loop {
// All the input starting at cur_start fits on the current line
if graphemes.len() - cur_start <= cur_max_chars {
- // trim trailing whitespaces
- let graphemes_minus_ws = if !fmt.trim_end {
- &graphemes[cur_start..]
- } else {
- match graphemes[cur_start..]
- .iter()
- .rposition(|grapheme| !is_whitespace(grapheme))
- {
- Some(index) => &graphemes[cur_start..=cur_start + index],
- None => &graphemes[cur_start..],
- }
- };
- if is_bareline_ok {
- // new lines don't need to start with line_start
- result.push_str(&graphemes_minus_ws.join(""));
- } else {
- // new lines need to be indented and prefixed with line_start
- for grapheme in graphemes_minus_ws {
- if is_line_feed(grapheme) {
- // take care of blank lines
- if fmt.trim_end && result.ends_with(' ') {
- result = result.trim_right().to_string();
- }
- result.push_str(&indent_with_newline);
+ for (i, grapheme) in graphemes[cur_start..].iter().enumerate() {
+ if is_line_feed(grapheme) {
+ // take care of blank lines
+ result = trim_right_but_line_feed(fmt.trim_end, result);
+ result.push_str("\n");
+ if !is_bareline_ok && cur_start + i + 1 < graphemes.len() {
+ result.push_str(&indent_without_newline);
result.push_str(fmt.line_start);
- } else {
- result.push_str(grapheme);
}
+ } else {
+ result.push_str(grapheme);
}
}
+ result = trim_right_but_line_feed(fmt.trim_end, result);
break;
}
// The input starting at cur_start needs to be broken
- match break_string(cur_max_chars, fmt.trim_end, &graphemes[cur_start..]) {
+ match break_string(
+ cur_max_chars,
+ fmt.trim_end,
+ fmt.line_end,
+ &graphemes[cur_start..],
+ ) {
SnippetState::LineEnd(line, len) => {
result.push_str(&line);
result.push_str(fmt.line_end);
wrap_str(result, fmt.config.max_width(), fmt.shape)
}
+/// Returns the index to the end of the url if the given string includes an
+/// URL or alike. Otherwise, returns None;
+fn detect_url(s: &[&str], index: usize) -> Option<usize> {
+ let start = match s[..=index].iter().rposition(|g| is_whitespace(g)) {
+ Some(pos) => pos + 1,
+ None => 0,
+ };
+ if s.len() < start + 8 {
+ return None;
+ }
+ let prefix = s[start..start + 8].join("");
+ if prefix.starts_with("https://")
+ || prefix.starts_with("http://")
+ || prefix.starts_with("ftp://")
+ || prefix.starts_with("file://")
+ {
+ match s[index..].iter().position(|g| is_whitespace(g)) {
+ Some(pos) => Some(index + pos - 1),
+ None => Some(s.len() - 1),
+ }
+ } else {
+ None
+ }
+}
+
+/// Trims whitespaces to the right except for the line feed character.
+fn trim_right_but_line_feed(trim_end: bool, result: String) -> String {
+ let whitespace_except_line_feed = |c: char| c.is_whitespace() && c != '\n';
+ if trim_end && result.ends_with(whitespace_except_line_feed) {
+ result
+ .trim_right_matches(whitespace_except_line_feed)
+ .to_string()
+ } else {
+ result
+ }
+}
+
/// Result of breaking a string so it fits in a line and the state it ended in.
/// The state informs about what to do with the snippet and how to continue the breaking process.
#[derive(Debug, PartialEq)]
EndWithLineFeed(String, usize),
}
+fn not_whitespace_except_line_feed(g: &str) -> bool {
+ is_line_feed(g) || !is_whitespace(g)
+}
+
/// Break the input string at a boundary character around the offset `max_chars`. A boundary
/// character is either a punctuation or a whitespace.
-fn break_string(max_chars: usize, trim_end: bool, input: &[&str]) -> SnippetState {
+fn break_string(max_chars: usize, trim_end: bool, line_end: &str, input: &[&str]) -> SnippetState {
let break_at = |index /* grapheme at index is included */| {
// Take in any whitespaces to the left/right of `input[index]` while
// preserving line feeds
let index_minus_ws = input[0..=index]
.iter()
- .rposition(|grapheme| !is_whitespace(grapheme))
+ .rposition(|grapheme| not_whitespace_except_line_feed(grapheme))
.unwrap_or(index);
// Take into account newlines occuring in input[0..=index], i.e., the possible next new
// line. If there is one, then text after it could be rewritten in a way that the available
// space is fully used.
for (i, grapheme) in input[0..=index].iter().enumerate() {
if is_line_feed(grapheme) {
- if i < index_minus_ws || !trim_end {
- return SnippetState::EndWithLineFeed(input[0..=i].join("").to_string(), i + 1);
+ if i <= index_minus_ws {
+ let mut line = input[0..i].join("");
+ if trim_end {
+ line = line.trim_right().to_string();
+ }
+ return SnippetState::EndWithLineFeed(format!("{}\n", line), i + 1);
}
break;
}
input[0..=index + 1 + i].join("").to_string(),
index + 2 + i,
);
- } else if !is_whitespace(grapheme) {
+ } else if not_whitespace_except_line_feed(grapheme) {
index_plus_ws = index + i;
break;
}
};
// Find the position in input for breaking the string
+ if line_end.is_empty()
+ && trim_end
+ && !is_whitespace(input[max_chars - 1])
+ && is_whitespace(input[max_chars])
+ {
+ // At a breaking point already
+ // The line won't invalidate the rewriting because:
+ // - no extra space needed for the line_end character
+ // - extra whitespaces to the right can be trimmed
+ return break_at(max_chars - 1);
+ }
+ if let Some(url_index_end) = detect_url(input, max_chars) {
+ let index_plus_ws = url_index_end
+ + input[url_index_end..]
+ .iter()
+ .skip(1)
+ .position(|grapheme| not_whitespace_except_line_feed(grapheme))
+ .unwrap_or(0);
+ return if trim_end {
+ SnippetState::LineEnd(
+ input[..=url_index_end].join("").to_string(),
+ index_plus_ws + 1,
+ )
+ } else {
+ return SnippetState::LineEnd(
+ input[..=index_plus_ws].join("").to_string(),
+ index_plus_ws + 1,
+ );
+ };
+ }
match input[0..max_chars]
.iter()
.rposition(|grapheme| is_whitespace(grapheme))
#[cfg(test)]
mod test {
- use super::{break_string, rewrite_string, SnippetState, StringFormat};
+ use super::{break_string, detect_url, rewrite_string, SnippetState, StringFormat};
use config::Config;
use shape::{Indent, Shape};
use unicode_segmentation::UnicodeSegmentation;
let string = "Placerat felis. Mauris porta ante sagittis purus.";
let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
assert_eq!(
- break_string(20, false, &graphemes[..]),
+ break_string(20, false, "", &graphemes[..]),
SnippetState::LineEnd("Placerat felis. ".to_string(), 16)
);
assert_eq!(
- break_string(20, true, &graphemes[..]),
+ break_string(20, true, "", &graphemes[..]),
SnippetState::LineEnd("Placerat felis.".to_string(), 16)
);
}
let string = "Placerat_felis._Mauris_porta_ante_sagittis_purus.";
let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
assert_eq!(
- break_string(20, false, &graphemes[..]),
+ break_string(20, false, "", &graphemes[..]),
SnippetState::LineEnd("Placerat_felis.".to_string(), 15)
);
}
let string = "Venenatis_tellus_vel_tellus. Aliquam aliquam dolor at justo.";
let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
assert_eq!(
- break_string(20, false, &graphemes[..]),
+ break_string(20, false, "", &graphemes[..]),
SnippetState::LineEnd("Venenatis_tellus_vel_tellus. ".to_string(), 29)
);
assert_eq!(
- break_string(20, true, &graphemes[..]),
+ break_string(20, true, "", &graphemes[..]),
SnippetState::LineEnd("Venenatis_tellus_vel_tellus.".to_string(), 29)
);
}
let string = "Venenatis_tellus_vel_tellus";
let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
assert_eq!(
- break_string(20, false, &graphemes[..]),
+ break_string(20, false, "", &graphemes[..]),
SnippetState::EndOfInput("Venenatis_tellus_vel_tellus".to_string())
);
}
let string = "Neque in sem. \n Pellentesque tellus augue.";
let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
assert_eq!(
- break_string(15, false, &graphemes[..]),
+ break_string(15, false, "", &graphemes[..]),
SnippetState::EndWithLineFeed("Neque in sem. \n".to_string(), 20)
);
assert_eq!(
- break_string(25, false, &graphemes[..]),
+ break_string(25, false, "", &graphemes[..]),
SnippetState::EndWithLineFeed("Neque in sem. \n".to_string(), 20)
);
- // if `StringFormat::line_end` is true, then the line feed does not matter anymore
+
assert_eq!(
- break_string(15, true, &graphemes[..]),
- SnippetState::LineEnd("Neque in sem.".to_string(), 26)
+ break_string(15, true, "", &graphemes[..]),
+ SnippetState::LineEnd("Neque in sem.".to_string(), 19)
);
assert_eq!(
- break_string(25, true, &graphemes[..]),
- SnippetState::LineEnd("Neque in sem.".to_string(), 26)
+ break_string(25, true, "", &graphemes[..]),
+ SnippetState::EndWithLineFeed("Neque in sem.\n".to_string(), 20)
);
}
let string = "Neque in sem. Pellentesque tellus augue.";
let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
assert_eq!(
- break_string(20, false, &graphemes[..]),
+ break_string(20, false, "", &graphemes[..]),
SnippetState::LineEnd("Neque in sem. ".to_string(), 25)
);
assert_eq!(
- break_string(20, true, &graphemes[..]),
+ break_string(20, true, "", &graphemes[..]),
SnippetState::LineEnd("Neque in sem.".to_string(), 25)
);
}
let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
assert_eq!(
- break_string(25, false, &graphemes[..]),
+ break_string(25, false, "", &graphemes[..]),
SnippetState::EndWithLineFeed("Nulla\n".to_string(), 6)
);
assert_eq!(
- break_string(25, true, &graphemes[..]),
+ break_string(25, true, "", &graphemes[..]),
SnippetState::EndWithLineFeed("Nulla\n".to_string(), 6)
);
)
);
}
+
+ #[test]
+ fn retain_blank_lines() {
+ let config: Config = Default::default();
+ let fmt = StringFormat {
+ opener: "",
+ closer: "",
+ line_start: "// ",
+ line_end: "",
+ shape: Shape::legacy(20, Indent::from_width(&config, 4)),
+ trim_end: true,
+ config: &config,
+ };
+
+ let comment = "Aenean\n\nmetus. Vestibulum ac lacus.\n\n";
+ assert_eq!(
+ rewrite_string(comment, &fmt),
+ Some(
+ "Aenean\n //\n // metus. Vestibulum ac\n // lacus.\n //\n".to_string()
+ )
+ );
+
+ let comment = "Aenean\n\nmetus. Vestibulum ac lacus.\n";
+ assert_eq!(
+ rewrite_string(comment, &fmt),
+ Some("Aenean\n //\n // metus. Vestibulum ac\n // lacus.\n".to_string())
+ );
+
+ let comment = "Aenean\n \nmetus. Vestibulum ac lacus.";
+ assert_eq!(
+ rewrite_string(comment, &fmt),
+ Some("Aenean\n //\n // metus. Vestibulum ac\n // lacus.".to_string())
+ );
+ }
+
+ #[test]
+ fn boundary_on_edge() {
+ let config: Config = Default::default();
+ let mut fmt = StringFormat {
+ opener: "",
+ closer: "",
+ line_start: "// ",
+ line_end: "",
+ shape: Shape::legacy(13, Indent::from_width(&config, 4)),
+ trim_end: true,
+ config: &config,
+ };
+
+ let comment = "Aenean metus. Vestibulum ac lacus.";
+ assert_eq!(
+ rewrite_string(comment, &fmt),
+ Some("Aenean metus.\n // Vestibulum ac\n // lacus.".to_string())
+ );
+
+ fmt.trim_end = false;
+ let comment = "Vestibulum ac lacus.";
+ assert_eq!(
+ rewrite_string(comment, &fmt),
+ Some("Vestibulum \n // ac lacus.".to_string())
+ );
+
+ fmt.trim_end = true;
+ fmt.line_end = "\\";
+ let comment = "Vestibulum ac lacus.";
+ assert_eq!(
+ rewrite_string(comment, &fmt),
+ Some("Vestibulum\\\n // ac lacus.".to_string())
+ );
+ }
+
+ #[test]
+ fn detect_urls() {
+ let string = "aaa http://example.org something";
+ let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
+ assert_eq!(detect_url(&graphemes, 8), Some(21));
+
+ let string = "https://example.org something";
+ let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
+ assert_eq!(detect_url(&graphemes, 0), Some(18));
+
+ let string = "aaa ftp://example.org something";
+ let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
+ assert_eq!(detect_url(&graphemes, 8), Some(20));
+
+ let string = "aaa file://example.org something";
+ let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
+ assert_eq!(detect_url(&graphemes, 8), Some(21));
+
+ let string = "aaa http not an url";
+ let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
+ assert_eq!(detect_url(&graphemes, 6), None);
+
+ let string = "aaa file://example.org";
+ let graphemes = UnicodeSegmentation::graphemes(&*string, false).collect::<Vec<&str>>();
+ assert_eq!(detect_url(&graphemes, 8), Some(21));
+ }
}
projects / rust.git / blobdiff