chore: move split_line_* from repl/cmd.rs to utils/ (sigoden#190)

rooct · Nov 30, 2023 · 952aa3f · 952aa3f
1 parent c2b92fb
commit 952aa3f
Show file tree

Hide file tree

Showing 8 changed files with 225 additions and 222 deletions.
diff --git a/README.md b/README.md
@@ -56,7 +56,7 @@ temperature: 1.0                 # GPT temperature, between 0 and 2
 save: true                       # Whether to save the message
 highlight: true                  # Set false to turn highlight
 light_theme: false               # Whether to use a light theme
-wrap: no                         # Specify the text-wrapping mode (*no*, auto, <max-width>)
+wrap: no                         # Specify the text-wrapping mode (no, auto, <max-width>)
 wrap_code: false                 # Whether wrap code block
 auto_copy: false                 # Automatically copy the last output to the clipboard
 keybindings: emacs               # REPL keybindings. values: emacs, vi

diff --git a/config.example.yaml b/config.example.yaml
@@ -3,7 +3,7 @@ temperature: 1.0                 # GPT temperature, between 0 and 2
 save: true                       # Whether to save the message
 highlight: true                  # Set false to turn highlight
 light_theme: false               # Whether to use a light theme
-wrap: no                         # Specify the text-wrapping mode (no*, auto, <max-width>)
+wrap: no                         # Specify the text-wrapping mode (no, auto, <max-width>)
 wrap_code: false                 # Whether wrap code block
 auto_copy: false                 # Automatically copy the last output to the clipboard
 keybindings: emacs               # REPL keybindings. values: emacs, vi

diff --git a/src/cli.rs b/src/cli.rs
@@ -18,7 +18,7 @@ pub struct Cli {
     /// No stream output
     #[clap(short = 'S', long)]
     pub no_stream: bool,
-    /// Specify the text-wrapping mode (no*, auto, <max-width>)
+    /// Specify the text-wrapping mode (no, auto, <max-width>)
     #[clap(short = 'w', long)]
     pub wrap: Option<String>,
     /// Use light theme

diff --git a/src/config/mod.rs b/src/config/mod.rs
@@ -65,7 +65,7 @@ pub struct Config {
     pub dry_run: bool,
     /// Whether to use a light theme
     pub light_theme: bool,
-    /// Specify the text-wrapping mode (*no*, auto, <max-width>)
+    /// Specify the text-wrapping mode (no, auto, <max-width>)
     pub wrap: Option<String>,
     /// Whether wrap code block
     pub wrap_code: bool,

diff --git a/src/render/cmd.rs b/src/render/cmd.rs
@@ -2,6 +2,7 @@ use super::MarkdownRender;
 
 use crate::print_now;
 use crate::repl::{ReplyStreamEvent, SharedAbortSignal};
+use crate::utils::{spaces, split_line_sematic, split_line_tail};
 
 use anyhow::Result;
 use crossbeam::channel::Receiver;
@@ -65,220 +66,3 @@ pub fn cmd_render_stream(
     }
     Ok(())
 }
-
-fn split_line_sematic(text: &str) -> Option<(String, String)> {
-    let mut balance: Vec<Kind> = Vec::new();
-    let chars: Vec<char> = text.chars().collect();
-    let mut index = 0;
-    let len = chars.len();
-    while index < len - 1 {
-        let ch = chars[index];
-        if balance.is_empty()
-            && ((matches!(ch, ',' | '.' | ';') && chars[index + 1].is_whitespace())
-                || matches!(ch, '，' | '。' | '；'))
-        {
-            let (output, remain) = chars.split_at(index + 1);
-            return Some((output.iter().collect(), remain.iter().collect()));
-        }
-        if index + 2 < len && do_balance(&mut balance, &chars[index..=index + 2]) {
-            index += 3;
-            continue;
-        }
-        if do_balance(&mut balance, &chars[index..=index + 1]) {
-            index += 2;
-            continue;
-        }
-        do_balance(&mut balance, &chars[index..=index]);
-        index += 1;
-    }
-
-    None
-}
-
-pub(crate) fn split_line_tail(text: &str) -> (&str, &str) {
-    if let Some((head, tail)) = text.rsplit_once('\n') {
-        (head, tail)
-    } else {
-        ("", text)
-    }
-}
-
-fn spaces(n: usize) -> String {
-    " ".repeat(n)
-}
-
-#[derive(Debug, Clone, Copy, Eq, PartialEq)]
-enum Kind {
-    ParentheseStart,
-    ParentheseEnd,
-    BracketStart,
-    BracketEnd,
-    Asterisk,
-    Asterisk2,
-    SingleQuota,
-    DoubleQuota,
-    Tilde,
-    Tilde2,
-    Backtick,
-    Backtick3,
-}
-
-impl Kind {
-    fn from_chars(chars: &[char]) -> Option<Self> {
-        let kind = match chars.len() {
-            1 => match chars[0] {
-                '(' => Self::ParentheseStart,
-                ')' => Self::ParentheseEnd,
-                '[' => Self::BracketStart,
-                ']' => Self::BracketEnd,
-                '*' => Self::Asterisk,
-                '\'' => Self::SingleQuota,
-                '"' => Self::DoubleQuota,
-                '~' => Self::Tilde,
-                '`' => Self::Backtick,
-                _ => return None,
-            },
-            2 if chars[0] == chars[1] => match chars[0] {
-                '*' => Self::Asterisk2,
-                '~' => Self::Tilde2,
-                _ => return None,
-            },
-            3 => {
-                if chars == ['`', '`', '`'] {
-                    Self::Backtick3
-                } else {
-                    return None;
-                }
-            }
-            _ => return None,
-        };
-        Some(kind)
-    }
-}
-
-fn do_balance(balance: &mut Vec<Kind>, chars: &[char]) -> bool {
-    Kind::from_chars(chars).map_or(false, |kind| {
-        let last = balance.last();
-        match (kind, last) {
-            (Kind::ParentheseEnd, Some(&Kind::ParentheseStart))
-            | (Kind::BracketEnd, Some(&Kind::BracketStart))
-            | (Kind::Asterisk, Some(&Kind::Asterisk))
-            | (Kind::Asterisk2, Some(&Kind::Asterisk2))
-            | (Kind::SingleQuota, Some(&Kind::SingleQuota))
-            | (Kind::DoubleQuota, Some(&Kind::DoubleQuota))
-            | (Kind::Tilde, Some(&Kind::Tilde))
-            | (Kind::Tilde2, Some(&Kind::Tilde2))
-            | (Kind::Backtick, Some(&Kind::Backtick))
-            | (Kind::Backtick3, Some(&Kind::Backtick3)) => {
-                balance.pop();
-                true
-            }
-            (
-                Kind::ParentheseStart
-                | Kind::BracketStart
-                | Kind::Asterisk
-                | Kind::Asterisk2
-                | Kind::SingleQuota
-                | Kind::DoubleQuota
-                | Kind::Tilde
-                | Kind::Tilde2
-                | Kind::Backtick
-                | Kind::Backtick3,
-                _,
-            ) => {
-                balance.push(kind);
-                true
-            }
-            _ => false,
-        }
-    })
-}
-
-#[cfg(test)]
-mod tests {
-    use super::*;
-
-    macro_rules! assert_split_line {
-        ($a:literal, $b:literal, true) => {
-            assert_eq!(
-                split_line_sematic(&format!("{}{}", $a, $b)),
-                Some(($a.into(), $b.into()))
-            );
-        };
-        ($a:literal, $b:literal, false) => {
-            assert_eq!(split_line_sematic(&format!("{}{}", $a, $b)), None);
-        };
-    }
-
-    #[test]
-    fn test_split_line() {
-        assert_split_line!(
-            "Lorem ipsum dolor sit amet,",
-            " consectetur adipiscing elit.",
-            true
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit amet.",
-            " consectetur adipiscing elit.",
-            true
-        );
-        assert_split_line!("黃更室幼許刀知，", "波食小午足田世根候法。", true);
-        assert_split_line!("黃更室幼許刀知。", "波食小午足田世根候法。", true);
-        assert_split_line!("黃更室幼許刀知；", "波食小午足田世根候法。", true);
-        assert_split_line!(
-            "Lorem ipsum (dolor sit amet).",
-            " consectetur adipiscing elit.",
-            true
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit `amet,",
-            " consectetur` adipiscing elit.",
-            false
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit ```amet,",
-            " consectetur``` adipiscing elit.",
-            false
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit *amet,",
-            " consectetur* adipiscing elit.",
-            false
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit **amet,",
-            " consectetur** adipiscing elit.",
-            false
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit ~amet,",
-            " consectetur~ adipiscing elit.",
-            false
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit ~~amet,",
-            " consectetur~~ adipiscing elit.",
-            false
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit ``amet,",
-            " consectetur`` adipiscing elit.",
-            true
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit \"amet,",
-            " consectetur\" adipiscing elit.",
-            false
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit 'amet,",
-            " consectetur' adipiscing elit.",
-            false
-        );
-        assert_split_line!(
-            "Lorem ipsum dolor sit amet.",
-            "consectetur adipiscing elit.",
-            false
-        );
-    }
-}
diff --git a/src/render/repl.rs b/src/render/repl.rs
@@ -1,6 +1,7 @@
-use super::{cmd::split_line_tail, MarkdownRender};
+use super::MarkdownRender;
 
 use crate::repl::{ReplyStreamEvent, SharedAbortSignal};
+use crate::utils::split_line_tail;
 
 use anyhow::Result;
 use crossbeam::channel::Receiver;

diff --git a/src/utils/mod.rs b/src/utils/mod.rs
@@ -1,5 +1,7 @@
+mod split_line;
 mod tiktoken;
 
+pub use self::split_line::*;
 pub use self::tiktoken::cl100k_base_singleton;
 
 use chrono::prelude::*;