From 4f2d50e9b8279f209ba0d721d675e3338c3244e7 Mon Sep 17 00:00:00 2001
From: Alex Kotov <kotovalexarian@gmail.com>
Date: Sun, 2 Apr 2023 15:02:15 +0400
Subject: [PATCH] Filter input chars

---
 src/input.rs  | 35 ++++++++++++++++++++++++++++-------
 src/lexer.rs  |  8 ++++----
 src/models.rs | 22 ++++++++++++++++++----
 3 files changed, 50 insertions(+), 15 deletions(-)

diff --git a/src/input.rs b/src/input.rs
index 2eb7946..1661ce5 100644
--- a/src/input.rs
+++ b/src/input.rs
@@ -1,10 +1,16 @@
 use crate::models::{Char, Pos};
 
-pub use std::io::{Error, Result};
-
 use std::io::BufRead;
 use std::vec::IntoIter;
 
+pub type Result = std::result::Result<Char, Error>;
+
+#[derive(Debug)]
+pub enum Error {
+    IO(std::io::Error),
+    InvalidChar(char),
+}
+
 pub struct Input<'a> {
     read: &'a mut dyn BufRead,
     fin: bool,
@@ -12,6 +18,18 @@ pub struct Input<'a> {
     pos: Pos,
 }
 
+impl From<std::io::Error> for Error {
+    fn from(err: std::io::Error) -> Self {
+        Self::IO(err)
+    }
+}
+
+impl From<char> for Error {
+    fn from(chr: char) -> Self {
+        Self::InvalidChar(chr)
+    }
+}
+
 impl<'a> Input<'a> {
     pub fn new(read: &'a mut dyn BufRead) -> Self {
         Self {
@@ -27,7 +45,7 @@ impl<'a> Input<'a> {
 }
 
 impl Iterator for Input<'_> {
-    type Item = Result<Char>;
+    type Item = Result;
 
     fn next(&mut self) -> Option<Self::Item> {
         if self.fin {
@@ -35,10 +53,13 @@ impl Iterator for Input<'_> {
         }
 
         if let Some(chr) = self.iter.next() {
-            let result = Some(Ok(Char::new(chr, self.pos)));
+            let chr = Char::new(chr, self.pos);
             self.pos.index += 1;
             self.pos.col += 1;
-            return result;
+            return match chr {
+                Err(chr) => Some(Err(chr.into())),
+                Ok(chr) => Some(Ok(chr)),
+            };
         }
 
         let mut buffer = String::new();
@@ -53,7 +74,7 @@ impl Iterator for Input<'_> {
             }
             Err(err) => {
                 self.fin = true;
-                Some(Err(err))
+                Some(Err(err.into()))
             }
         }
     }
@@ -64,7 +85,7 @@ mod tests {
     use super::*;
 
     fn chr(value: char, index: usize, line: usize, col: usize) -> Char {
-        Char::new(value, Pos::new(index, line, col))
+        Char::new(value, Pos::new(index, line, col)).unwrap()
     }
 
     #[test]
diff --git a/src/lexer.rs b/src/lexer.rs
index 5ed0774..0d3884c 100644
--- a/src/lexer.rs
+++ b/src/lexer.rs
@@ -1,12 +1,12 @@
 use crate::input::{self, Input};
-use crate::models::{Char, Pos, Token};
+use crate::models::{Pos, Token};
 use crate::tokenizer::Tokenizer;
 
 pub type Result = std::result::Result<Token, Error>;
 
 pub struct Lexer<'a> {
     input: Input<'a>,
-    cur: Option<input::Result<Char>>,
+    cur: Option<input::Result>,
     pos: Pos,
     tokenizer: Tokenizer,
 }
@@ -47,14 +47,14 @@ impl Iterator for Lexer<'_> {
                 Ok(chr) => {
                     let chr = *chr;
                     self.cur = self.input.next();
-                    match self.tokenizer.iter(Some(chr.value)) {
+                    match self.tokenizer.iter(Some(chr.value())) {
                         None => self.next(),
                         Some(value) => {
                             let token = Token {
                                 value,
                                 pos: self.pos,
                             };
-                            self.pos = chr.pos;
+                            self.pos = chr.pos();
                             Some(Ok(token))
                         }
                     }
diff --git a/src/models.rs b/src/models.rs
index 636992c..4aa497d 100644
--- a/src/models.rs
+++ b/src/models.rs
@@ -1,7 +1,9 @@
+const MIN_CHAR: char = ' '; // ASCII code: 32
+
 #[derive(Clone, Copy, Debug, Eq, PartialEq)]
 pub struct Char {
-    pub value: char,
-    pub pos: Pos,
+    value: char,
+    pos: Pos,
 }
 
 #[derive(Clone, Copy, Debug, Eq, PartialEq)]
@@ -18,8 +20,20 @@ pub struct Token {
 }
 
 impl Char {
-    pub fn new(value: char, pos: Pos) -> Self {
-        Self { value, pos }
+    pub fn new(value: char, pos: Pos) -> Result<Self, char> {
+        if value >= MIN_CHAR || value == ' ' || value == '\n' {
+            Ok(Self { value, pos })
+        } else {
+            Err(value)
+        }
+    }
+
+    pub fn value(&self) -> char {
+        self.value
+    }
+
+    pub fn pos(&self) -> Pos {
+        self.pos
     }
 }