16 Commits
1_2 ... 1_3

Author SHA1 Message Date
62120898d2 Bump version: 1.3 2017-06-27 01:10:26 +02:00
dc1b9bc7ca PdfTextExtractor.JoinTextElements: Joins PdfTextElements when they are nearby. 2017-06-27 01:09:50 +02:00
d1ea41474b Reorder Code. 2017-06-27 01:03:27 +02:00
b11a2ac393 Simplify PdfFont.ParseSizes. 2017-06-26 22:17:46 +02:00
36fb20eb2e Remove VisualStudio2015 incompatibilities (Remove C#7.0-isms) 2017-06-26 08:25:30 +02:00
15fbec2470 FrmPdfInfo: Improve rendering, making more accurate the location of the glyphs. 2017-06-26 01:49:48 +02:00
52841de51b PdfFont: Convert "Zero" widths to default 0.5 2017-06-26 01:46:05 +02:00
d4c4615684 PdfTextExtractor: Rework text position calculations. 2017-06-26 01:45:34 +02:00
ae76cab45d PdfTextExtractor: Fix HasText method to match contained text, instead of full PdfTextElements. 2017-06-25 12:38:27 +02:00
8dc54105fd Refactorings 2017-06-25 12:03:41 +02:00
3469593a2a VAR.PdfTools.Workbench: Crude rendering of the parsed PDF. 2017-06-25 02:21:37 +02:00
ebff0c2028 Remove Visual Studio 2010 support 2017-06-11 16:29:24 +02:00
2fd074e041 Add Visual Studio 2017 support to NuGet Generation script. 2017-06-11 16:16:55 +02:00
4223619802 Set "Times-Roman" as default basefont. 2017-06-11 16:05:17 +02:00
771305f5d0 Refactor PdfFont creator. 2017-06-11 16:04:40 +02:00
90c7c5db92 Fix NuGet buid script 2017-04-13 08:23:19 +02:00
33 changed files with 907 additions and 782 deletions

View File

@@ -1,29 +0,0 @@
Microsoft Visual Studio Solution File, Format Version 11.00
# Visual Studio 2010
Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "VAR.PdfTools.Net35", "VAR.PdfTools\VAR.PdfTools.Net35.csproj", "{EB7E003A-6A95-4002-809F-926C7C8A11E9}"
EndProject
Project("{FAE04EC0-301F-11D3-BF4B-00C04F79EFBC}") = "VAR.PdfTools.Workbench.Net35", "VAR.PdfTools.Workbench\VAR.PdfTools.Workbench.Net35.csproj", "{A5825D8E-9F81-49E0-B610-8AE5E46D02EA}"
EndProject
Global
GlobalSection(SolutionConfigurationPlatforms) = preSolution
Debug|Any CPU = Debug|Any CPU
Release|Any CPU = Release|Any CPU
EndGlobalSection
GlobalSection(ProjectConfigurationPlatforms) = postSolution
{A5825D8E-9F81-49E0-B610-8AE5E46D02EA}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
{A5825D8E-9F81-49E0-B610-8AE5E46D02EA}.Debug|Any CPU.Build.0 = Debug|Any CPU
{A5825D8E-9F81-49E0-B610-8AE5E46D02EA}.Release|Any CPU.ActiveCfg = Release|Any CPU
{A5825D8E-9F81-49E0-B610-8AE5E46D02EA}.Release|Any CPU.Build.0 = Release|Any CPU
{EB7E003A-6A95-4002-809F-926C7C8A11E9}.Debug|Any CPU.ActiveCfg = Debug|Any CPU
{EB7E003A-6A95-4002-809F-926C7C8A11E9}.Debug|Any CPU.Build.0 = Debug|Any CPU
{EB7E003A-6A95-4002-809F-926C7C8A11E9}.Release|Any CPU.ActiveCfg = Release|Any CPU
{EB7E003A-6A95-4002-809F-926C7C8A11E9}.Release|Any CPU.Build.0 = Release|Any CPU
EndGlobalSection
GlobalSection(MonoDevelopProperties) = preSolution
StartupItem = VAR.PdfTools.Workbench\VAR.PdfTools.Workbench.Net35.csproj
EndGlobalSection
GlobalSection(SolutionProperties) = preSolution
HideSolutionNode = FALSE
EndGlobalSection
EndGlobal

View File

@@ -40,6 +40,7 @@
this.btnGetField = new System.Windows.Forms.Button();
this.txtText = new System.Windows.Forms.TextBox();
this.btnHasText = new System.Windows.Forms.Button();
this.btnRender = new System.Windows.Forms.Button();
this.SuspendLayout();
//
// lblOutputs
@@ -159,11 +160,23 @@
this.btnHasText.UseVisualStyleBackColor = true;
this.btnHasText.Click += new System.EventHandler(this.btnHasText_Click);
//
// btnRender
//
this.btnRender.Anchor = ((System.Windows.Forms.AnchorStyles)((System.Windows.Forms.AnchorStyles.Top | System.Windows.Forms.AnchorStyles.Right)));
this.btnRender.Location = new System.Drawing.Point(397, 52);
this.btnRender.Name = "btnRender";
this.btnRender.Size = new System.Drawing.Size(75, 23);
this.btnRender.TabIndex = 18;
this.btnRender.Text = "Render";
this.btnRender.UseVisualStyleBackColor = true;
this.btnRender.Click += new System.EventHandler(this.btnRender_Click);
//
// FrmPdfInfo
//
this.AutoScaleDimensions = new System.Drawing.SizeF(6F, 13F);
this.AutoScaleMode = System.Windows.Forms.AutoScaleMode.Font;
this.ClientSize = new System.Drawing.Size(484, 461);
this.Controls.Add(this.btnRender);
this.Controls.Add(this.txtText);
this.Controls.Add(this.btnHasText);
this.Controls.Add(this.txtFieldName);
@@ -199,5 +212,6 @@
private System.Windows.Forms.Button btnGetField;
private System.Windows.Forms.TextBox txtText;
private System.Windows.Forms.Button btnHasText;
private System.Windows.Forms.Button btnRender;
}
}

View File

@@ -1,7 +1,12 @@
using System;
using System.Collections.Generic;
using System.Drawing;
using System.Drawing.Drawing2D;
using System.Drawing.Imaging;
using System.IO;
using System.Linq;
using System.Windows.Forms;
using VAR.PdfTools.PdfElements;
namespace VAR.PdfTools.Workbench
{
@@ -89,7 +94,7 @@ namespace VAR.PdfTools.Workbench
{
lines.Add(string.Format("Text({0}, {1})({2}, {3})[{4}]: \"{5}\"",
textElement.Matrix.Matrix[0, 2], textElement.Matrix.Matrix[1, 2], textElement.VisibleWidth, textElement.VisibleHeight,
textElement.Font == null ? string.Empty : textElement.Font.Name,
textElement.Font == null ? "#NULL#" : textElement.Font.Name,
textElement.VisibleText));
}
}
@@ -154,5 +159,151 @@ namespace VAR.PdfTools.Workbench
}
txtOutput.Lines = lines.ToArray();
}
private void btnRender_Click(object sender, EventArgs e)
{
if (File.Exists(txtPdfPath.Text) == false)
{
MessageBox.Show("File does not exist");
return;
}
const int Scale = 5;
PdfDocument doc = PdfDocument.Load(txtPdfPath.Text);
string baseDocumentPath = Path.GetDirectoryName(txtPdfPath.Text);
string baseDocumentFilename = Path.GetFileNameWithoutExtension(txtPdfPath.Text);
List<string> lines = new List<string>();
lines.Add(string.Format("Filename : {0}", baseDocumentFilename));
lines.Add(string.Format("Number of Pages : {0}", doc.Pages.Count));
int pageNumber = 1;
foreach (PdfDocumentPage page in doc.Pages)
{
double pageXMin = double.MaxValue;
double pageYMin = double.MaxValue;
double pageXMax = double.MinValue;
double pageYMax = double.MinValue;
// Preprocess page to get max size
PdfTextExtractor extractor = new PdfTextExtractor(page);
foreach (PdfTextElement textElement in extractor.Elements)
{
double textElementXMin = textElement.GetX();
double textElementYMax = textElement.GetY();
double textElementXMax = textElementXMin + textElement.VisibleWidth;
double textElementYMin = textElementYMax - textElement.VisibleHeight;
if (textElementXMax > pageXMax) { pageXMax = textElementXMax; }
if (textElementYMax > pageYMax) { pageYMax = textElementYMax; }
if (textElementXMin < pageXMin) { pageXMin = textElementXMin; }
if (textElementYMin < pageYMin) { pageYMin = textElementYMin; }
}
lines.Add(string.Format("Page {0:0000} TextElements : {1}", pageNumber, extractor.Elements.Count));
// Prepare page image
int pageWidth = (int)Math.Ceiling(pageXMax - pageXMin);
int pageHeight = (int)Math.Ceiling(pageYMax - pageYMin);
using (Bitmap bmp = new Bitmap(pageWidth * Scale, pageHeight * Scale, PixelFormat.Format32bppArgb))
using (Graphics gc = Graphics.FromImage(bmp))
using (Pen penTextElem = new Pen(Color.Blue))
{
gc.Clear(Color.White);
// Draw text elements
foreach (PdfTextElement textElement in extractor.Elements)
{
DrawTextElement(textElement, gc, penTextElem, Scale, pageHeight, pageXMin, pageYMin);
}
// Save image to disk
string fileName = Path.Combine(baseDocumentPath, string.Format("{0}_{1:0000}.png", baseDocumentFilename, pageNumber));
bmp.Save(fileName, ImageFormat.Png);
}
pageNumber++;
}
txtOutput.Lines = lines.ToArray();
}
private static void DrawTextElement(PdfTextElement textElement, Graphics gc, Pen penTextElem, int Scale, int pageHeight, double pageXMin, double pageYMin)
{
double textElementX = textElement.GetX() - pageXMin;
double textElementY = textElement.GetY() - pageYMin;
double textElementWidth = textElement.VisibleWidth;
double textElementHeight = textElement.VisibleHeight;
string textElementText = textElement.VisibleText;
string textElementFontName = (textElement.Font == null ? string.Empty : textElement.Font.Name);
if (textElementHeight < 0.0001) { return; }
double textElementPageX = textElementX;
double textElementPageY = pageHeight - textElementY;
DrawRoundedRectangle(gc, penTextElem,
(int)(textElementPageX * Scale),
(int)(textElementPageY * Scale),
(int)(textElementWidth * Scale),
(int)(textElementHeight * Scale),
Scale);
using (Font font = new Font("Arial", (int)(textElementHeight * Scale), GraphicsUnit.Pixel))
{
foreach (PdfCharElement c in textElement.Characters)
{
gc.DrawString(c.Char,
font,
Brushes.Black,
(int)((textElementPageX + c.Displacement) * Scale),
(int)(textElementPageY * Scale));
gc.FillRectangle(Brushes.Red,
(int)((textElementPageX + c.Displacement) * Scale),
(int)(textElementPageY * Scale),
2, 2);
}
}
}
public static GraphicsPath RoundedRect(int x, int y, int width, int height, int radius)
{
int diameter = radius * 2;
Size size = new Size(diameter, diameter);
Rectangle arc = new Rectangle(x, y, diameter, diameter);
GraphicsPath path = new GraphicsPath();
// top left arc
path.AddArc(arc, 180, 90);
// top right arc
arc.X = (x + width) - diameter;
path.AddArc(arc, 270, 90);
// bottom right arc
arc.Y = (y + height) - diameter;
path.AddArc(arc, 0, 90);
// bottom left arc
arc.X = x;
path.AddArc(arc, 90, 90);
path.CloseFigure();
return path;
}
public static void DrawRoundedRectangle(Graphics graphics, Pen pen, int x, int y, int width, int height, int cornerRadius)
{
if (graphics == null)
throw new ArgumentNullException("graphics");
if (pen == null)
throw new ArgumentNullException("pen");
using (GraphicsPath path = RoundedRect(x, y, width, height, cornerRadius))
{
graphics.DrawPath(pen, path);
}
}
}
}

View File

@@ -1,120 +0,0 @@
<?xml version="1.0" encoding="utf-8"?>
<root>
<!--
Microsoft ResX Schema
Version 2.0
The primary goals of this format is to allow a simple XML format
that is mostly human readable. The generation and parsing of the
various data types are done through the TypeConverter classes
associated with the data types.
Example:
... ado.net/XML headers & schema ...
<resheader name="resmimetype">text/microsoft-resx</resheader>
<resheader name="version">2.0</resheader>
<resheader name="reader">System.Resources.ResXResourceReader, System.Windows.Forms, ...</resheader>
<resheader name="writer">System.Resources.ResXResourceWriter, System.Windows.Forms, ...</resheader>
<data name="Name1"><value>this is my long string</value><comment>this is a comment</comment></data>
<data name="Color1" type="System.Drawing.Color, System.Drawing">Blue</data>
<data name="Bitmap1" mimetype="application/x-microsoft.net.object.binary.base64">
<value>[base64 mime encoded serialized .NET Framework object]</value>
</data>
<data name="Icon1" type="System.Drawing.Icon, System.Drawing" mimetype="application/x-microsoft.net.object.bytearray.base64">
<value>[base64 mime encoded string representing a byte array form of the .NET Framework object]</value>
<comment>This is a comment</comment>
</data>
There are any number of "resheader" rows that contain simple
name/value pairs.
Each data row contains a name, and value. The row also contains a
type or mimetype. Type corresponds to a .NET class that support
text/value conversion through the TypeConverter architecture.
Classes that don't support this are serialized and stored with the
mimetype set.
The mimetype is used for serialized objects, and tells the
ResXResourceReader how to depersist the object. This is currently not
extensible. For a given mimetype the value must be set accordingly:
Note - application/x-microsoft.net.object.binary.base64 is the format
that the ResXResourceWriter will generate, however the reader can
read any of the formats listed below.
mimetype: application/x-microsoft.net.object.binary.base64
value : The object must be serialized with
: System.Runtime.Serialization.Formatters.Binary.BinaryFormatter
: and then encoded with base64 encoding.
mimetype: application/x-microsoft.net.object.soap.base64
value : The object must be serialized with
: System.Runtime.Serialization.Formatters.Soap.SoapFormatter
: and then encoded with base64 encoding.
mimetype: application/x-microsoft.net.object.bytearray.base64
value : The object must be serialized into a byte array
: using a System.ComponentModel.TypeConverter
: and then encoded with base64 encoding.
-->
<xsd:schema id="root" xmlns="" xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns:msdata="urn:schemas-microsoft-com:xml-msdata">
<xsd:import namespace="http://www.w3.org/XML/1998/namespace" />
<xsd:element name="root" msdata:IsDataSet="true">
<xsd:complexType>
<xsd:choice maxOccurs="unbounded">
<xsd:element name="metadata">
<xsd:complexType>
<xsd:sequence>
<xsd:element name="value" type="xsd:string" minOccurs="0" />
</xsd:sequence>
<xsd:attribute name="name" use="required" type="xsd:string" />
<xsd:attribute name="type" type="xsd:string" />
<xsd:attribute name="mimetype" type="xsd:string" />
<xsd:attribute ref="xml:space" />
</xsd:complexType>
</xsd:element>
<xsd:element name="assembly">
<xsd:complexType>
<xsd:attribute name="alias" type="xsd:string" />
<xsd:attribute name="name" type="xsd:string" />
</xsd:complexType>
</xsd:element>
<xsd:element name="data">
<xsd:complexType>
<xsd:sequence>
<xsd:element name="value" type="xsd:string" minOccurs="0" msdata:Ordinal="1" />
<xsd:element name="comment" type="xsd:string" minOccurs="0" msdata:Ordinal="2" />
</xsd:sequence>
<xsd:attribute name="name" type="xsd:string" use="required" msdata:Ordinal="1" />
<xsd:attribute name="type" type="xsd:string" msdata:Ordinal="3" />
<xsd:attribute name="mimetype" type="xsd:string" msdata:Ordinal="4" />
<xsd:attribute ref="xml:space" />
</xsd:complexType>
</xsd:element>
<xsd:element name="resheader">
<xsd:complexType>
<xsd:sequence>
<xsd:element name="value" type="xsd:string" minOccurs="0" msdata:Ordinal="1" />
</xsd:sequence>
<xsd:attribute name="name" type="xsd:string" use="required" />
</xsd:complexType>
</xsd:element>
</xsd:choice>
</xsd:complexType>
</xsd:element>
</xsd:schema>
<resheader name="resmimetype">
<value>text/microsoft-resx</value>
</resheader>
<resheader name="version">
<value>2.0</value>
</resheader>
<resheader name="reader">
<value>System.Resources.ResXResourceReader, System.Windows.Forms, Version=2.0.0.0, Culture=neutral, PublicKeyToken=b77a5c561934e089</value>
</resheader>
<resheader name="writer">
<value>System.Resources.ResXResourceWriter, System.Windows.Forms, Version=2.0.0.0, Culture=neutral, PublicKeyToken=b77a5c561934e089</value>
</resheader>
</root>

View File

@@ -11,4 +11,4 @@ using System.Runtime.InteropServices;
[assembly: AssemblyCulture("")]
[assembly: ComVisible(false)]
[assembly: Guid("a5825d8e-9f81-49e0-b610-8ae5e46d02ea")]
[assembly: AssemblyVersion("1.2.*")]
[assembly: AssemblyVersion("1.3.*")]

View File

@@ -1,8 +1,13 @@
@echo off
:: MSBuild and tools path
if exist "%ProgramFiles%\MSBuild\14.0\bin" set PATH=%ProgramFiles%\MSBuild\14.0\bin;%PATH%
if exist "%ProgramFiles(x86)%\MSBuild\14.0\bin" set PATH=%ProgramFiles(x86)%\MSBuild\14.0\bin;%PATH%
if exist "%windir%\Microsoft.Net\Framework\v4.0.30319" set MsBuildPath=%windir%\Microsoft.NET\Framework\v4.0.30319
if exist "%windir%\Microsoft.Net\Framework64\v4.0.30319" set MsBuildPath=%windir%\Microsoft.NET\Framework64\v4.0.30319
if exist "C:\Program Files (x86)\MSBuild\14.0\Bin" set MsBuildPath=C:\Program Files (x86)\MSBuild\14.0\Bin
if exist "C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\MSBuild\15.0\Bin" set MsBuildPath=C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\MSBuild\15.0\Bin
set PATH=%MsBuildPath%;%PATH%
echo %MsBuildPath%
:: NuGet
set nuget="nuget"
@@ -18,7 +23,7 @@ msbuild VAR.PdfTools.csproj /t:Build /p:Configuration="Release .Net 4.6.1" /p:Pl
:: Packing Nuget
Title Packing Nuget
%nuget% pack VAR.PdfTools.csproj -Verbosity detailed -OutputDir "NuGet" -MSBuildVersion "14.0" -Properties Configuration="Release .Net 4.6.1" -Prop Platform=AnyCPU
%nuget% pack VAR.PdfTools.csproj -Verbosity detailed -OutputDir "NuGet" -Properties Configuration="Release .Net 4.6.1" -Prop Platform=AnyCPU
title Finished
pause

View File

@@ -0,0 +1,121 @@
using System;
namespace VAR.PdfTools.Maths
{
public class Matrix3x3
{
#region Declarations
public double[,] _matrix = new double[3, 3];
#endregion
#region Properties
public double[,] Matrix { get { return _matrix; } }
#endregion
#region Creator
public Matrix3x3()
{
Idenity();
}
public Matrix3x3(double a, double b, double c, double d, double e, double f)
{
Set(a, b, c, d, e, f);
}
#endregion
#region Public methods
public void Idenity()
{
_matrix[0, 0] = 1.0;
_matrix[0, 1] = 0.0;
_matrix[0, 2] = 0.0;
_matrix[1, 0] = 0.0;
_matrix[1, 1] = 1.0;
_matrix[1, 2] = 0.0;
_matrix[2, 0] = 0.0;
_matrix[2, 1] = 0.0;
_matrix[2, 2] = 1.0;
}
public void Set(double a, double b, double c, double d, double e, double f)
{
_matrix[0, 0] = a;
_matrix[1, 0] = b;
_matrix[2, 0] = 0;
_matrix[0, 1] = c;
_matrix[1, 1] = d;
_matrix[2, 1] = 0;
_matrix[0, 2] = e;
_matrix[1, 2] = f;
_matrix[2, 2] = 1;
}
public Vector3D Multiply(Vector3D vect)
{
Vector3D vectResult = new Vector3D();
vectResult.Vector[0] = (vect.Vector[0] * _matrix[0, 0]) + (vect.Vector[1] * _matrix[0, 1]) + (vect.Vector[2] * _matrix[0, 2]);
vectResult.Vector[1] = (vect.Vector[0] * _matrix[1, 0]) + (vect.Vector[1] * _matrix[1, 1]) + (vect.Vector[2] * _matrix[1, 2]);
vectResult.Vector[2] = (vect.Vector[0] * _matrix[2, 0]) + (vect.Vector[1] * _matrix[2, 1]) + (vect.Vector[2] * _matrix[2, 2]);
return vectResult;
}
public Matrix3x3 Multiply(Matrix3x3 matrix)
{
Matrix3x3 newMatrix = new Matrix3x3();
newMatrix._matrix[0, 0] = (_matrix[0, 0] * matrix._matrix[0, 0]) + (_matrix[1, 0] * matrix._matrix[0, 1]) + (_matrix[2, 0] * matrix._matrix[0, 2]);
newMatrix._matrix[0, 1] = (_matrix[0, 1] * matrix._matrix[0, 0]) + (_matrix[1, 1] * matrix._matrix[0, 1]) + (_matrix[2, 1] * matrix._matrix[0, 2]);
newMatrix._matrix[0, 2] = (_matrix[0, 2] * matrix._matrix[0, 0]) + (_matrix[1, 2] * matrix._matrix[0, 1]) + (_matrix[2, 2] * matrix._matrix[0, 2]);
newMatrix._matrix[1, 0] = (_matrix[0, 0] * matrix._matrix[1, 0]) + (_matrix[1, 0] * matrix._matrix[1, 1]) + (_matrix[2, 0] * matrix._matrix[1, 2]);
newMatrix._matrix[1, 1] = (_matrix[0, 1] * matrix._matrix[1, 0]) + (_matrix[1, 1] * matrix._matrix[1, 1]) + (_matrix[2, 1] * matrix._matrix[1, 2]);
newMatrix._matrix[1, 2] = (_matrix[0, 2] * matrix._matrix[1, 0]) + (_matrix[1, 2] * matrix._matrix[1, 1]) + (_matrix[2, 2] * matrix._matrix[1, 2]);
newMatrix._matrix[2, 0] = (_matrix[0, 0] * matrix._matrix[2, 0]) + (_matrix[1, 0] * matrix._matrix[2, 1]) + (_matrix[2, 0] * matrix._matrix[2, 2]);
newMatrix._matrix[2, 1] = (_matrix[0, 1] * matrix._matrix[2, 0]) + (_matrix[1, 1] * matrix._matrix[2, 1]) + (_matrix[2, 1] * matrix._matrix[2, 2]);
newMatrix._matrix[2, 2] = (_matrix[0, 2] * matrix._matrix[2, 0]) + (_matrix[1, 2] * matrix._matrix[2, 1]) + (_matrix[2, 2] * matrix._matrix[2, 2]);
return newMatrix;
}
public Matrix3x3 Copy()
{
Matrix3x3 newMatrix = new Matrix3x3();
newMatrix._matrix[0, 0] = _matrix[0, 0];
newMatrix._matrix[0, 1] = _matrix[0, 1];
newMatrix._matrix[0, 2] = _matrix[0, 2];
newMatrix._matrix[1, 0] = _matrix[1, 0];
newMatrix._matrix[1, 1] = _matrix[1, 1];
newMatrix._matrix[1, 2] = _matrix[1, 2];
newMatrix._matrix[2, 0] = _matrix[2, 0];
newMatrix._matrix[2, 1] = _matrix[2, 1];
newMatrix._matrix[2, 2] = _matrix[2, 2];
return newMatrix;
}
public bool IsCollinear(Matrix3x3 otherMatrix, double horizontalDelta = 0.00001, double verticalDelta = 0.00001)
{
double epsilon = 0.00001;
return (
Math.Abs(_matrix[0, 0] - otherMatrix.Matrix[0, 0]) <= epsilon &&
Math.Abs(_matrix[1, 0] - otherMatrix.Matrix[1, 0]) <= epsilon &&
Math.Abs(_matrix[0, 1] - otherMatrix.Matrix[0, 1]) <= epsilon &&
Math.Abs(_matrix[1, 1] - otherMatrix.Matrix[1, 1]) <= epsilon &&
Math.Abs(_matrix[0, 2] - otherMatrix.Matrix[0, 2]) <= horizontalDelta &&
Math.Abs(_matrix[1, 2] - otherMatrix.Matrix[1, 2]) <= verticalDelta &&
true);
}
#endregion
}
}

View File

@@ -0,0 +1,33 @@
namespace VAR.PdfTools.Maths
{
public class Vector3D
{
#region Declarations
public double[] _vector = new double[3];
#endregion
#region Properties
public double[] Vector { get { return _vector; } }
#endregion
#region Creator
public Vector3D()
{
Init();
}
public void Init()
{
_vector[0] = 0.0;
_vector[1] = 0.0;
_vector[2] = 1.0;
}
#endregion
}
}

View File

@@ -1,4 +1,5 @@
using System.Collections.Generic;
using VAR.PdfTools.PdfElements;
namespace VAR.PdfTools
{

View File

@@ -2,6 +2,7 @@
using System.Collections.Generic;
using System.IO;
using System.Linq;
using VAR.PdfTools.PdfElements;
namespace VAR.PdfTools
{

View File

@@ -1,5 +1,6 @@
using System;
using System.Collections.Generic;
using VAR.PdfTools.PdfElements;
namespace VAR.PdfTools
{

View File

@@ -1,202 +0,0 @@
using System.Collections.Generic;
using System.IO;
namespace VAR.PdfTools
{
public enum PdfElementTypes
{
Undefined,
Boolean,
Integer,
Real,
String,
Name,
Array,
Dictionary,
Null,
ObjectReference,
Object,
Stream,
};
public interface IPdfElement
{
PdfElementTypes Type { get; }
}
public class PdfBoolean : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Boolean; } }
public bool Value { get; set; }
}
public class PdfInteger : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Integer; } }
public long Value { get; set; }
}
public class PdfReal : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Real; } }
public double Value { get; set; }
}
public class PdfString : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.String; } }
public string Value { get; set; }
}
public class PdfName : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Name; } }
public string Value { get; set; }
}
public class PdfArray : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Array; } }
private List<IPdfElement> _values = new List<IPdfElement>();
public List<IPdfElement> Values { get { return _values; } }
}
public class PdfDictionary : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Dictionary; } }
private Dictionary<string, IPdfElement> _values = new Dictionary<string, IPdfElement>();
public Dictionary<string, IPdfElement> Values { get { return _values; } }
public string GetParamAsString(string name)
{
if (Values.ContainsKey(name) == false) { return null; }
IPdfElement value = Values[name];
if (value is PdfArray)
{
value = ((PdfArray)value).Values[0];
}
if (value is PdfName)
{
return ((PdfName)value).Value;
}
if (value is PdfString)
{
return ((PdfString)value).Value;
}
return null;
}
public long? GetParamAsInt(string name)
{
if (Values.ContainsKey(name) == false) { return null; }
IPdfElement value = Values[name];
if (value is PdfArray)
{
value = ((PdfArray)value).Values[0];
}
if (value is PdfInteger)
{
return ((PdfInteger)value).Value;
}
return null;
}
public byte[] GetParamAsStream(string name)
{
if (Values.ContainsKey(name) == false) { return null; }
IPdfElement value = Values[name];
if (value is PdfArray)
{
PdfArray array = value as PdfArray;
MemoryStream memStream = new MemoryStream();
foreach (IPdfElement elem in array.Values)
{
PdfStream stream = elem as PdfStream;
if (stream == null) { continue; }
memStream.Write(stream.Data, 0, stream.Data.Length);
}
if (memStream.Length > 0)
{
return memStream.ToArray();
}
return null;
}
if (value is PdfStream)
{
return ((PdfStream)value).Data;
}
return null;
}
}
public class PdfNull : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Null; } }
}
public class PdfObjectReference : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.ObjectReference; } }
public int ObjectID { get; set; }
public int ObjectGeneration { get; set; }
}
public class PdfStream : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Stream; } }
public PdfDictionary Dictionary { get; set; }
public byte[] Data { get; set; }
public byte[] OriginalData { get; set; }
public IPdfElement OriginalFilter { get; set; }
}
public class PdfObject : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Object; } }
public int ObjectID { get; set; }
public int ObjectGeneration { get; set; }
public IPdfElement Data { get; set; }
public int UsageCount { get; set; }
}
public static class PdfElementUtils
{
public static double GetReal(IPdfElement elem, double defaultValue)
{
if (elem == null)
{
return defaultValue;
}
if (elem is PdfInteger)
{
return ((PdfInteger)elem).Value;
}
if (elem is PdfReal)
{
return ((PdfReal)elem).Value;
}
return defaultValue;
}
public static long GetInt(IPdfElement elem, long defaultValue)
{
if (elem == null)
{
return defaultValue;
}
if (elem is PdfInteger)
{
return ((PdfInteger)elem).Value;
}
if (elem is PdfReal)
{
return (long)((PdfReal)elem).Value;
}
return defaultValue;
}
}
}

View File

@@ -0,0 +1,7 @@
namespace VAR.PdfTools.PdfElements
{
public interface IPdfElement
{
PdfElementTypes Type { get; }
}
}

View File

@@ -0,0 +1,11 @@
using System.Collections.Generic;
namespace VAR.PdfTools.PdfElements
{
public class PdfArray : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Array; } }
private List<IPdfElement> _values = new List<IPdfElement>();
public List<IPdfElement> Values { get { return _values; } }
}
}

View File

@@ -0,0 +1,8 @@
namespace VAR.PdfTools.PdfElements
{
public class PdfBoolean : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Boolean; } }
public bool Value { get; set; }
}
}

View File

@@ -0,0 +1,77 @@
using System.Collections.Generic;
using System.IO;
namespace VAR.PdfTools.PdfElements
{
public class PdfDictionary : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Dictionary; } }
private Dictionary<string, IPdfElement> _values = new Dictionary<string, IPdfElement>();
public Dictionary<string, IPdfElement> Values { get { return _values; } }
public string GetParamAsString(string name)
{
if (Values.ContainsKey(name) == false) { return null; }
IPdfElement value = Values[name];
if (value is PdfArray)
{
value = ((PdfArray)value).Values[0];
}
if (value is PdfName)
{
return ((PdfName)value).Value;
}
if (value is PdfString)
{
return ((PdfString)value).Value;
}
return null;
}
public long? GetParamAsInt(string name)
{
if (Values.ContainsKey(name) == false) { return null; }
IPdfElement value = Values[name];
if (value is PdfArray)
{
value = ((PdfArray)value).Values[0];
}
if (value is PdfInteger)
{
return ((PdfInteger)value).Value;
}
return null;
}
public byte[] GetParamAsStream(string name)
{
if (Values.ContainsKey(name) == false) { return null; }
IPdfElement value = Values[name];
if (value is PdfArray)
{
PdfArray array = value as PdfArray;
MemoryStream memStream = new MemoryStream();
foreach (IPdfElement elem in array.Values)
{
PdfStream stream = elem as PdfStream;
if (stream == null) { continue; }
memStream.Write(stream.Data, 0, stream.Data.Length);
}
if (memStream.Length > 0)
{
return memStream.ToArray();
}
return null;
}
if (value is PdfStream)
{
return ((PdfStream)value).Data;
}
return null;
}
}
}

View File

@@ -0,0 +1,18 @@
namespace VAR.PdfTools.PdfElements
{
public enum PdfElementTypes
{
Undefined,
Boolean,
Integer,
Real,
String,
Name,
Array,
Dictionary,
Null,
ObjectReference,
Object,
Stream,
};
}

View File

@@ -0,0 +1,56 @@
namespace VAR.PdfTools.PdfElements
{
public static class PdfElementUtils
{
public static double GetReal(IPdfElement elem, double defaultValue)
{
if (elem == null)
{
return defaultValue;
}
if (elem is PdfInteger)
{
return ((PdfInteger)elem).Value;
}
if (elem is PdfReal)
{
return ((PdfReal)elem).Value;
}
return defaultValue;
}
public static long GetInt(IPdfElement elem, long defaultValue)
{
if (elem == null)
{
return defaultValue;
}
if (elem is PdfInteger)
{
return ((PdfInteger)elem).Value;
}
if (elem is PdfReal)
{
return (long)((PdfReal)elem).Value;
}
return defaultValue;
}
public static string GetString(IPdfElement elem, string defaultValue)
{
if (elem == null)
{
return defaultValue;
}
if (elem is PdfString)
{
return ((PdfString)elem).Value;
}
if (elem is PdfName)
{
return ((PdfName)elem).Value;
}
return defaultValue;
}
}
}

View File

@@ -0,0 +1,8 @@
namespace VAR.PdfTools.PdfElements
{
public class PdfInteger : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Integer; } }
public long Value { get; set; }
}
}

View File

@@ -0,0 +1,8 @@
namespace VAR.PdfTools.PdfElements
{
public class PdfName : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Name; } }
public string Value { get; set; }
}
}

View File

@@ -0,0 +1,7 @@
namespace VAR.PdfTools.PdfElements
{
public class PdfNull : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Null; } }
}
}

View File

@@ -0,0 +1,11 @@
namespace VAR.PdfTools.PdfElements
{
public class PdfObject : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Object; } }
public int ObjectID { get; set; }
public int ObjectGeneration { get; set; }
public IPdfElement Data { get; set; }
public int UsageCount { get; set; }
}
}

View File

@@ -0,0 +1,9 @@
namespace VAR.PdfTools.PdfElements
{
public class PdfObjectReference : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.ObjectReference; } }
public int ObjectID { get; set; }
public int ObjectGeneration { get; set; }
}
}

View File

@@ -0,0 +1,8 @@
namespace VAR.PdfTools.PdfElements
{
public class PdfReal : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Real; } }
public double Value { get; set; }
}
}

View File

@@ -0,0 +1,12 @@
namespace VAR.PdfTools.PdfElements
{
public class PdfStream : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.Stream; } }
public PdfDictionary Dictionary { get; set; }
public byte[] Data { get; set; }
public byte[] OriginalData { get; set; }
public IPdfElement OriginalFilter { get; set; }
}
}

View File

@@ -0,0 +1,8 @@
namespace VAR.PdfTools.PdfElements
{
public class PdfString : IPdfElement
{
public PdfElementTypes Type { get { return PdfElementTypes.String; } }
public string Value { get; set; }
}
}

View File

@@ -1,5 +1,5 @@
using System;
using System.Collections.Generic;
using System.Collections.Generic;
using VAR.PdfTools.PdfElements;
namespace VAR.PdfTools
{
@@ -45,6 +45,19 @@ namespace VAR.PdfTools
_tainted = true;
}
PrepareSizes(baseData);
}
#endregion
#region Private methods
private void PrepareSizes(PdfDictionary baseData)
{
// Set "Times-Roman" as default basefont sizes
_widths = PdfStandar14FontMetrics.Times_Roman.Widths;
_height = PdfStandar14FontMetrics.Times_Roman.ApproxHeight;
if (baseData.Values.ContainsKey("ToUnicode"))
{
byte[] toUnicodeStream = ((PdfStream)baseData.Values["ToUnicode"]).Data;
@@ -52,7 +65,19 @@ namespace VAR.PdfTools
_toUnicode = parser.ParseToUnicode();
}
string baseFont = _baseData.GetParamAsString("BaseFont");
if (string.IsNullOrEmpty(baseFont))
{
SetBaseFontSizes(baseFont);
}
if (_baseData.Values.ContainsKey("FirstChar") && _baseData.Values.ContainsKey("LastChar") && _baseData.Values.ContainsKey("Widths"))
{
ParseSizes();
}
}
private void ParseSizes()
{
double glyphSpaceToTextSpace = 1000.0; // FIXME: SubType:Type3 Uses a FontMatrix that may not correspond to 1/1000th
_widths = new Dictionary<char, double>();
@@ -62,26 +87,14 @@ namespace VAR.PdfTools
char actualChar = firstChar;
foreach (IPdfElement elem in widths.Values)
{
PdfReal widthReal = elem as PdfReal;
if (widthReal != null)
{
_widths.Add(actualChar, widthReal.Value / glyphSpaceToTextSpace);
_widths.Add(actualChar, PdfElementUtils.GetReal(elem, 500) / glyphSpaceToTextSpace);
actualChar++;
continue;
}
PdfInteger widthInt = elem as PdfInteger;
if (widthInt != null)
{
_widths.Add(actualChar, widthInt.Value / glyphSpaceToTextSpace);
actualChar++;
continue;
}
}
// FIMXE: Calculate real height
}
else
private void SetBaseFontSizes(string baseFont)
{
string baseFont = _baseData.GetParamAsString("BaseFont");
if (baseFont == "Times-Roman")
{
_widths = PdfStandar14FontMetrics.Times_Roman.Widths;
@@ -153,7 +166,6 @@ namespace VAR.PdfTools
_height = PdfStandar14FontMetrics.ZapfDingbats.ApproxHeight;
}
}
}
#endregion
@@ -177,15 +189,23 @@ namespace VAR.PdfTools
public double GetCharWidth(char character)
{
double charWidth = 0;
if (_widths == null)
{
return 0;
return charWidth;
}
if (_widths.ContainsKey(character))
{
return _widths[character];
charWidth = _widths[character];
}
return 0;
// NOTE: Convert "Zero" to default width of 0.5
if (charWidth <= 0.0001)
{
charWidth = 0.5;
}
return charWidth;
}
#endregion

View File

@@ -4,6 +4,7 @@ using System.Globalization;
using System.IO;
using System.Linq;
using System.Text;
using VAR.PdfTools.PdfElements;
namespace VAR.PdfTools
{

View File

@@ -2,155 +2,15 @@
using System.Collections.Generic;
using System.Linq;
using System.Text;
using VAR.PdfTools.Maths;
using VAR.PdfTools.PdfElements;
namespace VAR.PdfTools
{
public class Vector3D
public struct PdfCharElement
{
#region Declarations
public double[] _vector = new double[3];
#endregion
#region Properties
public double[] Vector { get { return _vector; } }
#endregion
#region Creator
public Vector3D()
{
Init();
}
public void Init()
{
_vector[0] = 0.0;
_vector[1] = 0.0;
_vector[2] = 1.0;
}
#endregion
}
public class Matrix3x3
{
#region Declarations
public double[,] _matrix = new double[3, 3];
#endregion
#region Properties
public double[,] Matrix { get { return _matrix; } }
#endregion
#region Creator
public Matrix3x3()
{
Idenity();
}
public Matrix3x3(double a, double b, double c, double d, double e, double f)
{
Set(a, b, c, d, e, f);
}
#endregion
#region Public methods
public void Idenity()
{
_matrix[0, 0] = 1.0;
_matrix[0, 1] = 0.0;
_matrix[0, 2] = 0.0;
_matrix[1, 0] = 0.0;
_matrix[1, 1] = 1.0;
_matrix[1, 2] = 0.0;
_matrix[2, 0] = 0.0;
_matrix[2, 1] = 0.0;
_matrix[2, 2] = 1.0;
}
public void Set(double a, double b, double c, double d, double e, double f)
{
_matrix[0, 0] = a;
_matrix[1, 0] = b;
_matrix[2, 0] = 0;
_matrix[0, 1] = c;
_matrix[1, 1] = d;
_matrix[2, 1] = 0;
_matrix[0, 2] = e;
_matrix[1, 2] = f;
_matrix[2, 2] = 1;
}
public Vector3D Multiply(Vector3D vect)
{
Vector3D vectResult = new Vector3D();
vectResult.Vector[0] = (vect.Vector[0] * _matrix[0, 0]) + (vect.Vector[1] * _matrix[0, 1]) + (vect.Vector[2] * _matrix[0, 2]);
vectResult.Vector[1] = (vect.Vector[0] * _matrix[1, 0]) + (vect.Vector[1] * _matrix[1, 1]) + (vect.Vector[2] * _matrix[1, 2]);
vectResult.Vector[2] = (vect.Vector[0] * _matrix[2, 0]) + (vect.Vector[1] * _matrix[2, 1]) + (vect.Vector[2] * _matrix[2, 2]);
return vectResult;
}
public Matrix3x3 Multiply(Matrix3x3 matrix)
{
Matrix3x3 newMatrix = new Matrix3x3();
newMatrix._matrix[0, 0] = (_matrix[0, 0] * matrix._matrix[0, 0]) + (_matrix[1, 0] * matrix._matrix[0, 1]) + (_matrix[2, 0] * matrix._matrix[0, 2]);
newMatrix._matrix[0, 1] = (_matrix[0, 1] * matrix._matrix[0, 0]) + (_matrix[1, 1] * matrix._matrix[0, 1]) + (_matrix[2, 1] * matrix._matrix[0, 2]);
newMatrix._matrix[0, 2] = (_matrix[0, 2] * matrix._matrix[0, 0]) + (_matrix[1, 2] * matrix._matrix[0, 1]) + (_matrix[2, 2] * matrix._matrix[0, 2]);
newMatrix._matrix[1, 0] = (_matrix[0, 0] * matrix._matrix[1, 0]) + (_matrix[1, 0] * matrix._matrix[1, 1]) + (_matrix[2, 0] * matrix._matrix[1, 2]);
newMatrix._matrix[1, 1] = (_matrix[0, 1] * matrix._matrix[1, 0]) + (_matrix[1, 1] * matrix._matrix[1, 1]) + (_matrix[2, 1] * matrix._matrix[1, 2]);
newMatrix._matrix[1, 2] = (_matrix[0, 2] * matrix._matrix[1, 0]) + (_matrix[1, 2] * matrix._matrix[1, 1]) + (_matrix[2, 2] * matrix._matrix[1, 2]);
newMatrix._matrix[2, 0] = (_matrix[0, 0] * matrix._matrix[2, 0]) + (_matrix[1, 0] * matrix._matrix[2, 1]) + (_matrix[2, 0] * matrix._matrix[2, 2]);
newMatrix._matrix[2, 1] = (_matrix[0, 1] * matrix._matrix[2, 0]) + (_matrix[1, 1] * matrix._matrix[2, 1]) + (_matrix[2, 1] * matrix._matrix[2, 2]);
newMatrix._matrix[2, 2] = (_matrix[0, 2] * matrix._matrix[2, 0]) + (_matrix[1, 2] * matrix._matrix[2, 1]) + (_matrix[2, 2] * matrix._matrix[2, 2]);
return newMatrix;
}
public Matrix3x3 Copy()
{
Matrix3x3 newMatrix = new Matrix3x3();
newMatrix._matrix[0, 0] = _matrix[0, 0];
newMatrix._matrix[0, 1] = _matrix[0, 1];
newMatrix._matrix[0, 2] = _matrix[0, 2];
newMatrix._matrix[1, 0] = _matrix[1, 0];
newMatrix._matrix[1, 1] = _matrix[1, 1];
newMatrix._matrix[1, 2] = _matrix[1, 2];
newMatrix._matrix[2, 0] = _matrix[2, 0];
newMatrix._matrix[2, 1] = _matrix[2, 1];
newMatrix._matrix[2, 2] = _matrix[2, 2];
return newMatrix;
}
public bool IsCollinear(Matrix3x3 otherMatrix, double horizontalDelta = 0.00001, double verticalDelta = 0.00001)
{
double epsilon = 0.00001;
return (
Math.Abs(_matrix[0, 0] - otherMatrix.Matrix[0, 0]) <= epsilon &&
Math.Abs(_matrix[1, 0] - otherMatrix.Matrix[1, 0]) <= epsilon &&
Math.Abs(_matrix[0, 1] - otherMatrix.Matrix[0, 1]) <= epsilon &&
Math.Abs(_matrix[1, 1] - otherMatrix.Matrix[1, 1]) <= epsilon &&
Math.Abs(_matrix[0, 2] - otherMatrix.Matrix[0, 2]) <= horizontalDelta &&
Math.Abs(_matrix[1, 2] - otherMatrix.Matrix[1, 2]) <= verticalDelta &&
true);
}
#endregion
public string Char;
public double Displacement;
}
public class PdfTextElement
@@ -171,8 +31,9 @@ namespace VAR.PdfTools
public double VisibleHeight { get; set; }
private List<PdfTextElement> _childs = new List<PdfTextElement>();
public List<PdfTextElement> Childs { get { return _childs; } }
public List<PdfCharElement> Characters { get; set; }
public List<PdfTextElement> Childs { get; set; }
#endregion
@@ -206,15 +67,17 @@ namespace VAR.PdfTools
// Text state
private PdfFont _font = null;
private double _fontSize = 1;
private double _charSpacing = 0;
private double _wordSpacing = 0;
private double _textLeading = 0;
// Text object state
private bool inText = false;
private Matrix3x3 _textMatrix = new Matrix3x3();
private Matrix3x3 _textMatrixCurrent = new Matrix3x3();
private StringBuilder _sbText = new StringBuilder();
private double _textWidth = 0;
PdfTextElement _currentTextElement = null;
private List<PdfCharElement> _listCharacters = new List<PdfCharElement>();
#endregion
@@ -231,7 +94,8 @@ namespace VAR.PdfTools
public PdfTextExtractor(PdfDocumentPage page)
{
_page = page;
ProcessPage();
ProcessPageContent();
JoinTextElements();
}
#endregion
@@ -258,42 +122,24 @@ namespace VAR.PdfTools
PdfTextElement textElem = new PdfTextElement();
textElem.Font = _font;
textElem.FontSize = _fontSize;
textElem.Matrix = _textMatrix.Multiply(_graphicsMatrix);
textElem.Matrix = _textMatrixCurrent.Multiply(_graphicsMatrix);
textElem.RawText = _sbText.ToString();
textElem.VisibleText = PdfString_ToUnicode(textElem.RawText, _font);
textElem.VisibleWidth = _textWidth * textElem.Matrix.Matrix[0, 0];
textElem.VisibleHeight = (_font.Height * _fontSize) * textElem.Matrix.Matrix[1, 1];
textElem.Characters = new List<PdfCharElement>();
foreach (PdfCharElement c in _listCharacters)
{
textElem.Characters.Add(new PdfCharElement
{
Char = c.Char,
Displacement = (c.Displacement * textElem.Matrix.Matrix[0, 0]),
});
}
textElem.Childs = new List<PdfTextElement>();
return textElem;
}
private void FlushTextElementSoft()
{
if (_sbText.Length == 0)
{
return;
}
PdfTextElement textElem = BuildTextElement();
if (_currentTextElement == null)
{
_currentTextElement = new PdfTextElement();
_currentTextElement.Font = null;
_currentTextElement.FontSize = -1;
_currentTextElement.Matrix = textElem.Matrix.Copy();
_currentTextElement.RawText = string.Empty;
_currentTextElement.VisibleText = string.Empty;
_currentTextElement.VisibleWidth = 0;
_currentTextElement.VisibleHeight = 0;
}
_currentTextElement.VisibleText += textElem.VisibleText;
_currentTextElement.VisibleWidth += textElem.VisibleWidth;
_currentTextElement.VisibleHeight = System.Math.Max(_currentTextElement.VisibleHeight, textElem.VisibleHeight);
_currentTextElement.Childs.Add(textElem);
_sbText = new StringBuilder();
_textWidth = 0;
}
private void AddTextElement(PdfTextElement textElement)
{
if (string.IsNullOrEmpty(textElement.VisibleText.Trim()))
@@ -307,27 +153,16 @@ namespace VAR.PdfTools
{
if (_sbText.Length == 0)
{
if (_currentTextElement != null)
{
AddTextElement(_currentTextElement);
_currentTextElement = null;
}
_textWidth = 0;
return;
}
if (_currentTextElement != null)
{
FlushTextElementSoft();
AddTextElement(_currentTextElement);
_currentTextElement = null;
}
else
{
PdfTextElement textElem = BuildTextElement();
AddTextElement(textElem);
}
_textMatrixCurrent.Matrix[0, 2] += _textWidth;
_sbText = new StringBuilder();
_listCharacters.Clear();
_textWidth = 0;
}
@@ -362,6 +197,21 @@ namespace VAR.PdfTools
return null;
}
private List<PdfTextElement> FindElementsContainingText(string text, bool fuzzy)
{
List<PdfTextElement> list = new List<PdfTextElement>();
string matchingText = fuzzy ? SimplifyText(text) : text;
foreach (PdfTextElement elem in _textElements)
{
string elemText = fuzzy ? SimplifyText(elem.VisibleText) : elem.VisibleText;
if (elemText.Contains(matchingText))
{
list.Add(elem);
}
}
return list;
}
private bool TextElementVerticalIntersection(PdfTextElement elem1, PdfTextElement elem2)
{
double elem1X1 = elem1.GetX();
@@ -391,36 +241,47 @@ namespace VAR.PdfTools
_graphicsMatrixStack.Add(_graphicsMatrix.Copy());
}
private void OpSetGraphMatrix(double a, double b, double c, double d, double e, double f)
{
_graphicsMatrix.Set(a, b, c, d, e, f);
}
private void OpPopGraphState()
{
_graphicsMatrix = _graphicsMatrixStack[_graphicsMatrixStack.Count - 1];
_graphicsMatrixStack.RemoveAt(_graphicsMatrixStack.Count - 1);
}
private void OpSetGraphMatrix(double a, double b, double c, double d, double e, double f)
{
_graphicsMatrix.Set(a, b, c, d, e, f);
}
private void OpBeginText()
{
_textMatrix.Idenity();
_textMatrixCurrent.Idenity();
inText = true;
}
private void OpEndText()
{
FlushTextElementSoft();
FlushTextElement();
inText = false;
}
private void OpTextFont(string fontName, double size)
{
FlushTextElementSoft();
FlushTextElement();
_font = _page.Fonts[fontName];
_fontSize = size;
}
private void OpTextCharSpacing(double charSpacing)
{
_charSpacing = charSpacing;
}
private void OpTextWordSpacing(double wordSpacing)
{
_wordSpacing = wordSpacing;
}
private void OpTextLeading(double textLeading)
{
_textLeading = textLeading;
@@ -433,6 +294,7 @@ namespace VAR.PdfTools
newMatrix.Matrix[0, 2] = x;
newMatrix.Matrix[1, 2] = y;
_textMatrix = newMatrix.Multiply(_textMatrix);
_textMatrixCurrent = _textMatrix.Copy();
}
private void OpTextLineFeed()
@@ -442,35 +304,10 @@ namespace VAR.PdfTools
private void OpSetTextMatrix(double a, double b, double c, double d, double e, double f)
{
double halfSpaceWidth = 0;
double horizontalDelta = 0;
Matrix3x3 newMatrix = new Matrix3x3(a, b, c, d, e, f);
if (_font != null)
{
halfSpaceWidth = _font.GetCharWidth(' ') * _fontSize;
}
horizontalDelta = (_textWidth + halfSpaceWidth);
if (_textMatrix.IsCollinear(newMatrix, horizontalDelta: horizontalDelta))
{
return;
}
if (_currentTextElement != null)
{
if (_currentTextElement.Font != null)
{
halfSpaceWidth = _currentTextElement.Font.GetCharWidth(' ') * _currentTextElement.FontSize;
}
horizontalDelta = (_currentTextElement.VisibleWidth + halfSpaceWidth);
if (_currentTextElement.Matrix.IsCollinear(newMatrix, horizontalDelta: horizontalDelta))
{
FlushTextElementSoft();
_textMatrix = newMatrix;
return;
}
}
FlushTextElement();
_textMatrix = newMatrix;
_textMatrixCurrent = _textMatrix.Copy();
}
private void OpTextPut(string text)
@@ -481,7 +318,12 @@ namespace VAR.PdfTools
{
foreach (char c in text)
{
_textWidth += _font.GetCharWidth(c) * _fontSize;
string realChar = _font.ToUnicode(c);
if (realChar == "\0") { continue; }
_listCharacters.Add(new PdfCharElement { Char = _font.ToUnicode(c), Displacement = _textWidth, });
double charWidth = _font.GetCharWidth(c) * _fontSize;
_textWidth += charWidth;
_textWidth += ((c == 0x20) ? _wordSpacing : _charSpacing);
}
}
}
@@ -491,17 +333,16 @@ namespace VAR.PdfTools
if (inText == false) { return; }
foreach (IPdfElement elem in array.Values)
{
if(elem is PdfString)
if (elem is PdfString)
{
OpTextPut(((PdfString)elem).Value);
}
else if(elem is PdfInteger || elem is PdfReal)
else if (elem is PdfInteger || elem is PdfReal)
{
double spacing = PdfElementUtils.GetReal(elem, 0);
// FIXME: Apply correctly spacing
//_textWidth += spacing;
_textWidth -= (spacing / 1000) * _fontSize;
}
else if(elem is PdfArray)
else if (elem is PdfArray)
{
OpTextPutMultiple(((PdfArray)elem));
}
@@ -512,13 +353,14 @@ namespace VAR.PdfTools
#region Private methods
private void ProcessPage()
private void ProcessPageContent()
{
int unknowCount = 0;
for (int i = 0; i < _page.ContentActions.Count; i++)
{
PdfContentAction action = _page.ContentActions[i];
// Graphics Operations
// Special graphics state
if (action.Token == "q")
{
OpPushGraphState();
@@ -549,11 +391,13 @@ namespace VAR.PdfTools
}
else if (action.Token == "Tc")
{
// FIXME: Char spacing
double charSpacing = PdfElementUtils.GetReal(action.Parameters[0], 0);
OpTextCharSpacing(charSpacing);
}
else if (action.Token == "Tw")
{
// FIXME: Word spacing
double wordSpacing = PdfElementUtils.GetReal(action.Parameters[0], 0);
OpTextWordSpacing(wordSpacing);
}
else if (action.Token == "Tz")
{
@@ -561,7 +405,7 @@ namespace VAR.PdfTools
}
else if (action.Token == "Tf")
{
string fontName = ((PdfName)action.Parameters[0]).Value;
string fontName = PdfElementUtils.GetString(action.Parameters[0], string.Empty);
double fontSize = PdfElementUtils.GetReal(action.Parameters[1], 0);
OpTextFont(fontName, fontSize);
}
@@ -607,18 +451,23 @@ namespace VAR.PdfTools
}
else if (action.Token == "Tj")
{
OpTextPut(((PdfString)action.Parameters[0]).Value);
string text = PdfElementUtils.GetString(action.Parameters[0], string.Empty);
OpTextPut(text);
}
else if (action.Token == "'")
{
string text = PdfElementUtils.GetString(action.Parameters[0], string.Empty);
OpTextLineFeed();
OpTextPut(((PdfString)action.Parameters[0]).Value);
OpTextPut(text);
}
else if (action.Token == "\"")
{
double wordSpacing = PdfElementUtils.GetReal(action.Parameters[0], 0);
double charSpacing = PdfElementUtils.GetReal(action.Parameters[1], 0);
OpTextPut(((PdfString)action.Parameters[2]).Value);
string text = PdfElementUtils.GetString(action.Parameters[0], string.Empty);
OpTextCharSpacing(charSpacing);
OpTextWordSpacing(wordSpacing);
OpTextPut(text);
}
else if (action.Token == "TJ")
{
@@ -672,6 +521,89 @@ namespace VAR.PdfTools
FlushTextElement();
}
private void JoinTextElements()
{
var textElementsCondensed = new List<PdfTextElement>();
while (_textElements.Count > 0)
{
PdfTextElement elem = _textElements[0];
_textElements.Remove(elem);
double blockY = elem.GetY();
double blockXMin = elem.GetX();
double blockXMax = blockXMin + elem.VisibleWidth;
// Prepare first neighbour
var textElementNeighbours = new List<PdfTextElement>();
textElementNeighbours.Add(elem);
// Search Neighbours
int i = 0;
while (i < _textElements.Count)
{
PdfTextElement neighbour = _textElements[i];
double neighbourY = neighbour.GetY();
if (Math.Abs(neighbourY - blockY) > 0.001) { i++; continue; }
double neighbourXMin = neighbour.GetX();
double neighbourXMax = neighbourXMin + neighbour.VisibleWidth;
double auxBlockXMin = blockXMin - elem.FontSize;
double auxBlockXMax = blockXMax + elem.FontSize;
if (auxBlockXMax >= neighbourXMin && neighbourXMax >= auxBlockXMin)
{
_textElements.Remove(neighbour);
textElementNeighbours.Add(neighbour);
if (blockXMax < neighbourXMax) { blockXMax = neighbourXMax; }
if (blockXMin > neighbourXMin) { blockXMin = neighbourXMin; }
i = 0;
continue;
}
i++;
}
if(textElementNeighbours.Count == 0)
{
textElementsCondensed.Add(elem);
continue;
}
// Join neighbours
var chars = new List<PdfCharElement>();
foreach (PdfTextElement neighbour in textElementNeighbours)
{
double neighbourXMin = neighbour.GetX();
foreach(PdfCharElement c in neighbour.Characters)
{
chars.Add(new PdfCharElement
{
Char = c.Char,
Displacement = (c.Displacement + neighbourXMin) - blockXMin,
});
}
}
chars = chars.OrderBy(c => c.Displacement).ToList();
var sbText = new StringBuilder();
foreach(PdfCharElement c in chars)
{
sbText.Append(c.Char);
}
PdfTextElement blockElem = new PdfTextElement
{
Font = null,
FontSize = elem.FontSize,
Matrix = elem.Matrix.Copy(),
RawText = sbText.ToString(),
VisibleText = sbText.ToString(),
VisibleWidth = blockXMax - blockXMin,
VisibleHeight = elem.VisibleHeight,
Characters = chars,
Childs = textElementNeighbours,
};
blockElem.Matrix.Matrix[0, 2] = blockXMin;
textElementsCondensed.Add(blockElem);
}
_textElements = textElementsCondensed;
}
#endregion
#region Public methods
@@ -684,7 +616,7 @@ namespace VAR.PdfTools
public List<string> GetColumn(string column, bool fuzzy)
{
PdfTextElement columnHead = FindElementByText(column, fuzzy);
if(columnHead == null)
if (columnHead == null)
{
return new List<string>();
}
@@ -697,7 +629,7 @@ namespace VAR.PdfTools
double extentX2 = double.MaxValue;
foreach (PdfTextElement elem in _textElements)
{
if(elem == columnHead){continue;}
if (elem == columnHead) { continue; }
if (TextElementHorizontalIntersection(columnHead, elem) == false) { continue; }
double elemX1 = elem.GetX();
double elemX2 = elemX1 + elem.VisibleWidth;
@@ -778,7 +710,7 @@ namespace VAR.PdfTools
fieldData.Add(elem);
}
if(fieldData.Count == 0)
if (fieldData.Count == 0)
{
return null;
}
@@ -793,12 +725,8 @@ namespace VAR.PdfTools
public bool HasText(string text, bool fuzzy)
{
PdfTextElement fieldTitle = FindElementByText(text, fuzzy);
if (fieldTitle == null)
{
return false;
}
return true;
List<PdfTextElement> list = FindElementsContainingText(text, fuzzy);
return (list.Count > 0);
}
#endregion

View File

@@ -11,4 +11,4 @@ using System.Runtime.InteropServices;
[assembly: AssemblyCulture("")]
[assembly: ComVisible(false)]
[assembly: Guid("eb7e003a-6a95-4002-809f-926c7c8a11e9")]
[assembly: AssemblyVersion("1.2.*")]
[assembly: AssemblyVersion("1.3.*")]

View File

@@ -1,63 +0,0 @@
<?xml version="1.0" encoding="utf-8"?>
<Project ToolsVersion="4.0" DefaultTargets="Build" xmlns="http://schemas.microsoft.com/developer/msbuild/2003">
<Import Project="$(MSBuildExtensionsPath)\$(MSBuildToolsVersion)\Microsoft.Common.props" Condition="Exists('$(MSBuildExtensionsPath)\$(MSBuildToolsVersion)\Microsoft.Common.props')" />
<PropertyGroup>
<Configuration Condition=" '$(Configuration)' == '' ">Debug</Configuration>
<Platform Condition=" '$(Platform)' == '' ">AnyCPU</Platform>
<ProjectGuid>{EB7E003A-6A95-4002-809F-926C7C8A11E9}</ProjectGuid>
<OutputType>Library</OutputType>
<AppDesignerFolder>Properties</AppDesignerFolder>
<RootNamespace>VAR.PdfTools</RootNamespace>
<AssemblyName>VAR.PdfTools</AssemblyName>
<TargetFrameworkVersion>v3.5</TargetFrameworkVersion>
<FileAlignment>512</FileAlignment>
<TargetFrameworkProfile />
<ProductVersion>10.0.0</ProductVersion>
<SchemaVersion>2.0</SchemaVersion>
</PropertyGroup>
<PropertyGroup Condition=" '$(Configuration)|$(Platform)' == 'Debug|AnyCPU' ">
<DebugSymbols>true</DebugSymbols>
<DebugType>full</DebugType>
<Optimize>false</Optimize>
<OutputPath>bin\Debug\</OutputPath>
<DefineConstants>DEBUG;TRACE</DefineConstants>
<ErrorReport>prompt</ErrorReport>
<WarningLevel>4</WarningLevel>
</PropertyGroup>
<PropertyGroup Condition=" '$(Configuration)|$(Platform)' == 'Release|AnyCPU' ">
<DebugType>pdbonly</DebugType>
<Optimize>true</Optimize>
<OutputPath>bin\Release\</OutputPath>
<DefineConstants>TRACE</DefineConstants>
<ErrorReport>prompt</ErrorReport>
<WarningLevel>4</WarningLevel>
</PropertyGroup>
<ItemGroup>
<Reference Include="System" />
<Reference Include="System.Core" />
<Reference Include="System.Xml.Linq" />
<Reference Include="System.Data.DataSetExtensions" />
<Reference Include="System.Data" />
<Reference Include="System.Xml" />
</ItemGroup>
<ItemGroup>
<Compile Include="PdfContentAction.cs" />
<Compile Include="PdfDocument.cs" />
<Compile Include="PdfDocumentPage.cs" />
<Compile Include="PdfElements.cs" />
<Compile Include="PdfFilters.cs" />
<Compile Include="PdfFont.cs" />
<Compile Include="PdfParser.cs" />
<Compile Include="PdfStandar14FontMetrics.cs" />
<Compile Include="PdfTextExtractor.cs" />
<Compile Include="Properties\AssemblyInfo.cs" />
</ItemGroup>
<Import Project="$(MSBuildToolsPath)\Microsoft.CSharp.targets" />
<!-- To modify your build process, add your task inside one of the targets below and uncomment it.
Other similar extension points exist, see Microsoft.Common.targets.
<Target Name="BeforeBuild">
</Target>
<Target Name="AfterBuild">
</Target>
-->
</Project>

View File

@@ -60,16 +60,31 @@
<Reference Include="System.Xml" />
</ItemGroup>
<ItemGroup>
<Compile Include="Maths\Matrix3x3.cs" />
<Compile Include="PdfContentAction.cs" />
<Compile Include="PdfDocument.cs" />
<Compile Include="PdfDocumentPage.cs" />
<Compile Include="PdfElements.cs" />
<Compile Include="PdfElements\IPdfElement.cs" />
<Compile Include="PdfElements\PdfArray.cs" />
<Compile Include="PdfElements\PdfBoolean.cs" />
<Compile Include="PdfElements\PdfDictionary.cs" />
<Compile Include="PdfElements\PdfElementTypes.cs" />
<Compile Include="PdfElements\PdfElementUtils.cs" />
<Compile Include="PdfFilters.cs" />
<Compile Include="PdfFont.cs" />
<Compile Include="PdfElements\PdfInteger.cs" />
<Compile Include="PdfElements\PdfName.cs" />
<Compile Include="PdfElements\PdfNull.cs" />
<Compile Include="PdfElements\PdfObject.cs" />
<Compile Include="PdfElements\PdfObjectReference.cs" />
<Compile Include="PdfElements\PdfReal.cs" />
<Compile Include="PdfElements\PdfStream.cs" />
<Compile Include="PdfElements\PdfString.cs" />
<Compile Include="PdfParser.cs" />
<Compile Include="PdfStandar14FontMetrics.cs" />
<Compile Include="PdfTextExtractor.cs" />
<Compile Include="Properties\AssemblyInfo.cs" />
<Compile Include="Maths\Vector3D.cs" />
</ItemGroup>
<ItemGroup>
<None Include="NuGet\keep.txt" />