group 3 decoding

Author: hhrutter

README.md

@@ -13,15 +13,14 @@ It provides both an API and a CLI. Supported are all versions up to PDF 1.7 (ISO
 
 ## Status
 
-Version: 0.1.16
-
-* Introducing preliminary support for the CCITTFaxDecode filter.
-* Supported is the decoding of CCITTGroup 4 compressed data (K < 0) - encoding and group 3 support yet to come.
-* pdfcpu now extracts group 4 compressed black and white images to PNG.
-* pdfcpu's tiff reader is also supporting group 4 compressed images as of this release.
-* #38 extraction of metadata: `pdfcpu extract -mode meta` extracts optional XML metadata.
-* Refactored package structure.
-* Bugfixes #40, #41.
+Version: 0.1.17
+
+* Improved support for the CCITTFaxDecode filter.
+* Supported is the decoding of CCITTGroup 4 compressed data (for K<0)
+* Supported is also the 1D decoding of CCITTGroup 3 compressed data (for K=0)
+* CCITT Group3 2D(mixed) decoding is pending available test images (K>0).
+* pdfcpu now extracts group 3 and 4 encoded black and white images to PNG.
+* pdfcpu's tiff reader is also supporting CCITT group 3-1D and group 4 encoded images as of this release.
 
 ## Motivation
 

ccitt/ccitt_test.go

@@ -96,7 +96,7 @@ func compare(t *testing.T, fileName string, img1, img2 image.Image) {
 	}
 }
 
-func testFile(t *testing.T, fileName string, w, h int, inverse, align bool) {
+func testFile(t *testing.T, fileName string, mode, w, h int, inverse, align bool) {
 
 	f, err := os.Open(fileName)
 	if err != nil {
@@ -105,8 +105,8 @@ func testFile(t *testing.T, fileName string, w, h int, inverse, align bool) {
 	}
 	defer f.Close()
 
-	// Read & decode CCITT Group 4 file into buf.
-	r := NewReader(f, w, inverse, align)
+	// Read a CCITT encoded file and decode it into buf.
+	r := NewReader(f, mode, w, inverse, align)
 	buf, err := ioutil.ReadAll(r)
 	if err != nil {
 		t.Errorf("%s: %v", fileName, err)
@@ -126,32 +126,37 @@ func testFile(t *testing.T, fileName string, w, h int, inverse, align bool) {
 		return
 	}
 
-	// Compare images
+	// Compare images.
 	compare(t, fnNoExt, img1, img2)
 }
 
 func TestCCITT(t *testing.T) {
 
-	// Test Group 4 encoding
-
 	for _, tt := range []struct {
 		fileName string
+		mode     int
 		w, h     int
 		inverse  bool
 		align    bool
 	}{
-		{"testdata/amt.gr4", 43, 38, false, false},
-		{"testdata/lc.gr4", 154, 154, false, false},
-		{"testdata/do.gr4", 613, 373, true, false}, // <BlackIs1, true>
-		{"testdata/t6diagram.gr4", 1163, 2433, false, false},
-		{"testdata/Wonderwall.gr4", 2312, 3307, false, false}, // extracted from pdfcpu/pkg/api/testdata/Wonderwall.pdf
-		{"testdata/hoare.gr4", 2550, 3300, false, false},
-		{"testdata/jphys.gr4", 3440, 5200, false, false},
-		{"testdata/hl.gr4", 2548, 3300, false, true}, // <EncodedByteAlign, true> <EndOfLine, false>
-		{"testdata/ho2.gr4", 2040, 2640, false, false},
-		{"testdata/sie.gr4", 3310, 8672, false, false},
+
+		// Test Group 3 decoding
+		{"testdata/scan1.gr3", Group3, 2480, 3508, false, false},
+		{"testdata/scan2.gr3", Group3, 1656, 2339, false, false},
+
+		// // Test Group 4 decoding
+		{"testdata/amt.gr4", Group4, 43, 38, false, false},
+		{"testdata/lc.gr4", Group4, 154, 154, false, false},
+		{"testdata/do.gr4", Group4, 613, 373, true, false}, // <BlackIs1, true>
+		{"testdata/t6diagram.gr4", Group4, 1163, 2433, false, false},
+		{"testdata/Wonderwall.gr4", Group4, 2312, 3307, false, false},
+		{"testdata/hoare.gr4", Group4, 2550, 3300, false, false},
+		{"testdata/jphys.gr4", Group4, 3440, 5200, false, false},
+		{"testdata/hl.gr4", Group4, 2548, 3300, false, true}, // <EncodedByteAlign, true>
+		{"testdata/ho2.gr4", Group4, 2040, 2640, false, false},
+		{"testdata/sie.gr4", Group4, 3310, 8672, false, false},
 	} {
-		testFile(t, tt.fileName, tt.w, tt.h, tt.inverse, tt.align)
+		testFile(t, tt.fileName, tt.mode, tt.w, tt.h, tt.inverse, tt.align)
 	}
 
 }

ccitt/codes.go

@@ -17,17 +17,18 @@ limitations under the License.
 package ccitt
 
 const (
-	mP   = "0001"                     // pass mode
-	mH   = "001"                      // horizontal mode
-	mV0  = "1"                        // vertical mode, a1 = b1
-	mVR1 = "011"                      // vertical mode, a1 = b1 + 1
-	mVR2 = "000011"                   // vertical mode, a1 = b1 + 2
-	mVR3 = "0000011"                  // vertical mode, a1 = b1 + 3
-	mVL1 = "010"                      // vertical mode, a1 = b1 - 1
-	mVL2 = "000010"                   // vertical mode, a1 = b1 - 2
-	mVL3 = "0000010"                  // vertical mode, a1 = b1 - 3
-	ext  = "0000001111"               // extension mode, enter uncompressed mode
-	eofb = "000000000001000000000001" // end of facsimile block
+	mP   = "0001"         // pass mode
+	mH   = "001"          // horizontal mode
+	mV0  = "1"            // vertical mode, a1 = b1
+	mVR1 = "011"          // vertical mode, a1 = b1 + 1
+	mVR2 = "000011"       // vertical mode, a1 = b1 + 2
+	mVR3 = "0000011"      // vertical mode, a1 = b1 + 3
+	mVL1 = "010"          // vertical mode, a1 = b1 - 1
+	mVL2 = "000010"       // vertical mode, a1 = b1 - 2
+	mVL3 = "0000010"      // vertical mode, a1 = b1 - 3
+	ext  = "0000001111"   // extension mode, enter uncompressed mode
+	eol  = "000000000001" // end of line
+	eofb = eol + eol      // end of facsimile block
 )
 
 var codes = []string{mP, mH, mV0, mVR1, mVR2, mVR3, mVL1, mVL2, mVL3, ext, eofb}