberOfChunks(n_str.to_string()))?;
|
||||
Ok(Self::Lines(num_chunks))
|
||||
}
|
||||
["l", k_str, n_str] => {
|
||||
let num_chunks = n_str
|
||||
.parse()
|
||||
let num_chunks = parse_size(n_str)
|
||||
.map_err(|_| NumberTypeError::NumberOfChunks(n_str.to_string()))?;
|
||||
let chunk_number = k_str
|
||||
.parse()
|
||||
let chunk_number = parse_size(k_str)
|
||||
.map_err(|_| NumberTypeError::ChunkNumber(k_str.to_string()))?;
|
||||
if is_invalid_chunk(chunk_number, num_chunks) {
|
||||
return Err(NumberTypeError::ChunkNumber(k_str.to_string()));
|
||||
}
|
||||
Ok(Self::KthLines(chunk_number, num_chunks))
|
||||
}
|
||||
["r", n_str] => {
|
||||
let num_chunks = n_str
|
||||
.parse()
|
||||
let num_chunks = parse_size(n_str)
|
||||
.map_err(|_| NumberTypeError::NumberOfChunks(n_str.to_string()))?;
|
||||
Ok(Self::RoundRobin(num_chunks))
|
||||
}
|
||||
["r", k_str, n_str] => {
|
||||
let num_chunks = n_str
|
||||
.parse()
|
||||
let num_chunks = parse_size(n_str)
|
||||
.map_err(|_| NumberTypeError::NumberOfChunks(n_str.to_string()))?;
|
||||
let chunk_number = k_str
|
||||
.parse()
|
||||
let chunk_number = parse_size(k_str)
|
||||
.map_err(|_| NumberTypeError::ChunkNumber(k_str.to_string()))?;
|
||||
if is_invalid_chunk(chunk_number, num_chunks) {
|
||||
return Err(NumberTypeError::ChunkNumber(k_str.to_string()));
|
||||
}
|
||||
Ok(Self::KthRoundRobin(chunk_number, num_chunks))
|
||||
}
|
||||
_ => Err(NumberTypeError::NumberOfChunks(s.to_string())),
|
||||
|
|
@ -395,7 +589,7 @@ impl fmt::Display for StrategyError {
|
|||
|
||||
impl Strategy {
|
||||
/// Parse a strategy from the command-line arguments.
|
||||
fn from(matches: &ArgMatches) -> Result<Self, StrategyError> {
|
||||
fn from(matches: &ArgMatches, obs_lines: &Option<String>) -> Result<Self, StrategyError> {
|
||||
fn get_and_parse(
|
||||
matches: &ArgMatches,
|
||||
option: &str,
|
||||
|
|
@ -403,7 +597,7 @@ impl Strategy {
|
|||
error: fn(ParseSizeError) -> StrategyError,
|
||||
) -> Result<Strategy, StrategyError> {
|
||||
let s = matches.get_one::<String>(option).unwrap();
|
||||
let n = parse_size(s).map_err(error)?;
|
||||
let n = parse_size_max(s).map_err(error)?;
|
||||
if n > 0 {
|
||||
Ok(strategy(n))
|
||||
} else {
|
||||
|
|
@ -413,28 +607,40 @@ impl Strategy {
|
|||
// Check that the user is not specifying more than one strategy.
|
||||
//
|
||||
// Note: right now, this exact behavior cannot be handled by
|
||||
// `ArgGroup` since `ArgGroup` considers a default value `Arg`
|
||||
// as "defined".
|
||||
// overrides_with_all() due to obsolete lines value option
|
||||
match (
|
||||
obs_lines,
|
||||
matches.value_source(OPT_LINES) == Some(ValueSource::CommandLine),
|
||||
matches.value_source(OPT_BYTES) == Some(ValueSource::CommandLine),
|
||||
matches.value_source(OPT_LINE_BYTES) == Some(ValueSource::CommandLine),
|
||||
matches.value_source(OPT_NUMBER) == Some(ValueSource::CommandLine),
|
||||
) {
|
||||
(false, false, false, false) => Ok(Self::Lines(1000)),
|
||||
(true, false, false, false) => {
|
||||
(Some(v), false, false, false, false) => {
|
||||
let v = parse_size_max(v).map_err(|_| {
|
||||
StrategyError::Lines(ParseSizeError::ParseFailure(v.to_string()))
|
||||
})?;
|
||||
if v > 0 {
|
||||
Ok(Self::Lines(v))
|
||||
} else {
|
||||
Err(StrategyError::Lines(ParseSizeError::ParseFailure(
|
||||
v.to_string(),
|
||||
)))
|
||||
}
|
||||
}
|
||||
(None, false, false, false, false) => Ok(Self::Lines(1000)),
|
||||
(None, true, false, false, false) => {
|
||||
get_and_parse(matches, OPT_LINES, Self::Lines, StrategyError::Lines)
|
||||
}
|
||||
(false, true, false, false) => {
|
||||
(None, false, true, false, false) => {
|
||||
get_and_parse(matches, OPT_BYTES, Self::Bytes, StrategyError::Bytes)
|
||||
}
|
||||
(false, false, true, false) => get_and_parse(
|
||||
(None, false, false, true, false) => get_and_parse(
|
||||
matches,
|
||||
OPT_LINE_BYTES,
|
||||
Self::LineBytes,
|
||||
StrategyError::Bytes,
|
||||
),
|
||||
(false, false, false, true) => {
|
||||
(None, false, false, false, true) => {
|
||||
let s = matches.get_one::<String>(OPT_NUMBER).unwrap();
|
||||
let number_type = NumberType::from(s).map_err(StrategyError::NumberType)?;
|
||||
Ok(Self::Number(number_type))
|
||||
|
|
@ -553,7 +759,7 @@ impl fmt::Display for SettingsError {
|
|||
|
||||
impl Settings {
|
||||
/// Parse a strategy from the command-line arguments.
|
||||
fn from(matches: &ArgMatches) -> Result<Self, SettingsError> {
|
||||
fn from(matches: &ArgMatches, obs_lines: &Option<String>) -> Result<Self, SettingsError> {
|
||||
let additional_suffix = matches
|
||||
.get_one::<String>(OPT_ADDITIONAL_SUFFIX)
|
||||
.unwrap()
|
||||
|
|
@ -561,7 +767,7 @@ impl Settings {
|
|||
if additional_suffix.contains('/') {
|
||||
return Err(SettingsError::SuffixContainsSeparator(additional_suffix));
|
||||
}
|
||||
let strategy = Strategy::from(matches).map_err(SettingsError::Strategy)?;
|
||||
let strategy = Strategy::from(matches, obs_lines).map_err(SettingsError::Strategy)?;
|
||||
let (suffix_type, suffix_start) = suffix_type_from(matches)?;
|
||||
let suffix_length_str = matches.get_one::<String>(OPT_SUFFIX_LENGTH).unwrap();
|
||||
let suffix_length: usize = suffix_length_str
|
||||
|
|
@ -1089,7 +1295,7 @@ where
|
|||
// If we would have written zero chunks of output, then terminate
|
||||
// immediately. This happens on `split -e -n 3 /dev/null`, for
|
||||
// example.
|
||||
if num_chunks == 0 {
|
||||
if num_chunks == 0 || num_bytes == 0 {
|
||||
return Ok(());
|
||||
}
|
||||
|
||||
|
|
@ -1144,6 +1350,93 @@ where
|
|||
}
|
||||
}
|
||||
|
||||
/// Print the k-th chunk of a file to stdout, splitting by byte.
|
||||
///
|
||||
/// This function is like [`split_into_n_chunks_by_byte`], but instead
|
||||
/// of writing each chunk to its own file, it only writes to stdout
|
||||
/// the contents of the chunk identified by `chunk_number`
|
||||
///
|
||||
/// # Errors
|
||||
///
|
||||
/// This function returns an error if there is a problem reading from
|
||||
/// `reader` or writing to stdout.
|
||||
fn kth_chunks_by_byte<R>(
|
||||
settings: &Settings,
|
||||
reader: &mut R,
|
||||
chunk_number: u64,
|
||||
num_chunks: u64,
|
||||
) -> UResult<()>
|
||||
where
|
||||
R: BufRead,
|
||||
{
|
||||
// Get the size of the input file in bytes and compute the number
|
||||
// of bytes per chunk.
|
||||
//
|
||||
// If the requested number of chunks exceeds the number of bytes
|
||||
// in the file - just write empty byte string to stdout
|
||||
// NOTE: the `elide_empty_files` parameter is ignored here
|
||||
// as we do not generate any files
|
||||
// and instead writing to stdout
|
||||
let metadata = metadata(&settings.input).map_err(|_| {
|
||||
USimpleError::new(1, format!("{}: cannot determine file size", settings.input))
|
||||
})?;
|
||||
|
||||
let num_bytes = metadata.len();
|
||||
// If input file is empty and we would have written zero chunks of output,
|
||||
// then terminate immediately.
|
||||
// This happens on `split -e -n 3 /dev/null`, for example.
|
||||
if num_bytes == 0 {
|
||||
return Ok(());
|
||||
}
|
||||
|
||||
// Write to stdout instead of to a file.
|
||||
let stdout = std::io::stdout();
|
||||
let mut writer = stdout.lock();
|
||||
|
||||
let chunk_size = (num_bytes / (num_chunks)).max(1);
|
||||
let mut num_bytes: usize = num_bytes.try_into().unwrap();
|
||||
|
||||
let mut i = 1;
|
||||
loop {
|
||||
let buf: &mut Vec<u8> = &mut vec![];
|
||||
if num_bytes > 0 {
|
||||
// Read `chunk_size` bytes from the reader into `buf`
|
||||
// except the last.
|
||||
//
|
||||
// The last chunk gets all remaining bytes so that if the number
|
||||
// of bytes in the input file was not evenly divisible by
|
||||
// `num_chunks`, we don't leave any bytes behind.
|
||||
let limit = {
|
||||
if i == num_chunks {
|
||||
num_bytes.try_into().unwrap()
|
||||
} else {
|
||||
chunk_size
|
||||
}
|
||||
};
|
||||
let n_bytes_read = reader.by_ref().take(limit).read_to_end(buf);
|
||||
match n_bytes_read {
|
||||
Ok(n_bytes) => {
|
||||
num_bytes -= n_bytes;
|
||||
}
|
||||
Err(error) => {
|
||||
return Err(USimpleError::new(
|
||||
1,
|
||||
format!("{}: cannot read from input : {}", settings.input, error),
|
||||
));
|
||||
}
|
||||
}
|
||||
if i == chunk_number {
|
||||
writer.write_all(buf)?;
|
||||
break;
|
||||
}
|
||||
i += 1;
|
||||
} else {
|
||||
break;
|
||||
}
|
||||
}
|
||||
Ok(())
|
||||
}
|
||||
|
||||
/// Split a file into a specific number of chunks by line.
|
||||
///
|
||||
/// This function always creates one output file for each chunk, even
|
||||
|
|
@ -1319,6 +1612,50 @@ where
|
|||
Ok(())
|
||||
}
|
||||
|
||||
/// Print the k-th chunk of a file, splitting by line, but
|
||||
/// assign lines via round-robin to the specified number of output
|
||||
/// chunks, but output only the *k*th chunk.
|
||||
///
|
||||
/// This function is like [`kth_chunk_by_line`], as it only writes to stdout and
|
||||
/// prints out only *k*th chunk
|
||||
/// It is also like [`split_into_n_chunks_by_line_round_robin`], as it is assigning chunks
|
||||
/// using round robin distribution
|
||||
///
|
||||
/// # Errors
|
||||
///
|
||||
/// This function returns an error if there is a problem reading from
|
||||
/// `reader` or writing to one of the output files.
|
||||
///
|
||||
/// # See also
|
||||
///
|
||||
/// * [`split_into_n_chunks_by_line_round_robin`], which splits its input in the
|
||||
/// same way, but writes each chunk to its own file.
|
||||
fn kth_chunk_by_line_round_robin<R>(
|
||||
_settings: &Settings,
|
||||
reader: &mut R,
|
||||
chunk_number: u64,
|
||||
num_chunks: u64,
|
||||
) -> UResult<()>
|
||||
where
|
||||
R: BufRead,
|
||||
{
|
||||
// Write to stdout instead of to a file.
|
||||
let stdout = std::io::stdout();
|
||||
let mut writer = stdout.lock();
|
||||
|
||||
let num_chunks: usize = num_chunks.try_into().unwrap();
|
||||
let chunk_number: usize = chunk_number.try_into().unwrap();
|
||||
for (i, line_result) in reader.lines().enumerate() {
|
||||
let line = line_result?;
|
||||
let bytes = line.as_bytes();
|
||||
if (i % num_chunks) == chunk_number {
|
||||
writer.write_all(bytes)?;
|
||||
writer.write_all(b"\n")?;
|
||||
}
|
||||
}
|
||||
Ok(())
|
||||
}
|
||||
|
||||
fn split(settings: &Settings) -> UResult<()> {
|
||||
let mut reader = BufReader::new(if settings.input == "-" {
|
||||
Box::new(stdin()) as Box<dyn Read>
|
||||
|
|
@ -1336,6 +1673,9 @@ fn split(settings: &Settings) -> UResult<()> {
|
|||
Strategy::Number(NumberType::Bytes(num_chunks)) => {
|
||||
split_into_n_chunks_by_byte(settings, &mut reader, num_chunks)
|
||||
}
|
||||
Strategy::Number(NumberType::KthBytes(chunk_number, num_chunks)) => {
|
||||
kth_chunks_by_byte(settings, &mut reader, chunk_number, num_chunks)
|
||||
}
|
||||
Strategy::Number(NumberType::Lines(num_chunks)) => {
|
||||
split_into_n_chunks_by_line(settings, &mut reader, num_chunks)
|
||||
}
|
||||
|
|
@ -1348,7 +1688,12 @@ fn split(settings: &Settings) -> UResult<()> {
|
|||
Strategy::Number(NumberType::RoundRobin(num_chunks)) => {
|
||||
split_into_n_chunks_by_line_round_robin(settings, &mut reader, num_chunks)
|
||||
}
|
||||
Strategy::Number(_) => Err(USimpleError::new(1, "-n mode not yet fully implemented")),
|
||||
Strategy::Number(NumberType::KthRoundRobin(chunk_number, num_chunks)) => {
|
||||
// The chunk number is given as a 1-indexed number, but it
|
||||
// is a little easier to deal with a 0-indexed number.
|
||||
let chunk_number = chunk_number - 1;
|
||||
kth_chunk_by_line_round_robin(settings, &mut reader, chunk_number, num_chunks)
|
||||
}
|
||||
Strategy::Lines(chunk_size) => {
|
||||
let mut writer = LineChunkWriter::new(chunk_size, settings)?;
|
||||
match std::io::copy(&mut reader, &mut writer) {
|
||||
|
|
@ -1451,6 +1796,18 @@ mod tests {
|
|||
NumberType::from("l/abc/456").unwrap_err(),
|
||||
NumberTypeError::ChunkNumber("abc".to_string())
|
||||
);
|
||||
assert_eq!(
|
||||
NumberType::from("l/456/123").unwrap_err(),
|
||||
NumberTypeError::ChunkNumber("456".to_string())
|
||||
);
|
||||
assert_eq!(
|
||||
NumberType::from("r/456/123").unwrap_err(),
|
||||
NumberTypeError::ChunkNumber("456".to_string())
|
||||
);
|
||||
assert_eq!(
|
||||
NumberType::from("456/123").unwrap_err(),
|
||||
NumberTypeError::ChunkNumber("456".to_string())
|
||||
);
|
||||
// In GNU split, the number of chunks get precedence:
|
||||
//
|
||||
// $ split -n l/abc/xyz
|
||||
|
|
@ -1486,6 +1843,7 @@ mod tests {
|
|||
#[test]
|
||||
fn test_number_type_num_chunks() {
|
||||
assert_eq!(NumberType::from("123").unwrap().num_chunks(), 123);
|
||||
assert_eq!(NumberType::from("123/456").unwrap().num_chunks(), 456);
|
||||
assert_eq!(NumberType::from("l/123").unwrap().num_chunks(), 123);
|
||||
assert_eq!(NumberType::from("l/123/456").unwrap().num_chunks(), 456);
|
||||
assert_eq!(NumberType::from("r/123").unwrap().num_chunks(), 123);
|
||||
|
|
|
|||
|
|
@ -7,6 +7,7 @@
|
|||
|
||||
use std::error::Error;
|
||||
use std::fmt;
|
||||
use std::num::IntErrorKind;
|
||||
|
||||
use crate::display::Quotable;
|
||||
|
||||
|
|
@ -201,8 +202,10 @@ impl<'parser> Parser<'parser> {
|
|||
radix: u32,
|
||||
original_size: &str,
|
||||
) -> Result<u64, ParseSizeError> {
|
||||
u64::from_str_radix(numeric_string, radix)
|
||||
.map_err(|_| ParseSizeError::ParseFailure(original_size.to_string()))
|
||||
u64::from_str_radix(numeric_string, radix).map_err(|e| match e.kind() {
|
||||
IntErrorKind::PosOverflow => ParseSizeError::size_too_big(original_size),
|
||||
_ => ParseSizeError::ParseFailure(original_size.to_string()),
|
||||
})
|
||||
}
|
||||
}
|
||||
|
||||
|
|
@ -232,6 +235,23 @@ pub fn parse_size(size: &str) -> Result<u64, ParseSizeError> {
|
|||
Parser::default().parse(size)
|
||||
}
|
||||
|
||||
/// Same as `parse_size()`, except returns `u64::MAX` on overflow
|
||||
/// GNU lib/coreutils include similar functionality
|
||||
/// and GNU test suite checks this behavior for some utils
|
||||
pub fn parse_size_max(size: &str) -> Result<u64, ParseSizeError> {
|
||||
let result = Parser::default().parse(size);
|
||||
match result {
|
||||
Ok(_) => result,
|
||||
Err(error) => {
|
||||
if let ParseSizeError::SizeTooBig(_) = error {
|
||||
Ok(u64::MAX)
|
||||
} else {
|
||||
Err(error)
|
||||
}
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
#[derive(Debug, PartialEq, Eq)]
|
||||
pub enum ParseSizeError {
|
||||
InvalidSuffix(String), // Suffix
|
||||
|
|
@ -392,6 +412,14 @@ mod tests {
|
|||
);
|
||||
}
|
||||
|
||||
#[test]
|
||||
#[cfg(not(target_pointer_width = "128"))]
|
||||
fn overflow_to_max_x64() {
|
||||
assert_eq!(Ok(u64::MAX), parse_size_max("18446744073709551616"));
|
||||
assert_eq!(Ok(u64::MAX), parse_size_max("10000000000000000000000"));
|
||||
assert_eq!(Ok(u64::MAX), parse_size_max("1Y"));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn invalid_suffix() {
|
||||
let test_strings = ["5mib", "1eb", "1H"];
|
||||
|
|
|
|||
|
|
@ -2,7 +2,7 @@
|
|||
//
|
||||
// For the full copyright and license information, please view the LICENSE
|
||||
// file that was distributed with this source code.
|
||||
// spell-checker:ignore xzaaa sixhundredfiftyonebytes ninetyonebytes threebytes asciilowercase fghij klmno pqrst uvwxyz fivelines twohundredfortyonebytes onehundredlines nbbbb dxen
|
||||
// spell-checker:ignore xzaaa sixhundredfiftyonebytes ninetyonebytes threebytes asciilowercase fghij klmno pqrst uvwxyz fivelines twohundredfortyonebytes onehundredlines nbbbb dxen ncccc
|
||||
|
||||
use crate::common::util::{AtPath, TestScenario};
|
||||
use rand::{thread_rng, Rng, SeedableRng};
|
||||
|
|
@ -170,6 +170,22 @@ fn test_split_str_prefixed_chunks_by_bytes() {
|
|||
assert_eq!(glob.collate(), at.read_bytes(name));
|
||||
}
|
||||
|
||||
/// Test short bytes option concatenated with value
|
||||
#[test]
|
||||
fn test_split_by_bytes_short_concatenated_with_value() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let name = "split_by_bytes_short_concatenated_with_value";
|
||||
RandomFile::new(&at, name).add_bytes(10000);
|
||||
ucmd.args(&["-b1000", name]).succeeds();
|
||||
|
||||
let glob = Glob::new(&at, ".", r"x[[:alpha:]][[:alpha:]]$");
|
||||
assert_eq!(glob.count(), 10);
|
||||
for filename in glob.collect() {
|
||||
assert_eq!(glob.directory.metadata(&filename).len(), 1000);
|
||||
}
|
||||
assert_eq!(glob.collate(), at.read_bytes(name));
|
||||
}
|
||||
|
||||
// This is designed to test what happens when the desired part size is not a
|
||||
// multiple of the buffer size and we hopefully don't overshoot the desired part
|
||||
// size.
|
||||
|
|
@ -238,6 +254,18 @@ fn test_additional_suffix_no_slash() {
|
|||
.usage_error("invalid suffix 'a/b', contains directory separator");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_split_additional_suffix_hyphen_value() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let name = "split_additional_suffix";
|
||||
RandomFile::new(&at, name).add_lines(2000);
|
||||
ucmd.args(&["--additional-suffix", "-300", name]).succeeds();
|
||||
|
||||
let glob = Glob::new(&at, ".", r"x[[:alpha:]][[:alpha:]]-300$");
|
||||
assert_eq!(glob.count(), 2);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name));
|
||||
}
|
||||
|
||||
// note: the test_filter* tests below are unix-only
|
||||
// windows support has been waived for now because of the difficulty of getting
|
||||
// the `cmd` call right
|
||||
|
|
@ -312,12 +340,292 @@ fn test_split_lines_number() {
|
|||
.succeeds()
|
||||
.no_stderr()
|
||||
.no_stdout();
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--lines", "0", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_only("split: invalid number of lines: 0\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-0", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_only("split: invalid number of lines: 0\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--lines", "2fb", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_only("split: invalid number of lines: '2fb'\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--lines", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_only("split: invalid number of lines: 'file'\n");
|
||||
}
|
||||
|
||||
/// Test short lines option with value concatenated
|
||||
#[test]
|
||||
fn test_split_lines_short_concatenated_with_value() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let name = "split_num_prefixed_chunks_by_lines";
|
||||
RandomFile::new(&at, name).add_lines(10000);
|
||||
ucmd.args(&["-l1000", name]).succeeds();
|
||||
|
||||
let glob = Glob::new(&at, ".", r"x[[:alpha:]][[:alpha:]]$");
|
||||
assert_eq!(glob.count(), 10);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name));
|
||||
}
|
||||
|
||||
/// Test for obsolete lines option standalone
|
||||
#[test]
|
||||
fn test_split_obs_lines_standalone() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let name = "obs-lines-standalone";
|
||||
RandomFile::new(&at, name).add_lines(4);
|
||||
ucmd.args(&["-2", name]).succeeds().no_stderr().no_stdout();
|
||||
let glob = Glob::new(&at, ".", r"x[[:alpha:]][[:alpha:]]$");
|
||||
assert_eq!(glob.count(), 2);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name));
|
||||
}
|
||||
|
||||
/// Test for obsolete lines option standalone overflow
|
||||
#[test]
|
||||
fn test_split_obs_lines_standalone_overflow() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let name = "obs-lines-standalone";
|
||||
RandomFile::new(&at, name).add_lines(4);
|
||||
ucmd.args(&["-99999999999999999991", name])
|
||||
.succeeds()
|
||||
.no_stderr()
|
||||
.no_stdout();
|
||||
let glob = Glob::new(&at, ".", r"x[[:alpha:]][[:alpha:]]$");
|
||||
assert_eq!(glob.count(), 1);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name));
|
||||
}
|
||||
|
||||
/// Test for obsolete lines option as part of invalid combined short options
|
||||
#[test]
|
||||
fn test_split_obs_lines_within_invalid_combined_shorts() {
|
||||
let scene = TestScenario::new(util_name!());
|
||||
let at = &scene.fixtures;
|
||||
at.touch("file");
|
||||
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-2fb", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("error: unexpected argument '-f' found\n");
|
||||
}
|
||||
|
||||
/// Test for obsolete lines option as part of combined short options
|
||||
#[test]
|
||||
fn test_split_obs_lines_within_combined_shorts() {
|
||||
let scene = TestScenario::new(util_name!());
|
||||
let at = &scene.fixtures;
|
||||
let name = "obs-lines-within-shorts";
|
||||
RandomFile::new(&at, name).add_lines(400);
|
||||
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-x200de", name])
|
||||
.succeeds()
|
||||
.no_stderr()
|
||||
.no_stdout();
|
||||
let glob = Glob::new(&at, ".", r"x\d\d$");
|
||||
assert_eq!(glob.count(), 2);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name))
|
||||
}
|
||||
|
||||
/// Test for obsolete lines option as part of combined short options with tailing suffix length with value
|
||||
#[test]
|
||||
fn test_split_obs_lines_within_combined_shorts_tailing_suffix_length() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let name = "obs-lines-combined-shorts-tailing-suffix-length";
|
||||
RandomFile::new(&at, name).add_lines(1000);
|
||||
ucmd.args(&["-d200a4", name]).succeeds();
|
||||
|
||||
let glob = Glob::new(&at, ".", r"x\d\d\d\d$");
|
||||
assert_eq!(glob.count(), 5);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name));
|
||||
}
|
||||
|
||||
/// Test for obsolete lines option starts as part of combined short options
|
||||
#[test]
|
||||
fn test_split_obs_lines_starts_combined_shorts() {
|
||||
let scene = TestScenario::new(util_name!());
|
||||
let at = &scene.fixtures;
|
||||
let name = "obs-lines-starts-shorts";
|
||||
RandomFile::new(&at, name).add_lines(400);
|
||||
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-200xd", name])
|
||||
.succeeds()
|
||||
.no_stderr()
|
||||
.no_stdout();
|
||||
let glob = Glob::new(&at, ".", r"x\d\d$");
|
||||
assert_eq!(glob.count(), 2);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name))
|
||||
}
|
||||
|
||||
/// Test for using both obsolete lines (standalone) option and short/long lines option simultaneously
|
||||
#[test]
|
||||
fn test_split_both_lines_and_obs_lines_standalone() {
|
||||
// This test will ensure that:
|
||||
// if both lines option '-l' or '--lines' (with value) and obsolete lines option '-100' are used - it fails
|
||||
// if standalone lines option is used incorrectly and treated as a hyphen prefixed value of other option - it fails
|
||||
let scene = TestScenario::new(util_name!());
|
||||
let at = &scene.fixtures;
|
||||
at.touch("file");
|
||||
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-l", "2", "-2", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: cannot split in more than one way\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--lines", "2", "-2", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: cannot split in more than one way\n");
|
||||
}
|
||||
|
||||
/// Test for using obsolete lines option incorrectly, so it is treated as a hyphen prefixed value of other option
|
||||
#[test]
|
||||
fn test_split_obs_lines_as_other_option_value() {
|
||||
// This test will ensure that:
|
||||
// if obsolete lines option is used incorrectly and treated as a hyphen prefixed value of other option - it fails
|
||||
let scene = TestScenario::new(util_name!());
|
||||
let at = &scene.fixtures;
|
||||
at.touch("file");
|
||||
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--lines", "-200", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: invalid number of lines: '-200'\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-l", "-200", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: invalid number of lines: '-200'\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-a", "-200", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: invalid suffix length: '-200'\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--suffix-length", "-d200e", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: invalid suffix length: '-d200e'\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-C", "-200", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: invalid number of bytes: '-200'\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--line-bytes", "-x200a4", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: invalid number of bytes: '-x200a4'\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-b", "-200", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: invalid number of bytes: '-200'\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--bytes", "-200xd", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: invalid number of bytes: '-200xd'\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-n", "-200", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: invalid number of chunks: -200\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--number", "-e200", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: invalid number of chunks: -e200\n");
|
||||
}
|
||||
|
||||
/// Test for using more than one obsolete lines option (standalone)
|
||||
/// last one wins
|
||||
#[test]
|
||||
fn test_split_multiple_obs_lines_standalone() {
|
||||
let scene = TestScenario::new(util_name!());
|
||||
let at = &scene.fixtures;
|
||||
let name = "multiple-obs-lines";
|
||||
RandomFile::new(&at, name).add_lines(400);
|
||||
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-3000", "-200", name])
|
||||
.succeeds()
|
||||
.no_stderr()
|
||||
.no_stdout();
|
||||
let glob = Glob::new(&at, ".", r"x[[:alpha:]][[:alpha:]]$");
|
||||
assert_eq!(glob.count(), 2);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name))
|
||||
}
|
||||
|
||||
/// Test for using more than one obsolete lines option within combined shorts
|
||||
/// last one wins
|
||||
#[test]
|
||||
fn test_split_multiple_obs_lines_within_combined() {
|
||||
let scene = TestScenario::new(util_name!());
|
||||
let at = &scene.fixtures;
|
||||
let name = "multiple-obs-lines";
|
||||
RandomFile::new(&at, name).add_lines(400);
|
||||
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-d5000x", "-e200d", name])
|
||||
.succeeds()
|
||||
.no_stderr()
|
||||
.no_stdout();
|
||||
let glob = Glob::new(&at, ".", r"x\d\d$");
|
||||
assert_eq!(glob.count(), 2);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name))
|
||||
}
|
||||
|
||||
/// Test for using both obsolete lines option within combined shorts with conflicting -n option simultaneously
|
||||
#[test]
|
||||
fn test_split_obs_lines_within_combined_with_number() {
|
||||
let scene = TestScenario::new(util_name!());
|
||||
let at = &scene.fixtures;
|
||||
at.touch("file");
|
||||
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-3dxen", "4", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: cannot split in more than one way\n");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["-dxe30n", "4", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_contains("split: cannot split in more than one way\n");
|
||||
}
|
||||
|
||||
#[test]
|
||||
|
|
@ -327,14 +635,6 @@ fn test_split_invalid_bytes_size() {
|
|||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_only("split: invalid number of bytes: '1024R'\n");
|
||||
#[cfg(not(target_pointer_width = "128"))]
|
||||
new_ucmd!()
|
||||
.args(&["-b", "1Y"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_only(
|
||||
"split: invalid number of bytes: '1Y': Value too large for defined data type\n",
|
||||
);
|
||||
#[cfg(target_pointer_width = "32")]
|
||||
{
|
||||
let sizes = ["1000G", "10T"];
|
||||
|
|
@ -345,19 +645,29 @@ fn test_split_invalid_bytes_size() {
|
|||
}
|
||||
|
||||
#[test]
|
||||
fn test_split_overflow_bytes_size() {
|
||||
#[cfg(not(target_pointer_width = "128"))]
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let name = "test_split_overflow_bytes_size";
|
||||
RandomFile::new(&at, name).add_bytes(1000);
|
||||
ucmd.args(&["-b", "1Y", name]).succeeds();
|
||||
let glob = Glob::new(&at, ".", r"x[[:alpha:]][[:alpha:]]$");
|
||||
assert_eq!(glob.count(), 1);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name));
|
||||
}
|
||||
|
||||
#[test]
|
||||
#[cfg(target_pointer_width = "32")]
|
||||
fn test_split_chunks_num_chunks_oversized_32() {
|
||||
#[cfg(target_pointer_width = "32")]
|
||||
{
|
||||
let scene = TestScenario::new(util_name!());
|
||||
let at = &scene.fixtures;
|
||||
at.touch("file");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--number", "5000000000", "file"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_only("split: Number of chunks too big\n");
|
||||
}
|
||||
let scene = TestScenario::new(util_name!());
|
||||
let at = &scene.fixtures;
|
||||
at.touch("file");
|
||||
scene
|
||||
.ucmd()
|
||||
.args(&["--number", "5000000000", "sixhundredfiftyonebytes.txt"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_only("split: Number of chunks too big\n");
|
||||
}
|
||||
|
||||
#[test]
|
||||
|
|
@ -369,6 +679,15 @@ fn test_split_stdin_num_chunks() {
|
|||
.stderr_only("split: -: cannot determine file size\n");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_split_stdin_num_kth_chunk() {
|
||||
new_ucmd!()
|
||||
.args(&["--number=1/2"])
|
||||
.fails()
|
||||
.code_is(1)
|
||||
.stderr_only("split: -: cannot determine file size\n");
|
||||
}
|
||||
|
||||
fn file_read(at: &AtPath, filename: &str) -> String {
|
||||
let mut s = String::new();
|
||||
at.open(filename).read_to_string(&mut s).unwrap();
|
||||
|
|
@ -471,7 +790,7 @@ creating file 'xaf'
|
|||
}
|
||||
|
||||
#[test]
|
||||
fn test_number() {
|
||||
fn test_number_n() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let file_read = |f| {
|
||||
let mut s = String::new();
|
||||
|
|
@ -484,6 +803,98 @@ fn test_number() {
|
|||
assert_eq!(file_read("xac"), "klmno");
|
||||
assert_eq!(file_read("xad"), "pqrst");
|
||||
assert_eq!(file_read("xae"), "uvwxyz\n");
|
||||
#[cfg(unix)]
|
||||
new_ucmd!()
|
||||
.args(&["--number=100", "/dev/null"])
|
||||
.succeeds()
|
||||
.stdout_only("");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_number_kth_of_n() {
|
||||
new_ucmd!()
|
||||
.args(&["--number=3/5", "asciilowercase.txt"])
|
||||
.succeeds()
|
||||
.stdout_only("klmno");
|
||||
new_ucmd!()
|
||||
.args(&["--number=5/5", "asciilowercase.txt"])
|
||||
.succeeds()
|
||||
.stdout_only("uvwxyz\n");
|
||||
new_ucmd!()
|
||||
.args(&["-e", "--number=99/100", "asciilowercase.txt"])
|
||||
.succeeds()
|
||||
.stdout_only("");
|
||||
#[cfg(unix)]
|
||||
new_ucmd!()
|
||||
.args(&["--number=3/10", "/dev/null"])
|
||||
.succeeds()
|
||||
.stdout_only("");
|
||||
#[cfg(target_pointer_width = "64")]
|
||||
new_ucmd!()
|
||||
.args(&[
|
||||
"--number=r/9223372036854775807/18446744073709551615",
|
||||
"asciilowercase.txt",
|
||||
])
|
||||
.succeeds()
|
||||
.stdout_only("");
|
||||
new_ucmd!()
|
||||
.args(&["--number=0/5", "asciilowercase.txt"])
|
||||
.fails()
|
||||
.stderr_contains("split: invalid chunk number: 0");
|
||||
new_ucmd!()
|
||||
.args(&["--number=10/5", "asciilowercase.txt"])
|
||||
.fails()
|
||||
.stderr_contains("split: invalid chunk number: 10");
|
||||
#[cfg(target_pointer_width = "64")]
|
||||
new_ucmd!()
|
||||
.args(&[
|
||||
"--number=9223372036854775807/18446744073709551616",
|
||||
"asciilowercase.txt",
|
||||
])
|
||||
.fails()
|
||||
.stderr_contains("split: invalid number of chunks: 18446744073709551616");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_number_kth_of_n_round_robin() {
|
||||
new_ucmd!()
|
||||
.args(&["--number", "r/2/3", "fivelines.txt"])
|
||||
.succeeds()
|
||||
.stdout_only("2\n5\n");
|
||||
new_ucmd!()
|
||||
.args(&["--number", "r/1/4", "fivelines.txt"])
|
||||
.succeeds()
|
||||
.stdout_only("1\n5\n");
|
||||
new_ucmd!()
|
||||
.args(&["-e", "--number", "r/7/7", "fivelines.txt"])
|
||||
.succeeds()
|
||||
.stdout_only("");
|
||||
#[cfg(target_pointer_width = "64")]
|
||||
new_ucmd!()
|
||||
.args(&[
|
||||
"--number",
|
||||
"r/9223372036854775807/18446744073709551615",
|
||||
"fivelines.txt",
|
||||
])
|
||||
.succeeds()
|
||||
.stdout_only("");
|
||||
#[cfg(target_pointer_width = "64")]
|
||||
new_ucmd!()
|
||||
.args(&[
|
||||
"--number",
|
||||
"r/9223372036854775807/18446744073709551616",
|
||||
"fivelines.txt",
|
||||
])
|
||||
.fails()
|
||||
.stderr_contains("split: invalid number of chunks: 18446744073709551616");
|
||||
new_ucmd!()
|
||||
.args(&["--number", "r/0/3", "fivelines.txt"])
|
||||
.fails()
|
||||
.stderr_contains("split: invalid chunk number: 0");
|
||||
new_ucmd!()
|
||||
.args(&["--number", "r/10/3", "fivelines.txt"])
|
||||
.fails()
|
||||
.stderr_contains("split: invalid chunk number: 10");
|
||||
}
|
||||
|
||||
#[test]
|
||||
|
|
@ -524,6 +935,19 @@ fn test_invalid_suffix_length() {
|
|||
.stderr_contains("invalid suffix length: 'xyz'");
|
||||
}
|
||||
|
||||
/// Test short suffix length option with value concatenated
|
||||
#[test]
|
||||
fn test_split_suffix_length_short_concatenated_with_value() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let name = "split_num_prefixed_chunks_by_lines";
|
||||
RandomFile::new(&at, name).add_lines(10000);
|
||||
ucmd.args(&["-a4", name]).succeeds();
|
||||
|
||||
let glob = Glob::new(&at, ".", r"x[[:alpha:]][[:alpha:]][[:alpha:]][[:alpha:]]$");
|
||||
assert_eq!(glob.count(), 10);
|
||||
assert_eq!(glob.collate(), at.read_bytes(name));
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_include_newlines() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
|
|
@ -542,6 +966,19 @@ fn test_include_newlines() {
|
|||
assert_eq!(s, "5\n");
|
||||
}
|
||||
|
||||
/// Test short number of chunks option concatenated with value
|
||||
#[test]
|
||||
fn test_split_number_chunks_short_concatenated_with_value() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
ucmd.args(&["-n3", "threebytes.txt"])
|
||||
.succeeds()
|
||||
.no_stdout()
|
||||
.no_stderr();
|
||||
assert_eq!(at.read("xaa"), "a");
|
||||
assert_eq!(at.read("xab"), "b");
|
||||
assert_eq!(at.read("xac"), "c");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_allow_empty_files() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
|
|
@ -616,6 +1053,25 @@ fn test_line_bytes() {
|
|||
assert_eq!(at.read("xad"), "ee\n");
|
||||
}
|
||||
|
||||
#[test]
|
||||
#[cfg(target_pointer_width = "64")]
|
||||
fn test_line_bytes_overflow() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
ucmd.args(&["-C", "18446744073709551616", "letters.txt"])
|
||||
.succeeds();
|
||||
assert_eq!(at.read("xaa"), "aaaaaaaaa\nbbbb\ncccc\ndd\nee\n");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_line_bytes_concatenated_with_value() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
ucmd.args(&["-C8", "letters.txt"]).succeeds();
|
||||
assert_eq!(at.read("xaa"), "aaaaaaaa");
|
||||
assert_eq!(at.read("xab"), "a\nbbbb\n");
|
||||
assert_eq!(at.read("xac"), "cccc\ndd\n");
|
||||
assert_eq!(at.read("xad"), "ee\n");
|
||||
}
|
||||
|
||||
#[test]
|
||||
fn test_line_bytes_no_final_newline() {
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
|
|
@ -970,3 +1426,49 @@ fn test_split_invalid_input() {
|
|||
.no_stdout()
|
||||
.stderr_contains("split: invalid number of chunks: 0");
|
||||
}
|
||||
|
||||
/// Test if there are invalid (non UTF-8) in the arguments - unix
|
||||
/// clap is expected to fail/panic
|
||||
#[test]
|
||||
#[cfg(unix)]
|
||||
fn test_split_non_utf8_argument_unix() {
|
||||
use std::ffi::OsStr;
|
||||
use std::os::unix::ffi::OsStrExt;
|
||||
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let name = "test_split_non_utf8_argument";
|
||||
let opt = OsStr::from_bytes("--additional-suffix".as_bytes());
|
||||
RandomFile::new(&at, name).add_lines(2000);
|
||||
// Here, the values 0x66 and 0x6f correspond to 'f' and 'o'
|
||||
// respectively. The value 0x80 is a lone continuation byte, invalid
|
||||
// in a UTF-8 sequence.
|
||||
let opt_value = [0x66, 0x6f, 0x80, 0x6f];
|
||||
let opt_value = OsStr::from_bytes(&opt_value[..]);
|
||||
let name = OsStr::from_bytes(name.as_bytes());
|
||||
ucmd.args(&[opt, opt_value, name])
|
||||
.fails()
|
||||
.stderr_contains("error: invalid UTF-8 was detected in one or more arguments");
|
||||
}
|
||||
|
||||
/// Test if there are invalid (non UTF-8) in the arguments - windows
|
||||
/// clap is expected to fail/panic
|
||||
#[test]
|
||||
#[cfg(windows)]
|
||||
fn test_split_non_utf8_argument_windows() {
|
||||
use std::ffi::OsString;
|
||||
use std::os::windows::ffi::OsStringExt;
|
||||
|
||||
let (at, mut ucmd) = at_and_ucmd!();
|
||||
let name = "test_split_non_utf8_argument";
|
||||
let opt = OsString::from("--additional-suffix");
|
||||
RandomFile::new(&at, name).add_lines(2000);
|
||||
// Here the values 0x0066 and 0x006f correspond to 'f' and 'o'
|
||||
// respectively. The value 0xD800 is a lone surrogate half, invalid
|
||||
// in a UTF-16 sequence.
|
||||
let opt_value = [0x0066, 0x006f, 0xD800, 0x006f];
|
||||
let opt_value = OsString::from_wide(&opt_value[..]);
|
||||
let name = OsString::from(name);
|
||||
ucmd.args(&[opt, opt_value, name])
|
||||
.fails()
|
||||
.stderr_contains("error: invalid UTF-8 was detected in one or more arguments");
|
||||
}
|
||||
|
|
|
|||
Loading…
Reference in a new issue