scanscalar.cpp 4.16 KB
Newer Older
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
#include "crt.h"
#include "scanscalar.h"
#include "scanner.h"
#include "exp.h"
#include "exceptions.h"
#include "token.h"

namespace YAML
{
	// ScanScalar
	// . This is where the scalar magic happens.
	//
	// . We do the scanning in three phases:
	//   1. Scan until newline
	//   2. Eat newline
	//   3. Scan leading blanks.
	//
	// . Depending on the parameters given, we store or stop
	//   and different places in the above flow.
	std::string ScanScalar(Stream& INPUT, ScanScalarParams& params)
	{
		bool foundNonEmptyLine = false, pastOpeningBreak = false;
		bool emptyLine = false, moreIndented = false;
		std::string scalar;
		params.leadingSpaces = false;

		while(INPUT) {
			// ********************************
			// Phase #1: scan until line ending
			while(!params.end.Matches(INPUT) && !Exp::Break.Matches(INPUT)) {
				if(INPUT.peek() == EOF)
					break;

				// document indicator?
				if(INPUT.column == 0 && Exp::DocIndicator.Matches(INPUT)) {
					if(params.onDocIndicator == BREAK)
						break;
					else if(params.onDocIndicator == THROW)
						throw ParserException(INPUT.line, INPUT.column, ErrorMsg::DOC_IN_SCALAR);
				}

				foundNonEmptyLine = true;
				pastOpeningBreak = true;

				// escaped newline? (only if we're escaping on slash)
				if(params.escape == '\\' && Exp::EscBreak.Matches(INPUT)) {
					int n = Exp::EscBreak.Match(INPUT);
					INPUT.eat(n);
					continue;
				}

				// escape this?
				if(INPUT.peek() == params.escape) {
					scalar += Exp::Escape(INPUT);
					continue;
				}

				// otherwise, just add the damn character
				scalar += INPUT.get();
			}

			// eof? if we're looking to eat something, then we throw
			if(INPUT.peek() == EOF) {
				if(params.eatEnd)
					throw ParserException(INPUT.line, INPUT.column, ErrorMsg::EOF_IN_SCALAR);
				break;
			}

			// doc indicator?
			if(params.onDocIndicator == BREAK && INPUT.column == 0 && Exp::DocIndicator.Matches(INPUT))
				break;

			// are we done via character match?
			int n = params.end.Match(INPUT);
			if(n >= 0) {
				if(params.eatEnd)
					INPUT.eat(n);
				break;
			}

			// ********************************
			// Phase #2: eat line ending
			n = Exp::Break.Match(INPUT);
			INPUT.eat(n);

			// ********************************
			// Phase #3: scan initial spaces

			// first the required indentation
			while(INPUT.peek() == ' ' && (INPUT.column < params.indent || (params.detectIndent && !foundNonEmptyLine)))
				INPUT.eat(1);

			// update indent if we're auto-detecting
			if(params.detectIndent && !foundNonEmptyLine)
				params.indent = std::max(params.indent, INPUT.column);

			// and then the rest of the whitespace
			while(Exp::Blank.Matches(INPUT)) {
				// we check for tabs that masquerade as indentation
				if(INPUT.peek() == '\t'&& INPUT.column < params.indent && params.onTabInIndentation == THROW)
					throw ParserException(INPUT.line, INPUT.column, ErrorMsg::TAB_IN_INDENTATION);

				if(!params.eatLeadingWhitespace)
					break;

				INPUT.eat(1);
			}

			// was this an empty line?
			bool nextEmptyLine = Exp::Break.Matches(INPUT);
			bool nextMoreIndented = (INPUT.peek() == ' ');

			// for block scalars, we always start with a newline, so we should ignore it (not fold or keep)
			bool useNewLine = pastOpeningBreak;
			// and for folded scalars, we don't fold the very last newline to a space
			if(params.fold && !emptyLine && INPUT.column < params.indent)
				useNewLine = false;

			if(useNewLine) {
				if(params.fold && !emptyLine && !nextEmptyLine && !moreIndented && !nextMoreIndented)
					scalar += " ";
				else
					scalar += "\n";
			}

			emptyLine = nextEmptyLine;
			moreIndented = nextMoreIndented;
			pastOpeningBreak = true;

			// are we done via indentation?
			if(!emptyLine && INPUT.column < params.indent) {
				params.leadingSpaces = true;
				break;
			}
		}

		// post-processing
		if(params.trimTrailingSpaces) {
			unsigned pos = scalar.find_last_not_of(' ');
			if(pos < scalar.size())
				scalar.erase(pos + 1);
		}

		if(params.chomp <= 0) {
			unsigned pos = scalar.find_last_not_of('\n');
			if(params.chomp == 0 && pos + 1 < scalar.size())
				scalar.erase(pos + 2);
			else if(params.chomp == -1 && pos < scalar.size())
				scalar.erase(pos + 1);
		}

		return scalar;
	}
}