From 2f2d7cf8a8389d714be771ac06e83fe570c894b8 Mon Sep 17 00:00:00 2001
From: Tom Maenan Read Cutting <readcuttingt@gmail.com>
Date: Sat, 27 Aug 2022 09:59:05 +0100
Subject: [PATCH] Fix carriage return crash in multiline string

Follow the guidance of #38:

> However CR directly before NL is interpreted as only a newline and not part of the multiline string. zig fmt will delete the CR.

Zig fmt already had code for deleting carriage returns, but would still
crash - now it no longer does so. Carriage returns encountered before
line-feeds are now appropriately removed on program compilation as well.
---
 doc/langref.html.in       |  3 ++-
 lib/std/zig/tokenizer.zig |  2 +-
 src/AstGen.zig            | 26 ++++++++++++++++++++++----
 3 files changed, 25 insertions(+), 6 deletions(-)

diff --git a/doc/langref.html.in b/doc/langref.html.in
index 162ba447003b..13c4c4f39651 100644
--- a/doc/langref.html.in
+++ b/doc/langref.html.in
@@ -11708,7 +11708,8 @@ fn readU32Be() u32 {}
       </p>
       <p>
       Each LF may be immediately preceded by a single CR (byte value 0x0d, code point U+000d, {#syntax#}'\r'{#endsyntax#})
-      to form a Windows style line ending, but this is discouraged.
+      to form a Windows style line ending, but this is discouraged. Note that in mulitline strings, CRLF sequences will
+      be encoded as LF when compiled into a zig program.
       A CR in any other context is not allowed.
       </p>
       <p>
diff --git a/lib/std/zig/tokenizer.zig b/lib/std/zig/tokenizer.zig
index f43aa9d79336..9718ebc3cadb 100644
--- a/lib/std/zig/tokenizer.zig
+++ b/lib/std/zig/tokenizer.zig
@@ -935,7 +935,7 @@ pub const Tokenizer = struct {
                         self.index += 1;
                         break;
                     },
-                    '\t' => {},
+                    '\t', '\r' => {},
                     else => self.checkLiteralCharacter(),
                 },
 
diff --git a/src/AstGen.zig b/src/AstGen.zig
index 1502b970178a..1687e2a7a79f 100644
--- a/src/AstGen.zig
+++ b/src/AstGen.zig
@@ -2227,7 +2227,7 @@ fn blockExprStmts(gz: *GenZir, parent_scope: *Scope, statements: []const Ast.Nod
             .assign_add_wrap => try assignOp(gz, scope, statement, .addwrap),
             .assign_mul      => try assignOp(gz, scope, statement, .mul),
             .assign_mul_wrap => try assignOp(gz, scope, statement, .mulwrap),
-            
+
             .grouped_expression => {
                 inner_node = node_data[statement].lhs;
                 continue;
@@ -9977,16 +9977,34 @@ fn strLitNodeAsString(astgen: *AstGen, node: Ast.Node.Index) !IndexSlice {
     {
         const slice = tree.tokenSlice(tok_i);
         const line_bytes = slice[2 .. slice.len - 1];
-        try string_bytes.appendSlice(gpa, line_bytes);
+        const carriage_return_count = mem.count(u8, line_bytes, "\r");
+        if (carriage_return_count > 0) {
+            try string_bytes.ensureUnusedCapacity(gpa, line_bytes.len - carriage_return_count);
+            for (line_bytes) |line_byte| {
+                if (line_byte == '\r') continue;
+                string_bytes.appendAssumeCapacity(line_byte);
+            }
+        } else {
+            try string_bytes.appendSlice(gpa, line_bytes);
+        }
         tok_i += 1;
     }
     // Following lines: each line prepends a newline.
     while (tok_i <= end) : (tok_i += 1) {
         const slice = tree.tokenSlice(tok_i);
         const line_bytes = slice[2 .. slice.len - 1];
-        try string_bytes.ensureUnusedCapacity(gpa, line_bytes.len + 1);
+
+        const carriage_return_count = mem.count(u8, line_bytes, "\r");
+        try string_bytes.ensureUnusedCapacity(gpa, line_bytes.len - carriage_return_count + 1);
         string_bytes.appendAssumeCapacity('\n');
-        string_bytes.appendSliceAssumeCapacity(line_bytes);
+        if (carriage_return_count > 0) {
+            for (line_bytes) |line_byte| {
+                if (line_byte == '\r') continue;
+                string_bytes.appendAssumeCapacity(line_byte);
+            }
+        } else {
+            string_bytes.appendSliceAssumeCapacity(line_bytes);
+        }
     }
     const len = string_bytes.items.len - str_index;
     try string_bytes.append(gpa, 0);