diff options
Diffstat (limited to 'src')
-rw-r--r-- | src/core/utils/BufferedCharReader.cpp | 4 | ||||
-rw-r--r-- | src/core/utils/BufferedCharReader.hpp | 6 | ||||
-rw-r--r-- | src/core/utils/Tokenizer.cpp | 113 | ||||
-rw-r--r-- | src/core/utils/Tokenizer.hpp | 48 |
4 files changed, 148 insertions, 23 deletions
diff --git a/src/core/utils/BufferedCharReader.cpp b/src/core/utils/BufferedCharReader.cpp index 0377015..c13628f 100644 --- a/src/core/utils/BufferedCharReader.cpp +++ b/src/core/utils/BufferedCharReader.cpp @@ -1,6 +1,6 @@ /* - SCAENEA IDL Compiler (scidlc) - Copyright (C) 2014 Andreas Stöckel + Ousía + Copyright (C) 2014, 2015 Benjamin Paaßen, Andreas Stöckel This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by diff --git a/src/core/utils/BufferedCharReader.hpp b/src/core/utils/BufferedCharReader.hpp index 86f43b5..b13cde6 100644 --- a/src/core/utils/BufferedCharReader.hpp +++ b/src/core/utils/BufferedCharReader.hpp @@ -1,6 +1,6 @@ /* - SCAENEA IDL Compiler (scidlc) - Copyright (C) 2014 Andreas Stöckel + Ousía + Copyright (C) 2014, 2015 Benjamin Paaßen, Andreas Stöckel This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by @@ -236,5 +236,5 @@ public: } } -#endif /* _OUSISA_UTILS_BUFFERED_CHAR_READER_H_ */ +#endif /* _OUSIA_UTILS_BUFFERED_CHAR_READER_H_ */ diff --git a/src/core/utils/Tokenizer.cpp b/src/core/utils/Tokenizer.cpp index 38f7585..2c36438 100644 --- a/src/core/utils/Tokenizer.cpp +++ b/src/core/utils/Tokenizer.cpp @@ -16,6 +16,8 @@ along with this program. If not, see <http://www.gnu.org/licenses/>. */ +#include <sstream> + #include "Tokenizer.hpp" namespace ousia { @@ -52,10 +54,10 @@ static std::map<char, TokenTreeNode> buildChildren( static int buildId(const std::map<std::string, int> &inputs) { - int tokenId = -1; + int tokenId = TOKEN_NONE; for (auto &e : inputs) { if (e.first.empty()) { - if (tokenId != -1) { + if (tokenId != TOKEN_NONE) { throw TokenizerException{std::string{"Ambigous token found: "} + std::to_string(e.second)}; } else { @@ -68,8 +70,115 @@ static int buildId(const std::map<std::string, int> &inputs) TokenTreeNode::TokenTreeNode(const std::map<std::string, int> &inputs) : children(buildChildren(inputs)), tokenId(buildId(inputs)) +{ +} + +Tokenizer::Tokenizer(BufferedCharReader &input, const TokenTreeNode &root) + : input(input), root(root) +{ +} + +bool Tokenizer::prepare() +{ + std::stringstream buffer; + char c; + const int startColumn = input.getColumn(); + const int startLine = input.getLine(); + bool bufEmpty = true; + while (input.peek(&c)) { + if (root.children.find(c) != root.children.end()) { + // if there might be a special token, keep peeking forward + // until we find the token (or we don't). + TokenTreeNode const *n = &root; + std::stringstream tBuf; + int match = TOKEN_NONE; + while (true) { + tBuf << c; + n = &(n->children.at(c)); + if (n->tokenId != TOKEN_NONE) { + // from here on we found a token. If we have something + // in our buffer already, we end the search now. + if (!bufEmpty) { + break; + } else { + // if we want to return this token ( = we have nothing + // in our buffer yet) we look greedily for the longest + // possible token we can construct. + input.consumePeek(); + } + } + if (!input.peek(&c)) { + // if we are at the end we break off the search. + break; + } + if (n->children.find(c) == root.children.end()) { + // if we do not find a possible continuation anymore, + // break off the search. + break; + } + } + // check if we did indeed find a special token. + if (match != TOKEN_NONE) { + input.resetPeek(); + if (bufEmpty) { + // if we did not have text before, construct that token. + peeked.push_back(Token{match, tBuf.str(), startColumn, + startLine, input.getColumn(), + input.getLine()}); + return true; + } else { + // otherwise we return the text before the token. + peeked.push_back(Token{TOKEN_TEXT, buffer.str(), + startColumn, startLine, + input.getColumn(), input.getLine()}); + return true; + } + } + } + buffer << c; + bufEmpty = false; + input.consumePeek(); + } + if (!bufEmpty) { + peeked.push_back(Token{TOKEN_TEXT, buffer.str(), startColumn, startLine, + input.getColumn(), input.getLine()}); + return true; + } + return false; +} +bool Tokenizer::next(Token &t) { + if (peeked.empty()) { + if (!prepare()) { + return false; + } + } + t = peeked.front(); + peeked.pop_front(); + resetPeek(); + return true; +} + +bool Tokenizer::peek(Token &t) +{ + if (peekCursor >= peeked.size()) { + if (!prepare()) { + return false; + } + } + t = peeked[peekCursor]; + return true; +} + +void Tokenizer::resetPeek() { peekCursor = 0; } + +void Tokenizer::consumePeek() +{ + while (peekCursor > 0) { + peeked.pop_front(); + peekCursor--; + } } } } diff --git a/src/core/utils/Tokenizer.hpp b/src/core/utils/Tokenizer.hpp index 24c4f30..924b670 100644 --- a/src/core/utils/Tokenizer.hpp +++ b/src/core/utils/Tokenizer.hpp @@ -21,7 +21,9 @@ #include <istream> #include <map> -#include <queue> +#include <deque> + +#include "BufferedCharReader.hpp" namespace ousia { namespace utils { @@ -44,33 +46,47 @@ public: }; struct Token { - const int tokenId; - const std::string content; - const int column; - const int line; - - Token(int tokenId, std::string content, int column, int line) - : tokenId(tokenId), content(content), column(column), line(line) + int tokenId; + std::string content; + int startColumn; + int startLine; + int endColumn; + int endLine; + + Token(int tokenId, std::string content, int startColumn, int startLine, + int endColumn, int endLine) + : tokenId(tokenId), + content(content), + startColumn(startColumn), + startLine(startLine), + endColumn(endColumn), + endLine(endLine) { } }; +static const int TOKEN_NONE = -1; +static const int TOKEN_TEXT = -2; + class Tokenizer { private: - const std::istream &input; - const TokenTreeNode root; - const std::queue<Token> peekQueue; + BufferedCharReader &input; + const TokenTreeNode &root; + std::deque<Token> peeked; + unsigned int peekCursor = 0; + + bool prepare(); public: - Tokenizer(const TokenTreeNode &root, std::istream &input); + Tokenizer(BufferedCharReader &input, const TokenTreeNode &root); - bool hasNext(); + bool next(Token &t); - const Token &next(); + bool peek(Token &t); - const Token &peek(); + void resetPeek(); - void reset(); + void consumePeek(); }; } } |