averello/swift-mirror
1
0
Fork 0
mirror of https://github.com/apple/swift.git synced 2025-12-21 12:14:44 +01:00
Code Issues Packages Projects Releases Wiki Activity

Fixed the escaping of strings in swift::json::Output.

Browse Source 
Instead of using llvm::raw_ostream::write_escaped (which does not produce valid
JSON strings), implemented custom escaping logic based on the JSON standard,
which only requires that the following characters be escaped:

  - Quotation mark (U+0022)
  - Reverse solidus (U+005C)
  - Control characters (U+0000 to U+001F)

Since these characters all fit within a single UTF8 byte, and will not be
present in a multi-byte UTF8 representation, simply check whether the current
byte needs to be escaped according to those requirements. If the current byte
needs to be escaped, then print out the escaped version of the byte; otherwise,
pass the current byte to the stream directly.

This fixes <rdar://problem/18266570>.

Swift SVN r21892
This commit is contained in:
Connor Wakamo
2014-09-11 22:20:05 +00:00
parent 4647eb9601
commit 725f501e9b
1 changed files with 63 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

64
lib/Basic/JSONSerialization.cpp
View File

@@ -11,6 +11,7 @@
//===----------------------------------------------------------------------===//
#include "swift/Basic/JSONSerialization.h"
#include "llvm/ADT/StringExtras.h"
#include "llvm/Support/Format.h"
using namespace swift::json;
@@ -154,7 +155,68 @@ void Output::endBitSetScalar() {
void Output::scalarString(StringRef &S, bool MustQuote) {
if (MustQuote) {
Stream << '"';
Stream.write_escaped(S);
for (unsigned char c : S) {
// According to the JSON standard, the following characters must be
// escaped:
// - Quotation mark (U+0022)
// - Reverse solidus (U+005C)
// - Control characters (U+0000 to U+001F)
// We need to check for these and escape them if present.
//
// Since these are represented by a single byte in UTF8 (and will not be
// present in any multi-byte UTF8 representations), we can just switch on
// the value of the current byte.
//
// Any other bytes present in the string should therefore be emitted
// as-is, without any escaping.
switch (c) {
// First, check for characters for which JSON has custom escape sequences.
case '"':
Stream << '\\' << '"';
break;
case '\\':
Stream << '\\' << '\\';
break;
case '/':
Stream << '\\' << '/';
break;
case '\b':
Stream << '\\' << 'b';
break;
case '\f':
Stream << '\\' << 'f';
break;
case '\n':
Stream << '\\' << 'n';
break;
case '\r':
Stream << '\\' << 'r';
break;
case '\t':
Stream << '\\' << 't';
break;
default:
// Otherwise, check to see if the current byte is a control character.
if (c >= '\x00' && c <= '\x1F') {
// Since we have a control character, we need to escape it using
// JSON's only valid escape sequence: \uxxxx (where x is a hex digit).
// The upper two digits for control characters are always 00.
Stream << "\\u00";
// Convert the current character into hexadecimal digits.
Stream << llvm::hexdigit((c >> 4) & 0xF);
Stream << llvm::hexdigit((c >> 0) & 0xF);
}
else {
// This isn't a control character, so we don't need to escape it.
// As a result, emit it directly; if it's part of a multi-byte UTF8
// representation, all bytes will be emitted in this fashion.
Stream << c;
}
break;
}
}
Stream << '"';
}
else