summaryrefslogtreecommitdiffstats
path: root/yaml-cpp/src/scanscalar.cpp
blob: 064c0867a59e652bd79373bfad8e4ea1764d27d2 (plain)
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
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
#include "scanscalar.h"
#include "scanner.h"
#include "exp.h"
#include "yaml-cpp/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;
		bool pastOpeningBreak = (params.fold == FOLD_FLOW);
		bool emptyLine = false, moreIndented = false;
		int foldedNewlineCount = 0;
		bool foldedNewlineStartedMoreIndented = false;
        std::size_t lastEscapedChar = std::string::npos;
		std::string scalar;
		params.leadingSpaces = false;

		while(INPUT) {
			// ********************************
			// Phase #1: scan until line ending
			
			std::size_t lastNonWhitespaceChar = scalar.size();
			bool escapedNewline = false;
			while(!params.end.Matches(INPUT) && !Exp::Break().Matches(INPUT)) {
				if(!INPUT)
					break;

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

				foundNonEmptyLine = true;
				pastOpeningBreak = true;

				// escaped newline? (only if we're escaping on slash)
				if(params.escape == '\\' && Exp::EscBreak().Matches(INPUT)) {
					// eat escape character and get out (but preserve trailing whitespace!)
					INPUT.get();
					lastNonWhitespaceChar = scalar.size();
                    lastEscapedChar = scalar.size();
					escapedNewline = true;
					break;
				}

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

				// otherwise, just add the damn character
				char ch = INPUT.get();
				scalar += ch;
				if(ch != ' ' && ch != '\t')
					lastNonWhitespaceChar = scalar.size();
			}

			// eof? if we're looking to eat something, then we throw
			if(!INPUT) {
				if(params.eatEnd)
					throw ParserException(INPUT.mark(), 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;
			}
			
			// do we remove trailing whitespace?
			if(params.fold == FOLD_FLOW)
				scalar.erase(lastNonWhitespaceChar);
			
			// ********************************
			// 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.mark(), ErrorMsg::TAB_IN_INDENTATION);

				if(!params.eatLeadingWhitespace)
					break;

				INPUT.eat(1);
			}

			// was this an empty line?
			bool nextEmptyLine = Exp::Break().Matches(INPUT);
			bool nextMoreIndented = Exp::Blank().Matches(INPUT);
			if(params.fold == FOLD_BLOCK && foldedNewlineCount == 0 && nextEmptyLine)
				foldedNewlineStartedMoreIndented = moreIndented;

			// for block scalars, we always start with a newline, so we should ignore it (not fold or keep)
			if(pastOpeningBreak) {
				switch(params.fold) {
					case DONT_FOLD:
						scalar += "\n";
						break;
					case FOLD_BLOCK:
						if(!emptyLine && !nextEmptyLine && !moreIndented && !nextMoreIndented && INPUT.column() >= params.indent)
							scalar += " ";
						else if(nextEmptyLine)
							foldedNewlineCount++;
						else
							scalar += "\n";
						
						if(!nextEmptyLine && foldedNewlineCount > 0) {
							scalar += std::string(foldedNewlineCount - 1, '\n');
							if(foldedNewlineStartedMoreIndented || nextMoreIndented | !foundNonEmptyLine)
								scalar += "\n";
							foldedNewlineCount = 0;
						}
						break;
					case FOLD_FLOW:
						if(nextEmptyLine)
							scalar += "\n";
						else if(!emptyLine && !nextEmptyLine && !escapedNewline)
							scalar += " ";
						break;
				}
			}

			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) {
			std::size_t pos = scalar.find_last_not_of(' ');
            if(lastEscapedChar != std::string::npos) {
                if(pos < lastEscapedChar || pos == std::string::npos)
                    pos = lastEscapedChar;
            }
			if(pos < scalar.size())
				scalar.erase(pos + 1);
		}

		switch(params.chomp) {
			case CLIP: {
				std::size_t pos = scalar.find_last_not_of('\n');
                if(lastEscapedChar != std::string::npos) {
                    if(pos < lastEscapedChar || pos == std::string::npos)
                        pos = lastEscapedChar;
                }
				if(pos == std::string::npos)
					scalar.erase();
				else if(pos + 1 < scalar.size())
					scalar.erase(pos + 2);
			} break;
			case STRIP: {
				std::size_t pos = scalar.find_last_not_of('\n');
                if(lastEscapedChar != std::string::npos) {
                    if(pos < lastEscapedChar || pos == std::string::npos)
                        pos = lastEscapedChar;
                }
				if(pos == std::string::npos)
					scalar.erase();
				else if(pos < scalar.size())
					scalar.erase(pos + 1);
			} break;
			default:
				break;
		}

		return scalar;
	}
}