From c68acd8b5b106dcffe1c99456165a5c30cf8c334 Mon Sep 17 00:00:00 2001
From: Lee Byron <lee@leebyron.com>
Date: Thu, 12 Sep 2019 08:42:42 -0700
Subject: [PATCH] RFC: Lexing is Greedy (#2163)

Adds the test cases described in https://github.com/graphql/graphql-spec/pull/599

Makes the lookahead restriction change necessary for the new tests to pass for numbers, all other tests are already passing.
---
 src/language/__tests__/lexer-test.js | 128 +++++++++++++++++++++++++++
 src/language/lexer.js                |  10 +++
 2 files changed, 138 insertions(+)

diff --git a/src/language/__tests__/lexer-test.js b/src/language/__tests__/lexer-test.js
index c43281266b..974f003408 100644
--- a/src/language/__tests__/lexer-test.js
+++ b/src/language/__tests__/lexer-test.js
@@ -49,6 +49,65 @@ describe('Lexer', () => {
     });
   });
 
+  it('tracks line breaks', () => {
+    expect(lexOne('foo')).to.contain({
+      kind: TokenKind.NAME,
+      start: 0,
+      end: 3,
+      line: 1,
+      column: 1,
+      value: 'foo',
+    });
+    expect(lexOne('\nfoo')).to.contain({
+      kind: TokenKind.NAME,
+      start: 1,
+      end: 4,
+      line: 2,
+      column: 1,
+      value: 'foo',
+    });
+    expect(lexOne('\rfoo')).to.contain({
+      kind: TokenKind.NAME,
+      start: 1,
+      end: 4,
+      line: 2,
+      column: 1,
+      value: 'foo',
+    });
+    expect(lexOne('\r\nfoo')).to.contain({
+      kind: TokenKind.NAME,
+      start: 2,
+      end: 5,
+      line: 2,
+      column: 1,
+      value: 'foo',
+    });
+    expect(lexOne('\n\rfoo')).to.contain({
+      kind: TokenKind.NAME,
+      start: 2,
+      end: 5,
+      line: 3,
+      column: 1,
+      value: 'foo',
+    });
+    expect(lexOne('\r\r\n\nfoo')).to.contain({
+      kind: TokenKind.NAME,
+      start: 4,
+      end: 7,
+      line: 4,
+      column: 1,
+      value: 'foo',
+    });
+    expect(lexOne('\n\n\r\rfoo')).to.contain({
+      kind: TokenKind.NAME,
+      start: 4,
+      end: 7,
+      line: 5,
+      column: 1,
+      value: 'foo',
+    });
+  });
+
   it('records line and column', () => {
     expect(lexOne('\n \r\n \r  foo\n')).to.contain({
       kind: TokenKind.NAME,
@@ -164,6 +223,13 @@ describe('Lexer', () => {
   });
 
   it('lexes strings', () => {
+    expect(lexOne('""')).to.contain({
+      kind: TokenKind.STRING,
+      start: 0,
+      end: 2,
+      value: '',
+    });
+
     expect(lexOne('"simple"')).to.contain({
       kind: TokenKind.STRING,
       start: 0,
@@ -210,6 +276,10 @@ describe('Lexer', () => {
   it('lex reports useful string errors', () => {
     expectSyntaxError('"', 'Unterminated string.', { line: 1, column: 2 });
 
+    expectSyntaxError('"""', 'Unterminated string.', { line: 1, column: 4 });
+
+    expectSyntaxError('""""', 'Unterminated string.', { line: 1, column: 5 });
+
     expectSyntaxError('"no end quote', 'Unterminated string.', {
       line: 1,
       column: 14,
@@ -287,6 +357,13 @@ describe('Lexer', () => {
   });
 
   it('lexes block strings', () => {
+    expect(lexOne('""""""')).to.contain({
+      kind: TokenKind.BLOCK_STRING,
+      start: 0,
+      end: 6,
+      value: '',
+    });
+
     expect(lexOne('"""simple"""')).to.contain({
       kind: TokenKind.BLOCK_STRING,
       start: 0,
@@ -538,6 +615,20 @@ describe('Lexer', () => {
       column: 2,
     });
 
+    expectSyntaxError('01', 'Invalid number, unexpected digit after 0: "1".', {
+      line: 1,
+      column: 2,
+    });
+
+    expectSyntaxError(
+      '01.23',
+      'Invalid number, unexpected digit after 0: "1".',
+      {
+        line: 1,
+        column: 2,
+      },
+    );
+
     expectSyntaxError('+1', 'Cannot parse the unexpected character "+".', {
       line: 1,
       column: 1,
@@ -548,6 +639,16 @@ describe('Lexer', () => {
       column: 3,
     });
 
+    expectSyntaxError('1e', 'Invalid number, expected digit but got: <EOF>.', {
+      line: 1,
+      column: 3,
+    });
+
+    expectSyntaxError('1E', 'Invalid number, expected digit but got: <EOF>.', {
+      line: 1,
+      column: 3,
+    });
+
     expectSyntaxError('1.e1', 'Invalid number, expected digit but got: "e".', {
       line: 1,
       column: 3,
@@ -578,6 +679,33 @@ describe('Lexer', () => {
       line: 1,
       column: 5,
     });
+
+    expectSyntaxError(
+      '1.2e3e',
+      'Invalid number, expected digit but got: "e".',
+      {
+        line: 1,
+        column: 6,
+      },
+    );
+
+    expectSyntaxError(
+      '1.2e3.4',
+      'Invalid number, expected digit but got: ".".',
+      {
+        line: 1,
+        column: 6,
+      },
+    );
+
+    expectSyntaxError(
+      '1.23.4',
+      'Invalid number, expected digit but got: ".".',
+      {
+        line: 1,
+        column: 5,
+      },
+    );
   });
 
   it('lexes punctuation', () => {
diff --git a/src/language/lexer.js b/src/language/lexer.js
index ea29105d69..f2f3dfa377 100644
--- a/src/language/lexer.js
+++ b/src/language/lexer.js
@@ -444,6 +444,16 @@ function readNumber(source, start, firstCode, line, col, prev): Token {
       code = body.charCodeAt(++position);
     }
     position = readDigits(source, position, code);
+    code = body.charCodeAt(position);
+  }
+
+  // Numbers cannot be followed by . or e
+  if (code === 46 || code === 69 || code === 101) {
+    throw syntaxError(
+      source,
+      position,
+      `Invalid number, expected digit but got: ${printCharCode(code)}.`,
+    );
   }
 
   return new Tok(