Skip to content

Preserve Pseudo C string contents and encoding #8579

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
plafosse merged 3 commits into from
Oct 1, 2026
Merged

Preserve Pseudo C string contents and encoding #8579

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
fcfd35a
Preserve escapes and format sequences when wrapping strings
plafosse Sep 23, 2026
95396bc
Preserve Unicode prefixes in Pseudo C constant strings
plafosse Sep 23, 2026
4925d41
Preserve every byte of counted Pseudo C copies
plafosse Sep 23, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
79 changes: 56 additions & 23 deletions formatter/generic/genericformatter.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -156,9 +156,12 @@ struct Item
if (!tokens.empty())
{
InstructionTextToken token = tokens.front();
string trimmedText = TrimLeadingWhitespace(token.text);
token.width -= token.text.size() - trimmedText.size();
token.text = trimmedText;
if (token.type != StringToken)
{
string trimmedText = TrimLeadingWhitespace(token.text);
token.width -= token.text.size() - trimmedText.size();
token.text = trimmedText;
}
output.emplace_back(token);
output.insert(output.end(), tokens.begin() + 1, tokens.end());
firstTokenOfLine = false;
Expand Down Expand Up @@ -364,7 +367,29 @@ static vector<InstructionTextToken> ParseStringToken(
size_t start = curEnd;
curEnd++; // consume '\'
if (curEnd < tail)
curEnd++; // consume escaped char
{
char escape = src[curEnd++];
if (escape == 'x')
{
// A hexadecimal escape includes its digits. Splitting
// after the x produces an invalid C string literal.
while (curEnd < tail && isxdigit((unsigned char)src[curEnd]))
curEnd++;
}
else if (escape == 'u' || escape == 'U')
{
size_t digits = escape == 'u' ? 4 : 8;
while (digits-- && curEnd < tail && isxdigit((unsigned char)src[curEnd]))
curEnd++;
}
else if (escape >= '0' && escape <= '7')
{
// The first of at most three octal digits was consumed.
size_t digits = 2;
while (digits-- && curEnd < tail && src[curEnd] >= '0' && src[curEnd] <= '7')
curEnd++;
}
}
ConstructToken(start, curEnd);
curStart = curEnd;
}
Expand Down Expand Up @@ -399,18 +424,11 @@ static vector<InstructionTextToken> ParseStringToken(
// Check if we've exceeded max parsing length
if (curEnd > maxParsingLength)
{
// Never cut in the middle of a grapheme cluster, which would leave both sides holding a piece
// of a character that neither renders nor measures on its own. Walking the clusters gives the
// last boundary that stays within the limit.
size_t splitPos = 0;
while (splitPos < src.size())
{
size_t next = Unicode::GetNextGraphemeClusterBoundary(unprocessedStringToken.text, splitPos);
// Always consume at least one cluster to guarantee making progress
if (splitPos > 0 && next > maxParsingLength)
break;
splitPos = next;
}
// Finish the current atom/character without moving back into tokens
// already emitted (in particular an escape spanning the limit).
size_t splitPos = curStart;
while (splitPos < curEnd)
splitPos = Unicode::GetNextGraphemeClusterBoundary(unprocessedStringToken.text, splitPos);

// Flush any pending token
flushToken(curStart, splitPos);
Expand All @@ -419,7 +437,8 @@ static vector<InstructionTextToken> ParseStringToken(
InstructionTextToken remainingToken = unprocessedStringToken;
remainingToken.text = string(src.substr(splitPos));
remainingToken.width = Unicode::GetDisplayWidth(remainingToken.text);
result.emplace_back(std::move(remainingToken));
if (!remainingToken.text.empty())
result.emplace_back(std::move(remainingToken));
return result;
}
}
Expand Down Expand Up @@ -1137,10 +1156,14 @@ vector<DisassemblyTextLine> GenericLineFormatter::FormatLines(
auto newLine = [&](const bool forString = false) {
if (!firstTokenOfLine)
{
string lastTokenText = outputLine.tokens.back().text;
string trimmedText = TrimTrailingWhitespace(lastTokenText);
outputLine.tokens.back().width -= lastTokenText.size() - trimmedText.size();
outputLine.tokens.back().text = trimmedText;
// Whitespace inside a literal is data, including at a wrap point.
if (outputLine.tokens.back().type != StringToken)
{
string lastTokenText = outputLine.tokens.back().text;
string trimmedText = TrimTrailingWhitespace(lastTokenText);
outputLine.tokens.back().width -= lastTokenText.size() - trimmedText.size();
outputLine.tokens.back().text = trimmedText;
}
if (forString && outputLine.tokens.back().type == StringToken)
{
outputLine.tokens.emplace_back(BraceToken, "\"");
Expand Down Expand Up @@ -1209,8 +1232,18 @@ vector<DisassemblyTextLine> GenericLineFormatter::FormatLines(
if (desiredContinuationWidth < settings.minimumContentLength)
desiredContinuationWidth = settings.minimumContentLength;

layoutStack.push({item->items, additionalContinuationIndentation, desiredWidth,
desiredContinuationWidth, desiredStringWidth, false});
if (item->tokens.empty())
{
layoutStack.push({item->items, additionalContinuationIndentation, desiredWidth,
desiredContinuationWidth, desiredStringWidth, false});
}
else
{
// Escape/format atoms store their text in tokens, not
// child items. Keep the atom that caused the wrap.
item->AppendAllTokens(outputLine.tokens, firstTokenOfLine);
currentWidth += item->width;
}
break;
}

Expand Down
12 changes: 11 additions & 1 deletion lang/c/pseudoc.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1393,6 +1393,15 @@ void PseudoCFunction::GetExprTextInternal(const HighLevelILInstruction& instr, H
bool nullTerminates = true;
switch (builtin)
{
case BuiltinMemcpy:
{
// A byte copy needs the complete buffer, including embedded NULs.
// Unicode annotations can stop early or abbreviate long data.
tokens.Append(BraceToken, "\"");
tokens.Append(StringToken, ConstDataTokenContext, db.ToEscapedString(false, true), instr.address, data.value);
tokens.Append(BraceToken, "\"");
break;
}
case BuiltinStrcpy:
case BuiltinStrncpy:
{
Expand Down Expand Up @@ -1422,7 +1431,8 @@ void PseudoCFunction::GetExprTextInternal(const HighLevelILInstruction& instr, H
{
if (auto unicode = GetFunction()->GetView()->StringifyUnicodeData(instr.function->GetArchitecture(), db, nullTerminates); unicode.has_value())
{
auto wideStringPrefix = (builtin == BuiltinWcscpy) ? "L" : "";
auto wideStringPrefix = (builtin == BuiltinWcscpy) ? "L" :
DisassemblyTextRenderer::GetStringLiteralPrefix(unicode.value().second);
auto tokenContext = (builtin == BuiltinWcscpy) ? ConstStringDataTokenContext : ConstDataTokenContext;
tokens.Append(BraceToken, wideStringPrefix + string("\""));
tokens.Append(StringToken, tokenContext, unicode.value().first, instr.address, data.value);
Expand Down
Loading