Chris@629: /* -*- c-basic-offset: 4 indent-tabs-mode: nil -*-  vi:set ts=8 sts=4 sw=4: */
Chris@629: 
Chris@629: /*
Chris@629:     Sonic Visualiser
Chris@629:     An audio file viewer and annotation editor.
Chris@629:     Centre for Digital Music, Queen Mary, University of London.
Chris@629:     
Chris@629:     This program is free software; you can redistribute it and/or
Chris@629:     modify it under the terms of the GNU General Public License as
Chris@629:     published by the Free Software Foundation; either version 2 of the
Chris@629:     License, or (at your option) any later version.  See the file
Chris@629:     COPYING included with this distribution for more information.
Chris@629: */
Chris@629: 
Chris@629: /*
Chris@629:    This is a modified version of a source file from the 
Chris@629:    Rosegarden MIDI and audio sequencer and notation editor.
Chris@629:    This file copyright 2000-2010 Chris Cannam.
Chris@629: */
Chris@629: 
Chris@629: #include "StringBits.h"
Chris@629: 
Chris@1022: #include "Debug.h"
Chris@1022: 
Chris@1022: using namespace std;
Chris@1022: 
Chris@629: double
Chris@629: StringBits::stringToDoubleLocaleFree(QString s, bool *ok)
Chris@629: {
Chris@629:     int dp = 0;
Chris@629:     int sign = 1;
Chris@629:     int i = 0;
Chris@629:     double result = 0.0;
Chris@629:     int len = s.length();
Chris@629: 
Chris@629:     result = 0.0;
Chris@629: 
Chris@629:     if (ok) *ok = true;
Chris@629: 
Chris@629:     while (i < len && s[i].isSpace()) ++i;
Chris@629:     if (i < len && s[i] == '-') sign = -1;
Chris@629: 
Chris@629:     while (i < len) {
Chris@629: 
Chris@1429:         QChar c = s[i];
Chris@629: 
Chris@629:         if (c.isDigit()) {
Chris@629: 
Chris@629:             double d = c.digitValue();
Chris@629: 
Chris@629:             if (dp > 0) {
Chris@629:                 for (int p = dp; p > 0; --p) d /= 10.0;
Chris@629:                 ++dp;
Chris@629:             } else {
Chris@629:                 result *= 10.0;
Chris@629:             }
Chris@629: 
Chris@629:             result += d;
Chris@629: 
Chris@629:         } else if (c == '.') {
Chris@629: 
Chris@629:             dp = 1;
Chris@629: 
Chris@629:         } else if (ok) {
Chris@629:             *ok = false;
Chris@629:         }
Chris@629: 
Chris@629:         ++i;
Chris@629:     }
Chris@629: 
Chris@629:     return result * sign;
Chris@629: }
Chris@629:     
Chris@629: QStringList
Chris@1854: StringBits::splitQuoted(QString s, QChar separator, EscapeMode escapeMode)
Chris@629: {
Chris@629:     QStringList tokens;
Chris@629:     QString tok;
Chris@629: 
Chris@1854:     // beg -> at beginning of line
Chris@1854:     // sep -> just seen a field separator
Chris@1022:     // unq -> in an unquoted field
Chris@1022:     // q1  -> in a single-quoted field
Chris@1022:     // q2  -> in a double-quoted field
Chris@1022: 
Chris@1854:     enum { beg, sep, unq, q1, q2 } mode = beg;
Chris@1854: 
Chris@1854:     bool use_doubling = (escapeMode == EscapeDoubling ||
Chris@1854:                          escapeMode == EscapeAny);
Chris@1854:     bool use_backslash = (escapeMode == EscapeBackslash ||
Chris@1854:                           escapeMode == EscapeAny);
Chris@629: 
Chris@629:     for (int i = 0; i < s.length(); ++i) {
Chris@1429:         
Chris@1429:         QChar c = s[i];
Chris@629: 
Chris@1429:         if (c == '\'') {
Chris@1429:             switch (mode) {
Chris@1854:             case beg: case sep: mode = q1; break;
Chris@1429:             case unq: case q2: tok += c; break;
Chris@1854:             case q1:
Chris@1854:                 if (use_doubling && i+1 < s.length() && s[i+1] == c) {
Chris@1854:                     tok += c; ++i; break;
Chris@1854:                 } else {
Chris@1854:                     mode = unq; break;
Chris@1854:                 }
Chris@1429:             }
Chris@629: 
Chris@1429:         } else if (c == '"') {
Chris@1429:             switch (mode) {
Chris@1854:             case beg: case sep: mode = q2; break;
Chris@1429:             case unq: case q1: tok += c; break;
Chris@1854:             case q2: 
Chris@1854:                 if (use_doubling && i+1 < s.length() && s[i+1] == c) {
Chris@1854:                     tok += c; ++i; break;
Chris@1854:                 } else {
Chris@1854:                     mode = unq; break;
Chris@1854:                 }
Chris@1429:             }
Chris@629: 
Chris@1429:         } else if (c == separator || (separator == ' ' && c.isSpace())) {
Chris@1429:             switch (mode) {
Chris@1854:             case beg: mode = sep; tokens << ""; break;
Chris@1429:             case sep: if (separator != ' ') tokens << ""; break;
Chris@1429:             case unq: mode = sep; tokens << tok; tok = ""; break;
Chris@1429:             case q1: case q2: tok += c; break;
Chris@1429:             }
Chris@629: 
Chris@1854:         } else if (c == '\\' && use_backslash) {
Chris@1429:             if (++i < s.length()) {
Chris@1429:                 c = s[i];
Chris@1429:                 switch (mode) {
Chris@1854:                 case beg: case sep: mode = unq; tok += c; break;
Chris@928:                 case unq: case q1: case q2: tok += c; break;
Chris@1429:                 }
Chris@1429:             }
Chris@629: 
Chris@1429:         } else {
Chris@1429:             switch (mode) {
Chris@1854:             case beg: case sep: mode = unq; tok += c; break;
Chris@928:             case unq: case q1: case q2: tok += c; break;
Chris@1429:             }
Chris@1429:         }
Chris@629:     }
Chris@629: 
Chris@1854:     if (tok != "" || mode != beg) {
Chris@1022:         if (mode == q1) {
Chris@1022:             tokens << ("'" + tok);  // turns out it wasn't quoted after all
Chris@1022:         } else if (mode == q2) {
Chris@1022:             tokens << ("\"" + tok);
Chris@1022:         } else {
Chris@1022:             tokens << tok;
Chris@1022:         }
Chris@1022:     }
Chris@1022: 
Chris@629:     return tokens;
Chris@629: }
Chris@629: 
Chris@629: QStringList
Chris@629: StringBits::split(QString line, QChar separator, bool quoted)
Chris@629: {
Chris@629:     if (quoted) {
Chris@629:         return splitQuoted(line, separator);
Chris@629:     } else {
Chris@629:         return line.split(separator,
Chris@629:                           separator == ' ' ? QString::SkipEmptyParts :
Chris@629:                           QString::KeepEmptyParts);
Chris@629:     }
Chris@629: }
Chris@629: 
Chris@1833: QString
Chris@1833: StringBits::joinDelimited(QVector<QString> row, QString delimiter)
Chris@1833: {
Chris@1833:     QString s;
Chris@1833:     for (auto col: row) {
Chris@1833:         if (s != "") {
Chris@1833:             s += delimiter;
Chris@1833:         }
Chris@1833:         if (col.contains(delimiter)) {
Chris@1833:             col.replace("\"", "\"\"");
Chris@1833:             col = "\"" + col + "\"";
Chris@1833:         }
Chris@1833:         s += col;
Chris@1833:     }
Chris@1833:     return s;    
Chris@1833: }
Chris@1833: 
Chris@1851: bool
Chris@1851: StringBits::isValidUtf8(const std::string &bytes, bool isTruncated)
Chris@1851: {
Chris@1851:     size_t len = bytes.length();
Chris@1851:     size_t mblen = 0;
Chris@1851:     unsigned char first = '\0';
Chris@1851:     
Chris@1851:     for (size_t i = 0; i < len; ++i) {
Chris@1851: 
Chris@1851: 	unsigned char c = bytes[i];
Chris@1851: 
Chris@1851: 	if (((c & 0xc0) == 0xc0) || !(c & 0x80)) {
Chris@1851: 
Chris@1851: 	    // 11xxxxxx or 0xxxxxxx: first byte of a character sequence
Chris@1851: 
Chris@1851: 	    if (mblen > 0) {
Chris@1851:                 
Chris@1851: 		// have we seen a valid sequence?
Chris@1851: 		size_t length = 
Chris@1851: 		    (!(first & 0x20)) ? 2 :
Chris@1851: 		    (!(first & 0x10)) ? 3 :
Chris@1851: 		    (!(first & 0x08)) ? 4 :
Chris@1851: 		    (!(first & 0x04)) ? 5 : 0;
Chris@1851: 
Chris@1851:                 if (length != 0 && mblen != length) {
Chris@1851:                     // previous multibyte sequence had invalid length
Chris@1851:                     return false;
Chris@1851:                 }
Chris@1851:             }
Chris@1851: 
Chris@1851:             mblen = 0;
Chris@1851:             first = c;
Chris@1851: 
Chris@1851:         } else {
Chris@1851:             
Chris@1851: 	    // second or subsequent byte
Chris@1851: 
Chris@1851: 	    if (mblen == 0) {
Chris@1851:                 // ... without a first byte!                
Chris@1851:                 return false;
Chris@1851:             }
Chris@1851:         }
Chris@1851:     }
Chris@1851: 
Chris@1851:     // at the end
Chris@1851: 
Chris@1851:     if (isTruncated) {
Chris@1851:         // can't trust any errors arising now
Chris@1851:         return true;
Chris@1851:     }
Chris@1851: 
Chris@1851:     if (mblen > 0) {
Chris@1851:                 
Chris@1851:         // have we seen a valid sequence?
Chris@1851:         size_t length = 
Chris@1851:             (!(first & 0x20)) ? 2 :
Chris@1851:             (!(first & 0x10)) ? 3 :
Chris@1851:             (!(first & 0x08)) ? 4 :
Chris@1851:             (!(first & 0x04)) ? 5 : 0;
Chris@1851:         
Chris@1851:         if (length != 0 && mblen != length) {
Chris@1851:             // final multibyte sequence had invalid length
Chris@1851:             return false;
Chris@1851:         }
Chris@1851:     }
Chris@1851: 
Chris@1851:     return true;
Chris@1851: }
Chris@1851: 
Chris@1851: 
Chris@1851:         
Chris@1851: