mimedatabase.cpp 67.4 KB
Newer Older
hjk's avatar
hjk committed
1
/****************************************************************************
con's avatar
con committed
2
**
3
** Copyright (C) 2013 Digia Plc and/or its subsidiary(-ies).
hjk's avatar
hjk committed
4
** Contact: http://www.qt-project.org/legal
con's avatar
con committed
5
**
hjk's avatar
hjk committed
6
** This file is part of Qt Creator.
con's avatar
con committed
7
**
hjk's avatar
hjk committed
8 9 10 11 12 13 14
** Commercial License Usage
** Licensees holding valid commercial Qt licenses may use this file in
** accordance with the commercial license agreement provided with the
** Software or, alternatively, in accordance with the terms contained in
** a written agreement between you and Digia.  For licensing terms and
** conditions see http://qt.digia.com/licensing.  For further information
** use the contact form at http://qt.digia.com/contact-us.
15
**
16
** GNU Lesser General Public License Usage
hjk's avatar
hjk committed
17 18 19 20 21 22 23 24 25
** Alternatively, 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.
**
** In addition, as a special exception, Digia gives you certain additional
** rights.  These rights are described in the Digia Qt LGPL Exception
con's avatar
con committed
26 27
** version 1.1, included in the file LGPL_EXCEPTION.txt in this package.
**
hjk's avatar
hjk committed
28
****************************************************************************/
hjk's avatar
hjk committed
29

con's avatar
con committed
30
#include "mimedatabase.h"
31
#include "coreconstants.h"
32
#include "icore.h"
con's avatar
con committed
33

hjk's avatar
hjk committed
34 35
#include <utils/qtcassert.h>

36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
#include <QByteArray>
#include <QCoreApplication>
#include <QDebug>
#include <QFile>
#include <QDir>
#include <QFileInfo>
#include <QLocale>
#include <QHash>
#include <QMultiHash>
#include <QRegExp>
#include <QSharedData>
#include <QSharedPointer>
#include <QStringList>
#include <QTextStream>
#include <QXmlStreamReader>
#include <QXmlStreamWriter>
con's avatar
con committed
52

Christian Kandeler's avatar
Christian Kandeler committed
53
#include <algorithm>
54
#include <functional>
Christian Kandeler's avatar
Christian Kandeler committed
55

con's avatar
con committed
56 57 58
enum { debugMimeDB = 0 };

// XML tags in mime files
59 60 61 62 63 64 65 66 67 68
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
69

70 71 72 73 74 75 76 77
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
78 79

// Types
80 81
static const char textTypeC[] = "text/plain";
static const char binaryTypeC[] = "application/octet-stream";
con's avatar
con committed
82 83 84 85 86 87

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

// Fallback priorities, must be low.
88 89 90 91
enum {
    BinaryMatchPriority = Core::MimeGlobPattern::MinWeight + 1,
    TextMatchPriority
};
con's avatar
con committed
92

93 94 95 96 97 98 99 100
/*!
    \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
101 102 103
*/

namespace Core {
104 105 106

typedef QSharedPointer<MagicRuleMatcher> MagicRuleMatcherPtr;

con's avatar
con committed
107 108
namespace Internal {

109 110 111 112 113 114 115 116 117 118 119 120
/*!
    \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
121

hjk's avatar
hjk committed
122 123
class FileMatchContext
{
124
    Q_DISABLE_COPY(FileMatchContext)
hjk's avatar
hjk committed
125

con's avatar
con committed
126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173
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;
}

174 175 176 177 178 179 180 181 182
/*!
    \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
183 184
class BinaryMatcher : public IMagicMatcher
{
con's avatar
con committed
185 186 187 188 189 190
public:
    BinaryMatcher() {}
    virtual bool matches(const QByteArray & /*data*/) const { return true; }
    virtual int priority() const  { return BinaryMatchPriority; }
};

191 192 193 194 195 196 197 198 199 200 201 202
/*!
    \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
203 204
class HeuristicTextMagicMatcher : public IMagicMatcher
{
con's avatar
con committed
205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235
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

236 237 238 239 240 241 242 243 244 245 246 247 248 249
/*!
    \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
 */

250 251
const QChar MagicRule::kColon(QLatin1Char(':'));

252
MagicRule::MagicRule(int startPos, int endPos) : m_startPos(startPos), m_endPos(endPos)
con's avatar
con committed
253 254 255
{
}

256 257 258 259 260 261 262 263 264 265 266 267 268 269
MagicRule::~MagicRule()
{
}

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

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

270 271 272 273 274 275 276 277 278 279 280 281
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());
}

282 283 284 285 286 287 288 289 290
/*!
    \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
 */

291
const QString MagicStringRule::kMatchType(QLatin1String("string"));
292

293 294 295 296 297 298 299 300 301
MagicStringRule::MagicStringRule(const QString &s, int startPos, int endPos) :
    MagicRule(startPos, endPos), m_pattern(s.toUtf8())
{
}

MagicStringRule::~MagicStringRule()
{
}

302 303 304 305 306 307 308
QString MagicStringRule::matchType() const
{
    return kMatchType;
}

QString MagicStringRule::matchValue() const
{
309
    return QLatin1String(m_pattern);
310 311
}

312
bool MagicStringRule::matches(const QByteArray &data) const
con's avatar
con committed
313 314
{
    // Quick check
315
    if ((startPos() + m_pattern.size()) > data.size())
con's avatar
con committed
316 317
        return false;
    // Most common: some string at position 0:
318
    if (startPos() == 0 && startPos() == endPos())
con's avatar
con committed
319 320
        return data.startsWith(m_pattern);
    // Range
321 322
    const int index = data.indexOf(m_pattern, startPos());
    const bool rc = index != -1 && index <= endPos();
323
    if (debugMimeDB)
324
        qDebug() << "Checking " << m_pattern << startPos() << endPos() << " returns " << rc;
325
    return rc;
con's avatar
con committed
326 327
}

328 329 330 331 332 333 334 335 336 337 338 339 340 341
/*!
    \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
 */

342 343
const QString MagicByteRule::kMatchType(QLatin1String("byte"));

344
MagicByteRule::MagicByteRule(const QString &s, int startPos, int endPos) :
345 346 347 348 349 350 351 352 353 354 355 356 357
    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
358
{
359
    // Expect an hex format value like this: \0x7f\0x45\0x4c\0x46
360 361
    const QStringList &byteSequence = sequence.split(QLatin1Char('\\'), QString::SkipEmptyParts);
    foreach (const QString &byte, byteSequence) {
362 363 364
        bool ok;
        const int hex = byte.toInt(&ok, 16);
        if (ok) {
365 366
            if (bytes)
                bytes->push_back(hex);
367
        } else {
368
            return false;
369 370
        }
    }
371
    return true;
372 373
}

374
QString MagicByteRule::matchType() const
375
{
376 377 378 379 380 381 382 383 384
    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;
385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407
}

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
408 409
}

410 411 412 413 414 415 416 417 418 419 420 421
/*!
    \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
422 423 424 425 426 427 428
MagicRuleMatcher::MagicRuleMatcher() :
     m_priority(65535)
{
}

void MagicRuleMatcher::add(const MagicRuleSharedPointer &rule)
{
429 430 431 432 433 434 435 436 437 438 439
    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
440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460
}

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;
}

461 462 463 464 465 466 467 468 469 470 471 472 473 474
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;
}

475 476 477 478 479 480 481 482 483
/*!
    \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
484
MimeGlobPattern::MimeGlobPattern(const QRegExp &regExp, unsigned weight) :
485 486 487 488 489 490 491 492 493 494 495 496 497
    m_regExp(regExp), m_weight(weight)
{
}

MimeGlobPattern::~MimeGlobPattern()
{
}

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

Friedemann Kleint's avatar
Friedemann Kleint committed
498
unsigned MimeGlobPattern::weight() const
499 500 501 502
{
    return m_weight;
}

503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534
/*!
    \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
535 536 537
class MimeTypeData : public QSharedData {
public:
    typedef QHash<QString,QString> LocaleHash;
538 539 540

    MimeTypeData();

con's avatar
con committed
541
    void clear();
542 543
    void assignSuffix(const QString &pattern);
    void assignSuffixes(const QStringList &patterns);
con's avatar
con committed
544 545
    void debug(QTextStream &str, int indent = 0) const;

546
    QRegExp suffixPattern;
547

con's avatar
con committed
548 549 550 551 552
    QString type;
    QString comment;

    LocaleHash localeComments;
    QStringList aliases;
553
    QList<MimeGlobPattern> globPatterns;
con's avatar
con committed
554 555 556
    QStringList subClassesOf;
    QString preferredSuffix;
    QStringList suffixes;
557
    IMagicMatcher::IMagicMatcherList magicMatchers;
con's avatar
con committed
558 559
};

560 561 562 563 564
MimeTypeData::MimeTypeData()
    // RE to match a suffix glob pattern: "*.ext" (and not sth like "Makefile" or
    // "*.log[1-9]"
    : suffixPattern(QLatin1String("^\\*\\.[\\w+]+$"))
{
565
    QTC_CHECK(suffixPattern.isValid());
566 567
}

con's avatar
con committed
568 569 570 571 572 573 574 575 576 577 578 579
void MimeTypeData::clear()
{
    type.clear();
    comment.clear();
    aliases.clear();
    globPatterns.clear();
    subClassesOf.clear();
    preferredSuffix.clear();
    suffixes.clear();
    magicMatchers.clear();
}

580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595
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
596 597 598 599 600 601 602 603 604 605 606 607 608
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: ";
609
        foreach (const MimeGlobPattern &gp, globPatterns)
Friedemann Kleint's avatar
Friedemann Kleint committed
610
            str << gp.regExp().pattern() << '(' << gp.weight() << ')';
con's avatar
con committed
611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720
        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;
}

721
QList<MimeGlobPattern> MimeType::globPatterns() const
con's avatar
con committed
722 723 724 725
{
    return m_d->globPatterns;
}

726
void MimeType::setGlobPatterns(const QList<MimeGlobPattern> &g)
con's avatar
con committed
727 728
{
    m_d->globPatterns = g;
729 730 731 732 733 734 735

    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
736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765
}

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;
}

766
QString MimeType::formatFilterString(const QString &description, const QList<MimeGlobPattern> &globs)
con's avatar
con committed
767 768 769 770 771 772 773 774 775 776 777 778 779
{
    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 << ' ';
780
                str << globs.at(i).regExp().pattern();
con's avatar
con committed
781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801
            }
            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);
802 803
    const unsigned suffixPriority = matchesFileBySuffix(context);
    if (suffixPriority >= MimeGlobPattern::MaxWeight)
804
        return suffixPriority;
805
    return qMax(suffixPriority, matchesFileByContent(context));
con's avatar
con committed
806 807
}

808
unsigned MimeType::matchesFileBySuffix(Internal::FileMatchContext &c) const
con's avatar
con committed
809 810
{
    // check globs
811
    foreach (const MimeGlobPattern &gp, m_d->globPatterns) {
812 813
        QRegExp regExp = gp.regExp();
        if (regExp.exactMatch(c.fileName()))
814
            return gp.weight();
815
    }
816 817
    return 0;
}
818

819 820
unsigned MimeType::matchesFileByContent(Internal::FileMatchContext &c) const
{
con's avatar
con committed
821
    // Nope, try magic matchers on context data
822
    if (m_d->magicMatchers.isEmpty())
823 824 825 826
        return 0;

    return matchesData(c.data());
}
827

828 829 830
unsigned MimeType::matchesData(const QByteArray &data) const
{
    unsigned priority = 0;
con's avatar
con committed
831
    if (!data.isEmpty()) {
832
        foreach (const IMagicMatcher::IMagicMatcherSharedPointer &matcher, m_d->magicMatchers) {
833 834 835
            const unsigned magicPriority = matcher->priority();
            if (magicPriority > priority && matcher->matches(data))
                priority = magicPriority;
836
        }
con's avatar
con committed
837
    }
838
    return priority;
con's avatar
con committed
839 840 841 842 843 844 845
}

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

846
void MimeType::addMagicMatcher(const IMagicMatcherSharedPointer &matcher)
con's avatar
con committed
847
{
848
    m_d->magicMatchers.push_back(matcher);
con's avatar
con committed
849 850
}

851
const MimeType::IMagicMatcherList &MimeType::magicMatchers() const
con's avatar
con committed
852
{
853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889
    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
890 891 892 893 894 895 896 897 898 899 900 901 902 903 904
}

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

namespace Internal {

905 906 907 908 909 910 911 912 913 914 915
/*!
    \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
916 917
class BaseMimeTypeParser
{
918
    Q_DISABLE_COPY(BaseMimeTypeParser)
hjk's avatar
hjk committed
919

con's avatar
con committed
920
public:
921
    BaseMimeTypeParser() {}
con's avatar
con committed
922 923 924 925 926 927 928 929
    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;

930
    void addGlobPattern(const QString &pattern, const QString &weight, MimeTypeData *d) const;
con's avatar
con committed
931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946

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

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

947
void BaseMimeTypeParser::addGlobPattern(const QString &pattern, const QString &weight, MimeTypeData *d) const
con's avatar
con committed
948 949 950 951 952 953 954 955 956 957 958 959
{
    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;
    }

960 961 962 963 964
    if (weight.isEmpty())
        d->globPatterns.push_back(MimeGlobPattern(wildCard));
    else
        d->globPatterns.push_back(MimeGlobPattern(wildCard, weight.toInt()));

965
    d->assignSuffix(pattern);
con's avatar
con committed
966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997
}

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;
998 999
        if (startElement == QLatin1String(matchTagC))
            return ParseMagicMatchRule;
con's avatar
con committed
1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016
        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) {
1017
        *errorMessage = QString::fromLatin1("Not a number '%1'.").arg(n);
con's avatar
con committed
1018 1019 1020 1021 1022 1023 1024 1025 1026
        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,
1027
                              const MagicRuleMatcherPtr  &ruleMatcher,
con's avatar
con committed
1028 1029 1030
                              QString *errorMessage)
{
    const QString type = atts.value(QLatin1String(matchTypeAttributeC)).toString();
1031 1032
    if (type != QLatin1String(matchStringTypeValueC) &&
        type != QLatin1String(matchByteTypeValueC)) {
con's avatar
con committed
1033 1034 1035 1036 1037
        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()) {
1038
        *errorMessage = QString::fromLatin1("Empty match value detected.");
con's avatar
con committed
1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050
        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;
1051 1052 1053 1054 1055

    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
1056 1057 1058 1059 1060 1061
    return true;
}

bool BaseMimeTypeParser::parse(QIODevice *dev, const QString &fileName, QString *errorMessage)
{
    MimeTypeData data;
1062
    MagicRuleMatcherPtr ruleMatcher;
con's avatar
con committed
1063 1064
    QXmlStreamReader reader(dev);
    ParseStage ps = ParseBeginning;
1065
    QXmlStreamAttributes atts;
con's avatar
con committed
1066 1067 1068 1069
    while (!reader.atEnd()) {
        switch (reader.readNext()) {
        case QXmlStreamReader::StartElement:
            ps = nextStage(ps, reader.name());
1070
            atts = reader.attributes();
con's avatar
con committed
1071 1072
            switch (ps) {
            case ParseMimeType: { // start parsing a type
1073
                const QString type = atts.value(QLatin1String(mimeTypeAttributeC)).toString();
1074
                if (type.isEmpty())
1075
                    reader.raiseError(QString::fromLatin1("Missing 'type'-attribute"));
1076
                else
con's avatar
con committed
1077 1078 1079 1080
                    data.type = type;
            }
                break;
            case ParseGlobPattern:
1081 1082
                addGlobPattern(atts.value(QLatin1String(patternAttributeC)).toString(),
                               atts.value(QLatin1String(weightAttributeC)).toString(), &data);
con's avatar
con committed
1083 1084
                break;
            case ParseSubClass: {
1085
                const QString inheritsFrom = atts.value(QLatin1String(mimeTypeAttributeC)).toString();
con's avatar
con committed
1086 1087 1088 1089 1090 1091
                if (!inheritsFrom.isEmpty())
                    data.subClassesOf.push_back(inheritsFrom);
            }
                break;
            case ParseComment: {
                // comments have locale attributes. We want the default, English one
1092
                QString locale = atts.value(QLatin1String(localeAttributeC)).toString();
1093
                const QString comment = QCoreApplication::translate("MimeType", reader.readElementText().toLatin1());
1094
                if (locale.isEmpty())
con's avatar
con committed
1095
                    data.comment = comment;
1096
                else
con's avatar
con committed
1097 1098 1099 1100
                    data.localeComments.insert(locale, comment);
            }
                break;
            case ParseAlias: {
1101
                const QString alias = atts.value(QLatin1String(mimeTypeAttributeC)).toString();
con's avatar
con committed
1102 1103 1104 1105 1106 1107
                if (!alias.isEmpty())
                    data.aliases.push_back(alias);
            }
                break;
            case ParseMagic: {
                int priority = 0;
1108
                const QString priorityS = atts.value(QLatin1String(priorityAttributeC)).toString();
con's avatar
con committed
1109 1110 1111 1112 1113
                if (!priorityS.isEmpty()) {
                    if (!parseNumber(priorityS, &priority, errorMessage))
                        return false;

                }
1114
                ruleMatcher = MagicRuleMatcherPtr(new MagicRuleMatcher);
con's avatar
con committed
1115 1116 1117 1118
                ruleMatcher->setPriority(priority);
            }
                break;
            case ParseMagicMatchRule:
1119
                QTC_ASSERT(!ruleMatcher.isNull(), return false);
1120
                if (!addMagicMatchRule(atts, ruleMatcher, errorMessage))
con's avatar
con committed
1121 1122 1123
                    return false;
                break;
            case ParseError:
1124
                reader.raiseError(QString::fromLatin1("Unexpected element <%1>").arg(reader.name().toString()));
con's avatar
con committed
1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137
                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
1138
                if (reader.name() == QLatin1String(magicTagC)) {
1139
                    QTC_ASSERT(!ruleMatcher.isNull(), return false);
1140 1141
                    data.magicMatchers.push_back(ruleMatcher);
                    ruleMatcher = MagicRuleMatcherPtr();
con's avatar
con committed
1142 1143 1144 1145 1146 1147 1148 1149 1150 1151
                }
            }
            break;

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

    if (reader.hasError()) {
1152
        *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
1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163
        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
1164 1165
struct MimeMapEntry
{
con's avatar
con committed
1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176
    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)
{
}

1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191
/*!
    \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
1192 1193 1194 1195 1196
    \li Must be robust in case of incomplete hierarchies, dangling entries
    \li Plugins will not load and register their mime types in order of inheritance.
    \li Multiple inheritance (several subClassesOf) can occur
    \li Provide quick lookup by name
    \li Provide quick lookup by file type.
1197 1198 1199 1200
    \endlist

    This basically rules out some pointer-based tree, so the structure chosen is:
    \list
1201 1202
    \li An alias map QString->QString for mapping aliases to types
    \li A Map QString->MimeMapEntry for the types (MimeMapEntry being a pair of
1203
       MimeType and (hierarchy) level.
1204
    \li A map  QString->QString representing parent->child relations (enabling
1205
       recursing over children)
1206
    \li Using strings avoids dangling pointers.
1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220
    \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
1221

hjk's avatar
hjk committed
1222 1223
class MimeDatabasePrivate
{
con's avatar
con committed
1224
    Q_DISABLE_COPY(MimeDatabasePrivate)
hjk's avatar
hjk committed
1225

con's avatar
con committed
1226 1227 1228 1229 1230 1231 1232 1233 1234
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;
1235
    MimeType findByData(const QByteArray &data) const;
con's avatar
con committed
1236

1237
    QStringList filterStrings() const;
con's avatar
con committed
1238 1239

    QStringList suffixes() const;
1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258
    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
1259 1260 1261 1262 1263 1264 1265 1266

    void debug(QTextStream &str) const;

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

1267 1268 1269 1270 1271
    static const QChar kSemiColon;
    static const QString kModifiedMimeTypesFile;
    static QString kModifiedMimeTypesPath;


con's avatar
con committed
1272 1273 1274
    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;
1275
    MimeType findByData(const QByteArray &data, unsigned *priority) const;
con's avatar
con committed
1276 1277 1278 1279 1280 1281 1282 1283 1284
    void determineLevels();
    void raiseLevelRecursion(MimeMapEntry &e, int level);

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

1285 1286 1287 1288
const QChar MimeDatabasePrivate::kSemiColon(QLatin1Char(';'));
const QString MimeDatabasePrivate::kModifiedMimeTypesFile(QLatin1String("modifiedmimetypes.xml"));
QString MimeDatabasePrivate::kModifiedMimeTypesPath;

con's avatar
con committed
1289 1290 1291
MimeDatabasePrivate::MimeDatabasePrivate() :
    m_maxLevel(-1)
{
1292
    // Assign here to avoid non-local static data initialization issues.
hjk's avatar
hjk committed
1293
    kModifiedMimeTypesPath = ICore::userResourcePath() + QLatin1String("/mimetypes/");
con's avatar
con committed
1294 1295
}

1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306
/*!
    \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
1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328
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)) {
1329
        *errorMessage = QString::fromLatin1("Cannot open %1: %2").arg(fileName, file.errorString());
con's avatar
con committed
1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351
        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))
1352
            mt.addMagicMatcher(QSharedPointer<IMagicMatcher>(new Internal::BinaryMatcher));
con's avatar
con committed
1353 1354 1355
    }
    // insert the type.
    m_typeMimeTypeMap.insert(type, MimeMapEntry(mt));
1356 1357
    // Register the children
    // Aliases will be resolved later once all mime types are known.
con's avatar
con committed
1358 1359 1360 1361
    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)
1362
            m_parentChildrenMap.insert(*soit, type);
con's avatar
con committed
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
    }
    // 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.
1389 1390 1391
    QStringList childTypes = m_parentChildrenMap.values(e.type.type());
    foreach (const QString &alias, e.type.aliases())
        childTypes.append(m_parentChildrenMap.values(alias));
con's avatar
con committed
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
    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->toUtf