--- /dev/null
+////////////////////////////////////////////////////////////////////////////////\r
+\r
+// Author: Andy Rushton\r
+// Copyright: (c) Southampton University 1999-2004\r
+// (c) Andy Rushton 2004-2009\r
+// License: BSD License, see ../docs/license.html\r
+\r
+////////////////////////////////////////////////////////////////////////////////\r
+#include "string_utilities.hpp"\r
+#include "string_basic.hpp"\r
+#include <stdlib.h>\r
+#include <ctype.h>\r
+#include <stdarg.h>\r
+#include <stdio.h>\r
+\r
+namespace stlplus\r
+{\r
+\r
+ // added as a local copy to break the dependency on the portability library\r
+ static std::string local_dformat(const char* format, ...) throw(std::invalid_argument)\r
+ {\r
+ std::string formatted;\r
+ va_list args;\r
+ va_start(args, format);\r
+#ifdef MSWINDOWS\r
+ int length = 0;\r
+ char* buffer = 0;\r
+ for(int buffer_length = 256; ; buffer_length*=2)\r
+ {\r
+ buffer = (char*)malloc(buffer_length);\r
+ if (!buffer) throw std::invalid_argument("string_utilities");\r
+ length = _vsnprintf(buffer, buffer_length-1, format, args);\r
+ if (length >= 0)\r
+ {\r
+ buffer[length] = 0;\r
+ formatted += std::string(buffer);\r
+ free(buffer);\r
+ break;\r
+ }\r
+ free(buffer);\r
+ }\r
+#else\r
+ char* buffer = 0;\r
+ int length = vasprintf(&buffer, format, args);\r
+ if (!buffer) throw std::invalid_argument("string_utilities");\r
+ if (length >= 0)\r
+ formatted += std::string(buffer);\r
+ free(buffer);\r
+#endif\r
+ va_end(args);\r
+ if (length < 0) throw std::invalid_argument("string_utilities");\r
+ return formatted;\r
+ }\r
+\r
+ ////////////////////////////////////////////////////////////////////////////////\r
+\r
+ std::string pad(const std::string& str, alignment_t alignment, unsigned width, char padch)\r
+ throw(std::invalid_argument)\r
+ {\r
+ std::string result = str;\r
+ switch(alignment)\r
+ {\r
+ case align_left:\r
+ {\r
+ unsigned padding = width>str.size() ? width - str.size() : 0;\r
+ unsigned i = 0;\r
+ while (i++ < padding)\r
+ result.insert(result.end(), padch);\r
+ break;\r
+ }\r
+ case align_right:\r
+ {\r
+ unsigned padding = width>str.size() ? width - str.size() : 0;\r
+ unsigned i = 0;\r
+ while (i++ < padding)\r
+ result.insert(result.begin(), padch);\r
+ break;\r
+ }\r
+ case align_centre:\r
+ {\r
+ unsigned padding = width>str.size() ? width - str.size() : 0;\r
+ unsigned i = 0;\r
+ while (i++ < padding/2)\r
+ result.insert(result.end(), padch);\r
+ i--;\r
+ while (i++ < padding)\r
+ result.insert(result.begin(), padch);\r
+ break;\r
+ }\r
+ default:\r
+ throw std::invalid_argument("invalid alignment value");\r
+ }\r
+ return result;\r
+ }\r
+\r
+ ////////////////////////////////////////////////////////////////////////////////\r
+\r
+ std::string trim_left(const std::string& val)\r
+ {\r
+ std::string result = val;\r
+ while (!result.empty() && isspace(result[0]))\r
+ result.erase(result.begin());\r
+ return result;\r
+ }\r
+\r
+ std::string trim_right(const std::string& val)\r
+ {\r
+ std::string result = val;\r
+ while (!result.empty() && isspace(result[result.size()-1]))\r
+ result.erase(result.end()-1);\r
+ return result;\r
+ }\r
+\r
+ std::string trim(const std::string& val)\r
+ {\r
+ std::string result = val;\r
+ while (!result.empty() && isspace(result[0]))\r
+ result.erase(result.begin());\r
+ while (!result.empty() && isspace(result[result.size()-1]))\r
+ result.erase(result.end()-1);\r
+ return result;\r
+ }\r
+\r
+ ////////////////////////////////////////////////////////////////////////////////\r
+\r
+ std::string lowercase(const std::string& val)\r
+ {\r
+ std::string text = val;\r
+ for (unsigned i = 0; i < text.size(); i++)\r
+ text[i] = tolower(text[i]);\r
+ return text;\r
+ }\r
+\r
+ std::string uppercase(const std::string& val)\r
+ {\r
+ std::string text = val;\r
+ for (unsigned i = 0; i < text.size(); i++)\r
+ text[i] = toupper(text[i]);\r
+ return text;\r
+ }\r
+\r
+ ////////////////////////////////////////////////////////////////////////////////\r
+\r
+ std::string translate(const std::string& input, const std::string& from_set, const std::string& to_set)\r
+ {\r
+ std::string result;\r
+ for (unsigned i = 0; i < input.size(); i++)\r
+ {\r
+ char ch = input[i];\r
+ // check to see if the character is in the from set\r
+ std::string::size_type found = from_set.find(ch);\r
+ if (found == std::string::npos)\r
+ {\r
+ // not found so just copy across\r
+ result += ch;\r
+ }\r
+ else if (found < to_set.size())\r
+ {\r
+ // found and in range so translate\r
+ result += to_set[found];\r
+ }\r
+ }\r
+ return result;\r
+ }\r
+\r
+ ////////////////////////////////////////////////////////////////////////////////\r
+ // WARNING: wheel re-invention follows\r
+ // Given that all shells perform wildcard matching, why don't the library writers put it in the C run-time????????\r
+ // The problem:\r
+ // * matches any number of characters - this is achieved by matching 1 and seeing if the remainder matches\r
+ // if not, try 2 characters and see if the remainder matches etc.\r
+ // this must be recursive, not iterative, so that multiple *s can appear in the same wildcard expression\r
+ // ? matches exactly one character so doesn't need the what-if approach\r
+ // \ escapes special characters such as *, ? and [\r
+ // [] matches exactly one character in the set - the difficulty is the set can contain ranges, e.g [a-zA-Z0-9]\r
+ // a set cannot be empty and the ] character can be included by making it the first character\r
+\r
+ // function for testing whether a character matches a set\r
+ // I can't remember the exact rules and I have no definitive references but:\r
+ // a set contains characters, escaped characters (I think) and ranges in the form a-z\r
+ // The character '-' can only appear at the start of the set where it is not interpreted as a range\r
+ // This is a horrible mess - blame the Unix folks for making a hash of wildcards\r
+\r
+ static bool match_set (const std::string& set, char match)\r
+ {\r
+ // first expand any ranges and remove escape characters to make life more palatable\r
+ std::string simple_set;\r
+ for (std::string::const_iterator i = set.begin(); i != set.end(); ++i)\r
+ {\r
+ switch(*i)\r
+ {\r
+ case '-':\r
+ {\r
+ if (i == set.begin())\r
+ {\r
+ simple_set += *i;\r
+ }\r
+ else if (i+1 == set.end())\r
+ {\r
+ return false;\r
+ }\r
+ else\r
+ {\r
+ // found a set. The first character is already in the result, so first remove it (the set might be empty)\r
+ simple_set.erase(simple_set.end()-1);\r
+ char last = *++i;\r
+ for (char ch = *(i-2); ch <= last; ch++)\r
+ {\r
+ simple_set += ch;\r
+ }\r
+ }\r
+ break;\r
+ }\r
+ case '\\':\r
+ if (i+1 == set.end()) {return false;}\r
+ simple_set += *++i;\r
+ break;\r
+ default:\r
+ simple_set += *i;\r
+ break;\r
+ }\r
+ }\r
+ std::string::size_type result = simple_set.find(match);\r
+ return result != std::string::npos;\r
+ }\r
+\r
+ // the recursive bit - basically whenever a * is found you recursively call this for each candidate substring match\r
+ // until either it succeeds or you run out of string to match\r
+ // for each * in the wildcard another level of recursion is created\r
+\r
+ static bool match_remainder (const std::string& wild, std::string::const_iterator wildi,\r
+ const std::string& match, std::string::const_iterator matchi)\r
+ {\r
+ //cerr << "match_remainder called at " << *matchi << " with wildcard " << *wildi << endl;\r
+ while (wildi != wild.end() && matchi != match.end())\r
+ {\r
+ //cerr << "trying to match " << *matchi << " with wildcard " << *wildi << endl;\r
+ switch(*wildi)\r
+ {\r
+ case '*':\r
+ {\r
+ ++wildi;\r
+ ++matchi;\r
+ for (std::string::const_iterator i = matchi; i != match.end(); ++i)\r
+ {\r
+ // deal with * at the end of the wildcard - there is no remainder then\r
+ if (wildi == wild.end())\r
+ {\r
+ if (i == match.end()-1)\r
+ return true;\r
+ }\r
+ else if (match_remainder(wild, wildi, match, i))\r
+ {\r
+ return true;\r
+ }\r
+ }\r
+ return false;\r
+ }\r
+ case '[':\r
+ {\r
+ // scan for the end of the set using a similar method for avoiding escaped characters\r
+ bool found = false;\r
+ std::string::const_iterator end = wildi + 1;\r
+ for (; !found && end != wild.end(); ++end)\r
+ {\r
+ switch(*end)\r
+ {\r
+ case ']':\r
+ {\r
+ // found the set, now match with its contents excluding the brackets\r
+ if (!match_set(wild.substr(wildi - wild.begin() + 1, end - wildi - 1), *matchi))\r
+ return false;\r
+ found = true;\r
+ break;\r
+ }\r
+ case '\\':\r
+ if (end == wild.end()-1)\r
+ return false;\r
+ ++end;\r
+ break;\r
+ default:\r
+ break;\r
+ }\r
+ }\r
+ if (!found)\r
+ return false;\r
+ ++matchi;\r
+ wildi = end;\r
+ break;\r
+ }\r
+ case '?':\r
+ ++wildi;\r
+ ++matchi;\r
+ break;\r
+ case '\\':\r
+ if (wildi == wild.end()-1)\r
+ return false;\r
+ ++wildi;\r
+ if (*wildi != *matchi)\r
+ return false;\r
+ ++wildi;\r
+ ++matchi;\r
+ break;\r
+ default:\r
+ if (*wildi != *matchi)\r
+ return false;\r
+ ++wildi;\r
+ ++matchi;\r
+ break;\r
+ }\r
+ }\r
+ bool result = wildi == wild.end() && matchi == match.end();\r
+ return result;\r
+ }\r
+\r
+ // like all recursions the exported function has a simpler interface than the\r
+ // recursive function and is just a 'seed' to the recursion itself\r
+\r
+ bool match_wildcard(const std::string& wild, const std::string& match)\r
+ {\r
+ return match_remainder(wild, wild.begin(), match, match.begin());\r
+ }\r
+\r
+ ////////////////////////////////////////////////////////////////////////////////\r
+\r
+ std::vector<std::string> split(const std::string& str, const std::string& splitter)\r
+ {\r
+ std::vector<std::string> result;\r
+ if (!str.empty())\r
+ {\r
+ for(std::string::size_type offset = 0;;)\r
+ {\r
+ std::string::size_type found = str.find(splitter, offset);\r
+ if (found != std::string::npos)\r
+ {\r
+ result.push_back(str.substr(offset, found-offset));\r
+ offset = found + splitter.size();\r
+ }\r
+ else\r
+ {\r
+ result.push_back(str.substr(offset, str.size()-offset));\r
+ break;\r
+ }\r
+ }\r
+ }\r
+ return result;\r
+ }\r
+\r
+ std::string join (const std::vector<std::string>& str,\r
+ const std::string& joiner,\r
+ const std::string& prefix,\r
+ const std::string& suffix)\r
+ {\r
+ std::string result = prefix;\r
+ for (unsigned i = 0; i < str.size(); i++)\r
+ {\r
+ if (i) result += joiner;\r
+ result += str[i];\r
+ }\r
+ result += suffix;\r
+ return result;\r
+ }\r
+\r
+ ////////////////////////////////////////////////////////////////////////////////\r
+\r
+ std::string display_bytes(long bytes)\r
+ {\r
+ std::string result;\r
+ if (bytes < 0)\r
+ {\r
+ result += '-';\r
+ bytes = -bytes;\r
+ }\r
+ static const long kB = 1024l;\r
+ static const long MB = kB * kB;\r
+ static const long GB = MB * kB;\r
+ if (bytes < kB)\r
+ result += local_dformat("%i", bytes);\r
+ else if (bytes < (10l * kB))\r
+ result += local_dformat("%.2fk", ((float)bytes / (float)kB));\r
+ else if (bytes < (100l * kB))\r
+ result += local_dformat("%.1fk", ((float)bytes / (float)kB));\r
+ else if (bytes < MB)\r
+ result += local_dformat("%.0fk", ((float)bytes / (float)kB));\r
+ else if (bytes < (10l * MB))\r
+ result += local_dformat("%.2fM", ((float)bytes / (float)MB));\r
+ else if (bytes < (100l * MB))\r
+ result += local_dformat("%.1fM", ((float)bytes / (float)MB));\r
+ else if (bytes < GB)\r
+ result += local_dformat("%.0fM", ((float)bytes / (float)MB));\r
+ else\r
+ result += local_dformat("%.2fG", ((float)bytes / (float)GB));\r
+ return result;\r
+ }\r
+\r
+ std::string display_time(time_t seconds)\r
+ {\r
+ unsigned minutes = (unsigned)seconds / 60;\r
+ seconds %= 60;\r
+ unsigned hours = minutes / 60;\r
+ minutes %= 60;\r
+ unsigned days = hours / 24;\r
+ hours %= 24;\r
+ unsigned weeks = days / 7;\r
+ days %= 7;\r
+ std::string result;\r
+ if (weeks > 0)\r
+ {\r
+ result += unsigned_to_string(weeks, 10, radix_none, 1);\r
+ result += "w ";\r
+ }\r
+ if (!result.empty() || days > 0)\r
+ {\r
+ result += unsigned_to_string(days, 10, radix_none, 1);\r
+ result += "d ";\r
+ }\r
+ if (!result.empty() || hours > 0)\r
+ {\r
+ result += unsigned_to_string(hours, 10, radix_none, 1);\r
+ result += ":";\r
+ }\r
+ if (!result.empty() || minutes > 0)\r
+ {\r
+ if (!result.empty())\r
+ result += unsigned_to_string(minutes, 10, radix_none, 2);\r
+ else\r
+ result += unsigned_to_string(minutes, 10, radix_none, 1);\r
+ result += ":";\r
+ }\r
+ if (!result.empty())\r
+ result += unsigned_to_string((unsigned)seconds, 10, radix_none, 2);\r
+ else\r
+ {\r
+ result += unsigned_to_string((unsigned)seconds, 10, radix_none, 1);\r
+ result += "s";\r
+ }\r
+ return result;\r
+ }\r
+\r
+ ////////////////////////////////////////////////////////////////////////////////\r
+\r
+} // end namespace stlplus\r