| 1 | // Copyright 2008 and onwards Google, Inc. |
| 2 | // |
| 3 | // Functions for splitting and parsing strings. Functions may be migrated |
| 4 | // to this file from strutil.h in the future. |
| 5 | // |
| 6 | #ifndef STRINGS_SPLIT_H_ |
| 7 | #define STRINGS_SPLIT_H_ |
| 8 | |
| 9 | #include <string> |
| 10 | using std::string; |
| 11 | |
| 12 | #include <vector> |
| 13 | using std::vector; |
| 14 | |
| 15 | #include <set> |
| 16 | using std::set; |
| 17 | using std::multiset; |
| 18 | |
| 19 | #include <utility> |
| 20 | using std::pair; |
| 21 | using std::make_pair; |
| 22 | |
| 23 | #include <hash_map> |
| 24 | using __gnu_cxx::hash_map; |
| 25 | |
| 26 | #include <hash_set> |
| 27 | using __gnu_cxx::hash_set; |
| 28 | |
| 29 | |
| 30 | using namespace std; |
| 31 | using namespace __gnu_cxx; |
| 32 | |
| 33 | // ---------------------------------------------------------------------- |
| 34 | // SplitStringAllowEmpty() |
| 35 | // SplitStringToHashsetAllowEmpty() |
| 36 | // SplitStringToSetAllowEmpty() |
| 37 | // SplitStringToHashmapAllowEmpty() |
| 38 | |
| 39 | // Split a string using one or more character delimiters, presented |
| 40 | // as a nul-terminated c string. Append the components to 'result'. |
| 41 | // If there are consecutive delimiters, this function will return |
| 42 | // corresponding empty strings. |
| 43 | // |
| 44 | // If "full" is the empty string, yields an empty string as the only value. |
| 45 | // ---------------------------------------------------------------------- |
| 46 | void SplitStringAllowEmpty(const string& full, const char* delim, |
| 47 | vector<string>* res); |
| 48 | void SplitStringToHashsetAllowEmpty(const string& full, const char* delim, |
| 49 | hash_set<string>* res); |
| 50 | void SplitStringToSetAllowEmpty(const string& full, const char* delim, |
| 51 | set<string>* res); |
| 52 | // The even-positioned (0-based) components become the keys for the |
| 53 | // odd-positioned components that follow them. When there is an odd |
| 54 | // number of components, the value for the last key will be unchanged |
| 55 | // if the key was already present in the hash table, or will be the |
| 56 | // empty string if the key is a newly inserted key. |
| 57 | void SplitStringToHashmapAllowEmpty(const string& full, const char* delim, |
| 58 | hash_map<string, string>* result); |
| 59 | |
| 60 | // ---------------------------------------------------------------------- |
| 61 | // SplitStringUsing() |
| 62 | // SplitStringToHashsetUsing() |
| 63 | // SplitStringToSetUsing() |
| 64 | // SplitStringToHashmapUsing() |
| 65 | |
| 66 | // Split a string using one or more character delimiters, presented |
| 67 | // as a nul-terminated c string. Append the components to 'result'. |
| 68 | // If there are consecutive delimiters, this function skips over |
| 69 | // all of them. |
| 70 | // ---------------------------------------------------------------------- |
| 71 | void SplitStringUsing(const string& full, const char* delim, |
| 72 | vector<string>* res); |
| 73 | void SplitStringToHashsetUsing(const string& full, const char* delim, |
| 74 | hash_set<string>* res); |
| 75 | void SplitStringToSetUsing(const string& full, const char* delim, |
| 76 | set<string>* res); |
| 77 | // The even-positioned (0-based) components become the keys for the |
| 78 | // odd-positioned components that follow them. When there is an odd |
| 79 | // number of components, the value for the last key will be unchanged |
| 80 | // if the key was already present in the hash table, or will be the |
| 81 | // empty string if the key is a newly inserted key. |
| 82 | void SplitStringToHashmapUsing(const string& full, const char* delim, |
| 83 | hash_map<string, string>* result); |
| 84 | |
| 85 | // ---------------------------------------------------------------------- |
| 86 | // SplitOneIntToken() |
| 87 | // SplitOneInt32Token() |
| 88 | // SplitOneUint32Token() |
| 89 | // SplitOneInt64Token() |
| 90 | // SplitOneUint64Token() |
| 91 | // SplitOneDoubleToken() |
| 92 | // SplitOneFloatToken() |
| 93 | // Parse a single "delim" delimited number from "*source" into "*value". |
| 94 | // Modify *source to point after the delimiter. |
| 95 | // If no delimiter is present after the number, set *source to NULL. |
| 96 | // |
| 97 | // If the start of *source is not an number, return false. |
| 98 | // If the int is followed by the null character, return true. |
| 99 | // If the int is not followed by a character from delim, return false. |
| 100 | // If *source is NULL, return false. |
| 101 | // |
| 102 | // They cannot handle decimal numbers with leading 0s, since they will be |
| 103 | // treated as octal. |
| 104 | // ---------------------------------------------------------------------- |
| 105 | bool SplitOneIntToken(const char** source, const char* delim, |
| 106 | int* value); |
| 107 | bool SplitOneInt32Token(const char** source, const char* delim, |
| 108 | int32* value); |
| 109 | bool SplitOneUint32Token(const char** source, const char* delim, |
| 110 | uint32* value); |
| 111 | bool SplitOneInt64Token(const char** source, const char* delim, |
| 112 | int64* value); |
| 113 | bool SplitOneUint64Token(const char** source, const char* delim, |
| 114 | uint64* value); |
| 115 | bool SplitOneDoubleToken(const char** source, const char* delim, |
| 116 | double* value); |
| 117 | bool SplitOneFloatToken(const char** source, const char* delim, |
| 118 | float* value); |
| 119 | |
| 120 | // Some aliases, so that the function names are standardized against the names |
| 121 | // of the reflection setters/getters in proto2. This makes it easier to use |
| 122 | // certain macros with reflection when creating custom text formats for protos. |
| 123 | |
| 124 | inline bool SplitOneUInt32Token(const char** source, const char* delim, |
| 125 | uint32* value) { |
| 126 | return SplitOneUint32Token(source, delim, value); |
| 127 | } |
| 128 | |
| 129 | inline bool SplitOneUInt64Token(const char** source, const char* delim, |
| 130 | uint64* value) { |
| 131 | return SplitOneUint64Token(source, delim, value); |
| 132 | } |
| 133 | |
| 134 | // ---------------------------------------------------------------------- |
| 135 | // SplitOneDecimalIntToken() |
| 136 | // SplitOneDecimalInt32Token() |
| 137 | // SplitOneDecimalUint32Token() |
| 138 | // SplitOneDecimalInt64Token() |
| 139 | // SplitOneDecimalUint64Token() |
| 140 | // Parse a single "delim"-delimited number from "*source" into "*value". |
| 141 | // Unlike SplitOneIntToken, etc., this function always interprets |
| 142 | // the numbers as decimal. |
| 143 | bool SplitOneDecimalIntToken(const char** source, const char* delim, |
| 144 | int* value); |
| 145 | bool SplitOneDecimalInt32Token(const char** source, const char* delim, |
| 146 | int32* value); |
| 147 | bool SplitOneDecimalUint32Token(const char** source, const char* delim, |
| 148 | uint32* value); |
| 149 | bool SplitOneDecimalInt64Token(const char** source, const char* delim, |
| 150 | int64* value); |
| 151 | bool SplitOneDecimalUint64Token(const char** source, const char* delim, |
| 152 | uint64* value); |
| 153 | |
| 154 | // ---------------------------------------------------------------------- |
| 155 | // SplitOneHexUint32Token() |
| 156 | // SplitOneHexUint64Token() |
| 157 | // Once more, for hexadecimal numbers (unsigned only). |
| 158 | bool SplitOneHexUint32Token(const char** source, const char* delim, |
| 159 | uint32* value); |
| 160 | bool SplitOneHexUint64Token(const char** source, const char* delim, |
| 161 | uint64* value); |
| 162 | |
| 163 | |
| 164 | #endif // STRINGS_SPLIT_H_ |
| 165 | |