Section:

PDF Data Extraction - Images, Text, Paths - Go Sample Code

Sample code for using Apryse SDK to extract text, paths, and images from a PDF. The sample also shows how to do color conversion, image normalization, and process changes in the graphics state. Sample code provided in Python, C++, C#, Java, Node.js (JavaScript), PHP, Ruby and VB.

Learn more about our full PDF Data Extraction SDK Capabilities.

To start your free trial, get stated with Server SDK.

1//---------------------------------------------------------------------------------------
2// Copyright (c) 2001-2021 by PDFTron Systems Inc. All Rights Reserved.
3// Consult LICENSE.txt regarding license information.
4//---------------------------------------------------------------------------------------
5
6package main
7import (
8	"fmt"
9    "os"
10    "strconv"
11	. "pdftron"
12)
13
14import  "pdftron/Samples/LicenseKey/GO"
15
16func ProcessPath(reader ElementReader, path Element){
17    if path.IsClippingPath(){
18        fmt.Println("This is a clipping path")
19    }
20
21    pathData := path.GetPathData()
22    data := pathData.GetPoints()
23    opr := pathData.GetOperators()
24
25    oprIndex := 0
26    oprEnd := int(opr.Size())
27    dataIndex := 0
28    //dataEnd := data.Size()
29    
30    // Use path.GetCTM() if you are interested in CTM (current transformation matrix).
31    
32    os.Stdout.Write([]byte("Path Data Points := \""))
33    x1, x2, x3, x4 := 0.0, 0.0, 0.0, 0.0
34    y1, y2, y3, y4 := 0.0, 0.0, 0.0, 0.0
35    for oprIndex < oprEnd{
36        if int(opr.Get(oprIndex)) == int(PathDataE_moveto){
37            x1 = data.Get(dataIndex) 
38            dataIndex = dataIndex + 1
39            y1 = data.Get(dataIndex)
40            dataIndex = dataIndex + 1
41            os.Stdout.Write([]byte("M" + fmt.Sprintf("%f", x1) + " " + fmt.Sprintf("%f", y1)))
42        }else if int(opr.Get(oprIndex)) == int(PathDataE_lineto){
43            x1 = data.Get(dataIndex) 
44            dataIndex = dataIndex + 1
45            y1 = data.Get(dataIndex)
46            dataIndex = dataIndex + 1
47            os.Stdout.Write([]byte(" L" + fmt.Sprintf("%f", x1) + " " + fmt.Sprintf("%f", y1)))
48        }else if int(opr.Get(oprIndex)) == int(PathDataE_cubicto){
49            x1 = data.Get(dataIndex)
50            dataIndex = dataIndex + 1
51            y1 = data.Get(dataIndex)
52            dataIndex = dataIndex + 1
53            x2 = data.Get(dataIndex)
54            dataIndex = dataIndex + 1
55            y2 = data.Get(dataIndex)
56            dataIndex = dataIndex + 1
57            x3 = data.Get(dataIndex)
58            dataIndex = dataIndex + 1
59            y3 = data.Get(dataIndex)
60            dataIndex = dataIndex + 1
61            os.Stdout.Write([]byte(" C" + fmt.Sprintf("%f", x1) + " " + fmt.Sprintf("%f", y1) + " " + fmt.Sprintf("%f", x2) + " " + fmt.Sprintf("%f", y2) + " " + fmt.Sprintf("%f", x3) + " " + fmt.Sprintf("%f", y3)))
62        }else if int(opr.Get(oprIndex)) == int(PathDataE_rect){
63            x1 = data.Get(dataIndex)
64            dataIndex = dataIndex + 1
65            y1 = data.Get(dataIndex)
66            dataIndex = dataIndex + 1
67            w := data.Get(dataIndex)
68            dataIndex = dataIndex + 1
69            h := data.Get(dataIndex)
70            dataIndex = dataIndex + 1
71            x2 = x1 + w
72            y2 = y1
73            x3 = x2
74            y3 = y1 + h
75            x4 = x1
76            y4 = y3
77            os.Stdout.Write([]byte("M" + fmt.Sprintf("%.2f", x1) + " " + fmt.Sprintf("%.2f", y1) + " L" + fmt.Sprintf("%.2f", x2) + " " + fmt.Sprintf("%.2f", y2) + " L" + fmt.Sprintf("%.2f", x3) + " " + fmt.Sprintf("%.2f", y3) + " L" + fmt.Sprintf("%.2f", x4) + " " + fmt.Sprintf("%.2f", y4) + " Z"))
78        }else if int(opr.Get(oprIndex)) == int(PathDataE_closepath){
79            fmt.Println(" Close Path")
80        }else{
81            //
82        }
83        oprIndex = oprIndex + 1
84    }
85
86    os.Stdout.Write([]byte("\" "))
87    gs := path.GetGState()
88    
89    // Set Path State 0 (stroke, fill, fill-rule) -----------------------------------
90    if path.IsStroked(){
91        fmt.Println("Stroke path")
92        
93        if (gs.GetStrokeColorSpace().GetType() == ColorSpaceE_pattern){
94            fmt.Println("Path has associated pattern")
95        }else{
96            // Get stroke color (you can use PDFNet color conversion facilities)
97            // rgb = gs.GetStrokeColorSpace().Convert2RGB(gs.GetStrokeColor())
98        }
99    }else{
100        // Do not stroke path
101    }
102
103    if path.IsFilled(){
104        fmt.Println("Fill path")
105        
106        if (gs.GetFillColorSpace().GetType() == ColorSpaceE_pattern){
107            fmt.Println("Path has associated pattern")
108        }else{
109            // rgb = gs.GetFillColorSpace().Convert2RGB(gs.GetFillColor())
110        }
111    }else{
112        // Do not fill path
113    }
114
115    // Process any changes in graphics state  ---------------------------------
116    gsItr := reader.GetChangesIterator()
117    for gsItr.HasNext(){
118        if int(gsItr.Current()) == int(GStateE_transform){
119            // Get transform matrix for this element. Unlike path.GetCTM() 
120            // that return full transformation matrix gs.GetTransform() return 
121            // only the transformation matrix that was installed for this element.
122            //
123            // gs.GetTransform()
124            
125        }else if int(gsItr.Current()) == int(GStateE_line_width){
126            // gs.GetLineWidth()
127            
128        }else if int(gsItr.Current()) == int(GStateE_line_cap){
129            // gs.GetLineCap()
130            
131        }else if int(gsItr.Current()) == int(GStateE_line_join){
132            // gs.GetLineJoin()
133            
134        }else if int(gsItr.Current()) == int(GStateE_flatness){
135            
136        }else if int(gsItr.Current()) == int(GStateE_miter_limit){
137            // gs.GetMiterLimit()
138            
139        }else if int(gsItr.Current()) == int(GStateE_dash_pattern){
140            // dashes = gs.GetDashes()
141            // gs.GetPhase()
142            
143        }else if int(gsItr.Current()) == int(GStateE_fill_color){
144            if (int(gs.GetFillColorSpace().GetType()) == int(ColorSpaceE_pattern) && int(gs.GetFillPattern().GetType()) != int(PatternColorE_shading) ){
145                // process the pattern data
146                reader.PatternBegin(true)
147                ProcessElements(reader)
148                reader.End()
149            }
150        }
151        gsItr.Next()
152    }
153    reader.ClearChangeList()
154}
155
156func ProcessText (pageReader ElementReader){
157    // Begin text element
158    fmt.Println("Begin Text Block:")
159    
160    element := pageReader.Next()
161    
162    for element.GetMp_elem().Swigcptr() != 0{
163        etype := element.GetType()
164        if etype == ElementE_text_end{
165            // Finish the text block
166            fmt.Println("End Text Block.")
167            return
168        }else if etype == ElementE_text{
169            gs := element.GetGState()
170            
171            //csFill := gs.GetFillColorSpace()
172            //fill := gs.GetFillColor()
173            
174            //out := csFill.Convert2RGB(fill)
175            
176            //csStroke := gs.GetStrokeColorSpace()
177            //stroke := gs.GetStrokeColor()
178            
179            font := gs.GetFont()
180            fmt.Println("Font Name: " + font.GetName())
181            // font.IsFixedWidth()
182            // font.IsSerif()
183            // font.IsSymbolic()
184            // font.IsItalic()
185            // ... 
186
187            // fontSize = gs.GetFontSize()
188            // wordSpacing = gs.GetWordSpacing()
189            // charSpacing = gs.GetCharSpacing()
190            // txt := element.GetTextString()
191            if font.GetType() == FontE_Type3{
192                // type 3 font, process its data
193                itr := element.GetCharIterator()
194                for itr.HasNext(){
195                    pageReader.Type3FontBegin(itr.Current())
196                    ProcessElements(pageReader)
197                    pageReader.End()
198                }
199            }else{
200                text_mtx := element.GetTextMatrix()
201                
202                itr := element.GetCharIterator()
203                for itr.HasNext(){
204                    charCode := itr.Current().GetChar_data()
205                    if *charCode >= 32 && *charCode <= 255 {     // Print if in ASCII range...
206                        a := font.MapToUnicode(uint(*charCode))
207                        os.Stdout.Write([]byte( a )) // Revisit: if sys.version_info.major < 3 else ascii(a[0]) ))
208                    }    
209                    pt := NewPoint()   
210                    pt.SetX(itr.Current().GetX())     // character positioning information
211                    pt.SetY(itr.Current().GetY())
212                    
213                    // Use element.GetCTM() if you are interested in the CTM 
214                    // (current transformation matrix).
215                    ctm := element.GetCTM()
216                    
217                    // To get the exact character positioning information you need to 
218                    // concatenate current text matrix with CTM and then multiply 
219                    // relative positioning coordinates with the resulting matrix.
220                    mtx := ctm.Multiply(text_mtx)
221                    mtx.Mult(pt)
222                    itr.Next()
223                }
224            }
225            fmt.Println("")
226        }
227        element = pageReader.Next()
228    }
229}
230
231func ProcessImage (image Element){
232    //imageMask := image.IsImageMask()
233    //interpolate := image.IsImageInterpolate()
234    width := image.GetImageWidth()
235    height := image.GetImageHeight()
236    outDataSz := width * height * 3
237    
238    fmt.Println("Image: width=\"" + fmt.Sprintf("%d", width) + "\"" + " height=\"" + fmt.Sprintf("%d", height)+ "\"" )
239    
240    // Matrix2D& mtx = image->GetCTM() // image matrix (page positioning info)
241
242    // You can use GetImageData to read the raw (decoded) image data
243    //image->GetBitsPerComponent()    
244    //image->GetImageData()    // get raw image data
245    // .... or use Image2RGB filter that converts every image to RGB format,
246    // This should save you time since you don't need to deal with color conversions, 
247    // image up-sampling, decoding etc.
248    
249    imgConv := NewImage2RGB(image)     // Extract and convert image to RGB 8-bps format
250    reader := NewFilterReader(imgConv)
251
252    //imageDataOut := reader.Read(int64(outDataSz))
253    reader.Read(int64(outDataSz))
254    
255    // Note that you don't need to read a whole image at a time. Alternatively
256    // you can read a chuck at a time by repeatedly calling reader.Read(buf, buf_sz) 
257    // until the function returns 0. 
258}
259
260func ProcessElements(reader ElementReader){
261    element := reader.Next()     // Read page contents
262    for element.GetMp_elem().Swigcptr() != 0{
263        etype := element.GetType()
264        if etype == ElementE_path{      // Process path data...
265            ProcessPath(reader, element)
266        }else if etype == ElementE_text_begin{      // Process text block...
267            ProcessText(reader)
268        }else if etype == ElementE_form{    // Process form XObjects
269            reader.FormBegin()
270            ProcessElements(reader)
271            reader.End()
272        }else if etype == ElementE_image{    // Process Images
273            ProcessImage(element)
274        }
275        element = reader.Next()
276    }
277}
278
279func main(){
280    PDFNetInitialize(PDFTronLicense.Key)
281    
282    // Relative path to the folder containing the test files.
283    inputPath := "../../TestFiles/"
284    //outputPath := "../../TestFiles/Output/"
285    
286    // Extract text data from all pages in the document
287    
288    fmt.Println("__________________________________________________")
289    fmt.Println("Extract page element information from all ")
290    fmt.Println("pages in the document.")
291    
292    doc := NewPDFDoc(inputPath + "newsletter.pdf")
293    doc.InitSecurityHandler()
294    //pgnum := doc.GetPageCount()
295    pageBegin := doc.GetPageIterator()
296    pageReader := NewElementReader()
297    
298    itr := pageBegin
299    for itr.HasNext(){    // Read every page
300        fmt.Println("Page " + strconv.Itoa(itr.Current().GetIndex()) + "----------------------------------------")
301        pageReader.Begin(itr.Current())
302        ProcessElements(pageReader)
303        pageReader.End()
304        itr.Next()
305    }
306    doc.Close()
307    PDFNetTerminate()
308    fmt.Println("Done.")
309}

Did you find this helpful?

Trial setup questions?

Ask experts on Discord

Need other help?

Contact Support

Pricing or product questions?

Contact Sales