-
-
Notifications
You must be signed in to change notification settings - Fork 2k
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Co-authored-by: cmdlineluser <99486669+cmdlineluser@users.noreply.github.com>
- Loading branch information
1 parent
4dd2a6d
commit 0d1b4a7
Showing
10 changed files
with
205 additions
and
94 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
152 changes: 109 additions & 43 deletions
152
crates/polars-ops/src/chunked_array/strings/substring.rs
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,51 +1,117 @@ | ||
| use arrow::array::Utf8Array; | ||
| use polars_core::prelude::arity::{binary_elementwise, ternary_elementwise, unary_elementwise}; | ||
| use polars_core::prelude::{Int64Chunked, StringChunked, UInt64Chunked}; | ||
|
|
||
| /// Returns a Utf8Array<O> with a substring starting from `start` and with optional length `length` of each of the elements in `array`. | ||
| /// `start` can be negative, in which case the start counts from the end of the string. | ||
| pub(super) fn utf8_substring( | ||
| array: &Utf8Array<i64>, | ||
| start: i64, | ||
| length: &Option<u64>, | ||
| ) -> Utf8Array<i64> { | ||
| let length = length.map(|v| v as usize); | ||
| fn substring_ternary( | ||
| opt_str_val: Option<&str>, | ||
| opt_offset: Option<i64>, | ||
| opt_length: Option<u64>, | ||
| ) -> Option<&str> { | ||
| match (opt_str_val, opt_offset) { | ||
| (Some(str_val), Some(offset)) => { | ||
| // If `offset` is negative, it counts from the end of the string. | ||
| let offset = if offset >= 0 { | ||
| offset as usize | ||
| } else { | ||
| let offset = (0i64 - offset) as usize; | ||
| str_val | ||
| .char_indices() | ||
| .rev() | ||
| .nth(offset) | ||
| .map(|(idx, _)| idx + 1) | ||
| .unwrap_or(0) | ||
| }; | ||
|
|
||
| let iter = array.values_iter().map(|str_val| { | ||
| // compute where we should start slicing this entry. | ||
| let start = if start >= 0 { | ||
| start as usize | ||
| } else { | ||
| let start = (0i64 - start) as usize; | ||
| str_val | ||
| .char_indices() | ||
| .rev() | ||
| .nth(start) | ||
| .map(|(idx, _)| idx + 1) | ||
| .unwrap_or(0) | ||
| }; | ||
| let mut iter_chars = str_val.char_indices(); | ||
| if let Some((offset_idx, _)) = iter_chars.nth(offset) { | ||
| let len_end = str_val.len() - offset_idx; | ||
|
|
||
| let mut iter_chars = str_val.char_indices(); | ||
| if let Some((start_idx, _)) = iter_chars.nth(start) { | ||
| // length of the str | ||
| let len_end = str_val.len() - start_idx; | ||
| // Slice to end of str if no length given. | ||
| let length = if let Some(length) = opt_length { | ||
| length as usize | ||
| } else { | ||
| len_end | ||
| }; | ||
|
|
||
| // length to slice | ||
| let length = length.unwrap_or(len_end); | ||
| if length == 0 { | ||
| return Some(""); | ||
| } | ||
|
|
||
| if length == 0 { | ||
| return ""; | ||
| } | ||
| // compute | ||
| let end_idx = iter_chars | ||
| .nth(length.saturating_sub(1)) | ||
| .map(|(idx, _)| idx) | ||
| .unwrap_or(str_val.len()); | ||
| let end_idx = iter_chars | ||
| .nth(length.saturating_sub(1)) | ||
| .map(|(idx, _)| idx) | ||
| .unwrap_or(str_val.len()); | ||
|
|
||
| &str_val[start_idx..end_idx] | ||
| } else { | ||
| "" | ||
| } | ||
| }); | ||
| Some(&str_val[offset_idx..end_idx]) | ||
| } else { | ||
| Some("") | ||
| } | ||
| }, | ||
| _ => None, | ||
| } | ||
| } | ||
|
|
||
| let new = Utf8Array::<i64>::from_trusted_len_values_iter(iter); | ||
| new.with_validity(array.validity().cloned()) | ||
| pub(super) fn substring( | ||
| ca: &StringChunked, | ||
| offset: &Int64Chunked, | ||
| length: &UInt64Chunked, | ||
| ) -> StringChunked { | ||
| match (ca.len(), offset.len(), length.len()) { | ||
| (1, 1, _) => { | ||
| // SAFETY: index `0` is in bound. | ||
| let str_val = unsafe { ca.get_unchecked(0) }; | ||
| // SAFETY: index `0` is in bound. | ||
| let offset = unsafe { offset.get_unchecked(0) }; | ||
| unary_elementwise(length, |length| substring_ternary(str_val, offset, length)) | ||
| .with_name(ca.name()) | ||
| }, | ||
| (_, 1, 1) => { | ||
| // SAFETY: index `0` is in bound. | ||
| let offset = unsafe { offset.get_unchecked(0) }; | ||
| // SAFETY: index `0` is in bound. | ||
| let length = unsafe { length.get_unchecked(0) }; | ||
| unary_elementwise(ca, |str_val| substring_ternary(str_val, offset, length)) | ||
| }, | ||
| (1, _, 1) => { | ||
| // SAFETY: index `0` is in bound. | ||
| let str_val = unsafe { ca.get_unchecked(0) }; | ||
| // SAFETY: index `0` is in bound. | ||
| let length = unsafe { length.get_unchecked(0) }; | ||
| unary_elementwise(offset, |offset| substring_ternary(str_val, offset, length)) | ||
| .with_name(ca.name()) | ||
| }, | ||
| (1, len_b, len_c) if len_b == len_c => { | ||
| // SAFETY: index `0` is in bound. | ||
| let str_val = unsafe { ca.get_unchecked(0) }; | ||
| binary_elementwise(offset, length, |offset, length| { | ||
| substring_ternary(str_val, offset, length) | ||
| }) | ||
| }, | ||
| (len_a, 1, len_c) if len_a == len_c => { | ||
| fn infer<F: for<'a> FnMut(Option<&'a str>, Option<u64>) -> Option<&'a str>>(f: F) -> F where | ||
| { | ||
| f | ||
| } | ||
| // SAFETY: index `0` is in bound. | ||
| let offset = unsafe { offset.get_unchecked(0) }; | ||
| binary_elementwise( | ||
| ca, | ||
| length, | ||
| infer(|str_val, length| substring_ternary(str_val, offset, length)), | ||
| ) | ||
| }, | ||
| (len_a, len_b, 1) if len_a == len_b => { | ||
| fn infer<F: for<'a> FnMut(Option<&'a str>, Option<i64>) -> Option<&'a str>>(f: F) -> F where | ||
| { | ||
| f | ||
| } | ||
| // SAFETY: index `0` is in bound. | ||
| let length = unsafe { length.get_unchecked(0) }; | ||
| binary_elementwise( | ||
| ca, | ||
| offset, | ||
| infer(|str_val, offset| substring_ternary(str_val, offset, length)), | ||
| ) | ||
| }, | ||
| _ => ternary_elementwise(ca, offset, length, substring_ternary), | ||
| } | ||
| } |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Oops, something went wrong.