Andrew Scull | 5e1ddfa | 2018-08-14 10:06:54 +0100 | [diff] [blame] | 1 | //===-- Regex.h - Regular Expression matcher implementation -*- C++ -*-----===// |
| 2 | // |
| 3 | // The LLVM Compiler Infrastructure |
| 4 | // |
| 5 | // This file is distributed under the University of Illinois Open Source |
| 6 | // License. See LICENSE.TXT for details. |
| 7 | // |
| 8 | //===----------------------------------------------------------------------===// |
| 9 | // |
| 10 | // This file implements a POSIX regular expression matcher. Both Basic and |
| 11 | // Extended POSIX regular expressions (ERE) are supported. EREs were extended |
| 12 | // to support backreferences in matches. |
| 13 | // This implementation also supports matching strings with embedded NUL chars. |
| 14 | // |
| 15 | //===----------------------------------------------------------------------===// |
| 16 | |
| 17 | #ifndef LLVM_SUPPORT_REGEX_H |
| 18 | #define LLVM_SUPPORT_REGEX_H |
| 19 | |
| 20 | #include <string> |
| 21 | |
| 22 | struct llvm_regex; |
| 23 | |
| 24 | namespace llvm { |
| 25 | class StringRef; |
| 26 | template<typename T> class SmallVectorImpl; |
| 27 | |
| 28 | class Regex { |
| 29 | public: |
| 30 | enum { |
| 31 | NoFlags=0, |
| 32 | /// Compile for matching that ignores upper/lower case distinctions. |
| 33 | IgnoreCase=1, |
| 34 | /// Compile for newline-sensitive matching. With this flag '[^' bracket |
| 35 | /// expressions and '.' never match newline. A ^ anchor matches the |
| 36 | /// null string after any newline in the string in addition to its normal |
| 37 | /// function, and the $ anchor matches the null string before any |
| 38 | /// newline in the string in addition to its normal function. |
| 39 | Newline=2, |
| 40 | /// By default, the POSIX extended regular expression (ERE) syntax is |
| 41 | /// assumed. Pass this flag to turn on basic regular expressions (BRE) |
| 42 | /// instead. |
| 43 | BasicRegex=4 |
| 44 | }; |
| 45 | |
| 46 | Regex(); |
| 47 | /// Compiles the given regular expression \p Regex. |
| 48 | Regex(StringRef Regex, unsigned Flags = NoFlags); |
| 49 | Regex(const Regex &) = delete; |
| 50 | Regex &operator=(Regex regex) { |
| 51 | std::swap(preg, regex.preg); |
| 52 | std::swap(error, regex.error); |
| 53 | return *this; |
| 54 | } |
| 55 | Regex(Regex &®ex); |
| 56 | ~Regex(); |
| 57 | |
| 58 | /// isValid - returns the error encountered during regex compilation, or |
| 59 | /// matching, if any. |
| 60 | bool isValid(std::string &Error) const; |
| 61 | |
| 62 | /// getNumMatches - In a valid regex, return the number of parenthesized |
| 63 | /// matches it contains. The number filled in by match will include this |
| 64 | /// many entries plus one for the whole regex (as element 0). |
| 65 | unsigned getNumMatches() const; |
| 66 | |
| 67 | /// matches - Match the regex against a given \p String. |
| 68 | /// |
| 69 | /// \param Matches - If given, on a successful match this will be filled in |
| 70 | /// with references to the matched group expressions (inside \p String), |
| 71 | /// the first group is always the entire pattern. |
| 72 | /// |
| 73 | /// This returns true on a successful match. |
| 74 | bool match(StringRef String, SmallVectorImpl<StringRef> *Matches = nullptr); |
| 75 | |
| 76 | /// sub - Return the result of replacing the first match of the regex in |
| 77 | /// \p String with the \p Repl string. Backreferences like "\0" in the |
| 78 | /// replacement string are replaced with the appropriate match substring. |
| 79 | /// |
| 80 | /// Note that the replacement string has backslash escaping performed on |
| 81 | /// it. Invalid backreferences are ignored (replaced by empty strings). |
| 82 | /// |
| 83 | /// \param Error If non-null, any errors in the substitution (invalid |
| 84 | /// backreferences, trailing backslashes) will be recorded as a non-empty |
| 85 | /// string. |
| 86 | std::string sub(StringRef Repl, StringRef String, |
| 87 | std::string *Error = nullptr); |
| 88 | |
Andrew Scull | cdfcccc | 2018-10-05 20:58:37 +0100 | [diff] [blame] | 89 | /// If this function returns true, ^Str$ is an extended regular |
Andrew Scull | 5e1ddfa | 2018-08-14 10:06:54 +0100 | [diff] [blame] | 90 | /// expression that matches Str and only Str. |
| 91 | static bool isLiteralERE(StringRef Str); |
| 92 | |
Andrew Scull | cdfcccc | 2018-10-05 20:58:37 +0100 | [diff] [blame] | 93 | /// Turn String into a regex by escaping its special characters. |
Andrew Scull | 5e1ddfa | 2018-08-14 10:06:54 +0100 | [diff] [blame] | 94 | static std::string escape(StringRef String); |
| 95 | |
| 96 | private: |
| 97 | struct llvm_regex *preg; |
| 98 | int error; |
| 99 | }; |
| 100 | } |
| 101 | |
| 102 | #endif // LLVM_SUPPORT_REGEX_H |