diff options
| -rw-r--r-- | helix-core/src/lib.rs | 3 | ||||
| -rw-r--r-- | helix-core/src/movement.rs | 157 | ||||
| -rw-r--r-- | helix-core/src/position.rs | 74 | ||||
| -rw-r--r-- | helix-core/src/state.rs | 222 | ||||
| -rw-r--r-- | helix-core/src/syntax.rs | 2 | ||||
| -rw-r--r-- | helix-term/src/commands.rs | 20 |
6 files changed, 245 insertions, 233 deletions
diff --git a/helix-core/src/lib.rs b/helix-core/src/lib.rs index 81a35f10..a742ca13 100644 --- a/helix-core/src/lib.rs +++ b/helix-core/src/lib.rs @@ -5,6 +5,7 @@ pub mod graphemes; mod history; pub mod indent; pub mod macros; +pub mod movement; pub mod object; mod position; pub mod register; @@ -48,7 +49,7 @@ pub use tendril::StrTendril as Tendril; #[doc(inline)] pub use {regex, tree_sitter}; -pub use position::Position; +pub use position::{coords_at_pos, pos_at_coords, Position}; pub use selection::Range; pub use selection::Selection; pub use syntax::Syntax; diff --git a/helix-core/src/movement.rs b/helix-core/src/movement.rs new file mode 100644 index 00000000..57abc498 --- /dev/null +++ b/helix-core/src/movement.rs @@ -0,0 +1,157 @@ +use crate::graphemes::{nth_next_grapheme_boundary, nth_prev_grapheme_boundary, RopeGraphemes}; +use crate::{coords_at_pos, pos_at_coords, ChangeSet, Position, Range, Rope, RopeSlice, Selection}; + +pub fn move_next_word_start(slice: RopeSlice, mut pos: usize, count: usize) -> usize { + for _ in 0..count { + if pos + 1 == slice.len_chars() { + return pos; + } + + let mut ch = slice.char(pos); + let next = slice.char(pos + 1); + + // if we're at the end of a word, or on whitespce right before new one + if categorize(ch) != categorize(next) { + pos += 1; + ch = next; + } + + if is_word(ch) { + skip_over_next(slice, &mut pos, is_word); + } else if ch.is_ascii_punctuation() { + skip_over_next(slice, &mut pos, |ch| ch.is_ascii_punctuation()); + } + + // TODO: don't include newline? + skip_over_next(slice, &mut pos, |ch| ch.is_ascii_whitespace()); + } + + pos +} + +pub fn move_prev_word_start(slice: RopeSlice, mut pos: usize, count: usize) -> usize { + for _ in 0..count { + if pos == 0 { + return pos; + } + + let ch = slice.char(pos); + let prev = slice.char(pos - 1); + + if categorize(ch) != categorize(prev) { + pos -= 1; + } + + // match (category c1, category c2) => { + // if c1 != c2 { + // } + // } + + // TODO: skip while eol + + // TODO: don't include newline? + skip_over_prev(slice, &mut pos, |ch| ch.is_ascii_whitespace()); + + // refetch + let ch = slice.char(pos); + + if is_word(ch) { + skip_over_prev(slice, &mut pos, is_word); + } else if ch.is_ascii_punctuation() { + skip_over_prev(slice, &mut pos, |ch| ch.is_ascii_punctuation()); + } + pos = pos.saturating_add(1) + } + + pos +} + +pub fn move_next_word_end(slice: RopeSlice, mut pos: usize, count: usize) -> usize { + for _ in 0..count { + if pos + 1 == slice.len_chars() { + return pos; + } + + let ch = slice.char(pos); + let next = slice.char(pos + 1); + + if categorize(ch) != categorize(next) { + pos += 1; + } + + // TODO: don't include newline? + skip_over_next(slice, &mut pos, |ch| ch.is_ascii_whitespace()); + + // refetch + let ch = slice.char(pos); + + if is_word(ch) { + skip_over_next(slice, &mut pos, is_word); + } else if ch.is_ascii_punctuation() { + skip_over_next(slice, &mut pos, |ch| ch.is_ascii_punctuation()); + } + pos -= 1 + } + + pos +} + +// ---- util ------------ + +// used for by-word movement + +fn is_word(ch: char) -> bool { + ch.is_alphanumeric() || ch == '_' +} + +#[derive(Debug, Eq, PartialEq)] +enum Category { + Whitespace, + Eol, + Word, + Punctuation, +} +fn categorize(ch: char) -> Category { + if ch == '\n' { + Category::Eol + } else if ch.is_ascii_whitespace() { + Category::Whitespace + } else if ch.is_ascii_punctuation() { + Category::Punctuation + } else if ch.is_ascii_alphanumeric() { + Category::Word + } else { + unreachable!() + } +} + +#[inline] +pub fn skip_over_next<F>(slice: RopeSlice, pos: &mut usize, fun: F) +where + F: Fn(char) -> bool, +{ + let mut chars = slice.chars_at(*pos); + + for ch in chars { + if !fun(ch) { + break; + } + *pos += 1; + } +} + +#[inline] +pub fn skip_over_prev<F>(slice: RopeSlice, pos: &mut usize, fun: F) +where + F: Fn(char) -> bool, +{ + // need to +1 so that prev() includes current char + let mut chars = slice.chars_at(*pos + 1); + + while let Some(ch) = chars.prev() { + if !fun(ch) { + break; + } + *pos = pos.saturating_sub(1); + } +} diff --git a/helix-core/src/position.rs b/helix-core/src/position.rs index 5e5f902b..346779b0 100644 --- a/helix-core/src/position.rs +++ b/helix-core/src/position.rs @@ -1,3 +1,8 @@ +use crate::{ + graphemes::{nth_next_grapheme_boundary, RopeGraphemes}, + Rope, RopeSlice, +}; + /// Represents a single point in a text buffer. Zero indexed. #[derive(Debug, Default, Clone, Copy, PartialEq, Eq, PartialOrd, Ord)] pub struct Position { @@ -44,6 +49,21 @@ impl From<Position> for tree_sitter::Point { Self::new(pos.row, pos.col) } } +/// Convert a character index to (line, column) coordinates. +pub fn coords_at_pos(text: RopeSlice, pos: usize) -> Position { + let line = text.char_to_line(pos); + let line_start = text.line_to_char(line); + let col = RopeGraphemes::new(text.slice(line_start..pos)).count(); + Position::new(line, col) +} + +/// Convert (line, column) coordinates to a character index. +pub fn pos_at_coords(text: RopeSlice, coords: Position) -> usize { + let Position { row, col } = coords; + let line_start = text.line_to_char(row); + // line_start + col + nth_next_grapheme_boundary(text, line_start, col) +} #[cfg(test)] mod test { @@ -54,4 +74,58 @@ mod test { // (0, 5) is less than (1, 0) assert!(Position::new(0, 5) < Position::new(1, 0)); } + + #[test] + fn test_coords_at_pos() { + let text = Rope::from("ḧëḷḷö\nẅöṛḷḋ"); + let slice = text.slice(..); + // assert_eq!(coords_at_pos(slice, 0), (0, 0).into()); + // assert_eq!(coords_at_pos(slice, 5), (0, 5).into()); // position on \n + // assert_eq!(coords_at_pos(slice, 6), (1, 0).into()); // position on w + // assert_eq!(coords_at_pos(slice, 7), (1, 1).into()); // position on o + // assert_eq!(coords_at_pos(slice, 10), (1, 4).into()); // position on d + + // test with grapheme clusters + let text = Rope::from("a̐éö̲\r\n"); + let slice = text.slice(..); + assert_eq!(coords_at_pos(slice, 0), (0, 0).into()); + assert_eq!(coords_at_pos(slice, 2), (0, 1).into()); + assert_eq!(coords_at_pos(slice, 4), (0, 2).into()); + assert_eq!(coords_at_pos(slice, 7), (0, 3).into()); + + let text = Rope::from("किमपि"); + let slice = text.slice(..); + assert_eq!(coords_at_pos(slice, 0), (0, 0).into()); + assert_eq!(coords_at_pos(slice, 2), (0, 1).into()); + assert_eq!(coords_at_pos(slice, 3), (0, 2).into()); + assert_eq!(coords_at_pos(slice, 5), (0, 3).into()); + } + + #[test] + fn test_pos_at_coords() { + let text = Rope::from("ḧëḷḷö\nẅöṛḷḋ"); + let slice = text.slice(..); + assert_eq!(pos_at_coords(slice, (0, 0).into()), 0); + assert_eq!(pos_at_coords(slice, (0, 5).into()), 5); // position on \n + assert_eq!(pos_at_coords(slice, (1, 0).into()), 6); // position on w + assert_eq!(pos_at_coords(slice, (1, 1).into()), 7); // position on o + assert_eq!(pos_at_coords(slice, (1, 4).into()), 10); // position on d + + // test with grapheme clusters + let text = Rope::from("a̐éö̲\r\n"); + let slice = text.slice(..); + assert_eq!(pos_at_coords(slice, (0, 0).into()), 0); + assert_eq!(pos_at_coords(slice, (0, 1).into()), 2); + assert_eq!(pos_at_coords(slice, (0, 2).into()), 4); + assert_eq!(pos_at_coords(slice, (0, 3).into()), 7); // \r\n is one char here + assert_eq!(pos_at_coords(slice, (0, 4).into()), 9); + let text = Rope::from("किमपि"); + // 2 - 1 - 2 codepoints + // TODO: delete handling as per https://news.ycombinator.com/item?id=20058454 + let slice = text.slice(..); + assert_eq!(pos_at_coords(slice, (0, 0).into()), 0); + assert_eq!(pos_at_coords(slice, (0, 1).into()), 2); + assert_eq!(pos_at_coords(slice, (0, 2).into()), 3); + assert_eq!(pos_at_coords(slice, (0, 3).into()), 5); // eol + } } diff --git a/helix-core/src/state.rs b/helix-core/src/state.rs index 4d4df489..9dcd5548 100644 --- a/helix-core/src/state.rs +++ b/helix-core/src/state.rs @@ -1,7 +1,5 @@ use crate::graphemes::{nth_next_grapheme_boundary, nth_prev_grapheme_boundary, RopeGraphemes}; -use crate::syntax::LOADER; -use crate::{ChangeSet, Diagnostic, Position, Range, Rope, RopeSlice, Selection, Syntax}; -use anyhow::Error; +use crate::{coords_at_pos, pos_at_coords, ChangeSet, Position, Range, Rope, RopeSlice, Selection}; /// A state represents the current editor state of a single buffer. #[derive(Clone)] @@ -83,96 +81,6 @@ impl State { Range::new(if extend { range.anchor } else { pos }, pos) } - pub fn move_next_word_start(slice: RopeSlice, mut pos: usize, count: usize) -> usize { - for _ in 0..count { - if pos + 1 == slice.len_chars() { - return pos; - } - - let mut ch = slice.char(pos); - let next = slice.char(pos + 1); - - // if we're at the end of a word, or on whitespce right before new one - if categorize(ch) != categorize(next) { - pos += 1; - ch = next; - } - - if is_word(ch) { - skip_over_next(slice, &mut pos, is_word); - } else if ch.is_ascii_punctuation() { - skip_over_next(slice, &mut pos, |ch| ch.is_ascii_punctuation()); - } - - // TODO: don't include newline? - skip_over_next(slice, &mut pos, |ch| ch.is_ascii_whitespace()); - } - - pos - } - - pub fn move_prev_word_start(slice: RopeSlice, mut pos: usize, count: usize) -> usize { - for _ in 0..count { - if pos == 0 { - return pos; - } - - let ch = slice.char(pos); - let prev = slice.char(pos - 1); - - if categorize(ch) != categorize(prev) { - pos -= 1; - } - - // TODO: skip while eol - - // TODO: don't include newline? - skip_over_prev(slice, &mut pos, |ch| ch.is_ascii_whitespace()); - - // refetch - let ch = slice.char(pos); - - if is_word(ch) { - skip_over_prev(slice, &mut pos, is_word); - } else if ch.is_ascii_punctuation() { - skip_over_prev(slice, &mut pos, |ch| ch.is_ascii_punctuation()); - } - pos = pos.saturating_add(1) - } - - pos - } - - pub fn move_next_word_end(slice: RopeSlice, mut pos: usize, count: usize) -> usize { - for _ in 0..count { - if pos + 1 == slice.len_chars() { - return pos; - } - - let ch = slice.char(pos); - let next = slice.char(pos + 1); - - if categorize(ch) != categorize(next) { - pos += 1; - } - - // TODO: don't include newline? - skip_over_next(slice, &mut pos, |ch| ch.is_ascii_whitespace()); - - // refetch - let ch = slice.char(pos); - - if is_word(ch) { - skip_over_next(slice, &mut pos, is_word); - } else if ch.is_ascii_punctuation() { - skip_over_next(slice, &mut pos, |ch| ch.is_ascii_punctuation()); - } - pos -= 1 - } - - pos - } - pub fn move_selection( &self, dir: Direction, @@ -194,22 +102,6 @@ impl State { } } -/// Convert a character index to (line, column) coordinates. -pub fn coords_at_pos(text: RopeSlice, pos: usize) -> Position { - let line = text.char_to_line(pos); - let line_start = text.line_to_char(line); - let col = RopeGraphemes::new(text.slice(line_start..pos)).count(); - Position::new(line, col) -} - -/// Convert (line, column) coordinates to a character index. -pub fn pos_at_coords(text: RopeSlice, coords: Position) -> usize { - let Position { row, col } = coords; - let line_start = text.line_to_char(row); - // line_start + col - nth_next_grapheme_boundary(text, line_start, col) -} - fn move_vertically( text: RopeSlice, dir: Direction, @@ -238,123 +130,11 @@ fn move_vertically( range } -// used for by-word movement - -fn is_word(ch: char) -> bool { - ch.is_alphanumeric() || ch == '_' -} - -#[derive(Debug, Eq, PartialEq)] -enum Category { - Whitespace, - Eol, - Word, - Punctuation, -} -fn categorize(ch: char) -> Category { - if ch == '\n' { - Category::Eol - } else if ch.is_ascii_whitespace() { - Category::Whitespace - } else if ch.is_ascii_punctuation() { - Category::Punctuation - } else if ch.is_ascii_alphanumeric() { - Category::Word - } else { - unreachable!() - } -} - -#[inline] -pub fn skip_over_next<F>(slice: RopeSlice, pos: &mut usize, fun: F) -where - F: Fn(char) -> bool, -{ - let mut chars = slice.chars_at(*pos); - - for ch in chars { - if !fun(ch) { - break; - } - *pos += 1; - } -} - -#[inline] -pub fn skip_over_prev<F>(slice: RopeSlice, pos: &mut usize, fun: F) -where - F: Fn(char) -> bool, -{ - // need to +1 so that prev() includes current char - let mut chars = slice.chars_at(*pos + 1); - - while let Some(ch) = chars.prev() { - if !fun(ch) { - break; - } - *pos = pos.saturating_sub(1); - } -} - #[cfg(test)] mod test { use super::*; #[test] - fn test_coords_at_pos() { - let text = Rope::from("ḧëḷḷö\nẅöṛḷḋ"); - let slice = text.slice(..); - // assert_eq!(coords_at_pos(slice, 0), (0, 0).into()); - // assert_eq!(coords_at_pos(slice, 5), (0, 5).into()); // position on \n - // assert_eq!(coords_at_pos(slice, 6), (1, 0).into()); // position on w - // assert_eq!(coords_at_pos(slice, 7), (1, 1).into()); // position on o - // assert_eq!(coords_at_pos(slice, 10), (1, 4).into()); // position on d - - // test with grapheme clusters - let text = Rope::from("a̐éö̲\r\n"); - let slice = text.slice(..); - assert_eq!(coords_at_pos(slice, 0), (0, 0).into()); - assert_eq!(coords_at_pos(slice, 2), (0, 1).into()); - assert_eq!(coords_at_pos(slice, 4), (0, 2).into()); - assert_eq!(coords_at_pos(slice, 7), (0, 3).into()); - - let text = Rope::from("किमपि"); - let slice = text.slice(..); - assert_eq!(coords_at_pos(slice, 0), (0, 0).into()); - assert_eq!(coords_at_pos(slice, 2), (0, 1).into()); - assert_eq!(coords_at_pos(slice, 3), (0, 2).into()); - assert_eq!(coords_at_pos(slice, 5), (0, 3).into()); - } - - #[test] - fn test_pos_at_coords() { - let text = Rope::from("ḧëḷḷö\nẅöṛḷḋ"); - let slice = text.slice(..); - assert_eq!(pos_at_coords(slice, (0, 0).into()), 0); - assert_eq!(pos_at_coords(slice, (0, 5).into()), 5); // position on \n - assert_eq!(pos_at_coords(slice, (1, 0).into()), 6); // position on w - assert_eq!(pos_at_coords(slice, (1, 1).into()), 7); // position on o - assert_eq!(pos_at_coords(slice, (1, 4).into()), 10); // position on d - - // test with grapheme clusters - let text = Rope::from("a̐éö̲\r\n"); - let slice = text.slice(..); - assert_eq!(pos_at_coords(slice, (0, 0).into()), 0); - assert_eq!(pos_at_coords(slice, (0, 1).into()), 2); - assert_eq!(pos_at_coords(slice, (0, 2).into()), 4); - assert_eq!(pos_at_coords(slice, (0, 3).into()), 7); // \r\n is one char here - assert_eq!(pos_at_coords(slice, (0, 4).into()), 9); - let text = Rope::from("किमपि"); - // 2 - 1 - 2 codepoints - // TODO: delete handling as per https://news.ycombinator.com/item?id=20058454 - let slice = text.slice(..); - assert_eq!(pos_at_coords(slice, (0, 0).into()), 0); - assert_eq!(pos_at_coords(slice, (0, 1).into()), 2); - assert_eq!(pos_at_coords(slice, (0, 2).into()), 3); - assert_eq!(pos_at_coords(slice, (0, 3).into()), 5); // eol - } - - #[test] fn test_vertical_move() { let text = Rope::from("abcd\nefg\nwrs"); let slice = text.slice(..); diff --git a/helix-core/src/syntax.rs b/helix-core/src/syntax.rs index 3d665953..e75dc724 100644 --- a/helix-core/src/syntax.rs +++ b/helix-core/src/syntax.rs @@ -355,7 +355,7 @@ pub struct LanguageLayer { pub(crate) tree: Option<Tree>, } -use crate::state::coords_at_pos; +use crate::coords_at_pos; use crate::transaction::{ChangeSet, Operation}; use crate::Tendril; diff --git a/helix-term/src/commands.rs b/helix-term/src/commands.rs index 1235a94e..e703388d 100644 --- a/helix-term/src/commands.rs +++ b/helix-term/src/commands.rs @@ -1,10 +1,10 @@ use helix_core::{ - comment, graphemes, + comment, coords_at_pos, graphemes, indent::TAB_WIDTH, - object, + movement, object, pos_at_coords, regex::{self, Regex}, register, search, selection, - state::{coords_at_pos, pos_at_coords, Direction, Granularity, State}, + state::{Direction, Granularity, State}, Change, ChangeSet, Position, Range, RopeSlice, Selection, Tendril, Transaction, }; @@ -152,7 +152,7 @@ pub fn move_next_word_start(cx: &mut Context) { let text = doc.text().slice(..); let selection = doc.selection().transform(|range| { - let pos = State::move_next_word_start(text, range.head, count); + let pos = movement::move_next_word_start(text, range.head, count); Range::new(pos, pos) }); @@ -165,7 +165,7 @@ pub fn move_prev_word_start(cx: &mut Context) { let text = doc.text().slice(..); let selection = doc.selection().transform(|range| { - let pos = State::move_prev_word_start(text, range.head, count); + let pos = movement::move_prev_word_start(text, range.head, count); Range::new(pos, pos) }); @@ -178,7 +178,7 @@ pub fn move_next_word_end(cx: &mut Context) { let text = doc.text().slice(..); let selection = doc.selection().transform(|range| { - let pos = State::move_next_word_end(text, range.head, count); + let pos = movement::move_next_word_end(text, range.head, count); Range::new(pos, pos) }); @@ -207,7 +207,7 @@ pub fn extend_next_word_start(cx: &mut Context) { let text = doc.text().slice(..); let selection = doc.selection().transform(|mut range| { - let pos = State::move_next_word_start(text, range.head, count); + let pos = movement::move_next_word_start(text, range.head, count); Range::new(range.anchor, pos) }); @@ -220,7 +220,7 @@ pub fn extend_prev_word_start(cx: &mut Context) { let text = doc.text().slice(..); let selection = doc.selection().transform(|mut range| { - let pos = State::move_prev_word_start(text, range.head, count); + let pos = movement::move_prev_word_start(text, range.head, count); Range::new(range.anchor, pos) }); doc.set_selection(selection); @@ -232,7 +232,7 @@ pub fn extend_next_word_end(cx: &mut Context) { let text = doc.text().slice(..); let selection = doc.selection().transform(|mut range| { - let pos = State::move_next_word_end(text, range.head, count); + let pos = movement::move_next_word_end(text, range.head, count); Range::new(range.anchor, pos) }); @@ -1199,7 +1199,7 @@ pub fn format_selections(cx: &mut Context) { } pub fn join_selections(cx: &mut Context) { - use helix_core::state::skip_over_next; + use movement::skip_over_next; let doc = cx.doc(); let text = doc.text(); let slice = doc.text().slice(..); |