1 /***************************************************************************
2 * Copyright (C) 2005-2018 by the Quassel Project *
3 * devel@quassel-irc.org *
5 * This program is free software; you can redistribute it and/or modify *
6 * it under the terms of the GNU General Public License as published by *
7 * the Free Software Foundation; either version 2 of the License, or *
8 * (at your option) version 3. *
10 * This program is distributed in the hope that it will be useful, *
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of *
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the *
13 * GNU General Public License for more details. *
15 * You should have received a copy of the GNU General Public License *
16 * along with this program; if not, write to the *
17 * Free Software Foundation, Inc., *
18 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. *
19 ***************************************************************************/
27 #include <QCoreApplication>
35 // MIBenum values from http://www.iana.org/assignments/character-sets/character-sets.xml#table-character-sets-1
36 static QList<int> utf8DetectionBlacklist = QList<int>()
37 << 39 /* ISO-2022-JP */;
39 QString nickFromMask(const QString &mask)
41 return mask.left(mask.indexOf('!'));
45 QString userFromMask(const QString &mask)
47 const int offset = mask.indexOf('!') + 1;
50 const int length = mask.indexOf('@', offset) - offset;
51 return mask.mid(offset, length >= 0 ? length : -1);
55 QString hostFromMask(const QString &mask)
57 const int excl = mask.indexOf('!');
60 const int offset = mask.indexOf('@', excl + 1) + 1;
61 return offset > 0 && offset < mask.size() ? mask.mid(offset) : QString{};
65 bool isChannelName(const QString &str)
69 static constexpr std::array<quint8, 4> prefixes{{'#', '&', '!', '+'}};
70 return std::any_of(prefixes.cbegin(), prefixes.cend(), [&str](quint8 c) { return c == str[0]; });
74 QString stripFormatCodes(QString message)
76 static QRegExp regEx{"\x03(\\d\\d?(,\\d\\d?)?)?|\x04([\\da-fA-F]{6}(,[\\da-fA-F]{6})?)?|[\x02\x0f\x11\x12\x16\x1d\x1e\x1f]"};
77 return message.remove(regEx);
81 QString stripAcceleratorMarkers(const QString &label_)
83 QString label = label_;
86 p = label.indexOf('&', p);
87 if (p < 0 || p + 1 >= label.length())
90 if (label.at(p + 1).isLetterOrNumber() || label.at(p + 1) == '&')
99 QString decodeString(const QByteArray &input, QTextCodec *codec)
101 if (codec && utf8DetectionBlacklist.contains(codec->mibEnum()))
102 return codec->toUnicode(input);
104 // First, we check if it's utf8. It is very improbable to encounter a string that looks like
105 // valid utf8, but in fact is not. This means that if the input string passes as valid utf8, it
106 // is safe to assume that it is.
107 // Q_ASSERT(sizeof(const char) == sizeof(quint8)); // In God we trust...
110 for (uchar c : input) {
112 // We check a part of a multibyte char. These need to be of the form 10yyyyyy.
113 if ((c & 0xc0) != 0x80) { isUtf8 = false; break; }
117 if ((c & 0x80) == 0x00) continue; // 7 bit is always ok
118 if ((c & 0xf8) == 0xf0) { cnt = 3; continue; } // 4-byte char 11110xxx 10yyyyyy 10zzzzzz 10vvvvvv
119 if ((c & 0xf0) == 0xe0) { cnt = 2; continue; } // 3-byte char 1110xxxx 10yyyyyy 10zzzzzz
120 if ((c & 0xe0) == 0xc0) { cnt = 1; continue; } // 2-byte char 110xxxxx 10yyyyyy
121 isUtf8 = false; break; // 8 bit char, but not utf8!
123 if (isUtf8 && cnt == 0) {
124 QString s = QString::fromUtf8(input);
125 //qDebug() << "Detected utf8:" << s;
128 //QTextCodec *codec = QTextCodec::codecForName(encoding.toLatin1());
129 if (!codec) return QString::fromLatin1(input);
130 return codec->toUnicode(input);
134 uint editingDistance(const QString &s1, const QString &s2)
136 uint n = s1.size()+1;
137 uint m = s2.size()+1;
138 QVector<QVector<uint> > matrix(n, QVector<uint>(m, 0));
140 for (uint i = 0; i < n; i++)
143 for (uint i = 0; i < m; i++)
147 for (uint i = 1; i < n; i++) {
148 for (uint j = 1; j < m; j++) {
149 uint deleteChar = matrix[i-1][j] + 1;
150 uint insertChar = matrix[i][j-1] + 1;
152 if (deleteChar < insertChar)
157 if (s1[i-1] == s2[j-1]) {
158 uint inheritChar = matrix[i-1][j-1];
159 if (inheritChar < min)
166 return matrix[n-1][m-1];
170 QString secondsToString(int timeInSeconds)
172 static QVector<std::pair<int, QString>> timeUnit {
173 std::make_pair(365*24*60*60, QCoreApplication::translate("Quassel::secondsToString()", "year")),
174 std::make_pair(24*60*60, QCoreApplication::translate("Quassel::secondsToString()", "day")),
175 std::make_pair(60*60, QCoreApplication::translate("Quassel::secondsToString()", "h")),
176 std::make_pair(60, QCoreApplication::translate("Quassel::secondsToString()", "min")),
177 std::make_pair(1, QCoreApplication::translate("Quassel::secondsToString()", "sec"))
180 if (timeInSeconds != 0) {
181 QStringList returnString;
182 for (const auto &tu : timeUnit) {
183 int n = timeInSeconds / tu.first;
185 returnString += QString("%1 %2").arg(QString::number(n), tu.second);
187 timeInSeconds = timeInSeconds % tu.first;
189 return returnString.join(", ");
192 return QString("%1 %2").arg(QString::number(timeInSeconds), timeUnit.last().second);
196 QByteArray prettyDigest(const QByteArray &digest)
198 QByteArray hexDigest = digest.toHex().toUpper();
199 QByteArray prettyDigest;
200 prettyDigest.fill(':', hexDigest.count() + (hexDigest.count() / 2) - 1);
202 for (int i = 0; i * 2 < hexDigest.count(); i++) {
203 prettyDigest.replace(i * 3, 2, hexDigest.mid(i * 2, 2));
209 QString formatCurrentDateTimeInString(const QString &formatStr)
211 // Work on a copy of the string to avoid modifying the input string
212 QString formattedStr = QString(formatStr);
214 // Exit early if there's nothing to format
215 if (formattedStr.isEmpty())
218 // Find %%<text>%% in string. Replace inside text formatted to QDateTime with the current
219 // timestamp, using %%%% as an escape for multiple %% signs.
221 // Simple: "All Quassel clients vanished from the face of the earth... %%hh:mm:ss%%"
222 // > Result: "All Quassel clients vanished from the face of the earth... 23:20:34"
223 // Complex: "Away since %%hh:mm%% on %%dd.MM%% - %%%% not here %%%%"
224 // > Result: "Away since 23:20 on 21.05 - %% not here %%"
226 // Match groups of double % signs - Some text %%inside here%%, and even %%%%:
228 // (...) marks a capturing group
229 // .* matches zero or more characters, not including newlines
230 // Note that '\' must be escaped as '\\'
231 // Helpful interactive website for debugging and explaining: https://regex101.com/
232 QRegExp regExpMatchTime("%%(.*)%%");
234 // Preserve the smallest groups possible to allow for multiple %%blocks%%
235 regExpMatchTime.setMinimal(true);
237 // NOTE: Move regExpMatchTime to a static regular expression if used anywhere that performance
240 // Don't allow a runaway regular expression to loop for too long. This might not happen.. but
241 // when dealing with user input, better to be safe..?
242 int numIterations = 0;
244 // Find each group of %%text here%% starting from the beginning
245 int index = regExpMatchTime.indexIn(formattedStr);
247 QString matchedFormat;
248 while (index >= 0 && numIterations < 512) {
249 // Get the total length of the matched expression
250 matchLength = regExpMatchTime.cap(0).length();
251 // Get the format string, e.g. "this text here" from "%%this text here%%"
252 matchedFormat = regExpMatchTime.cap(1);
253 // Check that there's actual characters inside. A quadruple % (%%%%) represents two %%
255 if (matchedFormat.length() > 0) {
256 // Format the string according to the current date and time. Invalid time format
257 // strings are ignored.
258 formattedStr.replace(index, matchLength,
259 QDateTime::currentDateTime().toString(matchedFormat));
260 // Subtract the length of the removed % signs
261 // E.g. "%%h:mm ap%%" turns into "h:mm ap", removing four % signs, thus -4. This is
262 // used below to determine how far to advance when looking for the next formatting code.
264 } else if (matchLength == 4) {
265 // Remove two of the four percent signs, so '%%%%' escapes to '%%'
266 formattedStr.remove(index, 2);
267 // Subtract the length of the removed % signs, this time removing two % signs, thus -2.
270 // If neither of these match, something went wrong. Don't modify it to be safe.
271 qDebug() << "Unexpected time format when parsing string, no matchedFormat, matchLength "
272 "should be 4, actually is" << matchLength;
275 // Find the next group of %%text here%% starting from where the last group ended
276 index = regExpMatchTime.indexIn(formattedStr, index + matchLength);
284 QString tryFormatUnixEpoch(const QString &possibleEpochDate, Qt::DateFormat dateFormat, bool useUTC)
286 // Does the string resemble a Unix epoch? Parse as 64-bit time
287 qint64 secsSinceEpoch = possibleEpochDate.toLongLong();
288 if (secsSinceEpoch == 0) {
289 // Parsing either failed, or '0' was sent. No need to distinguish; either way, it's not
291 // See https://doc.qt.io/qt-5/qstring.html#toLongLong
292 return possibleEpochDate;
295 // Time checks out, parse it
297 #if QT_VERSION >= 0x050800
298 date.setSecsSinceEpoch(secsSinceEpoch);
300 // toSecsSinceEpoch() was added in Qt 5.8. Manually downconvert to seconds for now.
301 // See https://doc.qt.io/qt-5/qdatetime.html#toMSecsSinceEpoch
302 date.setMSecsSinceEpoch(secsSinceEpoch * 1000);
305 // Return the localized date/time
308 if (dateFormat == Qt::DateFormat::ISODate) {
309 // Replace the "T" date/time separator with " " for readability. This isn't quite the
310 // ISO 8601 spec (it specifies omitting the "T" entirely), but RFC 3339 allows this.
311 // Go with RFC 3339 for human readability that's still machine-parseable, too.
313 // Before: 2018-06-21T21:35:52Z
314 // After: 2018-06-21 21:35:52Z
315 // ..........^ (10th character)
317 // See https://en.wikipedia.org/wiki/ISO_8601#cite_note-32
318 // And https://www.ietf.org/rfc/rfc3339.txt
319 return date.toUTC().toString(dateFormat).replace(10, 1, " ");
321 return date.toUTC().toString(dateFormat);
323 } else if (dateFormat == Qt::DateFormat::ISODate) {
324 // Add in ISO local timezone information via special handling below
325 // formatDateTimeToOffsetISO() handles converting "T" to " "
326 return formatDateTimeToOffsetISO(date);
329 return date.toString(dateFormat);
334 QString formatDateTimeToOffsetISO(const QDateTime &dateTime)
336 if (!dateTime.isValid()) {
337 // Don't try to do anything with invalid date/time
338 return "formatDateTimeToISO() invalid date/time";
341 // Replace the "T" date/time separator with " " for readability. This isn't quite the ISO 8601
342 // spec (it specifies omitting the "T" entirely), but RFC 3339 allows this. Go with RFC 3339
343 // for human readability that's still machine-parseable, too.
345 // Before: 2018-08-22T18:43:10-05:00
346 // After: 2018-08-22 18:43:10-05:00
347 // ..........^ (10th character)
349 // See https://en.wikipedia.org/wiki/ISO_8601#cite_note-32
350 // And https://www.ietf.org/rfc/rfc3339.txt
353 // The expected way to get a UTC offset on ISO 8601 dates
354 // Remove the "T" date/time separator
355 return dateTime.toTimeSpec(Qt::OffsetFromUTC).toString(Qt::ISODate).replace(10, 1, " ");
357 // Work around Qt bug that converts to UTC instead of including timezone information
358 // See https://bugreports.qt.io/browse/QTBUG-26161
360 // NOTE: Despite the bug report marking as fixed in Qt 5.2.0 (QT_VERSION >= 0x050200), this
361 // still appears broken in Qt 5.5.1.
363 // Credit to "user362638" for the solution below, modified to fit Quassel's needs
364 // https://stackoverflow.com/questions/18750569/qdatetime-isodate-with-timezone
366 // Get the local and UTC time
367 QDateTime local = QDateTime(dateTime);
368 QDateTime utc = local.toUTC();
369 utc.setTimeSpec(Qt::LocalTime);
371 // Find the UTC offset
372 int utcOffset = utc.secsTo(local);
374 // Force the local time to follow this offset
375 local.setUtcOffset(utcOffset);
376 // Now the output should be correct
377 // Remove the "T" date/time separator
378 return local.toString(Qt::ISODate).replace(10, 1, " ");