Merge pull request #218 from gunnsth/release/v3.3.1

Prepare unipdf release v3.3.1

Author: gunnsth

common/version.go

@@ -10,13 +10,13 @@ import (
 	"time"
 )
 
-const releaseYear = 2019
-const releaseMonth = 9
-const releaseDay = 7
-const releaseHour = 13
-const releaseMin = 10
+const releaseYear = 2020
+const releaseMonth = 1
+const releaseDay = 4
+const releaseHour = 19
+const releaseMin = 5
 
 // Version holds version information, when bumping this make sure to bump the released at stamp also.
-const Version = "3.2.0"
+const Version = "3.3.1"
 
 var ReleasedAt = time.Date(releaseYear, releaseMonth, releaseDay, releaseHour, releaseMin, 0, 0, time.UTC)

contentstream/encoding.go

@@ -88,7 +88,7 @@ func newFlateEncoderFromInlineImage(inlineImage *ContentStreamInlineImage, decod
 	if decodeParams == nil {
 		obj := inlineImage.DecodeParms
 		if obj != nil {
-			dp, isDict := obj.(*core.PdfObjectDictionary)
+			dp, isDict := core.GetDict(obj)
 			if !isDict {
 				common.Log.Debug("Error: DecodeParms not a dictionary (%T)", obj)
 				return nil, fmt.Errorf("invalid DecodeParms")
@@ -163,7 +163,7 @@ func newLZWEncoderFromInlineImage(inlineImage *ContentStreamInlineImage, decodeP
 	// If decodeParams not provided, see if we can get from the inline image directly.
 	if decodeParams == nil {
 		if inlineImage.DecodeParms != nil {
-			dp, isDict := inlineImage.DecodeParms.(*core.PdfObjectDictionary)
+			dp, isDict := core.GetDict(inlineImage.DecodeParms)
 			if !isDict {
 				common.Log.Debug("Error: DecodeParms not a dictionary (%T)", inlineImage.DecodeParms)
 				return nil, fmt.Errorf("invalid DecodeParms")

contentstream/inline-image.go

@@ -316,7 +316,7 @@ func (csp *ContentStreamParser) ParseInlineImage() (*ContentStreamInlineImage, e
 
 		if !isOperand {
 			// Not an operand.. Read key value properties..
-			param, ok := obj.(*core.PdfObjectName)
+			param, ok := core.GetName(obj)
 			if !ok {
 				common.Log.Debug("Invalid inline image property (expecting name) - %T", obj)
 				return nil, fmt.Errorf("invalid inline image property (expecting name) - %T", obj)

contentstream/processor.go

@@ -7,6 +7,7 @@ package contentstream
 
 import (
 	"errors"
+	"fmt"
 
 	"github.com/unidoc/unipdf/v3/common"
 	"github.com/unidoc/unipdf/v3/core"
@@ -141,7 +142,7 @@ func (proc *ContentStreamProcessor) getColorspace(name string, resources *model.
 
 	// Otherwise unsupported.
 	common.Log.Debug("Unknown colorspace requested: %s", name)
-	return nil, errors.New("unsupported colorspace")
+	return nil, fmt.Errorf("unsupported colorspace: %s", name)
 }
 
 // Get initial color for a given colorspace.

core/encoding.go

@@ -443,7 +443,7 @@ func (enc *FlateEncoder) EncodeBytes(data []byte) ([]byte, error) {
 		rowLength := int(enc.Columns)
 		rows := len(data) / rowLength
 		if len(data)%rowLength != 0 {
-			common.Log.Error("Invalid column length")
+			common.Log.Error("Invalid row length")
 			return nil, errors.New("invalid row length")
 		}
 
@@ -590,7 +590,7 @@ func newLZWEncoderFromStream(streamObj *PdfObjectStream, decodeParams *PdfObject
 
 	// If decodeParams not provided, see if we can get from the stream.
 	if decodeParams == nil {
-		obj := encDict.Get("DecodeParms")
+		obj := TraceToDirectObject(encDict.Get("DecodeParms"))
 		if obj != nil {
 			if dp, isDict := obj.(*PdfObjectDictionary); isDict {
 				decodeParams = dp
@@ -1751,7 +1751,7 @@ func newCCITTFaxEncoderFromStream(streamObj *PdfObjectStream, decodeParams *PdfO
 
 	// If decodeParams not provided, see if we can get from the stream.
 	if decodeParams == nil {
-		obj := encDict.Get("DecodeParms")
+		obj := TraceToDirectObject(encDict.Get("DecodeParms"))
 		if obj != nil {
 			switch t := obj.(type) {
 			case *PdfObjectDictionary:
@@ -2061,7 +2061,7 @@ func newJBIG2EncoderFromStream(streamObj *PdfObjectStream, decodeParams *PdfObje
 
 	if decodeParams != nil {
 		if globals := decodeParams.Get("JBIG2Globals"); globals != nil {
-			globalsStream, ok := globals.(*PdfObjectStream)
+			globalsStream, ok := GetStream(globals)
 			if !ok {
 				err := errors.New("the Globals stream should be an Object Stream")
 				common.Log.Debug("ERROR: %s", err.Error())

core/symbols.go

@@ -22,52 +22,26 @@ func IsFloatDigit(c byte) bool {
 
 // IsDecimalDigit checks if the character is a part of a decimal number string.
 func IsDecimalDigit(c byte) bool {
-	if c >= '0' && c <= '9' {
-		return true
-	}
-
-	return false
+	return '0' <= c && c <= '9'
 }
 
 // IsOctalDigit checks if a character can be part of an octal digit string.
 func IsOctalDigit(c byte) bool {
-	if c >= '0' && c <= '7' {
-		return true
-	}
-
-	return false
+	return '0' <= c && c <= '7'
 }
 
 // IsPrintable checks if a character is printable.
 // Regular characters that are outside the range EXCLAMATION MARK(21h)
 // (!) to TILDE (7Eh) (~) should be written using the hexadecimal notation.
-func IsPrintable(char byte) bool {
-	if char < 0x21 || char > 0x7E {
-		return false
-	}
-	return true
+func IsPrintable(c byte) bool {
+	return 0x21 <= c && c <= 0x7E
 }
 
 // IsDelimiter checks if a character represents a delimiter.
-func IsDelimiter(char byte) bool {
-	if char == '(' || char == ')' {
-		return true
-	}
-	if char == '<' || char == '>' {
-		return true
-	}
-	if char == '[' || char == ']' {
-		return true
-	}
-	if char == '{' || char == '}' {
-		return true
-	}
-	if char == '/' {
-		return true
-	}
-	if char == '%' {
-		return true
-	}
-
-	return false
+func IsDelimiter(c byte) bool {
+	return c == '(' || c == ')' ||
+		c == '<' || c == '>' ||
+		c == '[' || c == ']' ||
+		c == '{' || c == '}' ||
+		c == '/' || c == '%'
 }

creator/creator.go

@@ -346,9 +346,25 @@ func (c *Creator) Context() DrawContext {
 	return c.context
 }
 
-// Call before writing out. Takes care of adding headers and footers, as well
-// as generating front Page and table of contents.
-func (c *Creator) finalize() error {
+// Finalize renders all blocks to the creator pages. In addition, it takes care
+// of adding headers and footers, as well as generating the front page,
+// table of contents and outlines.
+// Finalize is automatically called before writing the document out. Calling the
+// method manually can be useful when adding external pages to the creator,
+// using the AddPage method, as it renders all creator blocks to the added
+// pages, without having to write the document out.
+// NOTE: TOC and outlines are generated only if the AddTOC and AddOutlines
+// fields of the creator are set to true (enabled by default). Furthermore, TOCs
+// and outlines without content are skipped. TOC and outline content is
+// added automatically when using the chapter component. TOCs and outlines can
+// also be set externally, using the SetTOC and SetOutlineTree methods.
+// Finalize should only be called once, after all draw calls have taken place,
+// as it will return immediately if the creator instance has been finalized.
+func (c *Creator) Finalize() error {
+	if c.finalized {
+		return nil
+	}
+
 	totPages := len(c.pages)
 
 	// Estimate number of additional generated pages and update TOC.
@@ -563,8 +579,12 @@ func (c *Creator) MoveDown(dy float64) {
 	c.context.Y += dy
 }
 
-// Draw draws the Drawable widget to the document.  This can span over 1 or more pages. Additional
-// pages are added if the contents go over the current Page.
+// Draw processes the specified Drawable widget and generates blocks that can
+// be rendered to the output document. The generated blocks can span over one
+// or more pages. Additional pages are added if the contents go over the current
+// page. Each generated block is assigned to the creator page it will be
+// rendered to. In order to render the generated blocks to the creator pages,
+// call Finalize, Write or WriteToFile.
 func (c *Creator) Draw(d Drawable) error {
 	if c.getActivePage() == nil {
 		// Add a new Page if none added already.
@@ -604,10 +624,8 @@ func (c *Creator) Draw(d Drawable) error {
 
 // Write output of creator to io.Writer interface.
 func (c *Creator) Write(ws io.Writer) error {
-	if !c.finalized {
-		if err := c.finalize(); err != nil {
-			return err
-		}
+	if err := c.Finalize(); err != nil {
+		return err
 	}
 
 	pdfWriter := model.NewPdfWriter()

creator/creator_test.go

@@ -822,7 +822,7 @@ func TestSubchaptersSimple(t *testing.T) {
 	c.AddTOC = true
 
 	lineStyle := c.NewTextStyle()
-	lineStyle.Font = model.NewStandard14FontMustCompile(model.HelveticaBoldName)
+	lineStyle.Font = model.NewStandard14FontMustCompile(model.HelveticaName)
 
 	toc := c.TOC()
 	toc.SetLineStyle(lineStyle)
@@ -898,9 +898,6 @@ func TestSubchaptersSimple(t *testing.T) {
 		toc.SetHeading("Table of Contents", style)
 
 		// Set style of TOC lines just before render.
-		lineStyle := c.NewTextStyle()
-		lineStyle.FontSize = 14
-
 		helveticaBold := model.NewStandard14FontMustCompile(model.HelveticaBoldName)
 
 		lines := toc.Lines()
@@ -923,7 +920,6 @@ func TestSubchapters(t *testing.T) {
 
 	lineStyle := c.NewTextStyle()
 	lineStyle.Font = model.NewStandard14FontMustCompile(model.HelveticaName)
-	lineStyle.FontSize = 14
 	lineStyle.Color = ColorRGBFromArithmetic(0.5, 0.5, 0.5)
 
 	toc := c.TOC()

creator/toc_line.go

@@ -155,7 +155,12 @@ func (tl *TOCLine) SetLink(page int64, x, y float64) {
 	tl.linkY = y
 	tl.linkPage = page
 
-	tl.SetStyle(tl.sp.defaultLinkStyle)
+	// Set the color of the line components to the default link color.
+	linkColor := tl.sp.defaultLinkStyle.Color
+	tl.Number.Style.Color = linkColor
+	tl.Title.Style.Color = linkColor
+	tl.Separator.Style.Color = linkColor
+	tl.Page.Style.Color = linkColor
 }
 
 // getLineLink returns a new annotation if the line has a link set.

extractor/text.go

@@ -61,7 +61,8 @@ func (e *Extractor) extractPageText(contents string, resources *model.PdfPageRes
 	pageText := &PageText{}
 	state := newTextState()
 	fontStack := fontStacker{}
-	var to *textObject
+	to := newTextObject(e, resources, contentstream.GraphicsState{}, &state, &fontStack)
+	var inTextObj bool
 
 	cstreamParser := contentstream.NewContentStreamParser(contents)
 	operations, err := cstreamParser.Parse()
@@ -102,16 +103,31 @@ func (e *Extractor) extractPageText(contents string, resources *model.PdfPageRes
 					state.tfont = fontStack.pop()
 				}
 			case "BT": // Begin text
-				// Begin a text object, initializing the text matrix, Tm, and the text line matrix,
-				// Tlm, to the identity matrix. Text objects shall not be nested; a second BT shall
-				// not appear before an ET.
-				if to != nil {
+				// Begin a text object, initializing the text matrix, Tm, and
+				// the text line matrix, Tlm, to the identity matrix. Text
+				// objects shall not be nested. A second BT shall not appear
+				// before an ET. However, if that happens, all existing marks
+				// are added to the  page marks, in order to avoid losing content.
+				if inTextObj {
 					common.Log.Debug("BT called while in a text object")
+					pageText.marks = append(pageText.marks, to.marks...)
 				}
+				inTextObj = true
 				to = newTextObject(e, resources, gs, &state, &fontStack)
 			case "ET": // End Text
+				// End text object, discarding text matrix. If the current
+				// text object contains text marks, they are added to the
+				// page text marks collection.
+				// The ET operator should always have a matching BT operator.
+				// However, if ET appears outside of a text object, the behavior
+				// does not change: the text matrices are discarded and all
+				// existing marks in the text object are added to the page marks.
+				if !inTextObj {
+					common.Log.Debug("ET called outside of a text object")
+				}
+				inTextObj = false
 				pageText.marks = append(pageText.marks, to.marks...)
-				to = nil
+				to.reset()
 			case "T*": // Move to start of next text line
 				to.nextLine()
 			case "Td": // Move text location
@@ -202,10 +218,6 @@ func (e *Extractor) extractPageText(contents string, resources *model.PdfPageRes
 				}
 				to.setCharSpacing(y)
 			case "Tf": // Set font.
-				if to == nil {
-					// This is needed for 26-Hazard-Thermal-environment.pdf
-					to = newTextObject(e, resources, gs, &state, &fontStack)
-				}
 				if ok, err := to.checkOp(op, 2, true); !ok {
 					common.Log.Debug("ERROR: Tf err=%v", err)
 					return err
@@ -659,6 +671,14 @@ func newTextObject(e *Extractor, resources *model.PdfPageResources, gs contentst
 	}
 }
 
+// reset sets the text matrix `Tm` and the text line matrix `Tlm` of the text
+// object to the identity matrix. In addition, the marks collection is cleared.
+func (to *textObject) reset() {
+	to.tm = transform.IdentityMatrix()
+	to.tlm = transform.IdentityMatrix()
+	to.marks = nil
+}
+
 // renderText processes and renders byte array `data` for extraction purposes.
 func (to *textObject) renderText(data []byte) error {
 	font := to.getCurrentFont()
@@ -1205,7 +1225,7 @@ func (pt *PageText) sortPosition(tol float64) {
 		if pt.marks[i-1].orient != pt.marks[i].orient {
 			cluster++
 		} else {
-			if pt.marks[i-1].orientedStart.Y - pt.marks[i].orientedStart.Y > tol {
+			if pt.marks[i-1].orientedStart.Y-pt.marks[i].orientedStart.Y > tol {
 				cluster++
 			}
 		}