cppcheck/lib/token.h

507 lines
17 KiB
C
Raw Normal View History

2008-12-18 22:28:57 +01:00
/*
* Cppcheck - A tool for static C/C++ code analysis
2012-01-01 00:05:37 +01:00
* Copyright (C) 2007-2012 Daniel Marjamäki and Cppcheck team.
2008-12-18 22:28:57 +01:00
*
* 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
* the Free Software Foundation, either version 3 of the License, or
* (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program. If not, see <http://www.gnu.org/licenses/>.
2008-12-18 22:28:57 +01:00
*/
#ifndef TokenH
#define TokenH
2008-12-18 22:28:57 +01:00
#include <string>
#include <vector>
#include <ostream>
2008-12-18 22:28:57 +01:00
/// @addtogroup Core
/// @{
2009-07-13 13:35:33 +02:00
/**
* @brief The token list that the TokenList generates is a linked-list of this class.
2009-07-13 13:35:33 +02:00
*
* Tokens are stored as strings. The "if", "while", etc are stored in plain text.
* The reason the Token class is needed (instead of using the string class) is that some extra functionality is also needed for tokens:
* - location of the token is stored (linenr, fileIndex)
* - functions for classifying the token (isName, isNumber, isBoolean, isStandardType)
*
* The Token class also has other functions for management of token list, matching tokens, etc.
*/
2011-10-13 20:53:06 +02:00
class Token {
private:
Token **tokensBack;
// Not implemented..
Token();
public:
enum Type {
eVariable, eType, eFunction, eName, // Names: Variable (varId), Type (typeId, later), Function (FuncId, later), Name (unknown identifier)
eNumber, eString, eChar, eBoolean, eLiteral, // Literals: Number, String, Character, User defined literal (C++11)
eArithmeticalOp, eComparisonOp, eAssignmentOp, eLogicalOp, eBitOp, eIncDecOp, eExtendedOp, // Operators: Arithmetical, Comparison, Assignment, Logical, Bitwise, ++/--, Extended
eBracket, // {, }, <, >: < and > only if link() is set. Otherwise they are comparison operators.
eOther,
eNone
};
explicit Token(Token **tokensBack);
~Token();
void str(const std::string &s);
2008-12-18 22:28:57 +01:00
2011-10-23 21:21:42 +02:00
/**
* Concatenate two (quoted) strings. Automatically cuts of the last/first character.
* Example: "hello ""world" -> "hello world". Used by the token simplifier.
2011-10-23 21:21:42 +02:00
*/
void concatStr(std::string const& b);
2011-10-13 20:53:06 +02:00
const std::string &str() const {
return _str;
}
2008-12-18 22:28:57 +01:00
/**
* Unlink and delete the next 'index' tokens.
2008-12-18 22:28:57 +01:00
*/
void deleteNext(unsigned long index = 1);
2008-12-18 22:28:57 +01:00
/**
* Returns token in given index, related to this token.
* For example index 1 would return next token, and 2
* would return next from that one.
*/
const Token *tokAt(int index) const;
2009-01-20 07:20:55 +01:00
Token *tokAt(int index);
2008-12-18 22:28:57 +01:00
/**
* Returns the link to the token in given index, related to this token.
* For example index 1 would return the link to next token.
*/
const Token *linkAt(int index) const;
Token *linkAt(int index);
const std::string &strAt(int index) const;
2008-12-18 22:28:57 +01:00
2008-12-23 22:51:54 +01:00
/**
* Match given token (or list of tokens) to a pattern list.
*
* Possible patterns
* "someRandomText" If token contains "someRandomText".
* @note Use Match() if you want to use flags in patterns
*
* The patterns can be also combined to compare to multiple tokens at once
* by separating tokens with a space, e.g.
* ") void {" will return true if first token is ')' next token
* is "void" and token after that is '{'. If even one of the tokens does
* not match its pattern, false is returned.
*
* @param tok List of tokens to be compared to the pattern
* @param pattern The pattern against which the tokens are compared,
* e.g. "const" or ") void {".
* @return true if given token matches with given pattern
* false if given token does not match with given pattern
*/
static bool simpleMatch(const Token *tok, const char pattern[]);
2008-12-18 22:28:57 +01:00
/**
* Match given token (or list of tokens) to a pattern list.
*
* Possible patterns
2010-03-13 22:16:06 +01:00
* - "%any%" any token
* - "%var%" any token which is a name or type e.g. "hello" or "int"
* - "%type%" Anything that can be a variable type, e.g. "int", but not "delete".
* - "%num%" Any numeric token, e.g. "23"
* - "%bool%" true or false
* - "%str%" Any token starting with &quot;-character (C-string).
* - "%varid%" Match with parameter varid
* - "%or%" A bitwise-or operator '|'
* - "%oror%" A logical-or operator '||'
2010-03-13 22:16:06 +01:00
* - "[abc]" Any of the characters 'a' or 'b' or 'c'
* - "int|void|char" Any of the strings, int, void or char
* - "int|void|char|" Any of the strings, int, void or char or empty string
* - "!!else" No tokens or any token that is not "else".
* - "someRandomText" If token contains "someRandomText".
2008-12-18 22:28:57 +01:00
*
* multi-compare patterns such as "int|void|char" can contain %or%, %oror% and %op%
* but it is not recommended to put such an %cmd% as the first pattern.
*
2008-12-18 22:28:57 +01:00
* The patterns can be also combined to compare to multiple tokens at once
* by separating tokens with a space, e.g.
* ") const|void {" will return true if first token is ')' next token is either
* "const" or "void" and token after that is '{'. If even one of the tokens does not
* match its pattern, false is returned.
*
* @todo pattern "%type%|%num%" should mean either a type or a num.
*
2008-12-18 22:28:57 +01:00
* @param tok List of tokens to be compared to the pattern
2008-12-23 22:51:54 +01:00
* @param pattern The pattern against which the tokens are compared,
* e.g. "const" or ") const|volatile| {".
2010-03-13 22:16:06 +01:00
* @param varid if %varid% is given in the pattern the Token::varId will be matched against this argument
2008-12-18 22:28:57 +01:00
* @return true if given token matches with given pattern
* false if given token does not match with given pattern
*/
static bool Match(const Token *tok, const char pattern[], unsigned int varid = 0);
2008-12-18 22:28:57 +01:00
/**
* Return length of C-string.
*
* Should be called for %str% tokens only.
*
* @param tok token with C-string
**/
static std::size_t getStrLength(const Token *tok);
Type type() const {
return _type;
2009-08-23 17:17:57 +02:00
}
void type(Type t) {
_type = t;
2011-03-08 02:04:25 +01:00
}
bool isName() const {
return _type == eName || _type == eType || _type == eVariable || _type == eFunction ||
_type == eBoolean; // TODO: "true"/"false" aren't really a name...
2009-08-23 17:17:57 +02:00
}
bool isNumber() const {
return _type == eNumber;
2011-03-08 02:04:25 +01:00
}
2011-10-13 20:53:06 +02:00
bool isArithmeticalOp() const {
return _type == eArithmeticalOp;
2011-04-09 15:54:36 +02:00
}
2011-10-13 20:53:06 +02:00
bool isOp() const {
2011-04-09 15:54:36 +02:00
return (isArithmeticalOp() ||
_type == eLogicalOp ||
_type == eComparisonOp ||
_type == eBitOp);
}
2011-10-13 20:53:06 +02:00
bool isExtendedOp() const {
2011-04-09 16:22:47 +02:00
return isOp() ||
_type == eExtendedOp;
}
2011-10-13 20:53:06 +02:00
bool isAssignmentOp() const {
return _type == eAssignmentOp;
}
2011-10-13 20:53:06 +02:00
bool isBoolean() const {
return _type == eBoolean;
2011-03-08 02:04:25 +01:00
}
2011-10-13 20:53:06 +02:00
bool isUnsigned() const {
return _isUnsigned;
}
2011-10-13 20:53:06 +02:00
void isUnsigned(bool sign) {
_isUnsigned = sign;
}
2011-10-13 20:53:06 +02:00
bool isSigned() const {
return _isSigned;
}
2011-10-13 20:53:06 +02:00
void isSigned(bool sign) {
_isSigned = sign;
}
2011-10-13 20:53:06 +02:00
bool isPointerCompare() const {
return _isPointerCompare;
}
2011-10-13 20:53:06 +02:00
void isPointerCompare(bool b) {
_isPointerCompare = b;
}
2011-10-13 20:53:06 +02:00
bool isLong() const {
return _isLong;
}
2011-10-13 20:53:06 +02:00
void isLong(bool size) {
_isLong = size;
}
2011-10-13 20:53:06 +02:00
bool isUnused() const {
return _isUnused;
}
2011-10-13 20:53:06 +02:00
void isUnused(bool used) {
_isUnused = used;
}
bool isStandardType() const {
return _isStandardType;
}
bool isExpandedMacro() const {
return _isExpandedMacro;
}
void setExpandedMacro(bool m) {
_isExpandedMacro = m;
}
static const Token *findsimplematch(const Token *tok, const char pattern[]);
static const Token *findsimplematch(const Token *tok, const char pattern[], const Token *end);
static const Token *findmatch(const Token *tok, const char pattern[], unsigned int varId = 0);
2010-07-26 16:46:37 +02:00
static const Token *findmatch(const Token *tok, const char pattern[], const Token *end, unsigned int varId = 0);
2008-12-18 22:28:57 +01:00
/**
* Needle is build from multiple alternatives. If one of
* them is equal to haystack, return value is 1. If there
* are no matches, but one alternative to needle is empty
* string, return value is 0. If needle was not found, return
* value is -1.
*
* @param haystack e.g. "one|two" or "|one|two"
* @param needle e.g. "one", "two" or "invalid"
2008-12-18 22:28:57 +01:00
* @return 1 if needle is found from the haystack
* 0 if needle was empty string
* -1 if needle was not found
*/
static int multiCompare(const char *haystack, const char *needle);
2008-12-18 22:28:57 +01:00
2011-10-13 20:53:06 +02:00
unsigned int linenr() const {
2009-08-23 17:17:57 +02:00
return _linenr;
}
2011-10-13 20:53:06 +02:00
void linenr(unsigned int lineNumber) {
_linenr = lineNumber;
2009-08-23 17:17:57 +02:00
}
2008-12-18 22:28:57 +01:00
2011-10-13 20:53:06 +02:00
unsigned int fileIndex() const {
2009-08-23 17:17:57 +02:00
return _fileIndex;
}
2011-10-13 20:53:06 +02:00
void fileIndex(unsigned int indexOfFile) {
_fileIndex = indexOfFile;
2009-08-23 17:17:57 +02:00
}
2008-12-18 22:28:57 +01:00
2011-10-13 20:53:06 +02:00
Token *next() const {
2009-08-23 17:17:57 +02:00
return _next;
}
2008-12-18 22:28:57 +01:00
/**
* Delete tokens between begin and end. E.g. if begin = 1
* and end = 5, tokens 2,3 and 4 would be erased.
*
* @param begin Tokens after this will be erased.
* @param end Tokens before this will be erased.
*/
static void eraseTokens(Token *begin, const Token *end);
2008-12-18 22:28:57 +01:00
/**
* Insert new token after this token. This function will handle
* relations between next and previous token also.
2010-04-09 21:40:37 +02:00
* @param tokenStr String for the new token.
2008-12-18 22:28:57 +01:00
*/
2010-04-09 21:40:37 +02:00
void insertToken(const std::string &tokenStr);
2008-12-18 22:28:57 +01:00
2011-10-13 20:53:06 +02:00
Token *previous() const {
2009-08-23 17:17:57 +02:00
return _previous;
}
2008-12-18 22:28:57 +01:00
2011-10-13 20:53:06 +02:00
unsigned int varId() const {
2009-08-23 17:17:57 +02:00
return _varId;
}
2011-10-13 20:53:06 +02:00
void varId(unsigned int id) {
2009-08-23 17:17:57 +02:00
_varId = id;
}
2008-12-18 22:28:57 +01:00
/**
* For debugging purposes, prints token and all tokens
* followed by it.
* @param title Title for the printout or use default parameter or 0
* for no title.
*/
void printOut(const char *title = 0) const;
2008-12-18 22:28:57 +01:00
/**
* For debugging purposes, prints token and all tokens
* followed by it.
* @param title Title for the printout or use default parameter or 0
* for no title.
* @param fileNames Prints out file name instead of file index.
* File index should match the index of the string in this vector.
*/
void printOut(const char *title, const std::vector<std::string> &fileNames) const;
/**
* Replace token replaceThis with tokens between start and end,
* including start and end. The replaceThis token is deleted.
* @param replaceThis This token will be deleted.
* @param start This will be in the place of replaceThis
* @param end This is also in the place of replaceThis
*/
static void replace(Token *replaceThis, Token *start, Token *end);
/**
* Stringify a token
* @param os The result is shifted into that output stream
* @param varid Print varids. (Style: "varname@id")
* @param attributes Print attributes of tokens like "unsigned" in front of it.
*/
void stringify(std::ostream& os, bool varid, bool attributes) const;
/**
* Stringify a list of token, from current instance on.
* @param varid Print varids. (Style: "varname@id")
* @param attributes Print attributes of tokens like "unsigned" in front of it.
* @param linenumbers Print line number in front of each line
* @param linebreaks Insert \n into string when line number changes
* @param files print Files as numbers or as names (if fileNames is given)
* @param fileNames Vector of filenames. Used (if given) to print filenames as strings instead of numbers.
* @param title Prints a title on top of output
* @param end Stringification ends before this token is reached. 0 to stringify until end of list.
* @return Stringified token list as a string
*/
std::string stringifyList(bool varid, bool attributes, bool linenumbers, bool linebreaks, bool files, const std::vector<std::string>* fileNames = 0, const Token* end = 0) const;
std::string stringifyList(const Token* end, bool attributes = true) const;
std::string stringifyList(bool varid = false) const;
/**
2011-10-16 08:25:11 +02:00
* Remove the contents for this token from the token list.
*
* The contents are replaced with the contents of the next token and
* the next token is unlinked and deleted from the token list.
*
* So this token will still be valid after the 'deleteThis()'.
*/
void deleteThis();
/**
* Create link to given token
* @param linkToToken The token where this token should link
* to.
*/
2011-10-13 20:53:06 +02:00
void link(Token *linkToToken) {
_link = linkToToken;
2009-08-23 17:17:57 +02:00
}
/**
* Return token where this token links to.
* Supported links are:
* "{" <-> "}"
2011-10-16 08:25:11 +02:00
* "(" <-> ")"
* "[" <-> "]"
*
* @return The token where this token links to.
*/
2011-10-13 20:53:06 +02:00
Token *link() const {
2009-08-23 17:17:57 +02:00
return _link;
}
/**
* Links two elements against each other.
**/
static void createMutualLinks(Token *begin, Token *end);
/**
* This can be called only for tokens that are strings, else
* the assert() is called. If Token is e.g. '"hello"', this will return
* 'hello' (removing the double quotes).
* @return String value
*/
std::string strValue() const;
/**
* Move srcStart and srcEnd tokens and all tokens between them
* into new a location. Only links between tokens are changed.
* @param srcStart This is the first token to be moved
* @param srcEnd The last token to be moved
* @param newLocation srcStart will be placed after this token.
*/
static void move(Token *srcStart, Token *srcEnd, Token *newLocation);
/** Get progressValue */
2011-10-13 20:53:06 +02:00
unsigned int progressValue() const {
return _progressValue;
}
/** Calculate progress values for all tokens */
2011-10-13 20:53:06 +02:00
void assignProgressValues() {
unsigned int total_count = 0;
for (Token *tok = this; tok; tok = tok->next())
++total_count;
unsigned int count = 0;
for (Token *tok = this; tok; tok = tok->next())
tok->_progressValue = count++ * 100 / total_count;
}
2011-10-23 11:23:48 +02:00
/**
* Returns the first token of the next argument. Does only work on argument
* lists. Returns 0, if there is no next argument
*/
Token* nextArgument() const;
2011-10-23 11:23:48 +02:00
/**
* Returns the closing bracket of opening '<'. Should only be used if link()
* is unavailable.
* @param closing The closing token is stored in that parameter
* @return success
*/
bool findClosingBracket(const Token*& closing) const;
bool findClosingBracket(Token*& closing) const {
const Token* tok;
bool retVal = findClosingBracket(tok);
closing = const_cast<Token*>(tok);
return retVal;
}
2008-12-18 22:28:57 +01:00
private:
2011-10-13 20:53:06 +02:00
void next(Token *nextToken) {
_next = nextToken;
2009-08-23 17:17:57 +02:00
}
2011-10-13 20:53:06 +02:00
void previous(Token *previousToken) {
_previous = previousToken;
2009-08-23 17:17:57 +02:00
}
2008-12-18 22:28:57 +01:00
/**
* Works almost like strcmp() except returns only true or false and
2010-03-13 22:16:06 +01:00
* if str has empty space &apos; &apos; character, that character is handled
* as if it were &apos;\\0&apos;
*/
static bool firstWordEquals(const char *str, const char *word);
/**
* Works almost like strchr() except
2010-03-13 22:16:06 +01:00
* if str has empty space &apos; &apos; character, that character is handled
* as if it were &apos;\\0&apos;
*/
static const char *chrInFirstWord(const char *str, char c);
/**
* Works almost like strlen() except
2010-03-13 22:16:06 +01:00
* if str has empty space &apos; &apos; character, that character is handled
* as if it were &apos;\\0&apos;
*/
static int firstWordLen(const char *str);
Token *_next;
Token *_previous;
Token *_link;
Type _type;
bool _isUnsigned;
bool _isSigned;
bool _isPointerCompare;
bool _isLong;
bool _isUnused;
bool _isStandardType;
bool _isExpandedMacro;
2008-12-18 22:28:57 +01:00
unsigned int _varId;
unsigned int _fileIndex;
unsigned int _linenr;
/** Updates internal property cache like _isName or _isBoolean.
Called after any _str() modification. */
void update_property_info();
/** Update internal property cache about isStandardType() */
void update_property_isStandardType();
/**
* A value from 0-100 that provides a rough idea about where in the token
* list this token is located.
*/
unsigned int _progressValue;
std::string _str;
2008-12-18 22:28:57 +01:00
};
/// @}
#endif // TokenH