mirror of
https://github.com/JakubMelka/PDF4QT.git
synced 2024-12-28 17:20:02 +01:00
1604 lines
54 KiB
C++
1604 lines
54 KiB
C++
// Copyright (C) 2019-2021 Jakub Melka
|
|
//
|
|
// This file is part of PDF4QT.
|
|
//
|
|
// PDF4QT is free software: you can redistribute it and/or modify
|
|
// it under the terms of the GNU Lesser General Public License as published by
|
|
// the Free Software Foundation, either version 3 of the License, or
|
|
// with the written consent of the copyright owner, any later version.
|
|
//
|
|
// PDF4QT is distributed in the hope that it will be useful,
|
|
// but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
|
// GNU Lesser General Public License for more details.
|
|
//
|
|
// You should have received a copy of the GNU Lesser General Public License
|
|
// along with PDF4QT. If not, see <https://www.gnu.org/licenses/>.
|
|
|
|
#include "pdftextlayout.h"
|
|
#include "pdfutils.h"
|
|
#include "pdfexecutionpolicy.h"
|
|
|
|
#include <QPainter>
|
|
|
|
#include <execution>
|
|
|
|
namespace pdf
|
|
{
|
|
|
|
/// Spatial 2D index for indexing of text characters. It is a R-tree like structure,
|
|
/// build over an array of text characters. Array is modified (structure is build over
|
|
/// array).
|
|
class PDFTextCharacterSpatialIndex
|
|
{
|
|
public:
|
|
explicit PDFTextCharacterSpatialIndex(TextCharacters* characters, size_t leafSize) :
|
|
m_characters(characters),
|
|
m_leafSize(leafSize),
|
|
m_epsilon(0.0)
|
|
{
|
|
m_nodes.reserve(2 * characters->size() / leafSize);
|
|
|
|
// Calculate epsilon from the bounding box. We must use epsilon to avoid empty
|
|
// rectangles, which can occur, if text is on a single line.
|
|
QRectF boundingBox = getBoundingBox(characters->begin(), characters->end());
|
|
if (boundingBox.isValid())
|
|
{
|
|
qreal edge = qMax(boundingBox.width(), boundingBox.height());
|
|
m_epsilon = edge * 0.001;
|
|
}
|
|
else
|
|
{
|
|
m_epsilon = 0.01;
|
|
}
|
|
|
|
build(characters->begin(), characters->end());
|
|
}
|
|
|
|
using Iterator = TextCharacters::iterator;
|
|
|
|
/// Builds structure over range of iterators. Array is build in O(n * log^2 (n)) time.
|
|
/// Index to internal nodes array is returned.
|
|
/// \param it1 Start iterator
|
|
/// \param it2 End iterator
|
|
size_t build(Iterator it1, Iterator it2);
|
|
|
|
/// Returns bounding box of character positions over given iterator range.
|
|
/// If iterator range is empty, then empty bounding box is returned.
|
|
/// \param it1 Start iterator
|
|
/// \param it2 End iterator
|
|
QRectF getBoundingBox(Iterator it1, Iterator it2) const;
|
|
|
|
/// Performs query on structure - finds all characters, which are in given
|
|
/// rectangle, and returns intersection size. If \p result parameter is set
|
|
/// to valid pointer, all intersected characters are inserted into the result
|
|
/// array.
|
|
/// \param rect Query rectangle
|
|
/// \param result Result of query (can be nullptr)
|
|
/// \returns Size of intersection
|
|
size_t query(const QRectF& rect, TextCharacters* result) const;
|
|
|
|
/// Finds character array, which contains at least \p minimalSize characters,
|
|
/// with some extra characters, which must be filtered out.
|
|
/// \param minimalSize Minimal size
|
|
/// \param sample Sample character
|
|
/// \param result Result
|
|
void queryNearestEstimate(size_t minimalSize, const TextCharacter& sample, TextCharacters* result) const;
|
|
|
|
private:
|
|
size_t queryImpl(size_t nodeIndex, const QRectF& rect, TextCharacters* result) const;
|
|
|
|
struct Node
|
|
{
|
|
bool isLeaf = false;
|
|
size_t index1 = 0;
|
|
size_t index2 = 0;
|
|
QRectF boundingBox;
|
|
};
|
|
|
|
using Nodes = std::vector<Node>;
|
|
|
|
TextCharacters* m_characters;
|
|
Nodes m_nodes;
|
|
size_t m_leafSize;
|
|
qreal m_epsilon;
|
|
};
|
|
|
|
size_t PDFTextCharacterSpatialIndex::build(Iterator it1, Iterator it2)
|
|
{
|
|
size_t nodeIndex = m_nodes.size();
|
|
|
|
if (size_t(std::distance(it1, it2)) < m_leafSize)
|
|
{
|
|
// Create leaf node
|
|
Node node;
|
|
node.isLeaf = true;
|
|
node.index1 = std::distance(m_characters->begin(), it1);
|
|
node.index2 = std::distance(m_characters->begin(), it2);
|
|
node.boundingBox = getBoundingBox(it1, it2);
|
|
m_nodes.push_back(qMove(node));
|
|
}
|
|
else
|
|
{
|
|
// Jakub Melka: split array of nodes into half, using larger side.
|
|
// It is like in R-tree structure.
|
|
m_nodes.push_back(Node());
|
|
|
|
QRectF boundingBox = getBoundingBox(it1, it2);
|
|
if (boundingBox.width() > boundingBox.height())
|
|
{
|
|
// Split using x-axis
|
|
std::sort(it1, it2, [](const TextCharacter& l, const TextCharacter& r) { return l.position.x() < r.position.x(); });
|
|
}
|
|
else
|
|
{
|
|
// Split using y-axis
|
|
std::sort(it1, it2, [](const TextCharacter& l, const TextCharacter& r) { return l.position.y() < r.position.y(); });
|
|
}
|
|
|
|
const size_t distance = std::distance(it1, it2);
|
|
Iterator itMid = std::next(it1, distance / 2);
|
|
|
|
const size_t index1 = build(it1, itMid);
|
|
const size_t index2 = build(itMid, it2);
|
|
|
|
Node& node = m_nodes[nodeIndex];
|
|
node.isLeaf = false;
|
|
node.index1 = index1;
|
|
node.index2 = index2;
|
|
node.boundingBox = boundingBox;
|
|
}
|
|
|
|
return nodeIndex;
|
|
}
|
|
|
|
QRectF PDFTextCharacterSpatialIndex::getBoundingBox(Iterator it1, Iterator it2) const
|
|
{
|
|
if (it1 != it2)
|
|
{
|
|
qreal x_min = qInf();
|
|
qreal x_max = -qInf();
|
|
qreal y_min = qInf();
|
|
qreal y_max = -qInf();
|
|
|
|
for (Iterator it = it1; it != it2; ++it)
|
|
{
|
|
const TextCharacter& character = *it;
|
|
x_min = qMin(x_min, character.position.x() - m_epsilon);
|
|
x_max = qMax(x_max, character.position.x() + m_epsilon);
|
|
y_min = qMin(y_min, character.position.y() - m_epsilon);
|
|
y_max = qMax(y_max, character.position.y() + m_epsilon);
|
|
}
|
|
|
|
return QRectF(x_min, y_min, x_max - x_min, y_max - y_min);
|
|
}
|
|
|
|
return QRectF();
|
|
}
|
|
|
|
size_t PDFTextCharacterSpatialIndex::query(const QRectF& rect, TextCharacters* result) const
|
|
{
|
|
if (!m_nodes.empty())
|
|
{
|
|
return queryImpl(0, rect, result);
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
void PDFTextCharacterSpatialIndex::queryNearestEstimate(size_t minimalSize, const TextCharacter& sample, TextCharacters* result) const
|
|
{
|
|
if (m_characters->size() <= minimalSize)
|
|
{
|
|
*result = *m_characters;
|
|
}
|
|
else
|
|
{
|
|
// Query result
|
|
qreal querySizeEstimate = qMax(qMax(m_nodes[0].boundingBox.width(), m_nodes[0].boundingBox.height()) * 0.01, sample.advance * minimalSize * 0.5);
|
|
|
|
QRectF rect(sample.position, QSizeF(querySizeEstimate, querySizeEstimate));
|
|
rect.translate(-querySizeEstimate * 0.5, -querySizeEstimate * 0.5);
|
|
|
|
while (query(rect, nullptr) < minimalSize)
|
|
{
|
|
qreal expansion = rect.width() * 0.5;
|
|
rect.adjust(-expansion, -expansion, expansion, expansion);
|
|
}
|
|
|
|
qreal expansion = rect.width() * (qSqrt(2.0) - 1.0);
|
|
rect.adjust(-expansion, -expansion, expansion, expansion);
|
|
query(rect, result);
|
|
}
|
|
}
|
|
|
|
size_t PDFTextCharacterSpatialIndex::queryImpl(size_t nodeIndex, const QRectF& rect, TextCharacters* result) const
|
|
{
|
|
const Node& node = m_nodes[nodeIndex];
|
|
|
|
if (!node.boundingBox.intersects(rect))
|
|
{
|
|
// Node is not intersected, just return
|
|
return 0;
|
|
}
|
|
|
|
if (!node.isLeaf)
|
|
{
|
|
return queryImpl(node.index1, rect, result) + queryImpl(node.index2, rect, result);
|
|
}
|
|
else
|
|
{
|
|
// Jakub Melka: it is a leaf...
|
|
auto isInside = [&rect](const TextCharacter& character)
|
|
{
|
|
return rect.contains(character.position);
|
|
};
|
|
|
|
auto itStart = std::next(m_characters->begin(), node.index1);
|
|
auto itEnd = std::next(m_characters->begin(), node.index2);
|
|
|
|
if (result)
|
|
{
|
|
const size_t oldSize = result->size();
|
|
std::copy_if(itStart, itEnd, std::back_inserter(*result), isInside);
|
|
return result->size() - oldSize;
|
|
}
|
|
|
|
return std::count_if(itStart, itEnd, isInside);
|
|
}
|
|
}
|
|
|
|
PDFTextLayout::PDFTextLayout()
|
|
{
|
|
|
|
}
|
|
|
|
void PDFTextLayout::addCharacter(const PDFTextCharacterInfo& info)
|
|
{
|
|
TextCharacter character;
|
|
|
|
// Fill the basic info. For computing the angle, we must consider, if we are
|
|
// in vertical writing system. If yes, take vertical edge of the character,
|
|
// otherwise take horizontal edge.
|
|
character.character = info.character;
|
|
character.position = info.matrix.map(QPointF(0.0, 0.0));
|
|
|
|
QLineF testLine(QPointF(0.0, 0.0), QPointF(info.isVerticalWritingSystem ? 0.0 : info.advance, !info.isVerticalWritingSystem ? 0.0 : info.advance));
|
|
QLineF mappedLine = info.matrix.map(testLine);
|
|
character.advance = mappedLine.length();
|
|
character.angle = qRound(mappedLine.angle());
|
|
|
|
QLineF fontTestLine(QPointF(0.0, 0.0), QPointF(0.0, info.fontSize));
|
|
QLineF fontMappedLine = info.matrix.map(fontTestLine);
|
|
character.fontSize = fontMappedLine.length();
|
|
|
|
QRectF boundingBox = info.outline.boundingRect();
|
|
character.boundingBox.addPolygon(info.matrix.map(boundingBox));
|
|
|
|
m_characters.emplace_back(qMove(character));
|
|
m_angles.insert(character.angle);
|
|
}
|
|
|
|
void PDFTextLayout::perform()
|
|
{
|
|
for (PDFReal angle : m_angles)
|
|
{
|
|
performDoLayout(angle);
|
|
}
|
|
}
|
|
|
|
void PDFTextLayout::optimize()
|
|
{
|
|
m_characters.shrink_to_fit();
|
|
}
|
|
|
|
qint64 PDFTextLayout::getMemoryConsumptionEstimate() const
|
|
{
|
|
qint64 estimate = sizeof(*this);
|
|
estimate += sizeof(decltype(m_characters)::value_type) * m_characters.capacity();
|
|
estimate += sizeof(decltype(m_angles)::value_type) * m_angles.size();
|
|
return estimate;
|
|
}
|
|
|
|
bool PDFTextLayout::isHoveringOverTextBlock(const QPointF& point) const
|
|
{
|
|
for (const PDFTextBlock& block : m_blocks)
|
|
{
|
|
if (block.getBoundingBox().contains(point))
|
|
{
|
|
return true;
|
|
}
|
|
}
|
|
|
|
return false;
|
|
}
|
|
|
|
PDFTextSelection PDFTextLayout::createTextSelection(PDFInteger pageIndex, const QPointF& point1, const QPointF& point2, QColor selectionColor)
|
|
{
|
|
PDFTextSelection selection;
|
|
|
|
// Jakub Melka: We must treat each block in its own coordinate system. Because texts can
|
|
// have different angles, we will treat each block separately.
|
|
|
|
size_t blockId = 0;
|
|
for (PDFTextBlock& block : m_blocks)
|
|
{
|
|
QMatrix angleMatrix;
|
|
angleMatrix.rotate(block.getAngle());
|
|
block.applyTransform(angleMatrix);
|
|
|
|
QPointF pointA = angleMatrix.map(point1);
|
|
QPointF pointB = angleMatrix.map(point2);
|
|
|
|
const qreal xMin = qMin(pointA.x(), pointB.x());
|
|
const qreal yMin = qMin(pointA.y(), pointB.y());
|
|
const qreal xMax = qMax(pointA.x(), pointB.x());
|
|
const qreal yMax = qMax(pointA.y(), pointB.y());
|
|
|
|
QRectF rect(xMin, yMin, xMax - xMin, yMax - yMin);
|
|
QPainterPath rectPath;
|
|
rectPath.addRect(rect);
|
|
const QPainterPath& boundingBoxPath = block.getBoundingBox();
|
|
QPainterPath intersectionPath = boundingBoxPath.intersected(rectPath);
|
|
if (!intersectionPath.isEmpty())
|
|
{
|
|
QRectF intersectionRect = intersectionPath.boundingRect();
|
|
Q_ASSERT(intersectionRect.isValid());
|
|
|
|
bool isTopPointAboveText = false;
|
|
bool isBottomPointBelowText = false;
|
|
|
|
const PDFTextLines& lines = block.getLines();
|
|
auto itLineA = std::find_if(lines.cbegin(), lines.cend(), [pointA](const PDFTextLine& line) { return line.getBoundingBox().contains(pointA); });
|
|
auto itLineB = std::find_if(lines.cbegin(), lines.cend(), [pointB](const PDFTextLine& line) { return line.getBoundingBox().contains(pointB); });
|
|
if (itLineA == itLineB && itLineA != lines.cend())
|
|
{
|
|
// Both points are in the same line. We consider point with lesser
|
|
// horizontal coordinate as start selection point, and point with greater
|
|
// horizontal coordinate as end selection point.
|
|
if (pointA.x() > pointB.x())
|
|
{
|
|
std::swap(pointA, pointB);
|
|
}
|
|
}
|
|
else
|
|
{
|
|
// Otherwise points are not in the same line. Then start point will be
|
|
// point top of the second point. Bottom point will mark end of selection.
|
|
if (pointA.y() < pointB.y())
|
|
{
|
|
std::swap(pointA, pointB);
|
|
}
|
|
|
|
QRectF rect = boundingBoxPath.controlPointRect();
|
|
|
|
// If start point is above the text block, move start point to the left.
|
|
if (rect.bottom() < pointA.y())
|
|
{
|
|
pointA.setX(rect.left());
|
|
isTopPointAboveText = true;
|
|
}
|
|
if (rect.top() > pointB.y())
|
|
{
|
|
pointB.setX(rect.right());
|
|
isBottomPointBelowText = true;
|
|
}
|
|
}
|
|
|
|
// Now, we have pointA as start point and pointB as end point. We must found
|
|
// nearest character to the right of point A, and nearest character to the
|
|
// left of point B (with respect to point A/B).
|
|
|
|
qreal maxDistanceA = std::numeric_limits<qreal>::infinity();
|
|
qreal maxDistanceB = std::numeric_limits<qreal>::infinity();
|
|
|
|
PDFCharacterPointer ptrA;
|
|
PDFCharacterPointer ptrB;
|
|
|
|
for (size_t lineId = 0, linesCount = lines.size(); lineId < linesCount; ++lineId)
|
|
{
|
|
const PDFTextLine& line = lines[lineId];
|
|
const TextCharacters& characters = line.getCharacters();
|
|
for (size_t characterId = 0, characterCount = characters.size(); characterId < characterCount; ++characterId)
|
|
{
|
|
const TextCharacter& character = characters[characterId];
|
|
QPointF characterCenter = character.boundingBox.boundingRect().center();
|
|
|
|
qreal distanceA = QLineF(pointA, characterCenter).length();
|
|
qreal distanceB = QLineF(pointB, characterCenter).length();
|
|
|
|
if (distanceA < maxDistanceA && characterCenter.x() > pointA.x())
|
|
{
|
|
maxDistanceA = distanceA;
|
|
ptrA.pageIndex = pageIndex;
|
|
ptrA.blockIndex = blockId;
|
|
ptrA.lineIndex = lineId;
|
|
ptrA.characterIndex = characterId;
|
|
}
|
|
|
|
if (distanceB < maxDistanceB && characterCenter.x() < pointB.x())
|
|
{
|
|
maxDistanceB = distanceB;
|
|
ptrB.pageIndex = pageIndex;
|
|
ptrB.blockIndex = blockId;
|
|
ptrB.lineIndex = lineId;
|
|
ptrB.characterIndex = characterId;
|
|
}
|
|
}
|
|
}
|
|
|
|
if (isTopPointAboveText && !lines.empty())
|
|
{
|
|
ptrA.pageIndex = pageIndex;
|
|
ptrA.blockIndex = blockId;
|
|
ptrA.lineIndex = 0;
|
|
ptrA.characterIndex = 0;
|
|
}
|
|
|
|
if (isBottomPointBelowText && !lines.empty())
|
|
{
|
|
ptrB.pageIndex = pageIndex;
|
|
ptrB.blockIndex = blockId;
|
|
ptrB.lineIndex = lines.size() - 1;
|
|
ptrB.characterIndex = lines.back().getCharacters().size() - 1;
|
|
}
|
|
|
|
// If we have filled the pointers, add them to the selection
|
|
if (ptrA.isValid() && ptrB.isValid())
|
|
{
|
|
if (ptrA < ptrB)
|
|
{
|
|
selection.addItems({ PDFTextSelectionItem(ptrA, ptrB) }, selectionColor);
|
|
}
|
|
else
|
|
{
|
|
selection.addItems({ PDFTextSelectionItem(ptrB, ptrA) }, selectionColor);
|
|
}
|
|
}
|
|
}
|
|
|
|
// Increment block index
|
|
++blockId;
|
|
|
|
// Apply backward transformation to restore original coordinate system
|
|
block.applyTransform(angleMatrix.inverted());
|
|
}
|
|
|
|
selection.build();
|
|
return selection;
|
|
}
|
|
|
|
QString PDFTextLayout::getTextFromSelection(PDFTextSelection::iterator itBegin, PDFTextSelection::iterator itEnd, PDFInteger pageIndex) const
|
|
{
|
|
QStringList text;
|
|
|
|
if (itBegin != itEnd)
|
|
{
|
|
PDFTextFlows flows = PDFTextFlow::createTextFlows(*this, PDFTextFlow::RemoveSoftHyphen, pageIndex);
|
|
Q_ASSERT(flows.size() < 2);
|
|
|
|
if (!flows.empty())
|
|
{
|
|
const PDFTextFlow& textFlow = flows.front();
|
|
for (auto it = itBegin; it != itEnd; ++it)
|
|
{
|
|
text << textFlow.getText(it->start, it->end);
|
|
}
|
|
}
|
|
}
|
|
|
|
return text.join("\n");
|
|
}
|
|
|
|
QString PDFTextLayout::getTextFromSelection(const PDFTextSelection& selection, PDFInteger pageIndex) const
|
|
{
|
|
return getTextFromSelection(selection.begin(pageIndex), selection.end(pageIndex), pageIndex);
|
|
}
|
|
|
|
QDataStream& operator>>(QDataStream& stream, PDFTextLayout& layout)
|
|
{
|
|
stream >> layout.m_characters;
|
|
stream >> layout.m_angles;
|
|
stream >> layout.m_settings;
|
|
stream >> layout.m_blocks;
|
|
return stream;
|
|
}
|
|
|
|
QDataStream& operator<<(QDataStream& stream, const PDFTextLayout& layout)
|
|
{
|
|
stream << layout.m_characters;
|
|
stream << layout.m_angles;
|
|
stream << layout.m_settings;
|
|
stream << layout.m_blocks;
|
|
return stream;
|
|
}
|
|
|
|
struct NearestCharacterInfo
|
|
{
|
|
size_t index = std::numeric_limits<size_t>::max();
|
|
PDFReal distance = std::numeric_limits<PDFReal>::infinity();
|
|
|
|
inline bool operator<(const NearestCharacterInfo& other) const { return distance < other.distance; }
|
|
};
|
|
|
|
void PDFTextLayout::performDoLayout(PDFReal angle)
|
|
{
|
|
// We will implement variation of 'docstrum' algorithm, we have divided characters by angles,
|
|
// for each angle we get characters for that particular angle, and run 'docstrum' algorithm.
|
|
// We will do following steps:
|
|
// 1) Rotate the plane with characters so that they are all in horizontal line
|
|
// 2) Find k-nearest characters for each character (so each character will have
|
|
// k pointers to the nearest characters)
|
|
// 3) Find text lines. We will do that by creating transitive closure of characters, i.e.
|
|
// characters, which are close and are on horizontal line, are marked as in one text line.
|
|
// Consider also font size and empty space size between different characters.
|
|
// 4) Merge text lines into text blocks using various criteria, such as overlap,
|
|
// distance between the lines, and also using again, transitive closure.
|
|
// 5) Sort blocks using topological ordering
|
|
TextCharacters characters = getCharactersForAngle(angle);
|
|
|
|
// Step 1) - rotate blocks
|
|
QMatrix angleMatrix;
|
|
angleMatrix.rotate(angle);
|
|
applyTransform(characters, angleMatrix);
|
|
|
|
// Create spatial index
|
|
PDFTextCharacterSpatialIndex spatialIndex(&characters, 16);
|
|
for (size_t i = 0, count = characters.size(); i < count; ++i)
|
|
{
|
|
characters[i].index = i;
|
|
}
|
|
|
|
// Step 2) - find k-nearest characters
|
|
const size_t characterCount = characters.size();
|
|
const size_t bucketSize = m_settings.samples + 1;
|
|
std::vector<NearestCharacterInfo> nearestCharacters(bucketSize * characters.size(), NearestCharacterInfo());
|
|
|
|
auto findNearestCharacters = [this, bucketSize, &characters, &spatialIndex, &nearestCharacters](size_t currentCharacterIndex)
|
|
{
|
|
// It will be iterator to the start of the nearest neighbour sequence
|
|
auto it = std::next(nearestCharacters.begin(), currentCharacterIndex * bucketSize);
|
|
auto itLast = std::next(it, m_settings.samples);
|
|
NearestCharacterInfo& insertInfo = *itLast;
|
|
QPointF currentPoint = characters[currentCharacterIndex].position;
|
|
|
|
TextCharacters nearestPointSamples;
|
|
spatialIndex.queryNearestEstimate(m_settings.samples, characters[currentCharacterIndex], &nearestPointSamples);
|
|
for (size_t i = 0, count = nearestPointSamples.size(); i < count; ++i)
|
|
{
|
|
if (nearestPointSamples[i].index == currentCharacterIndex)
|
|
{
|
|
continue;
|
|
}
|
|
|
|
insertInfo.index = nearestPointSamples[i].index;
|
|
insertInfo.distance = QLineF(currentPoint, nearestPointSamples[i].position).length();
|
|
|
|
// Now, use insert sort to sort the array of samples + 1 elements (#samples elements
|
|
// are sorted, we use only insert sort on the last element).
|
|
auto itLeft = std::prev(itLast);
|
|
auto itRight = itLast;
|
|
while (true)
|
|
{
|
|
if (*itRight < *itLeft)
|
|
{
|
|
std::swap(*itRight, *itLeft);
|
|
itRight = itLeft;
|
|
|
|
if (itLeft == it)
|
|
{
|
|
// We have reached the end
|
|
break;
|
|
}
|
|
|
|
--itLeft;
|
|
}
|
|
else
|
|
{
|
|
// We have proper order, break the cycle
|
|
break;
|
|
}
|
|
}
|
|
}
|
|
};
|
|
|
|
auto range = PDFIntegerRange<size_t>(0, characterCount);
|
|
PDFExecutionPolicy::execute(PDFExecutionPolicy::Scope::Content, range.begin(), range.end(), findNearestCharacters);
|
|
|
|
// Step 3) - detect lines
|
|
PDFUnionFindAlgorithm<size_t> textLinesUF(characterCount);
|
|
for (size_t i = 0; i < characterCount; ++i)
|
|
{
|
|
auto it = std::next(nearestCharacters.begin(), i * bucketSize);
|
|
auto itEnd = std::next(it, m_settings.samples);
|
|
|
|
for (; it != itEnd; ++it)
|
|
{
|
|
const NearestCharacterInfo& info = *it;
|
|
if (info.index == std::numeric_limits<size_t>::max())
|
|
{
|
|
// We have reached the end - or we do not have enough characters
|
|
break;
|
|
}
|
|
|
|
// Criteria:
|
|
// 1) Distance of characters is not too large
|
|
// 2) Characters are approximately at same line
|
|
// 3) Font size of characters are approximately equal
|
|
|
|
PDFReal fontSizeMax = qMax(characters[i].fontSize, characters[info.index].fontSize);
|
|
PDFReal fontSizeMin = qMin(characters[i].fontSize, characters[info.index].fontSize);
|
|
|
|
if (info.distance < m_settings.distanceSensitivity * characters[i].advance && // 1)
|
|
std::fabs(characters[i].position.y() - characters[info.index].position.y()) < fontSizeMin * m_settings.charactersOnLineSensitivity && // 2)
|
|
fontSizeMax / fontSizeMin < m_settings.fontSensitivity) // 3)
|
|
{
|
|
textLinesUF.unify(i, info.index);
|
|
}
|
|
}
|
|
}
|
|
|
|
std::map<size_t, TextCharacters> lineToCharactersMap;
|
|
for (size_t i = 0; i < characterCount; ++i)
|
|
{
|
|
lineToCharactersMap[textLinesUF.find(i)].push_back(characters[i]);
|
|
}
|
|
|
|
PDFTextLines lines;
|
|
lines.reserve(lineToCharactersMap.size());
|
|
for (auto& item : lineToCharactersMap)
|
|
{
|
|
lines.emplace_back(qMove(item.second));
|
|
}
|
|
|
|
// Step 4) - detect text blocks
|
|
const size_t lineCount = lines.size();
|
|
PDFUnionFindAlgorithm<size_t> textBlocksUF(lineCount);
|
|
for (size_t i = 0; i < lineCount; ++i)
|
|
{
|
|
for (size_t j = i + 1; j < lineCount; ++j)
|
|
{
|
|
QRectF bb1 = lines[i].getBoundingBox().boundingRect();
|
|
QRectF bb2 = lines[j].getBoundingBox().boundingRect();
|
|
|
|
// Jakub Melka: we will join two blocks, if these two conditions both holds:
|
|
// 1) bounding boxes overlap horizontally by large portion
|
|
// 2) vertical space between bounding boxes is not too large
|
|
|
|
QRectF bbUnion = bb1.united(bb2);
|
|
const PDFReal height = bbUnion.height();
|
|
const PDFReal heightLimit = (bb1.height() + bb2.height()) * m_settings.blockVerticalSensitivity;
|
|
const PDFReal overlap = qMax(0.0, bb1.width() + bb2.width() - bbUnion.width());
|
|
const PDFReal minimalOverlap = qMin(bb1.width(), bb2.width()) * m_settings.blockOverlapSensitivity;
|
|
if (height < heightLimit && overlap > minimalOverlap)
|
|
{
|
|
textBlocksUF.unify(i, j);
|
|
}
|
|
}
|
|
}
|
|
|
|
std::map<size_t, PDFTextLines> blockToLines;
|
|
for (size_t i = 0; i < lineCount; ++i)
|
|
{
|
|
blockToLines[textBlocksUF.find(i)].push_back(qMove(lines[i]));
|
|
}
|
|
|
|
PDFTextBlocks blocks;
|
|
blocks.reserve(blockToLines.size());
|
|
for (auto& item : blockToLines)
|
|
{
|
|
blocks.emplace_back(qMove(item.second));
|
|
}
|
|
|
|
// 5) Sort block by topological ordering. We will use approache described in paper
|
|
// "High Performance Document Layout Analysis", T.M. Breuel, 2003, where are described
|
|
// two rules, which are used to determine block precedence.
|
|
//
|
|
// Rule 1: a<b, if:
|
|
// - blocks a,b have overlap in x-axis
|
|
// - block a is above block b
|
|
//
|
|
// Rule 2: a<b, if:
|
|
// - block a is entirely on left side of block b
|
|
// - there doesn't exist block c, which is between a,b in y-axis
|
|
// and moreover, overlaps both a and b in x-axis.
|
|
|
|
auto isBeforeByRule1 = [&blocks](const size_t aIndex, const size_t bIndex)
|
|
{
|
|
QRectF aBB = blocks[aIndex].getBoundingBox().boundingRect();
|
|
QRectF bBB = blocks[bIndex].getBoundingBox().boundingRect();
|
|
|
|
const bool isOverlappedOnHorizontalAxis = isRectangleHorizontallyOverlapped(aBB, bBB);
|
|
const bool isAoverB = aBB.bottom() > bBB.top();
|
|
return isOverlappedOnHorizontalAxis && isAoverB;
|
|
};
|
|
auto isBeforeByRule2 = [&blocks](const size_t aIndex, const size_t bIndex)
|
|
{
|
|
QRectF aBB = blocks[aIndex].getBoundingBox().boundingRect();
|
|
QRectF bBB = blocks[bIndex].getBoundingBox().boundingRect();
|
|
QRectF abBB = aBB.united(bBB);
|
|
|
|
if (aBB.right() < bBB.left())
|
|
{
|
|
// Check, if 'c' block doesn't exist
|
|
for (size_t i = 0, count = blocks.size(); i < count; ++i)
|
|
{
|
|
if (i == aIndex || i == bIndex)
|
|
{
|
|
continue;
|
|
}
|
|
|
|
QRectF cBB = blocks[i].getBoundingBox().boundingRect();
|
|
if (cBB.top() >= abBB.top() && cBB.bottom() <= abBB.bottom())
|
|
{
|
|
const bool isAOverlappedOnHorizontalAxis = isRectangleHorizontallyOverlapped(aBB, cBB);
|
|
const bool isBOverlappedOnHorizontalAxis = isRectangleHorizontallyOverlapped(bBB, cBB);
|
|
if (isAOverlappedOnHorizontalAxis && isBOverlappedOnHorizontalAxis)
|
|
{
|
|
return false;
|
|
}
|
|
}
|
|
}
|
|
|
|
return true;
|
|
}
|
|
|
|
return false;
|
|
};
|
|
|
|
// Order blocks using topological sort (https://en.wikipedia.org/wiki/Topological_sorting,
|
|
// Kahn's algorithm is used)
|
|
std::set<size_t> workBlocks;
|
|
std::vector<size_t> ordering;
|
|
std::vector<std::set<size_t>> orderingEdges(blocks.size(), std::set<size_t>());
|
|
ordering.reserve(blocks.size());
|
|
for (size_t i = 0; i < blocks.size(); ++i)
|
|
{
|
|
workBlocks.insert(workBlocks.end(), i);
|
|
for (size_t j = 0; j < blocks.size(); ++j)
|
|
{
|
|
if (i != j && (isBeforeByRule1(j, i) || isBeforeByRule2(j, i)))
|
|
{
|
|
orderingEdges[i].insert(j);
|
|
}
|
|
}
|
|
}
|
|
|
|
// Topological sort
|
|
QMatrix invertedAngleMatrix = angleMatrix.inverted();
|
|
while (!workBlocks.empty())
|
|
{
|
|
auto it = std::min_element(workBlocks.begin(), workBlocks.end(), [&orderingEdges](const size_t l, const size_t r) { return orderingEdges[l].size() < orderingEdges[r].size(); });
|
|
ordering.push_back(*it);
|
|
for (std::set<size_t>& edges : orderingEdges)
|
|
{
|
|
edges.erase(*it);
|
|
}
|
|
|
|
blocks[*it].applyTransform(invertedAngleMatrix);
|
|
m_blocks.emplace_back(qMove(blocks[*it]));
|
|
workBlocks.erase(it);
|
|
}
|
|
}
|
|
|
|
TextCharacters PDFTextLayout::getCharactersForAngle(PDFReal angle) const
|
|
{
|
|
TextCharacters result;
|
|
std::copy_if(m_characters.cbegin(), m_characters.cend(), std::back_inserter(result), [angle](const TextCharacter& character) { return character.angle == angle; });
|
|
return result;
|
|
}
|
|
|
|
void PDFTextLayout::applyTransform(TextCharacters& characters, const QMatrix& matrix)
|
|
{
|
|
for (TextCharacter& character : characters)
|
|
{
|
|
character.position = matrix.map(character.position);
|
|
character.boundingBox = matrix.map(character.boundingBox);
|
|
}
|
|
}
|
|
|
|
PDFTextLine::PDFTextLine(TextCharacters characters) :
|
|
m_characters(qMove(characters))
|
|
{
|
|
std::sort(m_characters.begin(), m_characters.end(), [](const TextCharacter& l, const TextCharacter& r) { return l.position.x() < r.position.x(); });
|
|
|
|
QRectF boundingBox;
|
|
for (const TextCharacter& character : m_characters)
|
|
{
|
|
boundingBox = boundingBox.united(character.boundingBox.boundingRect());
|
|
}
|
|
m_boundingBox.addRect(boundingBox);
|
|
m_topLeft = boundingBox.topLeft();
|
|
}
|
|
|
|
PDFReal PDFTextLine::getAngle() const
|
|
{
|
|
if (!m_characters.empty())
|
|
{
|
|
return m_characters.front().angle;
|
|
}
|
|
|
|
return 0.0;
|
|
}
|
|
|
|
void PDFTextLine::applyTransform(const QMatrix& matrix)
|
|
{
|
|
m_boundingBox = matrix.map(m_boundingBox);
|
|
m_topLeft = matrix.map(m_topLeft);
|
|
for (TextCharacter& character : m_characters)
|
|
{
|
|
character.applyTransform(matrix);
|
|
}
|
|
}
|
|
|
|
QDataStream& operator>>(QDataStream& stream, PDFTextLine& line)
|
|
{
|
|
stream >> line.m_characters;
|
|
stream >> line.m_boundingBox;
|
|
stream >> line.m_topLeft;
|
|
return stream;
|
|
}
|
|
|
|
QDataStream& operator<<(QDataStream& stream, const PDFTextLine& line)
|
|
{
|
|
stream << line.m_characters;
|
|
stream << line.m_boundingBox;
|
|
stream << line.m_topLeft;
|
|
return stream;
|
|
}
|
|
|
|
PDFTextBlock::PDFTextBlock(PDFTextLines textLines) :
|
|
m_lines(qMove(textLines))
|
|
{
|
|
auto sortFunction = [](const PDFTextLine& l, const PDFTextLine& r)
|
|
{
|
|
QRectF bl = l.getBoundingBox().boundingRect();
|
|
QRectF br = r.getBoundingBox().boundingRect();
|
|
const PDFReal xL = bl.x();
|
|
const PDFReal xR = br.x();
|
|
const PDFReal yL = qRound(bl.y() * 100.0);
|
|
const PDFReal yR = qRound(br.y() * 100.0);
|
|
return std::make_pair(-yL, xL) < std::make_pair(-yR, xR);
|
|
};
|
|
std::sort(m_lines.begin(), m_lines.end(), sortFunction);
|
|
|
|
QRectF boundingBox;
|
|
for (const PDFTextLine& line : m_lines)
|
|
{
|
|
boundingBox = boundingBox.united(line.getBoundingBox().boundingRect());
|
|
}
|
|
m_boundingBox.addRect(boundingBox);
|
|
m_topLeft = boundingBox.topLeft();
|
|
}
|
|
|
|
PDFReal PDFTextBlock::getAngle() const
|
|
{
|
|
if (!m_lines.empty())
|
|
{
|
|
return m_lines.front().getAngle();
|
|
}
|
|
|
|
return 0.0;
|
|
}
|
|
|
|
void PDFTextBlock::applyTransform(const QMatrix& matrix)
|
|
{
|
|
m_boundingBox = matrix.map(m_boundingBox);
|
|
m_topLeft = matrix.map(m_topLeft);
|
|
for (PDFTextLine& textLine : m_lines)
|
|
{
|
|
textLine.applyTransform(matrix);
|
|
}
|
|
}
|
|
|
|
QDataStream& operator>>(QDataStream& stream, PDFTextBlock& block)
|
|
{
|
|
stream >> block.m_lines;
|
|
stream >> block.m_boundingBox;
|
|
stream >> block.m_topLeft;
|
|
return stream;
|
|
}
|
|
|
|
QDataStream& operator<<(QDataStream& stream, const PDFTextBlock& block)
|
|
{
|
|
stream << block.m_lines;
|
|
stream << block.m_boundingBox;
|
|
stream << block.m_topLeft;
|
|
return stream;
|
|
}
|
|
|
|
void TextCharacter::applyTransform(const QMatrix& matrix)
|
|
{
|
|
position = matrix.map(position);
|
|
boundingBox = matrix.map(boundingBox);
|
|
}
|
|
|
|
QDataStream& operator<<(QDataStream& stream, const TextCharacter& character)
|
|
{
|
|
stream << character.character;
|
|
stream << character.position;
|
|
stream << character.angle;
|
|
stream << character.fontSize;
|
|
stream << character.advance;
|
|
stream << character.boundingBox;
|
|
return stream;
|
|
}
|
|
|
|
QDataStream& operator>>(QDataStream& stream, TextCharacter& character)
|
|
{
|
|
stream >> character.character;
|
|
stream >> character.position;
|
|
stream >> character.angle;
|
|
stream >> character.fontSize;
|
|
stream >> character.advance;
|
|
stream >> character.boundingBox;
|
|
return stream;
|
|
}
|
|
|
|
PDFTextLayout PDFTextLayoutStorage::getTextLayout(PDFInteger pageIndex) const
|
|
{
|
|
PDFTextLayout result;
|
|
|
|
if (pageIndex >= 0 && pageIndex < static_cast<PDFInteger>(m_offsets.size()))
|
|
{
|
|
QDataStream layoutStream(const_cast<QByteArray*>(&m_textLayouts), QIODevice::ReadOnly);
|
|
layoutStream.skipRawData(m_offsets[pageIndex]);
|
|
|
|
QByteArray buffer;
|
|
layoutStream >> buffer;
|
|
buffer = qUncompress(buffer);
|
|
|
|
QDataStream stream(&buffer, QIODevice::ReadOnly);
|
|
stream >> result;
|
|
}
|
|
|
|
return result;
|
|
}
|
|
|
|
void PDFTextLayoutStorage::setTextLayout(PDFInteger pageIndex, const PDFTextLayout& layout, QMutex* mutex)
|
|
{
|
|
QByteArray result;
|
|
{
|
|
QDataStream stream(&result, QIODevice::WriteOnly);
|
|
stream << layout;
|
|
}
|
|
result = qCompress(result, 9);
|
|
|
|
QMutexLocker lock(mutex);
|
|
m_offsets[pageIndex] = m_textLayouts.size();
|
|
|
|
QDataStream layoutStream(&m_textLayouts, QIODevice::Append | QIODevice::WriteOnly);
|
|
layoutStream << result;
|
|
}
|
|
|
|
PDFFindResults PDFTextLayoutStorage::find(const QString& text, Qt::CaseSensitivity caseSensitivity, PDFTextFlow::FlowFlags flowFlags) const
|
|
{
|
|
PDFFindResults results;
|
|
|
|
QMutex resultsMutex;
|
|
auto findImpl = [this, flowFlags, caseSensitivity, &results, &resultsMutex, &text](size_t pageIndex)
|
|
{
|
|
PDFTextLayout textLayout = getTextLayout(pageIndex);
|
|
PDFTextFlows textFlows = PDFTextFlow::createTextFlows(textLayout, flowFlags, pageIndex);
|
|
for (const PDFTextFlow& textFlow : textFlows)
|
|
{
|
|
PDFFindResults flowResults = textFlow.find(text, caseSensitivity);
|
|
|
|
// Jakub Melka: Do not lock mutex, if we didn't find anything. In that case, just skip to next flow.
|
|
if (!flowResults.empty())
|
|
{
|
|
QMutexLocker lock(&resultsMutex);
|
|
results.insert(results.end(), flowResults.begin(), flowResults.end());
|
|
}
|
|
}
|
|
};
|
|
|
|
auto range = PDFIntegerRange<size_t>(0, m_offsets.size());
|
|
PDFExecutionPolicy::execute(PDFExecutionPolicy::Scope::Page, range.begin(), range.end(), findImpl);
|
|
|
|
std::sort(results.begin(), results.end());
|
|
return results;
|
|
}
|
|
|
|
PDFFindResults PDFTextLayoutStorage::find(const QRegularExpression& expression, PDFTextFlow::FlowFlags flowFlags) const
|
|
{
|
|
PDFFindResults results;
|
|
|
|
QMutex resultsMutex;
|
|
auto findImpl = [this, flowFlags, &results, &resultsMutex, &expression](size_t pageIndex)
|
|
{
|
|
PDFTextLayout textLayout = getTextLayout(pageIndex);
|
|
PDFTextFlows textFlows = PDFTextFlow::createTextFlows(textLayout, flowFlags, pageIndex);
|
|
for (const PDFTextFlow& textFlow : textFlows)
|
|
{
|
|
PDFFindResults flowResults = textFlow.find(expression);
|
|
|
|
// Jakub Melka: Do not lock mutex, if we didn't find anything. In that case, just skip to next flow.
|
|
if (!flowResults.empty())
|
|
{
|
|
QMutexLocker lock(&resultsMutex);
|
|
results.insert(results.end(), flowResults.begin(), flowResults.end());
|
|
}
|
|
}
|
|
};
|
|
|
|
auto range = PDFIntegerRange<size_t>(0, m_offsets.size());
|
|
PDFExecutionPolicy::execute(PDFExecutionPolicy::Scope::Page, range.begin(), range.end(), findImpl);
|
|
|
|
std::sort(results.begin(), results.end());
|
|
return results;
|
|
}
|
|
|
|
QDataStream& operator<<(QDataStream& stream, const PDFTextLayoutSettings& settings)
|
|
{
|
|
stream << settings.samples;
|
|
stream << settings.distanceSensitivity;
|
|
stream << settings.charactersOnLineSensitivity;
|
|
stream << settings.fontSensitivity;
|
|
stream << settings.blockVerticalSensitivity;
|
|
stream << settings.blockOverlapSensitivity;
|
|
return stream;
|
|
}
|
|
|
|
QDataStream& operator>>(QDataStream& stream, PDFTextLayoutSettings& settings)
|
|
{
|
|
stream >> settings.samples;
|
|
stream >> settings.distanceSensitivity;
|
|
stream >> settings.charactersOnLineSensitivity;
|
|
stream >> settings.fontSensitivity;
|
|
stream >> settings.blockVerticalSensitivity;
|
|
stream >> settings.blockOverlapSensitivity;
|
|
return stream;
|
|
}
|
|
|
|
void PDFTextSelection::addItems(const PDFTextSelectionItems& items, QColor color)
|
|
{
|
|
std::transform(items.cbegin(), items.cend(), std::back_inserter(m_items), [color] (const auto& item) { return PDFTextSelectionColoredItem(item.first, item.second, color); });
|
|
}
|
|
|
|
void PDFTextSelection::build()
|
|
{
|
|
std::sort(m_items.begin(), m_items.end());
|
|
}
|
|
|
|
PDFTextSelection::iterator PDFTextSelection::begin(PDFInteger pageIndex) const
|
|
{
|
|
Q_ASSERT(std::is_sorted(m_items.cbegin(), m_items.end()));
|
|
|
|
PDFCharacterPointer pointer;
|
|
pointer.pageIndex = pageIndex;
|
|
pointer.blockIndex = 0;
|
|
pointer.lineIndex = 0;
|
|
pointer.characterIndex = 0;
|
|
|
|
PDFTextSelectionColoredItem item;
|
|
item.start = pointer;
|
|
item.end = pointer;
|
|
|
|
return std::lower_bound(m_items.cbegin(), m_items.end(), item);
|
|
}
|
|
|
|
PDFTextSelection::iterator PDFTextSelection::end(PDFInteger pageIndex) const
|
|
{
|
|
Q_ASSERT(std::is_sorted(m_items.cbegin(), m_items.end()));
|
|
|
|
PDFCharacterPointer pointer;
|
|
pointer.pageIndex = pageIndex;
|
|
pointer.blockIndex = std::numeric_limits<decltype(pointer.blockIndex)>::max();
|
|
pointer.lineIndex = std::numeric_limits<decltype(pointer.lineIndex)>::max();
|
|
pointer.characterIndex = std::numeric_limits<decltype(pointer.characterIndex)>::max();
|
|
|
|
PDFTextSelectionColoredItem item;
|
|
item.start = pointer;
|
|
item.end = pointer;
|
|
|
|
return std::upper_bound(m_items.cbegin(), m_items.end(), item);
|
|
}
|
|
|
|
PDFTextSelection::iterator PDFTextSelection::nextPageRange(iterator currentPageRange) const
|
|
{
|
|
auto it = currentPageRange;
|
|
while (it != m_items.cend() && it->start.pageIndex == currentPageRange->start.pageIndex)
|
|
{
|
|
++it;
|
|
}
|
|
|
|
return it;
|
|
}
|
|
|
|
PDFFindResults PDFTextFlow::find(const QString& text, Qt::CaseSensitivity caseSensitivity) const
|
|
{
|
|
PDFFindResults results;
|
|
|
|
int index = m_text.indexOf(text, 0, caseSensitivity);
|
|
while (index != -1)
|
|
{
|
|
PDFFindResult result;
|
|
result.matched = text;
|
|
result.textSelectionItems = getTextSelectionItems(index, text.length());
|
|
result.context = getContext(index, text.length());
|
|
|
|
if (!result.textSelectionItems.empty())
|
|
{
|
|
results.emplace_back(qMove(result));
|
|
}
|
|
|
|
index = m_text.indexOf(text, index + 1, caseSensitivity);
|
|
}
|
|
|
|
return results;
|
|
}
|
|
|
|
PDFFindResults PDFTextFlow::find(const QRegularExpression& expression) const
|
|
{
|
|
PDFFindResults results;
|
|
|
|
QRegularExpressionMatchIterator iterator = expression.globalMatch(m_text, 0, QRegularExpression::NormalMatch, QRegularExpression::NoMatchOption);
|
|
while (iterator.hasNext())
|
|
{
|
|
QRegularExpressionMatch match = iterator.next();
|
|
|
|
Q_ASSERT(match.hasMatch());
|
|
const int index = match.capturedStart();
|
|
const int length = match.capturedLength();
|
|
|
|
PDFFindResult result;
|
|
result.matched = match.captured();
|
|
result.textSelectionItems = getTextSelectionItems(index, length);
|
|
result.context = getContext(index, length);
|
|
|
|
if (!result.textSelectionItems.empty())
|
|
{
|
|
results.emplace_back(qMove(result));
|
|
}
|
|
}
|
|
|
|
return results;
|
|
}
|
|
|
|
QString PDFTextFlow::getText(const PDFCharacterPointer& begin, const PDFCharacterPointer& end) const
|
|
{
|
|
auto it = std::find(m_characterPointers.cbegin(), m_characterPointers.cend(), begin);
|
|
auto itEnd = std::find(m_characterPointers.cbegin(), m_characterPointers.cend(), end);
|
|
|
|
const std::size_t startIndex = std::distance(m_characterPointers.cbegin(), it);
|
|
const std::size_t endIndex = std::distance(m_characterPointers.cbegin(), itEnd);
|
|
if (startIndex <= endIndex)
|
|
{
|
|
return m_text.mid(int(startIndex), int(endIndex - startIndex + 1));
|
|
}
|
|
|
|
return QString();
|
|
}
|
|
|
|
void PDFTextFlow::merge(const PDFTextFlow& next)
|
|
{
|
|
m_text += next.m_text;
|
|
m_boundingBox = m_boundingBox.united(next.m_boundingBox);
|
|
m_characterPointers.insert(m_characterPointers.end(), next.m_characterPointers.cbegin(), next.m_characterPointers.cend());
|
|
}
|
|
|
|
PDFTextFlows PDFTextFlow::createTextFlows(const PDFTextLayout& layout, FlowFlags flags, PDFInteger pageIndex)
|
|
{
|
|
PDFTextFlows result;
|
|
|
|
if (!flags.testFlag(SeparateBlocks))
|
|
{
|
|
result.emplace_back();
|
|
}
|
|
|
|
QString lineBreak(" ");
|
|
if (flags.testFlag(AddLineBreaks))
|
|
{
|
|
#if defined(Q_OS_WIN)
|
|
lineBreak = QString("\r\n");
|
|
#elif defined(Q_OS_UNIX)
|
|
linebreak = QString("\n");
|
|
#elif defined(Q_OS_MAC)
|
|
lineBreak = QString("\r");
|
|
#else
|
|
static_assert(false, "Fix this code!");
|
|
#endif
|
|
}
|
|
|
|
size_t textBlockIndex = 0;
|
|
for (const PDFTextBlock& textBlock : layout.getTextBlocks())
|
|
{
|
|
PDFTextFlow currentFlow;
|
|
currentFlow.m_boundingBox = textBlock.getBoundingBox().controlPointRect();
|
|
|
|
size_t textLineIndex = 0;
|
|
for (const PDFTextLine& textLine : textBlock.getLines())
|
|
{
|
|
const TextCharacters& characters = textLine.getCharacters();
|
|
for (size_t i = 0, characterCount = characters.size(); i < characterCount; ++i)
|
|
{
|
|
const TextCharacter& currentCharacter = characters[i];
|
|
if (i > 0 && !currentCharacter.character.isSpace())
|
|
{
|
|
// Jakub Melka: try to guess space between letters
|
|
const TextCharacter& previousCharacter = characters[i - 1];
|
|
if (!previousCharacter.character.isSpace() && QLineF(previousCharacter.position, currentCharacter.position).length() > previousCharacter.advance * 1.2)
|
|
{
|
|
currentFlow.m_text += QChar(' ');
|
|
currentFlow.m_characterPointers.emplace_back();
|
|
}
|
|
}
|
|
|
|
currentFlow.m_text += currentCharacter.character;
|
|
|
|
PDFCharacterPointer pointer;
|
|
pointer.pageIndex = pageIndex;
|
|
pointer.blockIndex = textBlockIndex;
|
|
pointer.lineIndex = textLineIndex;
|
|
pointer.characterIndex = i;
|
|
currentFlow.m_characterPointers.emplace_back(qMove(pointer));
|
|
}
|
|
|
|
// Remove soft hyphen, if it is enabled
|
|
if (flags.testFlag(RemoveSoftHyphen) && !characters.empty() && currentFlow.m_text.back() == QChar(QChar::SoftHyphen))
|
|
{
|
|
currentFlow.m_text.chop(1);
|
|
currentFlow.m_characterPointers.pop_back();
|
|
|
|
if (!flags.testFlag(AddLineBreaks))
|
|
{
|
|
// Do not add single empty space - because soft hypen probably breaks a word
|
|
++textLineIndex;
|
|
continue;
|
|
}
|
|
}
|
|
|
|
// Add line break
|
|
currentFlow.m_text += lineBreak;
|
|
currentFlow.m_characterPointers.insert(currentFlow.m_characterPointers.end(), lineBreak.length(), PDFCharacterPointer());
|
|
|
|
++textLineIndex;
|
|
}
|
|
|
|
// If we are producing separate blocks, then make flow for each
|
|
// text block, otherwise join flows.
|
|
if (flags.testFlag(SeparateBlocks))
|
|
{
|
|
result.emplace_back(qMove(currentFlow));
|
|
}
|
|
else
|
|
{
|
|
result.back().merge(currentFlow);
|
|
}
|
|
|
|
++textBlockIndex;
|
|
}
|
|
|
|
return result;
|
|
}
|
|
|
|
PDFTextSelectionItems PDFTextFlow::getTextSelectionItems(size_t index, size_t length) const
|
|
{
|
|
PDFTextSelectionItems items;
|
|
|
|
auto it = std::next(m_characterPointers.cbegin(), index);
|
|
auto itEnd = std::next(m_characterPointers.cbegin(), index + length);
|
|
while (it != itEnd)
|
|
{
|
|
// Skip invalid items, find first valid
|
|
if (!it->isValid())
|
|
{
|
|
++it;
|
|
continue;
|
|
}
|
|
|
|
auto itSelectionStart = it;
|
|
while (it != itEnd && it->isValid() && it->hasSameBlock(*itSelectionStart))
|
|
{
|
|
++it;
|
|
}
|
|
auto itSelectionEnd = std::prev(it);
|
|
items.emplace_back(*itSelectionStart, *itSelectionEnd);
|
|
}
|
|
|
|
std::sort(items.begin(), items.end());
|
|
return items;
|
|
}
|
|
|
|
QString PDFTextFlow::getContext(size_t index, size_t length) const
|
|
{
|
|
Q_ASSERT(length > 0);
|
|
|
|
const PDFCharacterPointer& frontComparatorItem = m_characterPointers[index];
|
|
while (index > 0 && (m_characterPointers[index - 1].hasSameLine(frontComparatorItem) || !m_characterPointers[index - 1].isValid()))
|
|
{
|
|
--index;
|
|
++length;
|
|
}
|
|
|
|
size_t currentEnd = index + length - 1;
|
|
size_t last = m_characterPointers.size() - 1;
|
|
const PDFCharacterPointer& backComparatorItem = m_characterPointers[currentEnd];
|
|
while (currentEnd < last && (m_characterPointers[currentEnd + 1].hasSameLine(backComparatorItem) || !m_characterPointers[currentEnd + 1].isValid()))
|
|
{
|
|
++currentEnd;
|
|
++length;
|
|
}
|
|
|
|
return m_text.mid(int(index), int(length)).trimmed();
|
|
}
|
|
|
|
bool PDFCharacterPointer::hasSameBlock(const PDFCharacterPointer& other) const
|
|
{
|
|
return pageIndex == other.pageIndex && blockIndex == other.blockIndex;
|
|
}
|
|
|
|
bool PDFCharacterPointer::hasSameLine(const PDFCharacterPointer& other) const
|
|
{
|
|
return hasSameBlock(other) && lineIndex == other.lineIndex;
|
|
}
|
|
|
|
bool PDFFindResult::operator<(const PDFFindResult& other) const
|
|
{
|
|
Q_ASSERT(!textSelectionItems.empty());
|
|
Q_ASSERT(!other.textSelectionItems.empty());
|
|
|
|
return textSelectionItems.front() < other.textSelectionItems.front();
|
|
}
|
|
|
|
PDFTextLayout PDFTextLayoutStorageGetter::getTextLayoutImpl() const
|
|
{
|
|
return m_storage ? m_storage->getTextLayout(m_pageIndex) : PDFTextLayout();
|
|
}
|
|
|
|
void PDFTextSelectionPainter::draw(QPainter* painter, PDFInteger pageIndex, PDFTextLayoutGetter& textLayoutGetter, const QMatrix& matrix)
|
|
{
|
|
Q_ASSERT(painter);
|
|
|
|
auto it = m_selection->begin(pageIndex);
|
|
auto itEnd = m_selection->end(pageIndex);
|
|
|
|
if (it == itEnd)
|
|
{
|
|
// Jakub Melka: no text is selected on current page; do nothing
|
|
return;
|
|
}
|
|
|
|
painter->save();
|
|
|
|
const PDFTextLayout& layout = textLayoutGetter;
|
|
const PDFTextBlocks& blocks = layout.getTextBlocks();
|
|
for (; it != itEnd; ++it)
|
|
{
|
|
const PDFTextSelectionColoredItem& item = *it;
|
|
const PDFCharacterPointer& start = item.start;
|
|
const PDFCharacterPointer& end = item.end;
|
|
|
|
Q_ASSERT(start.pageIndex == end.pageIndex);
|
|
Q_ASSERT(start.blockIndex == end.blockIndex);
|
|
|
|
if (start.blockIndex >= blocks.size())
|
|
{
|
|
// Selection is invalid, do nothing
|
|
continue;
|
|
}
|
|
|
|
PDFTextBlock block = blocks[start.blockIndex];
|
|
|
|
// Fix angle of block, so we will get correct selection rectangles (parallel to lines)
|
|
QMatrix angleMatrix;
|
|
angleMatrix.rotate(block.getAngle());
|
|
block.applyTransform(angleMatrix);
|
|
|
|
QPainterPath path;
|
|
|
|
const size_t lineStart = start.lineIndex;
|
|
const size_t lineEnd = end.lineIndex;
|
|
Q_ASSERT(lineEnd >= lineStart);
|
|
|
|
const PDFTextLines& lines = block.getLines();
|
|
for (size_t lineIndex = lineStart; lineIndex <= lineEnd; ++lineIndex)
|
|
{
|
|
if (lineIndex >= lines.size())
|
|
{
|
|
// Selection is invalid, do nothing
|
|
continue;
|
|
}
|
|
|
|
const PDFTextLine& line = lines[lineIndex];
|
|
const TextCharacters& characters = line.getCharacters();
|
|
|
|
if (characters.empty())
|
|
{
|
|
// Selection is invalid, do nothing
|
|
continue;
|
|
}
|
|
|
|
// First determine, which characters will be selected
|
|
size_t characterStart = 0;
|
|
size_t characterEnd = characters.size() - 1;
|
|
|
|
if (lineIndex == lineStart)
|
|
{
|
|
characterStart = start.characterIndex;
|
|
}
|
|
if (lineIndex == lineEnd)
|
|
{
|
|
characterEnd = end.characterIndex;
|
|
}
|
|
|
|
// Validate indices, then calculate bounding box
|
|
if (!(characterStart <= characterEnd && characterEnd < characters.size()))
|
|
{
|
|
continue;
|
|
}
|
|
|
|
QRectF boundingBox;
|
|
for (size_t i = characterStart; i <= characterEnd; ++i)
|
|
{
|
|
boundingBox = boundingBox.united(characters[i].boundingBox.boundingRect());
|
|
}
|
|
|
|
if (boundingBox.isValid())
|
|
{
|
|
// Enlarge height by some percent
|
|
PDFReal heightAdvance = boundingBox.height() * HEIGHT_INCREASE_FACTOR * 0.5;
|
|
boundingBox.adjust(0, -heightAdvance, 0, heightAdvance);
|
|
path.addRect(boundingBox);
|
|
}
|
|
}
|
|
|
|
QMatrix transformMatrix = angleMatrix.inverted() * matrix;
|
|
path = transformMatrix.map(path);
|
|
|
|
QColor penColor = item.color.darker();
|
|
QColor brushColor = item.color;
|
|
brushColor.setAlphaF(SELECTION_ALPHA);
|
|
|
|
painter->setPen(penColor);
|
|
painter->setBrush(QBrush(brushColor, Qt::SolidPattern));
|
|
painter->drawPath(path);
|
|
}
|
|
|
|
painter->restore();
|
|
}
|
|
|
|
QPainterPath PDFTextSelectionPainter::prepareGeometry(PDFInteger pageIndex, PDFTextLayoutGetter& textLayoutGetter, const QMatrix& matrix, QPolygonF* quadrilaterals)
|
|
{
|
|
QPainterPath path;
|
|
|
|
auto it = m_selection->begin(pageIndex);
|
|
auto itEnd = m_selection->end(pageIndex);
|
|
|
|
if (it == itEnd)
|
|
{
|
|
// Jakub Melka: no text is selected on current page; do nothing
|
|
return path;
|
|
}
|
|
|
|
const PDFTextLayout& layout = textLayoutGetter;
|
|
const PDFTextBlocks& blocks = layout.getTextBlocks();
|
|
for (; it != itEnd; ++it)
|
|
{
|
|
const PDFTextSelectionColoredItem& item = *it;
|
|
const PDFCharacterPointer& start = item.start;
|
|
const PDFCharacterPointer& end = item.end;
|
|
|
|
Q_ASSERT(start.pageIndex == end.pageIndex);
|
|
Q_ASSERT(start.blockIndex == end.blockIndex);
|
|
|
|
if (start.blockIndex >= blocks.size())
|
|
{
|
|
// Selection is invalid, do nothing
|
|
continue;
|
|
}
|
|
|
|
PDFTextBlock block = blocks[start.blockIndex];
|
|
|
|
// Fix angle of block, so we will get correct selection rectangles (parallel to lines)
|
|
QMatrix angleMatrix;
|
|
angleMatrix.rotate(block.getAngle());
|
|
block.applyTransform(angleMatrix);
|
|
|
|
QPainterPath currentPath;
|
|
QPolygonF currentPolygon;
|
|
|
|
const size_t lineStart = start.lineIndex;
|
|
const size_t lineEnd = end.lineIndex;
|
|
Q_ASSERT(lineEnd >= lineStart);
|
|
|
|
const PDFTextLines& lines = block.getLines();
|
|
for (size_t lineIndex = lineStart; lineIndex <= lineEnd; ++lineIndex)
|
|
{
|
|
if (lineIndex >= lines.size())
|
|
{
|
|
// Selection is invalid, do nothing
|
|
continue;
|
|
}
|
|
|
|
const PDFTextLine& line = lines[lineIndex];
|
|
const TextCharacters& characters = line.getCharacters();
|
|
|
|
if (characters.empty())
|
|
{
|
|
// Selection is invalid, do nothing
|
|
continue;
|
|
}
|
|
|
|
// First determine, which characters will be selected
|
|
size_t characterStart = 0;
|
|
size_t characterEnd = characters.size() - 1;
|
|
|
|
if (lineIndex == lineStart)
|
|
{
|
|
characterStart = start.characterIndex;
|
|
}
|
|
if (lineIndex == lineEnd)
|
|
{
|
|
characterEnd = end.characterIndex;
|
|
}
|
|
|
|
// Validate indices, then calculate bounding box
|
|
if (!(characterStart <= characterEnd && characterEnd < characters.size()))
|
|
{
|
|
continue;
|
|
}
|
|
|
|
QRectF boundingBox;
|
|
for (size_t i = characterStart; i <= characterEnd; ++i)
|
|
{
|
|
boundingBox = boundingBox.united(characters[i].boundingBox.boundingRect());
|
|
}
|
|
|
|
if (boundingBox.isValid())
|
|
{
|
|
// Enlarge height by some percent
|
|
PDFReal heightAdvance = boundingBox.height() * HEIGHT_INCREASE_FACTOR * 0.5;
|
|
boundingBox.adjust(0, -heightAdvance, 0, heightAdvance);
|
|
currentPath.addRect(boundingBox);
|
|
|
|
if (quadrilaterals)
|
|
{
|
|
currentPolygon.append({ boundingBox.topLeft(), boundingBox.topRight(), boundingBox.bottomLeft(), boundingBox.bottomRight() });
|
|
}
|
|
}
|
|
}
|
|
|
|
QMatrix transformMatrix = angleMatrix.inverted() * matrix;
|
|
currentPath = transformMatrix.map(currentPath);
|
|
|
|
if (quadrilaterals)
|
|
{
|
|
currentPolygon = transformMatrix.map(currentPolygon);
|
|
quadrilaterals->append(currentPolygon);
|
|
}
|
|
|
|
path.addPath(currentPath);
|
|
}
|
|
|
|
return path;
|
|
}
|
|
|
|
PDFTextLayoutCache::PDFTextLayoutCache(std::function<PDFTextLayout (PDFInteger)> textLayoutGetter) :
|
|
m_textLayoutGetter(qMove(textLayoutGetter)),
|
|
m_pageIndex(-1),
|
|
m_layout()
|
|
{
|
|
|
|
}
|
|
|
|
void PDFTextLayoutCache::clear()
|
|
{
|
|
m_pageIndex = -1;
|
|
m_layout = PDFTextLayout();
|
|
}
|
|
|
|
const PDFTextLayout& PDFTextLayoutCache::getTextLayout(PDFInteger pageIndex)
|
|
{
|
|
if (m_pageIndex != pageIndex)
|
|
{
|
|
m_pageIndex = pageIndex;
|
|
m_layout = m_textLayoutGetter(pageIndex);
|
|
}
|
|
|
|
return m_layout;
|
|
}
|
|
|
|
} // namespace pdf
|