mimedatabase.cpp 59.7 KB
Newer Older
1
/**************************************************************************
con's avatar
con committed
2 3 4
**
** This file is part of Qt Creator
**
con's avatar
con committed
5
** Copyright (c) 2011 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
**
con's avatar
con committed
9
** No Commercial Usage
10
**
con's avatar
con committed
11 12 13 14
** This file contains pre-release code and may not be distributed.
** You may use this file in accordance with the terms and conditions
** contained in the Technology Preview License Agreement accompanying
** this package.
15
**
16
** GNU Lesser General Public License Usage
17
**
18 19 20 21 22 23
** 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.
24
**
con's avatar
con committed
25 26 27 28 29 30
** In addition, as a special exception, Nokia gives you certain additional
** rights.  These rights are described in the Nokia Qt LGPL Exception
** version 1.1, included in the file LGPL_EXCEPTION.txt in this package.
**
** If you have questions regarding the use of this file, please contact
** Nokia at qt-info@nokia.com.
con's avatar
con committed
31
**
32
**************************************************************************/
hjk's avatar
hjk committed
33

con's avatar
con committed
34
#include "mimedatabase.h"
35
#include "coreconstants.h"
36
#include "icore.h"
con's avatar
con committed
37

hjk's avatar
hjk committed
38 39 40 41 42
#include <utils/qtcassert.h>

#include <QtCore/QByteArray>
#include <QtCore/QCoreApplication>
#include <QtCore/QDebug>
con's avatar
con committed
43
#include <QtCore/QFile>
44
#include <QtCore/QDir>
hjk's avatar
hjk committed
45 46
#include <QtCore/QFileInfo>
#include <QtCore/QLocale>
con's avatar
con committed
47
#include <QtCore/QMap>
48
#include <QtCore/QHash>
con's avatar
con committed
49 50 51 52
#include <QtCore/QMultiHash>
#include <QtCore/QRegExp>
#include <QtCore/QSharedData>
#include <QtCore/QSharedPointer>
hjk's avatar
hjk committed
53
#include <QtCore/QStringList>
con's avatar
con committed
54
#include <QtCore/QTextStream>
55
#include <QtCore/QMutexLocker>
56 57
#include <QtCore/QXmlStreamReader>
#include <QtCore/QXmlStreamWriter>
con's avatar
con committed
58

Christian Kandeler's avatar
Christian Kandeler committed
59
#include <algorithm>
60
#include <functional>
Christian Kandeler's avatar
Christian Kandeler committed
61

con's avatar
con committed
62 63 64 65 66 67 68 69 70 71 72
enum { debugMimeDB = 0 };

// XML tags in mime files
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";
73
static const char *weightAttributeC = "weight";
con's avatar
con committed
74 75 76 77 78 79 80 81
static const char *localeAttributeC = "xml:lang";

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";
82
static const char *matchByteTypeValueC = "byte";
con's avatar
con committed
83 84 85 86 87 88 89 90 91 92 93
static const char *matchOffsetAttributeC = "offset";

// Types
static const char *textTypeC = "text/plain";
static const char *binaryTypeC = "application/octet-stream";

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

// Fallback priorities, must be low.
94 95 96 97
enum {
    BinaryMatchPriority = Core::MimeGlobPattern::MinWeight + 1,
    TextMatchPriority
};
con's avatar
con committed
98 99 100 101 102 103 104

/* Parse sth like (<mime-info> being optional):
 *\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">
105
    <comment xml:lang="en">Qt qmake Profile</comment>
106
    <glob pattern="*.pro" weight="50"/>
con's avatar
con committed
107 108 109 110 111 112
  </mime-type>
</mime-info>
 *\endcode
*/

namespace Core {
113 114 115

typedef QSharedPointer<MagicRuleMatcher> MagicRuleMatcherPtr;

con's avatar
con committed
116 117 118 119 120 121 122 123
namespace Internal {

// FileMatchContext: Passed on to the mimetypes from the database
// 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).

class FileMatchContext {
124
    Q_DISABLE_COPY(FileMatchContext)
con's avatar
con committed
125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174
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;
}

// The binary fallback matcher for "application/octet-stream".
class BinaryMatcher : public IMagicMatcher {
175
    Q_DISABLE_COPY(BinaryMatcher)
con's avatar
con committed
176 177 178 179 180 181 182 183 184
public:
    BinaryMatcher() {}
    virtual bool matches(const QByteArray & /*data*/) const { return true; }
    virtual int priority() const  { return BinaryMatchPriority; }
};

// A heuristic text file matcher: If the data do not contain any character
// below tab (9), detect as text.
class HeuristicTextMagicMatcher : public IMagicMatcher {
185
    Q_DISABLE_COPY(HeuristicTextMagicMatcher)
con's avatar
con committed
186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217
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

// MagicRule
218 219
const QChar MagicRule::kColon(QLatin1Char(':'));

220
MagicRule::MagicRule(int startPos, int endPos) : m_startPos(startPos), m_endPos(endPos)
con's avatar
con committed
221 222 223
{
}

224 225 226 227 228 229 230 231 232 233 234 235 236 237
MagicRule::~MagicRule()
{
}

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

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

238 239 240 241 242 243 244 245 246 247 248 249 250 251
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());
}

const QString MagicStringRule::kMatchType("string");

252 253 254 255 256 257 258 259 260
MagicStringRule::MagicStringRule(const QString &s, int startPos, int endPos) :
    MagicRule(startPos, endPos), m_pattern(s.toUtf8())
{
}

MagicStringRule::~MagicStringRule()
{
}

261 262 263 264 265 266 267 268 269 270
QString MagicStringRule::matchType() const
{
    return kMatchType;
}

QString MagicStringRule::matchValue() const
{
    return m_pattern;
}

271
bool MagicStringRule::matches(const QByteArray &data) const
con's avatar
con committed
272 273
{
    // Quick check
274
    if ((startPos() + m_pattern.size()) > data.size())
con's avatar
con committed
275 276
        return false;
    // Most common: some string at position 0:
277
    if (startPos() == 0 && startPos() == endPos())
con's avatar
con committed
278 279
        return data.startsWith(m_pattern);
    // Range
280 281
    const int index = data.indexOf(m_pattern, startPos());
    const bool rc = index != -1 && index <= endPos();
282
    if (debugMimeDB)
283
        qDebug() << "Checking " << m_pattern << startPos() << endPos() << " returns " << rc;
284
    return rc;
con's avatar
con committed
285 286
}

287 288
const QString MagicByteRule::kMatchType(QLatin1String("byte"));

289
MagicByteRule::MagicByteRule(const QString &s, int startPos, int endPos) :
290 291 292 293 294 295 296 297 298 299 300 301 302
    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
303
{
304
    // Expect an hex format value like this: \0x7f\0x45\0x4c\0x46
305 306
    const QStringList &byteSequence = sequence.split(QLatin1Char('\\'), QString::SkipEmptyParts);
    foreach (const QString &byte, byteSequence) {
307 308 309
        bool ok;
        const int hex = byte.toInt(&ok, 16);
        if (ok) {
310 311
            if (bytes)
                bytes->push_back(hex);
312
        } else {
313
            return false;
314 315
        }
    }
316
    return true;
317 318
}

319
QString MagicByteRule::matchType() const
320
{
321 322 323 324 325 326 327 328 329
    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;
330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352
}

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
353 354 355 356 357 358 359 360 361 362
}

// List matcher
MagicRuleMatcher::MagicRuleMatcher() :
     m_priority(65535)
{
}

void MagicRuleMatcher::add(const MagicRuleSharedPointer &rule)
{
363 364 365 366 367 368 369 370 371 372 373
    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
374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394
}

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;
}

395 396 397 398 399 400 401 402 403 404 405 406 407 408
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;
}

409
// GlobPattern
Friedemann Kleint's avatar
Friedemann Kleint committed
410
MimeGlobPattern::MimeGlobPattern(const QRegExp &regExp, unsigned weight) :
411 412 413 414 415 416 417 418 419 420 421 422 423
    m_regExp(regExp), m_weight(weight)
{
}

MimeGlobPattern::~MimeGlobPattern()
{
}

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

Friedemann Kleint's avatar
Friedemann Kleint committed
424
unsigned MimeGlobPattern::weight() const
425 426 427 428
{
    return m_weight;
}

con's avatar
con committed
429 430 431 432
// ---------- MimeTypeData
class MimeTypeData : public QSharedData {
public:
    typedef QHash<QString,QString> LocaleHash;
433 434 435

    MimeTypeData();

con's avatar
con committed
436
    void clear();
437 438
    void assignSuffix(const QString &pattern);
    void assignSuffixes(const QStringList &patterns);
con's avatar
con committed
439 440
    void debug(QTextStream &str, int indent = 0) const;

441 442
    const QRegExp suffixPattern;

con's avatar
con committed
443 444 445 446 447
    QString type;
    QString comment;

    LocaleHash localeComments;
    QStringList aliases;
448
    QList<MimeGlobPattern> globPatterns;
con's avatar
con committed
449 450 451
    QStringList subClassesOf;
    QString preferredSuffix;
    QStringList suffixes;
452
    IMagicMatcher::IMagicMatcherList magicMatchers;
con's avatar
con committed
453 454
};

455 456 457 458 459 460 461 462
MimeTypeData::MimeTypeData()
    // RE to match a suffix glob pattern: "*.ext" (and not sth like "Makefile" or
    // "*.log[1-9]"
    : suffixPattern(QLatin1String("^\\*\\.[\\w+]+$"))
{
    QTC_ASSERT(suffixPattern.isValid(), /**/);
}

con's avatar
con committed
463 464 465 466 467 468 469 470 471 472 473 474
void MimeTypeData::clear()
{
    type.clear();
    comment.clear();
    aliases.clear();
    globPatterns.clear();
    subClassesOf.clear();
    preferredSuffix.clear();
    suffixes.clear();
    magicMatchers.clear();
}

475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490
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
491 492 493 494 495 496 497 498 499 500 501 502 503
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: ";
504
        foreach (const MimeGlobPattern &gp, globPatterns)
Friedemann Kleint's avatar
Friedemann Kleint committed
505
            str << gp.regExp().pattern() << '(' << gp.weight() << ')';
con's avatar
con committed
506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615
        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;
}

616
QList<MimeGlobPattern> MimeType::globPatterns() const
con's avatar
con committed
617 618 619 620
{
    return m_d->globPatterns;
}

621
void MimeType::setGlobPatterns(const QList<MimeGlobPattern> &g)
con's avatar
con committed
622 623
{
    m_d->globPatterns = g;
624 625 626 627 628 629 630

    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
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
}

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;
}

661
QString MimeType::formatFilterString(const QString &description, const QList<MimeGlobPattern> &globs)
con's avatar
con committed
662 663 664 665 666 667 668 669 670 671 672 673 674
{
    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 << ' ';
675
                str << globs.at(i).regExp().pattern();
con's avatar
con committed
676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696
            }
            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);
697 698
    const unsigned suffixPriority = matchesFileBySuffix(context);
    if (suffixPriority >= MimeGlobPattern::MaxWeight)
699
        return suffixPriority;
700
    return qMax(suffixPriority, matchesFileByContent(context));
con's avatar
con committed
701 702
}

703
unsigned MimeType::matchesFileBySuffix(Internal::FileMatchContext &c) const
con's avatar
con committed
704 705
{
    // check globs
706 707 708
    foreach (const MimeGlobPattern &gp, m_d->globPatterns) {
        if (gp.regExp().exactMatch(c.fileName()))
            return gp.weight();
709
    }
710 711
    return 0;
}
712

713 714
unsigned MimeType::matchesFileByContent(Internal::FileMatchContext &c) const
{
con's avatar
con committed
715
    // Nope, try magic matchers on context data
716
    if (m_d->magicMatchers.isEmpty())
717 718 719 720
        return 0;

    return matchesData(c.data());
}
721

722 723 724
unsigned MimeType::matchesData(const QByteArray &data) const
{
    unsigned priority = 0;
con's avatar
con committed
725
    if (!data.isEmpty()) {
726
        foreach (const IMagicMatcher::IMagicMatcherSharedPointer &matcher, m_d->magicMatchers) {
727 728 729
            const unsigned magicPriority = matcher->priority();
            if (magicPriority > priority && matcher->matches(data))
                priority = magicPriority;
730
        }
con's avatar
con committed
731
    }
732
    return priority;
con's avatar
con committed
733 734 735 736 737 738 739
}

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

740
void MimeType::addMagicMatcher(const IMagicMatcherSharedPointer &matcher)
con's avatar
con committed
741
{
742
    m_d->magicMatchers.push_back(matcher);
con's avatar
con committed
743 744
}

745
const MimeType::IMagicMatcherList &MimeType::magicMatchers() const
con's avatar
con committed
746
{
747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783
    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
784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801
}

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

namespace Internal {

//  MimeDatabase helpers: Generic parser for a sequence of <mime-type>.
//  Calls abstract handler function process for MimeType it finds.
class BaseMimeTypeParser {
802
    Q_DISABLE_COPY(BaseMimeTypeParser)
con's avatar
con committed
803
public:
804
    BaseMimeTypeParser() {}
con's avatar
con committed
805 806 807 808 809 810 811 812
    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;

813
    void addGlobPattern(const QString &pattern, const QString &weight, MimeTypeData *d) const;
con's avatar
con committed
814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829

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

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

830
void BaseMimeTypeParser::addGlobPattern(const QString &pattern, const QString &weight, MimeTypeData *d) const
con's avatar
con committed
831 832 833 834 835 836 837 838 839 840 841 842
{
    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;
    }

843 844 845 846 847
    if (weight.isEmpty())
        d->globPatterns.push_back(MimeGlobPattern(wildCard));
    else
        d->globPatterns.push_back(MimeGlobPattern(wildCard, weight.toInt()));

848
    d->assignSuffix(pattern);
con's avatar
con committed
849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880
}

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;
881 882
        if (startElement == QLatin1String(matchTagC))
            return ParseMagicMatchRule;
con's avatar
con committed
883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899
        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) {
900
        *errorMessage = QString::fromLatin1("Not a number '%1'.").arg(n);
con's avatar
con committed
901 902 903 904 905 906 907 908 909
        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,
910
                              const MagicRuleMatcherPtr  &ruleMatcher,
con's avatar
con committed
911 912 913
                              QString *errorMessage)
{
    const QString type = atts.value(QLatin1String(matchTypeAttributeC)).toString();
914 915
    if (type != QLatin1String(matchStringTypeValueC) &&
        type != QLatin1String(matchByteTypeValueC)) {
con's avatar
con committed
916 917 918 919 920
        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()) {
921
        *errorMessage = QString::fromLatin1("Empty match value detected.");
con's avatar
con committed
922 923 924 925 926 927 928 929 930 931 932 933
        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;
934 935 936 937 938

    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
939 940 941 942 943 944
    return true;
}

bool BaseMimeTypeParser::parse(QIODevice *dev, const QString &fileName, QString *errorMessage)
{
    MimeTypeData data;
945
    MagicRuleMatcherPtr ruleMatcher;
con's avatar
con committed
946 947
    QXmlStreamReader reader(dev);
    ParseStage ps = ParseBeginning;
948
    QXmlStreamAttributes atts;
con's avatar
con committed
949 950 951 952
    while (!reader.atEnd()) {
        switch (reader.readNext()) {
        case QXmlStreamReader::StartElement:
            ps = nextStage(ps, reader.name());
953
            atts = reader.attributes();
con's avatar
con committed
954 955
            switch (ps) {
            case ParseMimeType: { // start parsing a type
956
                const QString type = atts.value(QLatin1String(mimeTypeAttributeC)).toString();
con's avatar
con committed
957
                if (type.isEmpty()) {
958
                    reader.raiseError(QString::fromLatin1("Missing 'type'-attribute"));
con's avatar
con committed
959 960 961 962 963 964
                } else {
                    data.type = type;
                }
            }
                break;
            case ParseGlobPattern:
965 966
                addGlobPattern(atts.value(QLatin1String(patternAttributeC)).toString(),
                               atts.value(QLatin1String(weightAttributeC)).toString(), &data);
con's avatar
con committed
967 968
                break;
            case ParseSubClass: {
969
                const QString inheritsFrom = atts.value(QLatin1String(mimeTypeAttributeC)).toString();
con's avatar
con committed
970 971 972 973 974 975
                if (!inheritsFrom.isEmpty())
                    data.subClassesOf.push_back(inheritsFrom);
            }
                break;
            case ParseComment: {
                // comments have locale attributes. We want the default, English one
976
                QString locale = atts.value(QLatin1String(localeAttributeC)).toString();
977
                const QString comment = QCoreApplication::translate("MimeType", reader.readElementText().toAscii());
con's avatar
con committed
978 979 980 981 982 983 984 985
                if (locale.isEmpty()) {
                    data.comment = comment;
                } else {
                    data.localeComments.insert(locale, comment);
                }
            }
                break;
            case ParseAlias: {
986
                const QString alias = atts.value(QLatin1String(mimeTypeAttributeC)).toString();
con's avatar
con committed
987 988 989 990 991 992
                if (!alias.isEmpty())
                    data.aliases.push_back(alias);
            }
                break;
            case ParseMagic: {
                int priority = 0;
993
                const QString priorityS = atts.value(QLatin1String(priorityAttributeC)).toString();
con's avatar
con committed
994 995 996 997 998
                if (!priorityS.isEmpty()) {
                    if (!parseNumber(priorityS, &priority, errorMessage))
                        return false;

                }
999
                ruleMatcher = MagicRuleMatcherPtr(new MagicRuleMatcher);
con's avatar
con committed
1000 1001 1002 1003
                ruleMatcher->setPriority(priority);
            }
                break;
            case ParseMagicMatchRule:
1004
                QTC_ASSERT(!ruleMatcher.isNull(), return false)
1005
                if (!addMagicMatchRule(atts, ruleMatcher, errorMessage))
con's avatar
con committed
1006 1007 1008
                    return false;
                break;
            case ParseError:
1009
                reader.raiseError(QString::fromLatin1("Unexpected element <%1>").arg(reader.name().toString()));
con's avatar
con committed
1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022
                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
1023 1024 1025 1026
                if (reader.name() == QLatin1String(magicTagC)) {
                    QTC_ASSERT(!ruleMatcher.isNull(), return false)
                    data.magicMatchers.push_back(ruleMatcher);
                    ruleMatcher = MagicRuleMatcherPtr();
con's avatar
con committed
1027 1028 1029 1030 1031 1032 1033 1034 1035 1036
                }
            }
            break;

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

    if (reader.hasError()) {
1037
        *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
1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048
        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
1049 1050
struct MimeMapEntry
{
con's avatar
con committed
1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068
    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)
{
}

/* MimeDatabasePrivate: Requirements for storage:
 * - Must be robust in case of incomplete hierachies, dangling entries
 * - Plugins will not load and register their mime types in order
 *   of inheritance.
 * - Multiple inheritance (several subClassesOf) can occur
 * - Provide quick lookup by name
 * - Provide quick lookup by file type.
Tobias Hunger's avatar
Tobias Hunger committed
1069
 * This basically rules out some pointer-based tree, so the structure chosen
con's avatar
con committed
1070
 * is:
1071 1072
 * - An alias map QString->QString for mapping aliases to types
 * - A Map QString->MimeMapEntry for the types (MimeMapEntry being a pair of
con's avatar
con committed
1073
 *   MimeType and (hierarchy) level.
1074
 * - A map  QString->QString representing parent->child relations (enabling
con's avatar
con committed
1075 1076 1077 1078 1079 1080 1081 1082 1083 1084
 *   recursing over children)
 * Using strings avoids dangling pointers.
 * 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. */

hjk's avatar
hjk committed
1085 1086
class MimeDatabasePrivate
{
con's avatar
con committed
1087 1088 1089 1090 1091 1092 1093 1094 1095 1096
    Q_DISABLE_COPY(MimeDatabasePrivate)
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;
1097
    MimeType findByData(const QByteArray &data) const;
con's avatar
con committed
1098

1099
    QStringList filterStrings() const;
con's avatar
con committed
1100 1101

    QStringList suffixes() const;
1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120
    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
1121 1122 1123 1124 1125 1126 1127 1128

    void debug(QTextStream &str) const;

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

1129 1130 1131 1132 1133
    static const QChar kSemiColon;
    static const QString kModifiedMimeTypesFile;
    static QString kModifiedMimeTypesPath;


con's avatar
con committed
1134 1135 1136
    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;
1137
    MimeType findByData(const QByteArray &data, unsigned *priority) const;
con's avatar
con committed
1138 1139 1140 1141 1142 1143 1144 1145 1146
    void determineLevels();
    void raiseLevelRecursion(MimeMapEntry &e, int level);

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

1147 1148 1149 1150
const QChar MimeDatabasePrivate::kSemiColon(QLatin1Char(';'));
const QString MimeDatabasePrivate::kModifiedMimeTypesFile(QLatin1String("modifiedmimetypes.xml"));
QString MimeDatabasePrivate::kModifiedMimeTypesPath;

con's avatar
con committed
1151 1152 1153
MimeDatabasePrivate::MimeDatabasePrivate() :
    m_maxLevel(-1)
{
1154 1155
    // Assign here to avoid non-local static data initialization issues.
    kModifiedMimeTypesPath = ICore::instance()->userResourcePath() + QLatin1String("/mimetypes/");
con's avatar
con committed
1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179
}

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)) {
1180
        *errorMessage = QString::fromLatin1("Cannot open %1: %2").arg(fileName, file.errorString());
con's avatar
con committed
1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202
        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))
1203
            mt.addMagicMatcher(QSharedPointer<IMagicMatcher>(new Internal::BinaryMatcher));
con's avatar
con committed
1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312
    }
    // 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)
1313
        qDebug() << '>' << Q_FUNC_INFO << f.absoluteFilePath();
con's avatar
con committed
1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332
    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();
    }
1333 1334 1335 1336 1337 1338

    // 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.
con's avatar
con committed
1339
    *priorityPtr = 0;
1340
    MimeType candidate;
con's avatar
con committed
1341
    Internal::FileMatchContext context(f);
1342

1343
    // Pass 1) Try to match on suffix
con's avatar
con committed
1344
    const TypeMimeTypeMap::const_iterator cend = m_typeMimeTypeMap.constEnd();
1345
    for (int level = m_maxLevel; level >= 0 && candidate.isNull(); level--)
1346
        for (TypeMimeTypeMap::const_iterator it = m_typeMimeTypeMap.constBegin(); it != cend; ++it)
1347 1348 1349
            if (it.value().level == level) {
                const unsigned suffixPriority = it.value().type.matchesFileBySuffix(context);
                if (suffixPriority && suffixPriority > *priorityPtr) {
1350
                    *priorityPtr = suffixPriority;
1351 1352 1353
                    candidate = it.value().type;
                    if (suffixPriority >= MimeGlobPattern::MaxWeight)
                        return candidate;
1354
                }
1355 1356
            }

1357 1358
    // Pass 2) Match on content
    if (!f.isReadable())
1359
        return candidate;
con's avatar
con committed
1360 1361 1362
    for (int level = m_maxLevel; level >= 0; level--)
        for (TypeMimeTypeMap::const_iterator it = m_typeMimeTypeMap.constBegin(); it != cend; ++it)
            if (it.value().level == level) {
1363 1364 1365 1366
                const unsigned contentPriority = it.value().type.matchesFileByContent(context);
                if (contentPriority && contentPriority > *priorityPtr) {
                    *priorityPtr = contentPriority;
                    candidate = it.value().type;
1367
                }
con's avatar
con committed
1368
            }
1369 1370

    return candidate;
con's avatar
con committed
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
// Debugging wrapper around findByData()
MimeType MimeDatabasePrivate::findByData(const QByteArray &data) const
{
    unsigned priority = 0;
    if (debugMimeDB)
        qDebug() << '>' << Q_FUNC_INFO << data.left(20).toHex();
    const MimeType rc = findByData(data, &priority);
    if (debugMimeDB) {
        if (rc) {
            qDebug() << "<MimeDatabase::findByData: match prio=" << priority << rc.type();
        } else {
            qDebug() << "<MimeDatabase::findByData: no match";
        }
    }
    return rc;
}

// Returns a mime type or Null one if none found
MimeType MimeDatabasePrivate::findByData(const QByteArray &data, 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();
    }

    *priorityPtr = 0;
    MimeType candidate;

    const TypeMimeTypeMap::const_iterator cend = m_typeMimeTypeMap.constEnd();
    for (int level = m_maxLevel; level >= 0; level--)
        for (TypeMimeTypeMap::const_iterator it = m_typeMimeTypeMap.constBegin(); it != cend; ++it)
            if (it.value().level == level) {
                const unsigned contentPriority = it.value().type.matchesData(data);
                if (contentPriority && contentPriority > *priorityPtr) {
                    *priorityPtr = contentPriority;
                    candidate = it.value().type;
                }
            }

    return candidate;
}

con's avatar
con committed
1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429
// Return all known suffixes
QStringList MimeDatabasePrivate::suffixes() const
{
    QStringList rc;
    const TypeMimeTypeMap::const_iterator cend = m_typeMimeTypeMap.constEnd();
    for (TypeMimeTypeMap::const_iterator it = m_typeMimeTypeMap.constBegin(); it != cend; ++it)
        rc += it.value().type.suffixes();
    return rc;
}

QStringList MimeDatabasePrivate::filterStrings() const
{
    QStringList rc;
    const TypeMimeTypeMap::const_iterator cend = m_typeMimeTypeMap.constEnd();
1430 1431 1432 1433 1434
    for (TypeMimeTypeMap::const_iterator it = m_typeMimeTypeMap.constBegin(); it != cend; ++it) {
        const QString filterString = it.value().type.filterString();
        if (!filterString.isEmpty())
            rc += filterString;
    }
con's avatar
con committed
1435 1436 1437
    return rc;
}