From 3a8d833e6ad5b1ee1302e5a5a6206a9afb7031f4 Mon Sep 17 00:00:00 2001
From: Mikkel Kongsted <mtkongsted@gmail.com>
Date: Mon, 4 Nov 2024 14:54:55 +0100
Subject: [PATCH] lexer

---
 src/Lexer.ts | 96 ++++++++++++++++++++++++++++++++++++++++++++++++++--
 src/main.ts  | 16 +++++++++
 2 files changed, 110 insertions(+), 2 deletions(-)
 create mode 100644 src/main.ts

diff --git a/src/Lexer.ts b/src/Lexer.ts
index e69e070..6deae17 100644
--- a/src/Lexer.ts
+++ b/src/Lexer.ts
@@ -1,6 +1,6 @@
 import { Pos, Token } from "./Token.ts";
 
-class Lexer {
+export class Lexer {
     private index = 0;
     private line = 1;
     private col = 1;
@@ -8,7 +8,88 @@ class Lexer {
 
     public constructor (private text: string) {}
 
-    public next(): Token | null { return null }
+    public next(): Token | null {
+        if (this.done())
+            return null;
+        const pos = this.pos();
+        if (this.test(/[ \t\n]/)) {
+            while (!this.done() && this.test(/[ \t\n]/))
+                this.step();
+            return this.next();
+        }
+        if (this.test("/")) {
+            this.step()
+            if (this.test("/")) {
+                while (!this.done() && !this.test("\n"))
+                    this.step();
+                return this.token("//", pos)
+            }
+            return this.token("/", pos)
+        }
+        if (this.test(/[a-zA-Z_]/)) {
+            let value = "";
+            while (!this.done() && this.test(/[a-zA-Z0-9_]/)) {
+                value += this.current();
+                this.step();
+            }
+            switch (value) {
+                case "if":
+                    return this.token("if", pos);
+                case "else":
+                    return this.token("else", pos);
+                default:
+                    return { ...this.token("ident", pos), identValue: value };
+            }
+        }
+        if (this.test(/[0-9]/)) {
+            let textValue = "";
+            while (!this.done() && this.test(/[0-9]/)) {
+                textValue += this.current();
+                this.step();
+            }
+            return { ...this.token("int", pos), intValue: parseInt(textValue) };
+        }
+        if (this.test("\"")) {
+            this.step();
+            let value = "";
+            while (!this.done() && !this.test("\"")) {
+                if (this.test("\\")) {
+                    this.step();
+                    if (this.done())
+                        break;
+                    value += {
+                        "n": "\n",
+                        "t": "\t",
+                        "0": "\0",
+                    }[this.current()] ?? this.current();
+                } else {
+                    value += this.current();
+                }
+                this.step();
+            }
+            if (this.done() || !this.test("\"")) {
+                console.error(
+                    `Lexer: unclosed/malformed string`
+                        + ` at ${pos.line}:${pos.col}`,
+                );
+                return this.token("error", pos);
+            }
+            this.step();
+            return { ...this.token("string", pos), stringValue: value };
+        }
+        if (this.test(/[\+\{\};=]/)) {
+            const first = this.current();
+            this.step();
+            if (first === "=" && !this.done() && this.test("=")) {
+                this.step();
+                return this.token("==", pos);
+            }
+            return this.token(first, pos);
+        }
+        console.error(`Lexer: illegal character '${this.current()}' at ${pos.line}:${pos.col}`);
+        this.step();
+        return this.next();    
+    }
     
     private done(): boolean { return this.index >= this.text.length; }
     private current(): string { return this.text[this.index]; }
@@ -37,4 +118,15 @@ class Lexer {
         const length = this.index - pos.index;
         return { type, pos, length };
     }
+
+    private test(pattern: RegExp | string): boolean {
+        if (typeof pattern === "string")
+            return this.current() === pattern;
+        else
+            return pattern.test(this.current());
+    }
+
+
+
+
 }
diff --git a/src/main.ts b/src/main.ts
new file mode 100644
index 0000000..31de7e5
--- /dev/null
+++ b/src/main.ts
@@ -0,0 +1,16 @@
+import { Lexer } from "./Lexer.ts";
+
+const text = `
+    a1 123 +
+    // comment
+    "hello"
+    "escaped\\"\\nnewline"
+`;
+
+const lexer = new Lexer(text);
+let token = lexer.next();
+while (token !== null) {
+    const value = token.identValue ?? token.intValue ?? token.stringValue ?? "";
+    console.log(`Lexed ${token}(${value})`);
+    token = lexer.next();
+}