| 1 | // Licensed to the .NET Foundation under one or more agreements. |
|---|---|
| 2 | // The .NET Foundation licenses this file to you under the MIT license. |
| 3 | // See the LICENSE file in the project root for more information. |
| 4 | // ============================================================ |
| 5 | // |
| 6 | // StringLexer.cpp |
| 7 | // |
| 8 | |
| 9 | |
| 10 | // |
| 11 | // Implements the StringLexer class |
| 12 | // |
| 13 | // ============================================================ |
| 14 | |
| 15 | #define DISABLE_BINDER_DEBUG_LOGGING |
| 16 | |
| 17 | #include "stringlexer.hpp" |
| 18 | #include "utils.hpp" |
| 19 | |
| 20 | #include "ex.h" |
| 21 | |
| 22 | namespace BINDER_SPACE |
| 23 | { |
| 24 | StringLexer::LEXEME_TYPE |
| 25 | StringLexer::GetNextLexeme(SString ¤tString, BOOL fPermitUnescapedQuotes) |
| 26 | { |
| 27 | BOOL fIsEscaped = FALSE; |
| 28 | WCHAR wcCurrentChar = INVALID_CHARACTER; |
| 29 | BINDER_LOG_ENTER(L"StringLexer::GetNextLexeme"); |
| 30 | |
| 31 | // Remove any white spaces |
| 32 | do |
| 33 | { |
| 34 | wcCurrentChar = PopCharacter(&fIsEscaped); |
| 35 | } |
| 36 | while (IsWhitespace(wcCurrentChar)); |
| 37 | |
| 38 | // Determine lexeme type |
| 39 | LEXEME_TYPE kLexemeType = LEXEME_TYPE_INVALID; |
| 40 | if (!fIsEscaped) |
| 41 | { |
| 42 | kLexemeType = GetLexemeType(wcCurrentChar); |
| 43 | |
| 44 | if (kLexemeType != LEXEME_TYPE_STRING) |
| 45 | { |
| 46 | return kLexemeType; |
| 47 | } |
| 48 | } |
| 49 | |
| 50 | // First character of string lexeme; push it back |
| 51 | PushCharacter(wcCurrentChar, fIsEscaped); |
| 52 | kLexemeType = ParseString(currentString, fPermitUnescapedQuotes); |
| 53 | if (kLexemeType == LEXEME_TYPE_STRING) |
| 54 | { |
| 55 | BINDER_LOG_LEAVE_HR(L"StringLexer::GetNextLexeme(LEXEME_TYPE_STRING)", S_OK); |
| 56 | } |
| 57 | else |
| 58 | { |
| 59 | BINDER_LOG_LEAVE_HR(L"StringLexer::GetNextLexeme(LEXEME_TYPE_INVALID)", |
| 60 | S_FALSE); |
| 61 | } |
| 62 | |
| 63 | return kLexemeType; |
| 64 | } |
| 65 | |
| 66 | StringLexer::LEXEME_TYPE |
| 67 | StringLexer::ParseString(SString ¤tString, BOOL fPermitUnescapedQuotes) |
| 68 | { |
| 69 | BOOL fIsFirstCharacter = TRUE; |
| 70 | WCHAR wcCurrentChar = INVALID_CHARACTER; |
| 71 | WCHAR wcOpeningQuote = INVALID_CHARACTER; |
| 72 | |
| 73 | currentString.Clear(); |
| 74 | |
| 75 | // Read until we find another lexeme that's not a string character |
| 76 | for (;;) |
| 77 | { |
| 78 | BOOL fIsEscaped = FALSE; |
| 79 | wcCurrentChar = PopCharacter(&fIsEscaped); |
| 80 | |
| 81 | if (wcCurrentChar == INVALID_CHARACTER) |
| 82 | { |
| 83 | // Found invalid character encoding |
| 84 | BINDER_LOG(L"StringLexer::ParseString: Invalid character encoding"); |
| 85 | return LEXEME_TYPE_INVALID; |
| 86 | } |
| 87 | |
| 88 | if (IsEOS(wcCurrentChar)) |
| 89 | { |
| 90 | if (IsQuoteCharacter(wcOpeningQuote)) |
| 91 | { |
| 92 | // EOS and unclosed quotes is an error |
| 93 | BINDER_LOG(L"StringLexer::ParseString: EOS and unclosed quotes"); |
| 94 | return LEXEME_TYPE_INVALID; |
| 95 | } |
| 96 | else |
| 97 | { |
| 98 | // Reached end of input and therefore of string |
| 99 | break; |
| 100 | } |
| 101 | } |
| 102 | |
| 103 | if (fIsFirstCharacter) |
| 104 | { |
| 105 | fIsFirstCharacter = FALSE; |
| 106 | |
| 107 | // If first character is quote, then record its quoteness |
| 108 | if (IsQuoteCharacter(wcCurrentChar)) |
| 109 | { |
| 110 | wcOpeningQuote = wcCurrentChar; |
| 111 | continue; |
| 112 | } |
| 113 | } |
| 114 | |
| 115 | if (wcCurrentChar == wcOpeningQuote) |
| 116 | { |
| 117 | // We've found the closing quote for a quoted string |
| 118 | break; |
| 119 | } |
| 120 | |
| 121 | if (!fPermitUnescapedQuotes && !fIsEscaped && IsQuoteCharacter(wcCurrentChar) && !IsQuoteCharacter(wcOpeningQuote)) |
| 122 | { |
| 123 | // Unescaped quotes in the middle of the string are an error |
| 124 | BINDER_LOG(L"StringLexer::ParseString: Quote in the middle of a string"); |
| 125 | return LEXEME_TYPE_INVALID; |
| 126 | } |
| 127 | |
| 128 | if (IsSeparatorChar(wcCurrentChar) && !IsQuoteCharacter(wcOpeningQuote) && !fIsEscaped) |
| 129 | { |
| 130 | // Unescaped separator char terminates the string |
| 131 | PushCharacter(wcCurrentChar, fIsEscaped); |
| 132 | break; |
| 133 | } |
| 134 | |
| 135 | // Add character to current string |
| 136 | currentString.Append(wcCurrentChar); |
| 137 | } |
| 138 | |
| 139 | if (!IsQuoteCharacter(wcOpeningQuote)) |
| 140 | { |
| 141 | // Remove trailing white spaces from unquoted string |
| 142 | BINDER_LOG(L"StringLexer::ParseString: Trimming string"); |
| 143 | TrimTrailingWhiteSpaces(currentString); |
| 144 | } |
| 145 | |
| 146 | BINDER_LOG_STRING(L"string", currentString); |
| 147 | |
| 148 | return LEXEME_TYPE_STRING; |
| 149 | } |
| 150 | |
| 151 | void StringLexer::TrimTrailingWhiteSpaces(SString ¤tString) |
| 152 | { |
| 153 | SString::Iterator begin = currentString.Begin(); |
| 154 | SString::Iterator cursor = currentString.End() - 1; |
| 155 | BOOL fFoundWhiteSpace = FALSE; |
| 156 | |
| 157 | for (;;) |
| 158 | { |
| 159 | if ((cursor >= begin) && IsWhitespace(cursor[0])) |
| 160 | { |
| 161 | fFoundWhiteSpace = TRUE; |
| 162 | cursor--; |
| 163 | continue; |
| 164 | } |
| 165 | break; |
| 166 | } |
| 167 | |
| 168 | if (fFoundWhiteSpace) |
| 169 | { |
| 170 | currentString.Truncate(cursor + 1); |
| 171 | } |
| 172 | } |
| 173 | }; |
| 174 |