drop support for legacy \r newline sequence

Author: neverRare

dictionary/parser.ts

@@ -39,7 +39,7 @@ const WORDS = new RegExp(`[^${escapeRegex(RESERVED_SYMBOLS)}]`);
 function lex<T>(parser: Parser<T>): Parser<T> {
   return parser.skip(spaces);
 }
-const comment = match(/#[^\n\r]*/, "comment");
+const comment = match(/#[^\n]*\n?/, "comment");
 const spaces = sourceOnly(all(choiceOnlyOne(match(/\s/, "space"), comment)));
 const backtick = matchString("`", "backtick");
 const colon = matchString(":", "colon");

src/misc.ts

@@ -2,7 +2,7 @@ import { distinctBy } from "@std/collections/distinct-by";
 import { escape } from "@std/regexp/escape";
 import { Lazy } from "./cache.ts";
 
-export const NEWLINES = /\r\n|\n|\r/g;
+export const NEWLINES = /\r?\n/g;
 
 export function nullableAsArray<T>(value?: T): ReadonlyArray<NonNullable<T>> {
   if (value == null) {

src/parser/lexer.ts

@@ -42,8 +42,8 @@ import {
 } from "./ucsur.ts";
 import { throwError } from "../misc.ts";
 
-const spacesWithoutNewline = match(/[^\S\n\r]*/, "spaces");
-const newline = match(/[\n\r]\s*/, "newline");
+const spacesWithoutNewline = match(/[^\S\n]*?(?=\S|\r?\n|$)/, "spaces");
+const newline = match(/\r?\n\s*/, "newline");
 const spaces = sourceOnly(
   sequence(spacesWithoutNewline, choice(nothing, newline)),
 );

src/parser/parser_lib.ts

@@ -191,7 +191,7 @@ function describeSource(src: string): string {
   } else {
     const [token] = src.match(/\S*/)!;
     if (token === "") {
-      if (/^[\n\r]/.test(src)) {
+      if (/^\r?\n/.test(src)) {
         return "newline";
       } else {
         return "space";