mimedatabase.cpp 67.3 KB
Newer Older
1
/**************************************************************************
con's avatar
con committed
2 3 4
**
** This file is part of Qt Creator
**
hjk's avatar
hjk committed
5
** Copyright (c) 2012 Nokia Corporation and/or its subsidiary(-ies).
con's avatar
con committed
6
**
7
** Contact: Nokia Corporation (qt-info@nokia.com)
con's avatar
con committed
8
**
9
**
10
** GNU Lesser General Public License Usage
11
**
hjk's avatar
hjk committed
12 13 14 15 16 17
** This file may be used under the terms of the GNU Lesser General Public
** License version 2.1 as published by the Free Software Foundation and
** appearing in the file LICENSE.LGPL included in the packaging of this file.
** Please review the following information to ensure the GNU Lesser General
** Public License version 2.1 requirements will be met:
** http://www.gnu.org/licenses/old-licenses/lgpl-2.1.html.
18
**
con's avatar
con committed
19
** In addition, as a special exception, Nokia gives you certain additional
hjk's avatar
hjk committed
20
** rights. These rights are described in the Nokia Qt LGPL Exception
con's avatar
con committed
21 22
** version 1.1, included in the file LGPL_EXCEPTION.txt in this package.
**
hjk's avatar
hjk committed
23 24 25 26 27
** Other Usage
**
** Alternatively, this file may be used in accordance with the terms and
** conditions contained in a signed written agreement between you and Nokia.
**
con's avatar
con committed
28
** If you have questions regarding the use of this file, please contact
29
** Nokia at qt-info@nokia.com.
con's avatar
con committed
30
**
31
**************************************************************************/
hjk's avatar
hjk committed
32

con's avatar
con committed
33
#include "mimedatabase.h"
34
#include "coreconstants.h"
35
#include "icore.h"
con's avatar
con committed
36

hjk's avatar
hjk committed
37 38
#include <utils/qtcassert.h>

39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56
#include <QByteArray>
#include <QCoreApplication>
#include <QDebug>
#include <QFile>
#include <QDir>
#include <QFileInfo>
#include <QLocale>
#include <QMap>
#include <QHash>
#include <QMultiHash>
#include <QRegExp>
#include <QSharedData>
#include <QSharedPointer>
#include <QStringList>
#include <QTextStream>
#include <QMutexLocker>
#include <QXmlStreamReader>
#include <QXmlStreamWriter>
con's avatar
con committed
57

Christian Kandeler's avatar
Christian Kandeler committed
58
#include <algorithm>
59
#include <functional>
Christian Kandeler's avatar
Christian Kandeler committed
60

con's avatar
con committed
61 62 63
enum { debugMimeDB = 0 };

// XML tags in mime files
64 65 66 67 68 69 70 71 72 73
static const char mimeInfoTagC[] = "mime-info";
static const char mimeTypeTagC[] = "mime-type";
static const char mimeTypeAttributeC[] = "type";
static const char subClassTagC[] = "sub-class-of";
static const char commentTagC[] = "comment";
static const char globTagC[] = "glob";
static const char aliasTagC[] = "alias";
static const char patternAttributeC[] = "pattern";
static const char weightAttributeC[] = "weight";
static const char localeAttributeC[] = "xml:lang";
con's avatar
con committed
74

75 76 77 78 79 80 81 82
static const char magicTagC[] = "magic";
static const char priorityAttributeC[] = "priority";
static const char matchTagC[] = "match";
static const char matchValueAttributeC[] = "value";
static const char matchTypeAttributeC[] = "type";
static const char matchStringTypeValueC[] = "string";
static const char matchByteTypeValueC[] = "byte";
static const char matchOffsetAttributeC[] = "offset";
con's avatar
con committed
83 84

// Types
85 86
static const char textTypeC[] = "text/plain";
static const char binaryTypeC[] = "application/octet-stream";
con's avatar
con committed
87 88 89 90 91 92

// UTF16 byte order marks
static const char bigEndianByteOrderMarkC[] = "\xFE\xFF";
static const char littleEndianByteOrderMarkC[] = "\xFF\xFE";

// Fallback priorities, must be low.
93 94 95 96
enum {
    BinaryMatchPriority = Core::MimeGlobPattern::MinWeight + 1,
    TextMatchPriority
};
con's avatar
con committed
97

98 99 100 101 102 103 104 105
/*!
    \class Core::IMagicMatcher

    \brief Interface for a Mime type magic matcher (examinig file contents).

    \sa Core::MimeType, Core::MimeDatabase, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
con's avatar
con committed
106 107 108
*/

namespace Core {
109 110 111

typedef QSharedPointer<MagicRuleMatcher> MagicRuleMatcherPtr;

con's avatar
con committed
112 113
namespace Internal {

114 115 116 117 118 119 120 121 122 123 124 125
/*!
    \class Core::Internal::FileMatchContext

    \brief Context passed on to the mime types when looking for a file match.

    It exists to enable reading the file contents "on demand"
    (as opposed to each mime type trying to open and read while checking).

    \sa Core::MimeType, Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
*/
con's avatar
con committed
126

hjk's avatar
hjk committed
127 128
class FileMatchContext
{
129
    Q_DISABLE_COPY(FileMatchContext)
hjk's avatar
hjk committed
130

con's avatar
con committed
131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178
public:
    // Max data to be read from a file
    enum { MaxData = 2048 };

    explicit FileMatchContext(const QFileInfo &fi);

    inline QString fileName() const { return m_fileName; }
    // Return (cached) first MaxData bytes of file
    QByteArray data();

private:
    enum State {
        // File cannot be read/does not exist
        NoDataAvailable,
        // Not read yet
        DataNotRead,
        // Available
        DataRead };
    const QFileInfo m_fileInfo;
    const QString m_fileName;
    State m_state;
    QByteArray m_data;
};

FileMatchContext::FileMatchContext(const QFileInfo &fi) :
    m_fileInfo(fi),
    m_fileName(fi.fileName()),
    m_state(fi.isFile() && fi.isReadable() && fi.size() > 0 ? DataNotRead : NoDataAvailable)
{
}

QByteArray FileMatchContext::data()
{
    // Do we need to read?
    if (m_state == DataNotRead) {
        const QString fullName = m_fileInfo.absoluteFilePath();
        QFile file(fullName);
        if (file.open(QIODevice::ReadOnly)) {
            m_data = file.read(MaxData);
            m_state = DataRead;
        } else {
            qWarning("%s failed to open %s: %s\n", Q_FUNC_INFO, fullName.toUtf8().constData(), file.errorString().toUtf8().constData());
            m_state = NoDataAvailable;
        }
    }
    return m_data;
}

179 180 181 182 183 184 185 186 187
/*!
    \class Core::Internal::BinaryMatcher
    \brief The binary fallback matcher for mime type "application/octet-stream".

    \sa Core::MimeType, Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
*/

hjk's avatar
hjk committed
188 189
class BinaryMatcher : public IMagicMatcher
{
con's avatar
con committed
190 191 192 193 194 195
public:
    BinaryMatcher() {}
    virtual bool matches(const QByteArray & /*data*/) const { return true; }
    virtual int priority() const  { return BinaryMatchPriority; }
};

196 197 198 199 200 201 202 203 204 205 206 207
/*!
    \class Core::Internal::HeuristicTextMagicMatcher
    \brief Heuristic text file matcher for mime types.

    If the data do not contain any character below tab (9), detect as text.
    Additionally, check on UTF16 byte order markers.

    \sa Core::MimeType, Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
*/

hjk's avatar
hjk committed
208 209
class HeuristicTextMagicMatcher : public IMagicMatcher
{
con's avatar
con committed
210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240
public:
    HeuristicTextMagicMatcher() {}
    virtual bool matches(const QByteArray &data) const;
    virtual int priority() const  { return TextMatchPriority; }

    static bool isTextFile(const QByteArray &data);
};

bool HeuristicTextMagicMatcher::isTextFile(const QByteArray &data)
{
    const int size = data.size();
    for (int i = 0; i < size; i++) {
        const char c = data.at(i);
        if (c >= 0x01 && c < 0x09) // Sure-fire binary
            return false;
        if (c == 0)             // Check for UTF16
            return data.startsWith(bigEndianByteOrderMarkC) || data.startsWith(littleEndianByteOrderMarkC);
    }
    return true;
}

bool HeuristicTextMagicMatcher::matches(const QByteArray &data) const
{
    const bool rc = isTextFile(data);
    if (debugMimeDB)
        qDebug() << Q_FUNC_INFO << " on " << data.size() << " returns " << rc;
    return rc;
}

} // namespace Internal

241 242 243 244 245 246 247 248 249 250 251 252 253 254
/*!
    \class Core::MagicRule
    \brief Base class for standard Magic match rules based on contents
    and offset specification.

    Stores the offset and provides conversion helpers.
    Base class for implementations for "string" and "byte".
    (Others like little16, big16, etc. can be created whenever there is a need.)

    \sa Core::MimeType, Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
 */

255 256
const QChar MagicRule::kColon(QLatin1Char(':'));

257
MagicRule::MagicRule(int startPos, int endPos) : m_startPos(startPos), m_endPos(endPos)
con's avatar
con committed
258 259 260
{
}

261 262 263 264 265 266 267 268 269 270 271 272 273 274
MagicRule::~MagicRule()
{
}

int MagicRule::startPos() const
{
    return m_startPos;
}

int MagicRule::endPos() const
{
    return m_endPos;
}

275 276 277 278 279 280 281 282 283 284 285 286
QString MagicRule::toOffset(const QPair<int, int> &startEnd)
{
    return QString(QLatin1String("%1:%2")).arg(startEnd.first).arg(startEnd.second);
}

QPair<int, int> MagicRule::fromOffset(const QString &offset)
{
    const QStringList &startEnd = offset.split(kColon);
    Q_ASSERT(startEnd.size() == 2);
    return qMakePair(startEnd.at(0).toInt(), startEnd.at(1).toInt());
}

287 288 289 290 291 292 293 294 295
/*!
    \class Core::MagicStringRule
    \brief Match on a string.

    \sa Core::MimeType, Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
 */

296
const QString MagicStringRule::kMatchType(QLatin1String("string"));
297

298 299 300 301 302 303 304 305 306
MagicStringRule::MagicStringRule(const QString &s, int startPos, int endPos) :
    MagicRule(startPos, endPos), m_pattern(s.toUtf8())
{
}

MagicStringRule::~MagicStringRule()
{
}

307 308 309 310 311 312 313
QString MagicStringRule::matchType() const
{
    return kMatchType;
}

QString MagicStringRule::matchValue() const
{
314
    return QLatin1String(m_pattern);
315 316
}

317
bool MagicStringRule::matches(const QByteArray &data) const
con's avatar
con committed
318 319
{
    // Quick check
320
    if ((startPos() + m_pattern.size()) > data.size())
con's avatar
con committed
321 322
        return false;
    // Most common: some string at position 0:
323
    if (startPos() == 0 && startPos() == endPos())
con's avatar
con committed
324 325
        return data.startsWith(m_pattern);
    // Range
326 327
    const int index = data.indexOf(m_pattern, startPos());
    const bool rc = index != -1 && index <= endPos();
328
    if (debugMimeDB)
329
        qDebug() << "Checking " << m_pattern << startPos() << endPos() << " returns " << rc;
330
    return rc;
con's avatar
con committed
331 332
}

333 334 335 336 337 338 339 340 341 342 343 344 345 346
/*!
    \class Core::MagicByteRule
    \brief Match on a sequence of binary data.

    Format:
    \code
    \0x7f\0x45\0x4c\0x46
    \endcode

    \sa Core::MimeType, Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
 */

347 348
const QString MagicByteRule::kMatchType(QLatin1String("byte"));

349
MagicByteRule::MagicByteRule(const QString &s, int startPos, int endPos) :
350 351 352 353 354 355 356 357 358 359 360 361 362
    MagicRule(startPos, endPos), m_bytesSize(0)
{
    if (validateByteSequence(s, &m_bytes))
        m_bytesSize = m_bytes.size();
    else
        m_bytes.clear();
}

MagicByteRule::~MagicByteRule()
{
}

bool MagicByteRule::validateByteSequence(const QString &sequence, QList<int> *bytes)
con's avatar
con committed
363
{
364
    // Expect an hex format value like this: \0x7f\0x45\0x4c\0x46
365 366
    const QStringList &byteSequence = sequence.split(QLatin1Char('\\'), QString::SkipEmptyParts);
    foreach (const QString &byte, byteSequence) {
367 368 369
        bool ok;
        const int hex = byte.toInt(&ok, 16);
        if (ok) {
370 371
            if (bytes)
                bytes->push_back(hex);
372
        } else {
373
            return false;
374 375
        }
    }
376
    return true;
377 378
}

379
QString MagicByteRule::matchType() const
380
{
381 382 383 384 385 386 387 388 389
    return kMatchType;
}

QString MagicByteRule::matchValue() const
{
    QString value;
    foreach (int byte, m_bytes)
        value.append(QString(QLatin1String("\\0x%1")).arg(byte, 0, 16));
    return value;
390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412
}

bool MagicByteRule::matches(const QByteArray &data) const
{
    if (m_bytesSize == 0)
        return false;

    const int dataSize = data.size();
    for (int start = startPos(); start <= endPos(); ++start) {
        if ((start + m_bytesSize) > dataSize)
            return false;

        int matchAt = 0;
        while (matchAt < m_bytesSize) {
            if (data.at(start + matchAt) != m_bytes.at(matchAt))
                break;
            ++matchAt;
        }
        if (matchAt == m_bytesSize)
            return true;
    }

    return false;
con's avatar
con committed
413 414
}

415 416 417 418 419 420 421 422 423 424 425 426
/*!
    \class Core::MagicRuleMatcher

    \brief A Magic matcher that checks a number of rules based on operator "or".

    It is used for rules parsed from XML files.

    \sa Core::MimeType, Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
*/

con's avatar
con committed
427 428 429 430 431 432 433
MagicRuleMatcher::MagicRuleMatcher() :
     m_priority(65535)
{
}

void MagicRuleMatcher::add(const MagicRuleSharedPointer &rule)
{
434 435 436 437 438 439 440 441 442 443 444
    m_list.append(rule);
}

void MagicRuleMatcher::add(const MagicRuleList &ruleList)
{
    m_list.append(ruleList);
}

MagicRuleMatcher::MagicRuleList MagicRuleMatcher::magicRules() const
{
    return m_list;
con's avatar
con committed
445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465
}

bool MagicRuleMatcher::matches(const QByteArray &data) const
{
    const MagicRuleList::const_iterator cend = m_list.constEnd();
    for (MagicRuleList::const_iterator it = m_list.constBegin(); it != cend; ++it)
        if ( (*it)->matches(data))
            return true;
    return false;
}

int MagicRuleMatcher::priority() const
{
    return m_priority;
}

void MagicRuleMatcher::setPriority(int p)
{
    m_priority = p;
}

466 467 468 469 470 471 472 473 474 475 476 477 478 479
IMagicMatcher::IMagicMatcherList MagicRuleMatcher::createMatchers(
    const QHash<int, MagicRuleList> &rulesByPriority)
{
    IMagicMatcher::IMagicMatcherList matchers;
    QHash<int, MagicRuleList>::const_iterator ruleIt = rulesByPriority.begin();
    for (; ruleIt != rulesByPriority.end(); ++ruleIt) {
        MagicRuleMatcher *magicRuleMatcher = new MagicRuleMatcher();
        magicRuleMatcher->setPriority(ruleIt.key());
        magicRuleMatcher->add(ruleIt.value());
        matchers.append(IMagicMatcher::IMagicMatcherSharedPointer(magicRuleMatcher));
    }
    return matchers;
}

480 481 482 483 484 485 486 487 488
/*!
    \class Core::GlobPattern
    \brief Glob pattern for file names for mime type matching.

    \sa Core::MimeType, Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
*/

Friedemann Kleint's avatar
Friedemann Kleint committed
489
MimeGlobPattern::MimeGlobPattern(const QRegExp &regExp, unsigned weight) :
490 491 492 493 494 495 496 497 498 499 500 501 502
    m_regExp(regExp), m_weight(weight)
{
}

MimeGlobPattern::~MimeGlobPattern()
{
}

const QRegExp &MimeGlobPattern::regExp() const
{
    return m_regExp;
}

Friedemann Kleint's avatar
Friedemann Kleint committed
503
unsigned MimeGlobPattern::weight() const
504 505 506 507
{
    return m_weight;
}

508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539
/*!
    \class Core::MimeType

    \brief Mime type data used in Qt Creator.

    Contains most information from standard mime type XML database files.

    Currently, magic of types "string", "bytes" is supported. In addition,
    C++ classes, derived from Core::IMagicMatcher can be added to check
    on contents.

    In addition, the class provides a list of suffixes and a concept of the
    'preferred suffix' (derived from glob patterns). This is used for example
    to be able to configure the suffix used for C++-files in Qt Creator.

    Mime XML looks like:
    \code
    <?xml version="1.0" encoding="UTF-8"?>
    <mime-info xmlns="http://www.freedesktop.org/standards/shared-mime-info">
    <!-- Mime types must match the desktop file associations -->
      <mime-type type="application/vnd.nokia.qt.qmakeprofile">
        <comment xml:lang="en">Qt qmake Profile</comment>
        <glob pattern="*.pro" weight="50"/>
      </mime-type>
    </mime-info>
    \endcode

    \sa Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
*/

con's avatar
con committed
540 541 542
class MimeTypeData : public QSharedData {
public:
    typedef QHash<QString,QString> LocaleHash;
543 544 545

    MimeTypeData();

con's avatar
con committed
546
    void clear();
547 548
    void assignSuffix(const QString &pattern);
    void assignSuffixes(const QStringList &patterns);
con's avatar
con committed
549 550
    void debug(QTextStream &str, int indent = 0) const;

551 552
    const QRegExp suffixPattern;

con's avatar
con committed
553 554 555 556 557
    QString type;
    QString comment;

    LocaleHash localeComments;
    QStringList aliases;
558
    QList<MimeGlobPattern> globPatterns;
con's avatar
con committed
559 560 561
    QStringList subClassesOf;
    QString preferredSuffix;
    QStringList suffixes;
562
    IMagicMatcher::IMagicMatcherList magicMatchers;
con's avatar
con committed
563 564
};

565 566 567 568 569
MimeTypeData::MimeTypeData()
    // RE to match a suffix glob pattern: "*.ext" (and not sth like "Makefile" or
    // "*.log[1-9]"
    : suffixPattern(QLatin1String("^\\*\\.[\\w+]+$"))
{
570
    QTC_CHECK(suffixPattern.isValid());
571 572
}

con's avatar
con committed
573 574 575 576 577 578 579 580 581 582 583 584
void MimeTypeData::clear()
{
    type.clear();
    comment.clear();
    aliases.clear();
    globPatterns.clear();
    subClassesOf.clear();
    preferredSuffix.clear();
    suffixes.clear();
    magicMatchers.clear();
}

585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600
void MimeTypeData::assignSuffix(const QString &pattern)
{
    if (suffixPattern.exactMatch(pattern)) {
        const QString suffix = pattern.right(pattern.size() - 2);
        suffixes.push_back(suffix);
        if (preferredSuffix.isEmpty())
            preferredSuffix = suffix;
    }
}

void MimeTypeData::assignSuffixes(const QStringList &patterns)
{
    foreach (const QString &pattern, patterns)
        assignSuffix(pattern);
}

con's avatar
con committed
601 602 603 604 605 606 607 608 609 610 611 612 613
void MimeTypeData::debug(QTextStream &str, int indent) const
{
    const QString indentS = QString(indent, QLatin1Char(' '));
    const QString comma = QString(1, QLatin1Char(','));
    str << indentS << "Type: " << type;
    if (!aliases.empty())
        str << " Aliases: " << aliases.join(comma);
    str << ", magic: " << magicMatchers.size() << '\n';
    str << indentS << "Comment: " << comment << '\n';
    if (!subClassesOf.empty())
        str << indentS << "SubClassesOf: " << subClassesOf.join(comma) << '\n';
    if (!globPatterns.empty()) {
        str << indentS << "Glob: ";
614
        foreach (const MimeGlobPattern &gp, globPatterns)
Friedemann Kleint's avatar
Friedemann Kleint committed
615
            str << gp.regExp().pattern() << '(' << gp.weight() << ')';
con's avatar
con committed
616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725
        str << '\n';
        if (!suffixes.empty()) {
            str <<  indentS << "Suffixes: " << suffixes.join(comma)
                << " preferred: " << preferredSuffix << '\n';
        }
    }
    str << '\n';
}

// ---------------- MimeType
MimeType::MimeType() :
    m_d(new MimeTypeData)
{
}

MimeType::MimeType(const MimeType &rhs) :
    m_d(rhs.m_d)
{
}

MimeType &MimeType::operator=(const MimeType &rhs)
{
    if (this != &rhs)
        m_d = rhs.m_d;
    return *this;
}

MimeType::MimeType(const MimeTypeData &d) :
    m_d(new MimeTypeData(d))
{
}

MimeType::~MimeType()
{
}

void MimeType::clear()
{
    m_d->clear();
}

bool MimeType::isNull() const
{
    return m_d->type.isEmpty();
}

MimeType::operator bool() const
{
    return !isNull();
}

bool MimeType::isTopLevel() const
{
    return m_d->subClassesOf.empty();
}

QString MimeType::type() const
{
    return m_d->type;
}

void MimeType::setType(const QString &type)
{
    m_d->type = type;
}

QString MimeType::comment() const
{
    return m_d->comment;
}

void MimeType::setComment(const QString &comment)
{
    m_d->comment = comment;
}

// Return "en", "de", etc. derived from "en_US", de_DE".
static inline QString systemLanguage()
{
    QString name = QLocale::system().name();
    const int underScorePos = name.indexOf(QLatin1Char('_'));
    if (underScorePos != -1)
        name.truncate(underScorePos);
    return name;
}

QString MimeType::localeComment(const QString &localeArg) const
{
    const QString locale = localeArg.isEmpty() ? systemLanguage() : localeArg;
    const MimeTypeData::LocaleHash::const_iterator it = m_d->localeComments.constFind(locale);
    if (it == m_d->localeComments.constEnd())
        return m_d->comment;
    return it.value();
}

void MimeType::setLocaleComment(const QString &locale, const QString &comment)
{
     m_d->localeComments[locale] = comment;
}

QStringList MimeType::aliases() const
{
    return m_d->aliases;
}

void MimeType::setAliases(const QStringList &a)
{
     m_d->aliases = a;
}

726
QList<MimeGlobPattern> MimeType::globPatterns() const
con's avatar
con committed
727 728 729 730
{
    return m_d->globPatterns;
}

731
void MimeType::setGlobPatterns(const QList<MimeGlobPattern> &g)
con's avatar
con committed
732 733
{
    m_d->globPatterns = g;
734 735 736 737 738 739 740

    QString oldPrefferedSuffix = m_d->preferredSuffix;
    m_d->suffixes.clear();
    m_d->preferredSuffix.clear();
    m_d->assignSuffixes(MimeDatabase::fromGlobPatterns(g));
    if (m_d->preferredSuffix != oldPrefferedSuffix && m_d->suffixes.contains(oldPrefferedSuffix))
        m_d->preferredSuffix = oldPrefferedSuffix;
con's avatar
con committed
741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770
}

QStringList MimeType::subClassesOf() const
{
    return m_d->subClassesOf;
}

void MimeType::setSubClassesOf(const QStringList &s)
{
    m_d->subClassesOf = s;
}

QString MimeType::preferredSuffix() const
{
    return m_d->preferredSuffix;
}

bool MimeType::setPreferredSuffix(const QString &s)
{
    if (!m_d->suffixes.contains(s)) {
        qWarning("%s: Attempt to set preferred suffix to '%s', which is not in the list of suffixes: %s.",
                 m_d->type.toUtf8().constData(),
                 s.toUtf8().constData(),
                 m_d->suffixes.join(QLatin1String(",")).toUtf8().constData());
        return false;
    }
    m_d->preferredSuffix = s;
    return true;
}

771
QString MimeType::formatFilterString(const QString &description, const QList<MimeGlobPattern> &globs)
con's avatar
con committed
772 773 774 775 776 777 778 779 780 781 782 783 784
{
    QString rc;
    if (globs.empty())  // Binary files
        return rc;
    {
        QTextStream str(&rc);
        str << description;
        if (!globs.empty()) {
            str << " (";
            const int size = globs.size();
            for (int i = 0; i < size; i++) {
                if (i)
                    str << ' ';
785
                str << globs.at(i).regExp().pattern();
con's avatar
con committed
786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806
            }
            str << ')';
        }
    }
    return rc;
}

QString MimeType::filterString() const
{
    // @todo: Use localeComment() once creator is shipped with translations
    return formatFilterString(comment(), m_d->globPatterns);
}

bool MimeType::matchesType(const QString &type) const
{
    return m_d->type == type || m_d->aliases.contains(type);
}

unsigned MimeType::matchesFile(const QFileInfo &file) const
{
    Internal::FileMatchContext context(file);
807 808
    const unsigned suffixPriority = matchesFileBySuffix(context);
    if (suffixPriority >= MimeGlobPattern::MaxWeight)
809
        return suffixPriority;
810
    return qMax(suffixPriority, matchesFileByContent(context));
con's avatar
con committed
811 812
}

813
unsigned MimeType::matchesFileBySuffix(Internal::FileMatchContext &c) const
con's avatar
con committed
814 815
{
    // check globs
816 817 818
    foreach (const MimeGlobPattern &gp, m_d->globPatterns) {
        if (gp.regExp().exactMatch(c.fileName()))
            return gp.weight();
819
    }
820 821
    return 0;
}
822

823 824
unsigned MimeType::matchesFileByContent(Internal::FileMatchContext &c) const
{
con's avatar
con committed
825
    // Nope, try magic matchers on context data
826
    if (m_d->magicMatchers.isEmpty())
827 828 829 830
        return 0;

    return matchesData(c.data());
}
831

832 833 834
unsigned MimeType::matchesData(const QByteArray &data) const
{
    unsigned priority = 0;
con's avatar
con committed
835
    if (!data.isEmpty()) {
836
        foreach (const IMagicMatcher::IMagicMatcherSharedPointer &matcher, m_d->magicMatchers) {
837 838 839
            const unsigned magicPriority = matcher->priority();
            if (magicPriority > priority && matcher->matches(data))
                priority = magicPriority;
840
        }
con's avatar
con committed
841
    }
842
    return priority;
con's avatar
con committed
843 844 845 846 847 848 849
}

QStringList MimeType::suffixes() const
{
    return m_d->suffixes;
}

850
void MimeType::addMagicMatcher(const IMagicMatcherSharedPointer &matcher)
con's avatar
con committed
851
{
852
    m_d->magicMatchers.push_back(matcher);
con's avatar
con committed
853 854
}

855
const MimeType::IMagicMatcherList &MimeType::magicMatchers() const
con's avatar
con committed
856
{
857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893
    return m_d->magicMatchers;
}

void MimeType::setMagicMatchers(const IMagicMatcherList &matchers)
{
    m_d->magicMatchers = matchers;
}

namespace {
struct RemovePred : std::unary_function<MimeType::IMagicMatcherSharedPointer, bool>
{
    RemovePred(bool keepRuleBased) : m_keepRuleBase(keepRuleBased) {}
    bool m_keepRuleBase;

    bool operator()(const MimeType::IMagicMatcherSharedPointer &matcher) {
        if ((m_keepRuleBase && !dynamic_cast<MagicRuleMatcher *>(matcher.data()))
                || (!m_keepRuleBase && dynamic_cast<MagicRuleMatcher *>(matcher.data())))
            return true;
        return false;
    }
};
} // Anonymous

MimeType::IMagicMatcherList MimeType::magicRuleMatchers() const
{
    IMagicMatcherList ruleMatchers = m_d->magicMatchers;
    ruleMatchers.erase(std::remove_if(ruleMatchers.begin(), ruleMatchers.end(), RemovePred(true)),
                       ruleMatchers.end());
    return ruleMatchers;
}

void MimeType::setMagicRuleMatchers(const IMagicMatcherList &matchers)
{
    m_d->magicMatchers.erase(std::remove_if(m_d->magicMatchers.begin(), m_d->magicMatchers.end(),
                                            RemovePred(false)),
                             m_d->magicMatchers.end());
    m_d->magicMatchers.append(matchers);
con's avatar
con committed
894 895 896 897 898 899 900 901 902 903 904 905 906 907 908
}

QDebug operator<<(QDebug d, const MimeType &mt)
{
    QString s;
    {
        QTextStream str(&s);
        mt.m_d->debug(str);
    }
    d << s;
    return d;
}

namespace Internal {

909 910 911 912 913 914 915 916 917 918 919
/*!
    \class Core::Internal::BaseMimeTypeParser
    \brief Generic parser for a sequence of <mime-type>.

    Calls abstract handler function process for MimeType it finds.

    \sa Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::MimeTypeParser
*/

hjk's avatar
hjk committed
920 921
class BaseMimeTypeParser
{
922
    Q_DISABLE_COPY(BaseMimeTypeParser)
hjk's avatar
hjk committed
923

con's avatar
con committed
924
public:
925
    BaseMimeTypeParser() {}
con's avatar
con committed
926 927 928 929 930 931 932 933
    virtual ~BaseMimeTypeParser() {}

    bool parse(QIODevice *dev, const QString &fileName, QString *errorMessage);

private:
    // Overwrite to process the sequence of parsed data
    virtual bool process(const MimeType &t, QString *errorMessage) = 0;

934
    void addGlobPattern(const QString &pattern, const QString &weight, MimeTypeData *d) const;
con's avatar
con committed
935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950

    enum ParseStage { ParseBeginning,
                  ParseMimeInfo,
                  ParseMimeType,
                  ParseComment,
                  ParseGlobPattern,
                  ParseSubClass,
                  ParseAlias,
                  ParseMagic,
                  ParseMagicMatchRule,
                  ParseOtherMimeTypeSubTag,
                  ParseError };

    static ParseStage nextStage(ParseStage currentStage, const QStringRef &startElement);
};

951
void BaseMimeTypeParser::addGlobPattern(const QString &pattern, const QString &weight, MimeTypeData *d) const
con's avatar
con committed
952 953 954 955 956 957 958 959 960 961 962 963
{
    if (pattern.isEmpty())
        return;
    // Collect patterns as a QRegExp list and filter out the plain
    // suffix ones for our suffix list. Use first one as preferred
    const QRegExp wildCard(pattern, Qt::CaseSensitive, QRegExp::Wildcard);
    if (!wildCard.isValid()) {
        qWarning("%s: Invalid wildcard '%s'.",
                 Q_FUNC_INFO, pattern.toUtf8().constData());
        return;
    }

964 965 966 967 968
    if (weight.isEmpty())
        d->globPatterns.push_back(MimeGlobPattern(wildCard));
    else
        d->globPatterns.push_back(MimeGlobPattern(wildCard, weight.toInt()));

969
    d->assignSuffix(pattern);
con's avatar
con committed
970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001
}

BaseMimeTypeParser::ParseStage BaseMimeTypeParser::nextStage(ParseStage currentStage, const QStringRef &startElement)
{
    switch (currentStage) {
    case ParseBeginning:
        if (startElement == QLatin1String(mimeInfoTagC))
            return ParseMimeInfo;
        if (startElement == QLatin1String(mimeTypeTagC))
            return ParseMimeType;
        return ParseError;
    case ParseMimeInfo:
        return startElement == QLatin1String(mimeTypeTagC) ? ParseMimeType : ParseError;
    case ParseMimeType:
    case ParseComment:
    case ParseGlobPattern:
    case ParseSubClass:
    case ParseAlias:
    case ParseOtherMimeTypeSubTag:
    case ParseMagicMatchRule:
        if (startElement == QLatin1String(mimeTypeTagC)) // Sequence of <mime-type>
            return ParseMimeType;
        if (startElement == QLatin1String(commentTagC ))
            return ParseComment;
        if (startElement == QLatin1String(globTagC))
            return ParseGlobPattern;
        if (startElement == QLatin1String(subClassTagC))
            return ParseSubClass;
        if (startElement == QLatin1String(aliasTagC))
            return ParseAlias;
        if (startElement == QLatin1String(magicTagC))
            return ParseMagic;
1002 1003
        if (startElement == QLatin1String(matchTagC))
            return ParseMagicMatchRule;
con's avatar
con committed
1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020
        return ParseOtherMimeTypeSubTag;
    case ParseMagic:
        if (startElement == QLatin1String(matchTagC))
            return ParseMagicMatchRule;
        break;
    case ParseError:
        break;
    }
    return ParseError;
}

// Parse int number from an (attribute) string)
static bool parseNumber(const QString &n, int *target, QString *errorMessage)
{
    bool ok;
    *target = n.toInt(&ok);
    if (!ok) {
1021
        *errorMessage = QString::fromLatin1("Not a number '%1'.").arg(n);
con's avatar
con committed
1022 1023 1024 1025 1026 1027 1028 1029 1030
        return false;
    }
    return true;
}

// Evaluate a magic match rule like
//  <match value="must be converted with BinHex" type="string" offset="11"/>
//  <match value="0x9501" type="big16" offset="0:64"/>
static bool addMagicMatchRule(const QXmlStreamAttributes &atts,
1031
                              const MagicRuleMatcherPtr  &ruleMatcher,
con's avatar
con committed
1032 1033 1034
                              QString *errorMessage)
{
    const QString type = atts.value(QLatin1String(matchTypeAttributeC)).toString();
1035 1036
    if (type != QLatin1String(matchStringTypeValueC) &&
        type != QLatin1String(matchByteTypeValueC)) {
con's avatar
con committed
1037 1038 1039 1040 1041
        qWarning("%s: match type %s is not supported.", Q_FUNC_INFO, type.toUtf8().constData());
        return true;
    }
    const QString value = atts.value(QLatin1String(matchValueAttributeC)).toString();
    if (value.isEmpty()) {
1042
        *errorMessage = QString::fromLatin1("Empty match value detected.");
con's avatar
con committed
1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054
        return false;
    }
    // Parse for offset as "1" or "1:10"
    int startPos, endPos;
    const QString offsetS = atts.value(QLatin1String(matchOffsetAttributeC)).toString();
    const int colonIndex = offsetS.indexOf(QLatin1Char(':'));
    const QString startPosS = colonIndex == -1 ? offsetS : offsetS.mid(0, colonIndex);
    const QString endPosS   = colonIndex == -1 ? offsetS : offsetS.mid(colonIndex + 1);
    if (!parseNumber(startPosS, &startPos, errorMessage) || !parseNumber(endPosS, &endPos, errorMessage))
        return false;
    if (debugMimeDB)
        qDebug() << Q_FUNC_INFO << value << startPos << endPos;
1055 1056 1057 1058 1059

    if (type == QLatin1String(matchStringTypeValueC))
        ruleMatcher->add(QSharedPointer<MagicRule>(new MagicStringRule(value, startPos, endPos)));
    else
        ruleMatcher->add(QSharedPointer<MagicRule>(new MagicByteRule(value, startPos, endPos)));
con's avatar
con committed
1060 1061 1062 1063 1064 1065
    return true;
}

bool BaseMimeTypeParser::parse(QIODevice *dev, const QString &fileName, QString *errorMessage)
{
    MimeTypeData data;
1066
    MagicRuleMatcherPtr ruleMatcher;
con's avatar
con committed
1067 1068
    QXmlStreamReader reader(dev);
    ParseStage ps = ParseBeginning;
1069
    QXmlStreamAttributes atts;
con's avatar
con committed
1070 1071 1072 1073
    while (!reader.atEnd()) {
        switch (reader.readNext()) {
        case QXmlStreamReader::StartElement:
            ps = nextStage(ps, reader.name());
1074
            atts = reader.attributes();
con's avatar
con committed
1075 1076
            switch (ps) {
            case ParseMimeType: { // start parsing a type
1077
                const QString type = atts.value(QLatin1String(mimeTypeAttributeC)).toString();
con's avatar
con committed
1078
                if (type.isEmpty()) {
1079
                    reader.raiseError(QString::fromLatin1("Missing 'type'-attribute"));
con's avatar
con committed
1080 1081 1082 1083 1084 1085
                } else {
                    data.type = type;
                }
            }
                break;
            case ParseGlobPattern:
1086 1087
                addGlobPattern(atts.value(QLatin1String(patternAttributeC)).toString(),
                               atts.value(QLatin1String(weightAttributeC)).toString(), &data);
con's avatar
con committed
1088 1089
                break;
            case ParseSubClass: {
1090
                const QString inheritsFrom = atts.value(QLatin1String(mimeTypeAttributeC)).toString();
con's avatar
con committed
1091 1092 1093 1094 1095 1096
                if (!inheritsFrom.isEmpty())
                    data.subClassesOf.push_back(inheritsFrom);
            }
                break;
            case ParseComment: {
                // comments have locale attributes. We want the default, English one
1097
                QString locale = atts.value(QLatin1String(localeAttributeC)).toString();
1098
                const QString comment = QCoreApplication::translate("MimeType", reader.readElementText().toAscii());
con's avatar
con committed
1099 1100 1101 1102 1103 1104 1105 1106
                if (locale.isEmpty()) {
                    data.comment = comment;
                } else {
                    data.localeComments.insert(locale, comment);
                }
            }
                break;
            case ParseAlias: {
1107
                const QString alias = atts.value(QLatin1String(mimeTypeAttributeC)).toString();
con's avatar
con committed
1108 1109 1110 1111 1112 1113
                if (!alias.isEmpty())
                    data.aliases.push_back(alias);
            }
                break;
            case ParseMagic: {
                int priority = 0;
1114
                const QString priorityS = atts.value(QLatin1String(priorityAttributeC)).toString();
con's avatar
con committed
1115 1116 1117 1118 1119
                if (!priorityS.isEmpty()) {
                    if (!parseNumber(priorityS, &priority, errorMessage))
                        return false;

                }
1120
                ruleMatcher = MagicRuleMatcherPtr(new MagicRuleMatcher);
con's avatar
con committed
1121 1122 1123 1124
                ruleMatcher->setPriority(priority);
            }
                break;
            case ParseMagicMatchRule:
1125
                QTC_ASSERT(!ruleMatcher.isNull(), return false);
1126
                if (!addMagicMatchRule(atts, ruleMatcher, errorMessage))
con's avatar
con committed
1127 1128 1129
                    return false;
                break;
            case ParseError:
1130
                reader.raiseError(QString::fromLatin1("Unexpected element <%1>").arg(reader.name().toString()));
con's avatar
con committed
1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143
                break;
            default:
                break;
            } // switch nextStage
            break;
        // continue switch QXmlStreamReader::Token...
        case QXmlStreamReader::EndElement: // Finished element
            if (reader.name() == QLatin1String(mimeTypeTagC)) {
                if (!process(MimeType(data), errorMessage))
                    return false;
                data.clear();
            } else {
                // Finished a match sequence
1144
                if (reader.name() == QLatin1String(magicTagC)) {
1145
                    QTC_ASSERT(!ruleMatcher.isNull(), return false);
1146 1147
                    data.magicMatchers.push_back(ruleMatcher);
                    ruleMatcher = MagicRuleMatcherPtr();
con's avatar
con committed
1148 1149 1150 1151 1152 1153 1154 1155 1156 1157
                }
            }
            break;

        default:
            break;
        } // switch reader.readNext()
    }

    if (reader.hasError()) {
1158
        *errorMessage = QString::fromLatin1("An error has been encountered at line %1 of %2: %3:").arg(reader.lineNumber()).arg(fileName, reader.errorString());
con's avatar
con committed
1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169
        return false;
    }
    return true;
}

} // namespace Internal

// MimeMapEntry: Entry of a type map, consisting of type and level.

enum { Dangling = 32767 };

hjk's avatar
hjk committed
1170 1171
struct MimeMapEntry
{
con's avatar
con committed
1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182
    explicit MimeMapEntry(const MimeType &t = MimeType(), int aLevel = Dangling);
    MimeType type;
    int level; // hierachy level
};

MimeMapEntry::MimeMapEntry(const MimeType &t, int aLevel) :
    type(t),
    level(aLevel)
{
}

1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226
/*!
    \class Core::MimeDatabase
    \brief Mime data base to which the plugins can add the mime types they handle.

    The class is protected by a QMutex and can therefore be accessed by threads.

    A good testcase is to run it over \c '/usr/share/mime/<*>/<*>.xml' on Linux.

    When adding a "text/plain" to it, the mimetype will receive a magic matcher
    that checks for text files that do not match the globs by heuristics.

    \section1 Design Considerations

    Storage requirements:
    \list
    \o Must be robust in case of incomplete hierarchies, dangling entries
    \o Plugins will not load and register their mime types in order of inheritance.
    \o Multiple inheritance (several subClassesOf) can occur
    \o Provide quick lookup by name
    \o Provide quick lookup by file type.
    \endlist

    This basically rules out some pointer-based tree, so the structure chosen is:
    \list
    \o An alias map QString->QString for mapping aliases to types
    \o A Map QString->MimeMapEntry for the types (MimeMapEntry being a pair of
       MimeType and (hierarchy) level.
    \o A map  QString->QString representing parent->child relations (enabling
       recursing over children)
    \o Using strings avoids dangling pointers.
    \endlist

    The hierarchy level is used for mapping by file types. When findByFile()
    is first called after addMimeType() it recurses over the hierarchy and sets
    the hierarchy level of the entries accordingly (0 toplevel, 1 first
    order...). It then does several passes over the type map, checking the
    globs for maxLevel, maxLevel-1....until it finds a match (idea being to
    to check the most specific types first). Starting a recursion from the
    leaves is not suitable since it will hit parent nodes several times.

    \sa Core::MimeType, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::BaseMimeTypeParser, Core::Internal::MimeTypeParser
*/
con's avatar
con committed
1227

hjk's avatar
hjk committed
1228 1229
class MimeDatabasePrivate
{
con's avatar
con committed
1230
    Q_DISABLE_COPY(MimeDatabasePrivate)
hjk's avatar
hjk committed
1231

con's avatar
con committed
1232 1233 1234 1235 1236 1237 1238 1239 1240
public:
    MimeDatabasePrivate();

    bool addMimeTypes(const QString &fileName, QString *errorMessage);
    bool addMimeTypes(QIODevice *device, QString *errorMessage);
    bool addMimeType(MimeType mt);

    MimeType findByType(const QString &type) const;
    MimeType findByFile(const QFileInfo &f) const;
1241
    MimeType findByData(const QByteArray &data) const;
con's avatar
con committed
1242

1243
    QStringList filterStrings() const;
con's avatar
con committed
1244 1245

    QStringList suffixes() const;
1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264
    bool setPreferredSuffix(const QString &typeOrAlias, const QString &suffix);

    QList<MimeGlobPattern> globPatterns() const;
    void setGlobPatterns(const QString &typeOrAlias, const QList<MimeGlobPattern> &globPatterns);

    QList<QSharedPointer<IMagicMatcher> > magicMatchers() const;
    void setMagicMatchers(const QString &typeOrAlias,
                          const QList<QSharedPointer<IMagicMatcher> > &matchers);

    QList<MimeType> mimeTypes() const;

    void syncUserModifiedMimeTypes();
    static QList<MimeType> readUserModifiedMimeTypes();
    static void writeUserModifiedMimeTypes(const QList<MimeType> &mimeTypes);
    void clearUserModifiedMimeTypes();

    static QList<MimeGlobPattern> toGlobPatterns(const QStringList &patterns,
                                                 int weight = MimeGlobPattern::MaxWeight);
    static QStringList fromGlobPatterns(const QList<MimeGlobPattern> &globPatterns);
con's avatar
con committed
1265 1266 1267 1268 1269 1270 1271 1272

    void debug(QTextStream &str) const;

private:
    typedef QHash<QString, MimeMapEntry> TypeMimeTypeMap;
    typedef QHash<QString, QString> AliasMap;
    typedef QMultiHash<QString, QString> ParentChildrenMap;

1273 1274 1275 1276 1277
    static const QChar kSemiColon;
    static const QString kModifiedMimeTypesFile;
    static QString kModifiedMimeTypesPath;


con's avatar
con committed
1278 1279 1280
    bool addMimeTypes(QIODevice *device, const QString &fileName, QString *errorMessage);
    inline const QString &resolveAlias(const QString &name) const;
    MimeType findByFile(const QFileInfo &f, unsigned *priority) const;
1281
    MimeType findByData(const QByteArray &data, unsigned *priority) const;
con's avatar
con committed
1282 1283 1284 1285 1286 1287 1288 1289 1290
    void determineLevels();
    void raiseLevelRecursion(MimeMapEntry &e, int level);

    TypeMimeTypeMap m_typeMimeTypeMap;
    AliasMap m_aliasMap;
    ParentChildrenMap m_parentChildrenMap;
    int m_maxLevel;
};

1291 1292 1293 1294
const QChar MimeDatabasePrivate::kSemiColon(QLatin1Char(';'));
const QString MimeDatabasePrivate::kModifiedMimeTypesFile(QLatin1String("modifiedmimetypes.xml"));
QString MimeDatabasePrivate::kModifiedMimeTypesPath;

con's avatar
con committed
1295 1296 1297
MimeDatabasePrivate::MimeDatabasePrivate() :
    m_maxLevel(-1)
{
1298
    // Assign here to avoid non-local static data initialization issues.
hjk's avatar
hjk committed
1299
    kModifiedMimeTypesPath = ICore::userResourcePath() + QLatin1String("/mimetypes/");
con's avatar
con committed
1300 1301
}

1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312
/*!
    \class Core::Internal::MimeTypeParser
    \brief Mime type parser

    Populates Core::MimeDataBase

    \sa Core::MimeDatabase, Core::IMagicMatcher, Core::MagicRuleMatcher, Core::MagicRule, Core::MagicStringRule, Core::MagicByteRule, Core::GlobPattern
    \sa Core::Internal::FileMatchContext, Core::Internal::BinaryMatcher, Core::Internal::HeuristicTextMagicMatcher
    \sa Core::Internal::MimeTypeParser
*/

con's avatar
con committed
1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334
namespace Internal {
    // Parser that builds MimeDB hierarchy by adding to MimeDatabasePrivate
    class MimeTypeParser : public BaseMimeTypeParser {
    public:
        explicit MimeTypeParser(MimeDatabasePrivate &db) : m_db(db) {}
    private:
        virtual bool process(const MimeType &t, QString *) { m_db.addMimeType(t); return true; }

        MimeDatabasePrivate &m_db;
    };
} // namespace Internal

bool MimeDatabasePrivate::addMimeTypes(QIODevice *device, const QString &fileName, QString *errorMessage)
{
    Internal::MimeTypeParser parser(*this);
    return parser.parse(device, fileName, errorMessage);
}

bool MimeDatabasePrivate::addMimeTypes(const QString &fileName, QString *errorMessage)
{
    QFile file(fileName);
    if (!file.open(QIODevice::ReadOnly|QIODevice::Text)) {
1335
        *errorMessage = QString::fromLatin1("Cannot open %1: %2").arg(fileName, file.errorString());
con's avatar
con committed
1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357
        return false;
    }
    return addMimeTypes(&file, fileName, errorMessage);
}

bool MimeDatabasePrivate::addMimeTypes(QIODevice *device, QString *errorMessage)
{
    return addMimeTypes(device, QLatin1String("<stream>"), errorMessage);
}

bool MimeDatabasePrivate::addMimeType(MimeType mt)
{
    if (!mt)
        return false;

    const QString type = mt.type();
    // Hack: Add a magic text matcher to "text/plain" and the fallback matcher to
    // binary types "application/octet-stream"
    if (type == QLatin1String(textTypeC)) {
        mt.addMagicMatcher(QSharedPointer<IMagicMatcher>(new Internal::HeuristicTextMagicMatcher));
    } else {
        if (type == QLatin1String(binaryTypeC))
1358
            mt.addMagicMatcher(QSharedPointer<IMagicMatcher>(new Internal::BinaryMatcher));
con's avatar
con committed
1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467
    }
    // insert the type.
    m_typeMimeTypeMap.insert(type, MimeMapEntry(mt));
    // Register the children, resolved via alias map. Note that it is still
    // possible that aliases end up in the map if the parent classes are not inserted
    // at this point (thus their aliases not known).
    const QStringList subClassesOf = mt.subClassesOf();
    if (!subClassesOf.empty()) {
        const QStringList::const_iterator socend = subClassesOf.constEnd();
        for (QStringList::const_iterator soit = subClassesOf.constBegin(); soit !=  socend; ++soit)
            m_parentChildrenMap.insert(resolveAlias(*soit), type);
    }
    // register aliasses
    const QStringList aliases = mt.aliases();
    if (!aliases.empty()) {
        const QStringList::const_iterator cend = aliases.constEnd();
        for (QStringList::const_iterator it = aliases.constBegin(); it !=  cend; ++it)
            m_aliasMap.insert(*it, type);
    }
    m_maxLevel = -1; // Mark as dirty
    return true;
}

const QString &MimeDatabasePrivate::resolveAlias(const  QString &name) const
{
    const AliasMap::const_iterator aliasIt = m_aliasMap.constFind(name);
    return aliasIt == m_aliasMap.constEnd() ? name : aliasIt.value();
}

void MimeDatabasePrivate::raiseLevelRecursion(MimeMapEntry &e, int level)
{
    if (e.level == Dangling || e.level < level)
        e.level = level;
    if (m_maxLevel < level)
        m_maxLevel = level;
    // At all events recurse over children since nodes might have been
    // added.
    const QStringList childTypes = m_parentChildrenMap.values(e.type.type());
    if (childTypes.empty())
        return;
    // look them up in the type->mime type map
    const int nextLevel = level + 1;
    const TypeMimeTypeMap::iterator tm_end = m_typeMimeTypeMap.end();
    const QStringList::const_iterator cend = childTypes.constEnd();
    for (QStringList::const_iterator it = childTypes.constBegin(); it !=  cend; ++it) {
        const TypeMimeTypeMap::iterator tm_it = m_typeMimeTypeMap.find(resolveAlias(*it));
        if (tm_it == tm_end) {
            qWarning("%s: Inconsistent mime hierarchy detected, child %s of %s cannot be found.",
                     Q_FUNC_INFO, it->toUtf8().constData(), e.type.type().toUtf8().constData());
        } else {
            raiseLevelRecursion(*tm_it, nextLevel);
        }
    }
}

void MimeDatabasePrivate::determineLevels()
{
    // Loop over toplevels and recurse down their hierarchies.
    // Determine top levels by subtracting the children from the parent
    // set. Note that a toplevel at this point might have 'subclassesOf'
    // set to some class that is not in the DB, so, checking for an empty
    // 'subclassesOf' set is not sufficient to find the toplevels.
    // First, take the parent->child entries  whose parent exists and build
    // sets of parents/children
    QSet<QString> parentSet, childrenSet;
    const ParentChildrenMap::const_iterator pcend = m_parentChildrenMap.constEnd();
    for (ParentChildrenMap::const_iterator it =  m_parentChildrenMap.constBegin(); it !=  pcend; ++it)
        if (m_typeMimeTypeMap.contains(it.key())) {
            parentSet.insert(it.key());
            childrenSet.insert(it.value());
        }
    const QSet<QString> topLevels = parentSet.subtract(childrenSet);
    if (debugMimeDB)
        qDebug() << Q_FUNC_INFO << "top levels" << topLevels;
    const TypeMimeTypeMap::iterator tm_end = m_typeMimeTypeMap.end();
    const QSet<QString>::const_iterator tl_cend = topLevels.constEnd();
    for (QSet<QString>::const_iterator tl_it =  topLevels.constBegin(); tl_it !=  tl_cend; ++tl_it) {
        const TypeMimeTypeMap::iterator tm_it = m_typeMimeTypeMap.find(resolveAlias(*tl_it));
        if (tm_it == tm_end) {
            qWarning("%s: Inconsistent mime hierarchy detected, top level element %s cannot be found.",
                     Q_FUNC_INFO, tl_it->toUtf8().constData());
        } else {
            raiseLevelRecursion(tm_it.value(), 0);
        }
    }
}

bool MimeDatabasePrivate::setPreferredSuffix(const QString &typeOrAlias, const QString &suffix)
{
    TypeMimeTypeMap::iterator tit =  m_typeMimeTypeMap.find(resolveAlias(typeOrAlias));
    if (tit != m_typeMimeTypeMap.end())
        return tit.value().type.setPreferredSuffix(suffix);
    return false;
}

// Returns a mime type or Null one if none found
MimeType MimeDatabasePrivate::findByType(const QString &typeOrAlias) const
{
    const TypeMimeTypeMap::const_iterator tit =  m_typeMimeTypeMap.constFind(resolveAlias(typeOrAlias));
    if (tit != m_typeMimeTypeMap.constEnd())
        return tit.value().type;
    return MimeType();
}

// Debugging wrapper around findByFile()
MimeType MimeDatabasePrivate::findByFile(const QFileInfo &f) const
{
    unsigned priority = 0;
    if (debugMimeDB)
1468
        qDebug() << '>' << Q_FUNC_INFO << f.absoluteFilePath();
con's avatar
con committed
1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487
    const MimeType rc = findByFile(f, &priority);
    if (debugMimeDB) {
        if (rc) {
            qDebug() << "<MimeDatabase::findByFile: match prio=" << priority << rc.type();
        } else {
            qDebug() << "<MimeDatabase::findByFile: no match";
        }
    }
    return rc;
}

// Returns a mime type or Null one if none found
MimeType MimeDatabasePrivate::findByFile(const QFileInfo &f, unsigned *priorityPtr) const
{
    // Is the hierarchy set up in case we find several matches?
    if (m_maxLevel < 0) {
        MimeDatabasePrivate *db = const_cast<MimeDatabasePrivate *>(this);
        db->determineLevels();
    }
1488 1489 1490 1491 1492 1493

    // First, glob patterns are evaluated. If there is a match with max weight,
    // this one is selected and we are done. Otherwise, the file contents are
    // evaluated and the match with the highest value (either a magic priority or
    // a glob pattern weight) is selected. Matching starts from max level (most
    // specific) in both cases, even when there is already a suffix matching candidate.