'''' This code is part of Document Solutions for PDF .NET demos.'' Copyright (c) MESCIUS inc. All rights reserved.''Imports System.IOImports System.DrawingImports GrapeCity.Documents.PdfImports GrapeCity.Documents.Pdf.TextMapImports GrapeCity.Documents.TextImports GrapeCity.Documents.Pdf.Annotations
'' This sample loads an existing PDF, and imports a predefined list of key words,'' builds an alphabetical index of those words linked to pages where they occur'' in the document. The generated index pages are appended to the original document,'' and saved in a new PDF.'' The index is rendered in two balanced columns, imports the technique'' demonstrated in the BalancedColumns sample.'''' NOTE: if you download this sample and run it locally on your own system '' without a valid DsPdf license, only the first five pages of the sample PDF'' will be loaded, and the index will be generated for those five pages only.Public Class WordIndex
'' Font collection to hold the fonts we need: Private _fc As FontCollection = New FontCollection() '' Font family used throughout this sample (this is not case-sensitive): Const _fontFamily = "noto sans"
'' Main sample entry: Function CreatePDF(ByVal stream As Stream) As Integer '' Set up a font collection with the fonts we need: _fc.RegisterDirectory(Path.Combine("Resources", "Fonts"))
'' Get the PDF to add index to: Dim tfile = Path.Combine("Resources", "PDFs", "CompleteJavaScriptBook.pdf")
'' The list of words on which we will build the index: Dim words = _keywords.Distinct(StringComparer.InvariantCultureIgnoreCase).Where(Function(w_) Not String.IsNullOrEmpty(w_))
'' Load the PDF and add the index: Using fs = New FileStream(tfile, FileMode.Open, FileAccess.Read) Dim doc = New GcPdfDocument() doc.Load(fs) '' Dim origPageCount = doc.Pages.Count '' Build and add the index: AddWordIndex(doc, words) '' Open document on the first index page by default '' (may not work in browser viewers, but works in Acrobat): doc.OpenAction = New DestinationFit(origPageCount) '' Done: doc.Save(stream) Return doc.Pages.Count End Using End Function
'' The list of words to build the index on: Private ReadOnly _keywords() As String = { "JavaScript", "Framework", "MVC", "npm", "URL", "CDN", "HTML5", "CSS", "ES2015", "web", "Node.js", "API", "model", "view", "controller", "data management", "UI", "HTML", "API", "function", "var", "component", "design pattern", "React.js", "Angular", "AJAX", "DOM", "TypeScript", "ECMAScript", "CLI", "Wijmo", "CoffeeScript", "Elm", "plugin", "VueJS", "Knockout", "event", "AngularJS", "pure JS", "data binding", "OOP", "GrapeCity", "gauge", "JSX", "mobile", "desktop", "Vue", "template", "server-side", "client-side", "SPEC", "RAM", "ECMA" }
'' Calling FindText() on a document Or a page builds text maps for each page on the fly. '' Reusing cached text maps speeds things up a lot. Private Function FindTextPages(ByVal maps As ITextMap(), ByVal tp As FindTextParams) As SortedSet(Of Integer) Dim finds = New SortedSet(Of Integer) Dim currPageIdx = -1 For Each map In maps currPageIdx = map.Page.Index map.FindText(tp, Function(fp_) finds.Add(currPageIdx)) Next Return finds End Function
'' Adds a word index to the end of the passed document: Private Sub AddWordIndex(ByVal doc As GcPdfDocument, ByVal words As IEnumerable(Of String)) Dim tStart = Util.TimeNow()
'' Build text maps for all pages to speed up FindText() calls Dim textMaps(doc.Pages.Count - 1) As ITextMap For i = 0 To doc.Pages.Count - 1 textMaps(i) = doc.Pages(i).GetTextMap() Next
'' Words and page indices where they occur, sorted on words: Dim index = New SortedDictionary(Of String, List(Of Integer))()
'' Here the main loop building the index is on key words. '' An alternative would be to loop over the pages. '' Depending on the relative sizes of the keyword dictionary vs '' the number of pages in the document, one or the other might be better, '' but this is beyond the scope of this sample. For Each word In words Dim wholeWord As Boolean = word.IndexOf(" "c) = -1 Dim pgs = FindTextPages(textMaps, New FindTextParams(word, wholeWord, False)) '' A very simplistic way of also finding plurals: If wholeWord AndAlso Not word.EndsWith("s") Then pgs.UnionWith(FindTextPages(textMaps, New FindTextParams(word + "s", wholeWord, False))) End If If (pgs.Any()) Then index.Add(word, pgs.ToList()) End If Next
'' Prepare to render the index. The whole index is built '' in a single TextLayout instance, set up to render it '' in two columns per page. '' The main rendering loop uses the TextLayout.SplitAndBalance method '' imports the approach demonstrated in BalancedColumns sample. '' The complication here is that we need to associate a link to the '' relevant page with each page number rendered, see linkIndices below. '' Set up the TextLayout: Const margin = 72.0F Dim pageWidth = doc.PageSize.Width Dim pageHeight = doc.PageSize.Height Dim cW = pageWidth - margin * 2 '' Caption (index letter) format: Dim tfCap = New TextFormat() With { .FontName = _fontFamily, .FontBold = True, .FontSize = 16, .LineGap = 24 } '' Index word and pages format: Dim tfRun = New TextFormat() With { .FontName = _fontFamily, .FontSize = 10 } '' Page headers/footers: Dim tfHdr = New TextFormat() With { .FontName = _fontFamily, .FontItalic = True, .FontSize = 10 } '' FirstLineIndent = -18 sets up hanging indent: Dim tl = New TextLayout(72) With { .FontCollection = _fc, .FirstLineIndent = -18, .MaxWidth = pageWidth, .MaxHeight = pageHeight, .MarginLeft = margin, .MarginRight = margin, .MarginBottom = margin, .MarginTop = margin, .ColumnWidth = cW * 0.46F, .TextAlignment = TextAlignment.Leading, .ParagraphSpacing = 4, .LineGapBeforeFirstLine = False }
'' The list of text runs created for page numbers: Dim pgnumRuns = New List(Of Tuple(Of TextRun, Integer))() '' This loop builds the index on the TextLayout, saving the text runs '' created for each page number rendered. Note that at this point '' (prior to the PerformLayout(true) call) the text runs do not contain any info '' about their code points and render locations, so we can only save the text runs here. '' Later they will be used to add links to referenced pages in the PDF: Dim litera As Char = " " For Each kvp In index Dim word = kvp.Key Dim pageIndices = kvp.Value If Char.ToUpper(word(0)) <> litera Then litera = Char.ToUpper(word(0)) tl.Append($"{litera}{ChrW(&H2029)}", tfCap) End If tl.Append(word, tfRun) tl.Append(" ", tfRun) For i = 0 To pageIndices.Count - 1 Dim from_ = pageIndices(i) Dim tr = tl.Append((from_ + 1).ToString(), tfRun) pgnumRuns.Add(Tuple.Create(Of TextRun, Integer)(tr, from_)) '' We merge sequential pages into "..-M": Dim k = i For j = i + 1 To pageIndices.Count - 1 If pageIndices(j) <> pageIndices(j - 1) + 1 Then Exit For End If k = j Next If (k > i + 1) Then tl.Append("-", tfRun) Dim to_ = pageIndices(k) tr = tl.Append((to_ + 1).ToString(), tfRun) pgnumRuns.Add(Tuple.Create(Of TextRun, Integer)(tr, to_)) '' Fast forward: i = k End If If (i < pageIndices.Count - 1) Then tl.Append(", ", tfRun) Else tl.AppendLine(tfRun) End If Next Next '' This calculates the glyphs and lays out the whole index. '' The tl.SplitAndBalance() call in the loop below does not require redoing the layout: tl.PerformLayout(True)
'' '' Now we are ready to split and render the text layout, and also add links to page numbers. ''
'' Split areas and options - see BalancedColumns for details: Dim psas() As PageSplitArea = { New PageSplitArea(tl) With {.MarginLeft = tl.MarginLeft + (cW * 0.54F)} } Dim tso = New TextSplitOptions(tl) With { .KeepParagraphLinesTogether = True }
'' First original code point index in the current column: Dim cpiStart = 0 '' Max+1 original code point index in the current column: Dim cpiEnd = 0 '' Current index in pgnumRuns: Dim pgnumRunsIdx = 0
'' Method to add links to actual pages over page numbers in the current column: Dim linkIndices As Action(Of TextLayout, Page) = Sub(tl_, page_) cpiEnd += tl_.CodePointCount While pgnumRunsIdx < pgnumRuns.Count Dim run = pgnumRuns(pgnumRunsIdx) Dim textRun = run.Item1 Dim cpi = textRun.CodePointIndex If cpi >= cpiEnd Then Exit While End If cpi -= cpiStart Dim rects = tl_.GetTextRects(cpi, textRun.CodePointCount) Debug.Assert(rects.Count > 0) page_.Annotations.Add(New LinkAnnotation(rects(0).ToRectangleF(), New DestinationFit(run.Item2))) pgnumRunsIdx += 1 End While cpiStart += tl_.CodePointCount End Sub
'' Split and render the index in 2 columns: Dim page = doc.Pages.Add() While True Dim g = Page.Graphics '' Add a simple page header: g.DrawString($"Index generated by DsPdf on {tStart:R}", tfHdr, New RectangleF(margin, 0, pageWidth - margin * 2, margin), TextAlignment.Center, ParagraphAlignment.Center, False) '' 'rest' will accept the text that did not fit on this page: Dim rest As TextLayout = Nothing Dim splitResult = tl.SplitAndBalance(psas, tso, rest) '' Render text: g.DrawTextLayout(tl, PointF.Empty) g.DrawTextLayout(psas(0).TextLayout, PointF.Empty) '' Add links from page numbers to pages: linkIndices(tl, page) linkIndices(psas(0).TextLayout, page) '' Are we done yet? If splitResult <> SplitResult.Split Then Exit While End If tl = rest page = doc.Pages.Add() End While '' Done: End Sub
'' Creates a sample document with 100 pages of 'lorem ipsum': Private Function MakeDocumentToIndex() As String Const N = 100 Dim tfile = Path.GetTempFileName() Using fsOut = New FileStream(tfile, FileMode.Open, FileAccess.ReadWrite) Dim tdoc = New GcPdfDocument() '' See StartEndDoc for details on StartDoc/EndDoc mode: tdoc.StartDoc(fsOut) '' Prep a TextLayout to hold/format the text: Dim tl = New TextLayout(72) tl.FontCollection = _fc tl.DefaultFormat.FontName = _fontFamily tl.DefaultFormat.FontSize = 12 '' Use TextLayout to layout the whole page including margins: tl.MaxHeight = tdoc.PageSize.Height tl.MaxWidth = tdoc.PageSize.Width tl.MarginAll = 72 tl.FirstLineIndent = 72 / 2 '' Generate the document: For pageIdx = 0 To N - 1 tl.Append(Util.LoremIpsum(1)) tl.PerformLayout(True) tdoc.NewPage().Graphics.DrawTextLayout(tl, PointF.Empty) tl.Clear() Next tdoc.EndDoc() End Using Return tfile End FunctionEnd Class