mirror of
https://github.com/RPCS3/rpcs3.git
synced 2025-07-10 00:41:26 +12:00
Refactored common scalar scanning code (from plain, quoted, and block) to one function.
This commit is contained in:
Jesse Beder
parent
6c193d6fbd
commit
0d5a97bffe
6 changed files with 306 additions and 188 deletions
408
scanscalar.cpp
408
scanscalar.cpp
|
|
@ -75,74 +75,77 @@ namespace YAML
|
|||
// and in-line whitespace (which is kept) separately.
|
||||
template <> PlainScalarToken *Scanner::ScanToken(PlainScalarToken *pToken)
|
||||
{
|
||||
//// now eat and store the scalar
|
||||
//std::string scalar;
|
||||
//WhitespaceInfo info;
|
||||
|
||||
//while(INPUT) {
|
||||
// // doc start/end tokens
|
||||
// if(IsDocumentStart() || IsDocumentEnd())
|
||||
// break;
|
||||
|
||||
// // comment
|
||||
// if(Exp::Comment.Matches(INPUT))
|
||||
// break;
|
||||
|
||||
// // first eat non-blanks
|
||||
// while(INPUT && !Exp::BlankOrBreak.Matches(INPUT)) {
|
||||
// // illegal colon in flow context
|
||||
// if(m_flowLevel > 0 && Exp::IllegalColonInScalar.Matches(INPUT))
|
||||
// throw IllegalScalar();
|
||||
|
||||
// // characters that might end the scalar
|
||||
// if(m_flowLevel > 0 && Exp::EndScalarInFlow.Matches(INPUT))
|
||||
// break;
|
||||
// if(m_flowLevel == 0 && Exp::EndScalar.Matches(INPUT))
|
||||
// break;
|
||||
|
||||
// // finally, read the character!
|
||||
// scalar += GetChar();
|
||||
// }
|
||||
|
||||
// // did we hit a non-blank character that ended us?
|
||||
// if(!Exp::BlankOrBreak.Matches(INPUT))
|
||||
// break;
|
||||
|
||||
// // now eat blanks
|
||||
// while(INPUT && Exp::BlankOrBreak.Matches(INPUT)) {
|
||||
// if(Exp::Blank.Matches(INPUT)) {
|
||||
// // can't use tabs as indentation! only spaces!
|
||||
// if(INPUT.peek() == '\t' && info.leadingBlanks && m_column <= m_indents.top())
|
||||
// throw IllegalTabInScalar();
|
||||
|
||||
// info.AddBlank(GetChar());
|
||||
// } else {
|
||||
// // we know it's a line break; see how many characters to read
|
||||
// int n = Exp::Break.Match(INPUT);
|
||||
// std::string line = GetChar(n);
|
||||
// info.AddBreak(line);
|
||||
|
||||
// // and we can't continue a simple key to the next line
|
||||
// ValidateSimpleKey();
|
||||
// }
|
||||
// }
|
||||
|
||||
// // break if we're below the indentation level
|
||||
// if(m_flowLevel == 0 && m_column <= m_indents.top())
|
||||
// break;
|
||||
|
||||
// // finally join whitespace
|
||||
// scalar += info.Join();
|
||||
//}
|
||||
|
||||
RegEx end = (m_flowLevel > 0 ? Exp::EndScalarInFlow : Exp::EndScalar) || (RegEx(' ') + Exp::Comment);
|
||||
int indent = (m_flowLevel > 0 ? 0 : m_indents.top() + 1);
|
||||
|
||||
// insert a potential simple key
|
||||
if(m_simpleKeyAllowed)
|
||||
InsertSimpleKey();
|
||||
|
||||
pToken->value = ScanScalar(end, false, indent, 0, true, true, true, 0);
|
||||
|
||||
m_simpleKeyAllowed = false;
|
||||
|
||||
// now eat and store the scalar
|
||||
std::string scalar;
|
||||
WhitespaceInfo info;
|
||||
|
||||
while(INPUT) {
|
||||
// doc start/end tokens
|
||||
if(IsDocumentStart() || IsDocumentEnd())
|
||||
break;
|
||||
|
||||
// comment
|
||||
if(Exp::Comment.Matches(INPUT))
|
||||
break;
|
||||
|
||||
// first eat non-blanks
|
||||
while(INPUT && !Exp::BlankOrBreak.Matches(INPUT)) {
|
||||
// illegal colon in flow context
|
||||
if(m_flowLevel > 0 && Exp::IllegalColonInScalar.Matches(INPUT))
|
||||
throw IllegalScalar();
|
||||
|
||||
// characters that might end the scalar
|
||||
if(m_flowLevel > 0 && Exp::EndScalarInFlow.Matches(INPUT))
|
||||
break;
|
||||
if(m_flowLevel == 0 && Exp::EndScalar.Matches(INPUT))
|
||||
break;
|
||||
|
||||
// finally, read the character!
|
||||
scalar += GetChar();
|
||||
}
|
||||
|
||||
// did we hit a non-blank character that ended us?
|
||||
if(!Exp::BlankOrBreak.Matches(INPUT))
|
||||
break;
|
||||
|
||||
// now eat blanks
|
||||
while(INPUT && Exp::BlankOrBreak.Matches(INPUT)) {
|
||||
if(Exp::Blank.Matches(INPUT)) {
|
||||
// can't use tabs as indentation! only spaces!
|
||||
if(INPUT.peek() == '\t' && info.leadingBlanks && m_column <= m_indents.top())
|
||||
throw IllegalTabInScalar();
|
||||
|
||||
info.AddBlank(GetChar());
|
||||
} else {
|
||||
// we know it's a line break; see how many characters to read
|
||||
int n = Exp::Break.Match(INPUT);
|
||||
std::string line = GetChar(n);
|
||||
info.AddBreak(line);
|
||||
|
||||
// and we can't continue a simple key to the next line
|
||||
ValidateSimpleKey();
|
||||
}
|
||||
}
|
||||
|
||||
// break if we're below the indentation level
|
||||
if(m_flowLevel == 0 && m_column <= m_indents.top())
|
||||
break;
|
||||
|
||||
// finally join whitespace
|
||||
scalar += info.Join();
|
||||
}
|
||||
|
||||
// now modify our token
|
||||
pToken->value = scalar;
|
||||
if(info.leadingBlanks)
|
||||
if(true/*info.leadingBlanks*/)
|
||||
m_simpleKeyAllowed = true;
|
||||
|
||||
return pToken;
|
||||
|
|
@ -151,91 +154,92 @@ namespace YAML
|
|||
// QuotedScalarToken
|
||||
template <> QuotedScalarToken *Scanner::ScanToken(QuotedScalarToken *pToken)
|
||||
{
|
||||
// insert a potential simple key
|
||||
if(m_simpleKeyAllowed)
|
||||
InsertSimpleKey();
|
||||
m_simpleKeyAllowed = false;
|
||||
//// now eat and store the scalar
|
||||
//std::string scalar;
|
||||
//WhitespaceInfo info;
|
||||
|
||||
//while(INPUT) {
|
||||
// if(IsDocumentStart() || IsDocumentEnd())
|
||||
// throw DocIndicatorInQuote();
|
||||
|
||||
// if(INPUT.peek() == EOF)
|
||||
// throw EOFInQuote();
|
||||
|
||||
// // first eat non-blanks
|
||||
// while(INPUT && !Exp::BlankOrBreak.Matches(INPUT)) {
|
||||
// // escaped single quote?
|
||||
// if(pToken->single && Exp::EscSingleQuote.Matches(INPUT)) {
|
||||
// int n = Exp::EscSingleQuote.Match(INPUT);
|
||||
// scalar += GetChar(n);
|
||||
// continue;
|
||||
// }
|
||||
|
||||
// // is the quote ending?
|
||||
// if(INPUT.peek() == quote)
|
||||
// break;
|
||||
|
||||
// // escaped newline?
|
||||
// if(Exp::EscBreak.Matches(INPUT))
|
||||
// break;
|
||||
|
||||
// // other escape sequence
|
||||
// if(INPUT.peek() == '\\') {
|
||||
// int length = 0;
|
||||
// scalar += Exp::Escape(INPUT, length);
|
||||
// m_column += length;
|
||||
// continue;
|
||||
// }
|
||||
|
||||
// // and finally, just add the damn character
|
||||
// scalar += GetChar();
|
||||
// }
|
||||
|
||||
// // is the quote ending?
|
||||
// if(INPUT.peek() == quote) {
|
||||
// // eat and go
|
||||
// GetChar();
|
||||
// break;
|
||||
// }
|
||||
|
||||
// // now we eat blanks
|
||||
// while(Exp::BlankOrBreak.Matches(INPUT)) {
|
||||
// if(Exp::Blank.Matches(INPUT)) {
|
||||
// info.AddBlank(GetChar());
|
||||
// } else {
|
||||
// // we know it's a line break; see how many characters to read
|
||||
// int n = Exp::Break.Match(INPUT);
|
||||
// std::string line = GetChar(n);
|
||||
// info.AddBreak(line);
|
||||
|
||||
// // and we can't continue a simple key to the next line
|
||||
// ValidateSimpleKey();
|
||||
// }
|
||||
// }
|
||||
|
||||
// // and finally join the whitespace
|
||||
// scalar += info.Join();
|
||||
//}
|
||||
|
||||
// eat single or double quote
|
||||
char quote = GetChar();
|
||||
pToken->single = (quote == '\'');
|
||||
|
||||
// now eat and store the scalar
|
||||
std::string scalar;
|
||||
WhitespaceInfo info;
|
||||
RegEx end = (pToken->single ? RegEx(quote) && !Exp::EscSingleQuote : RegEx(quote));
|
||||
char escape = (pToken->single ? '\'' : '\\');
|
||||
|
||||
while(INPUT) {
|
||||
if(IsDocumentStart() || IsDocumentEnd())
|
||||
throw DocIndicatorInQuote();
|
||||
// insert a potential simple key
|
||||
if(m_simpleKeyAllowed)
|
||||
InsertSimpleKey();
|
||||
|
||||
if(INPUT.peek() == EOF)
|
||||
throw EOFInQuote();
|
||||
pToken->value = ScanScalar(end, true, 0, escape, true, true, false, 0);
|
||||
m_simpleKeyAllowed = false;
|
||||
|
||||
// first eat non-blanks
|
||||
while(INPUT && !Exp::BlankOrBreak.Matches(INPUT)) {
|
||||
// escaped single quote?
|
||||
if(pToken->single && Exp::EscSingleQuote.Matches(INPUT)) {
|
||||
int n = Exp::EscSingleQuote.Match(INPUT);
|
||||
scalar += GetChar(n);
|
||||
continue;
|
||||
}
|
||||
|
||||
// is the quote ending?
|
||||
if(INPUT.peek() == quote)
|
||||
break;
|
||||
|
||||
// escaped newline?
|
||||
if(Exp::EscBreak.Matches(INPUT))
|
||||
break;
|
||||
|
||||
// other escape sequence
|
||||
if(INPUT.peek() == '\\') {
|
||||
int length = 0;
|
||||
scalar += Exp::Escape(INPUT, length);
|
||||
m_column += length;
|
||||
continue;
|
||||
}
|
||||
|
||||
// and finally, just add the damn character
|
||||
scalar += GetChar();
|
||||
}
|
||||
|
||||
// is the quote ending?
|
||||
if(INPUT.peek() == quote) {
|
||||
// eat and go
|
||||
GetChar();
|
||||
break;
|
||||
}
|
||||
|
||||
// now we eat blanks
|
||||
while(Exp::BlankOrBreak.Matches(INPUT)) {
|
||||
if(Exp::Blank.Matches(INPUT)) {
|
||||
info.AddBlank(GetChar());
|
||||
} else {
|
||||
// we know it's a line break; see how many characters to read
|
||||
int n = Exp::Break.Match(INPUT);
|
||||
std::string line = GetChar(n);
|
||||
info.AddBreak(line);
|
||||
|
||||
// and we can't continue a simple key to the next line
|
||||
ValidateSimpleKey();
|
||||
}
|
||||
}
|
||||
|
||||
// and finally join the whitespace
|
||||
scalar += info.Join();
|
||||
}
|
||||
|
||||
pToken->value = scalar;
|
||||
return pToken;
|
||||
}
|
||||
|
||||
// BlockScalarToken
|
||||
template <> BlockScalarToken *Scanner::ScanToken(BlockScalarToken *pToken)
|
||||
{
|
||||
// simple keys always ok after block scalars (since we're gonna start a new line anyways)
|
||||
m_simpleKeyAllowed = true;
|
||||
|
||||
WhitespaceInfo info;
|
||||
|
||||
// eat block indicator ('|' or '>')
|
||||
|
|
@ -268,37 +272,13 @@ namespace YAML
|
|||
if(info.increment && m_indents.top() >= 0)
|
||||
indent += m_indents.top();
|
||||
|
||||
// finally, grab that scalar
|
||||
std::string scalar;
|
||||
while(INPUT) {
|
||||
// initialize indentation
|
||||
GetBlockIndentation(indent, info.trailingBreaks);
|
||||
GetBlockIndentation(indent, info.trailingBreaks);
|
||||
|
||||
// are we done with this guy (i.e. at a lower indentation?)
|
||||
if(m_column != indent)
|
||||
break;
|
||||
|
||||
bool trailingBlank = Exp::Blank.Matches(INPUT);
|
||||
scalar += info.Join();
|
||||
|
||||
bool leadingBlank = Exp::Blank.Matches(INPUT);
|
||||
|
||||
// now eat and save the line
|
||||
while(INPUT.peek() != EOF && !Exp::Break.Matches(INPUT))
|
||||
scalar += GetChar();
|
||||
|
||||
// we know it's a line break; see how many characters to read
|
||||
int n = Exp::Break.Match(INPUT);
|
||||
std::string line = GetChar(n);
|
||||
info.AddBreak(line);
|
||||
}
|
||||
|
||||
// one last whitespace join (with chompers this time)
|
||||
scalar += info.Join(true);
|
||||
|
||||
// finally set the scalar
|
||||
pToken->value = scalar;
|
||||
bool eatLeadingWhitespace = false;
|
||||
pToken->value = ScanScalar(RegEx(), false, indent, 0, info.fold, eatLeadingWhitespace, false, info.chomp);
|
||||
|
||||
// simple keys always ok after block scalars (since we're gonna start a new line anyways)
|
||||
m_simpleKeyAllowed = true;
|
||||
return pToken;
|
||||
}
|
||||
|
||||
|
|
@ -340,4 +320,104 @@ namespace YAML
|
|||
indent = 1;
|
||||
}
|
||||
}
|
||||
|
||||
// ScanScalar
|
||||
std::string Scanner::ScanScalar(RegEx end, bool eatEnd, int indent, char escape, bool fold, bool eatLeadingWhitespace, bool trimTrailingSpaces, int chomp)
|
||||
{
|
||||
bool emptyLine = false, moreIndented = false;
|
||||
std::string scalar;
|
||||
|
||||
while(INPUT) {
|
||||
// ********************************
|
||||
// Phase #1: scan until line ending
|
||||
while(!end.Matches(INPUT) && !Exp::Break.Matches(INPUT)) {
|
||||
if(INPUT.peek() == EOF)
|
||||
break;
|
||||
|
||||
// escaped newline? (only if we're escaping on slash)
|
||||
if(escape == '\\' && Exp::EscBreak.Matches(INPUT)) {
|
||||
int n = Exp::EscBreak.Match(INPUT);
|
||||
Eat(n);
|
||||
continue;
|
||||
}
|
||||
|
||||
// escape this?
|
||||
if(INPUT.peek() == escape) {
|
||||
int length = 0;
|
||||
scalar += Exp::Escape(INPUT, length);
|
||||
m_column += length;
|
||||
continue;
|
||||
}
|
||||
|
||||
// otherwise, just add the damn character
|
||||
scalar += GetChar();
|
||||
}
|
||||
|
||||
// eof? if we're looking to eat something, then we throw
|
||||
if(INPUT.peek() == EOF) {
|
||||
if(eatEnd)
|
||||
throw EOFInQuote();
|
||||
break;
|
||||
}
|
||||
|
||||
// are we done via character match?
|
||||
int n = end.Match(INPUT);
|
||||
if(n >= 0) {
|
||||
if(eatEnd)
|
||||
Eat(n);
|
||||
break;
|
||||
}
|
||||
|
||||
// ********************************
|
||||
// Phase #2: eat line ending
|
||||
n = Exp::Break.Match(INPUT);
|
||||
Eat(n);
|
||||
|
||||
// ********************************
|
||||
// Phase #3: scan initial spaces
|
||||
|
||||
// first the required indentation
|
||||
while(INPUT.peek() == ' ' && m_column < indent)
|
||||
Eat(1);
|
||||
|
||||
// and then the rest of the whitespace
|
||||
if(eatLeadingWhitespace) {
|
||||
while(Exp::Blank.Matches(INPUT))
|
||||
Eat(1);
|
||||
}
|
||||
|
||||
// was this an empty line?
|
||||
bool nextEmptyLine = Exp::Break.Matches(INPUT);
|
||||
bool nextMoreIndented = (INPUT.peek() == ' ');
|
||||
|
||||
if(fold && !emptyLine && !nextEmptyLine && !moreIndented && !nextMoreIndented)
|
||||
scalar += " ";
|
||||
else
|
||||
scalar += "\n";
|
||||
|
||||
emptyLine = nextEmptyLine;
|
||||
moreIndented = nextMoreIndented;
|
||||
|
||||
// are we done via indentation?
|
||||
if(!emptyLine && m_column < indent)
|
||||
break;
|
||||
}
|
||||
|
||||
// post-processing
|
||||
if(trimTrailingSpaces) {
|
||||
unsigned pos = scalar.find_last_not_of(' ');
|
||||
if(pos < scalar.size())
|
||||
scalar.erase(pos + 1);
|
||||
}
|
||||
|
||||
if(chomp <= 0) {
|
||||
unsigned pos = scalar.find_last_not_of('\n');
|
||||
if(chomp == 0 && pos + 1 < scalar.size())
|
||||
scalar.erase(pos + 2);
|
||||
else if(chomp == -1 && pos < scalar.size())
|
||||
scalar.erase(pos + 1);
|
||||
}
|
||||
|
||||
return scalar;
|
||||
}
|
||||
}
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue