introduced FontSubsetter interface

Author: cdillond

font.go

@@ -4,26 +4,36 @@ import (
 	"io"
 	"os"
 
+	"github.com/cdillond/gdf/font"
 	"golang.org/x/image/font/sfnt"
 	"golang.org/x/text/encoding"
 	"golang.org/x/text/encoding/charmap"
 )
 
-// A FontSubsetFunc is intended to give the user control over how a Font is subset when it is embedded in the output PDF.
-// By default, gdf uses the TTFSubset function from github.com/cdillond/gdf/font, but there are good reasons to choose a different
-// implementation. For example, Harfbuzz's hb-subset tool and Microsoft's Win32 CreateFontPackage are robust alternatives
-// written in C++ that can be wrapped by a user-defined FontSubsetFunc. If you do not want the embedded font to be subset
-// at all, you can define a function that simply returns src and a nil error. A FontSubsetFunc should not alter the glyph
-// ID of any rune in the cutset. It must also be sure to include a .notdef glyph.
-type FontSubsetFunc func(f *sfnt.Font, src []byte, cutset map[rune]struct{}) ([]byte, error)
+// A FontSubsetter is intended to give the user control over how a Font is subset when it is embedded in the output PDF.
+// By default, gdf uses the DefaultSubsetter, which is equivalent to the BasicSubsetter type from github.com/cdillond/gdf/font,
+// but there are good reasons to choose a different implementation (the gdf/font package provides several).
+// For example, Harfbuzz's hb-subset tool and Microsoft's Win32 CreateFontPackage are robust alternatives written in C++
+// that can be wrapped by a user-defined FontSubsetter.
+// If you do not want the embedded font to be subset at all, you can set the font's FontSubsetter to nil.
+// A FontSubsetter should not alter the glyph ID of any rune in the cutset. It must also be sure to include a .notdef glyph.
+type FontSubsetter interface {
+	Init(SFNT *sfnt.Font, src []byte, path string)
+	Subset(cutset map[rune]struct{}) ([]byte, error)
+}
+
+type DefaultSubsetter struct {
+	font.BasicSubsetter
+}
 
 const ppem = 1000
 
 // A Font represents a TrueType/OpenType font. Any given Font struct should be used on at most 1 PDF. To use the same underlying
 // font on multiple PDF files, derive a new Font struct from the source font file or bytes for each PDF.
 type Font struct {
-	SFNT           *sfnt.Font // The source TrueType or OpenType font.
-	FontSubsetFunc            // If nil, gdf's default is used.
+	SFNT      *sfnt.Font // The source TrueType or OpenType font.
+	Subsetter FontSubsetter
+
 	*simpleFD
 
 	refnum    int
@@ -38,22 +48,22 @@ type Font struct {
 	source    *stream
 	buf       *sfnt.Buffer
 	srcb      []byte
+	srcPath   string
 }
 
 // LoadTrueType returns a *Font object, which can be used for drawing text to a ContentStream or XObject, and an error.
 func LoadTrueType(b []byte, flag FontFlag) (*Font, error) {
-	//b2 := make([]byte, len(b))
-	//copy(b2, b)
 	b2 := b
 	fnt, err := sfnt.Parse(b)
 	if err != nil {
 		return nil, err
 	}
 	out := &Font{
-		subtype: "/TrueType",
-		encName: "/WinAnsiEncoding",
-		charset: make(map[rune]int),
-		enc:     charmap.Windows1252.NewEncoder(),
+		Subsetter: new(DefaultSubsetter),
+		subtype:   "/TrueType",
+		encName:   "/WinAnsiEncoding",
+		charset:   make(map[rune]int),
+		enc:       charmap.Windows1252.NewEncoder(),
 		source: &stream{
 			Filter: Flate,
 		},
@@ -79,7 +89,11 @@ func LoadTrueTypeFile(path string, flag FontFlag) (*Font, error) {
 	if err != nil {
 		return nil, err
 	}
-	return LoadTrueType(b, flag)
+	f, err := LoadTrueType(b, flag)
+	if err != nil {
+		f.srcPath = path
+	}
+	return f, err
 }
 
 type simpleFD struct {

font/hb.go

@@ -7,38 +7,34 @@ import (
 	"fmt"
 	"os/exec"
 	"strconv"
-
-	"golang.org/x/image/font/sfnt"
 )
 
-// HBSubsetPath returns a func that can be used as a gdf.FontSubsetFunc. The returned func
-// is similar to HBSubset and has the same limitations. The path parameter represents the
-// path of the source font.
-func HBSubsetPath(path string) func(*sfnt.Font, []byte, map[rune]struct{}) ([]byte, error) {
-	return func(_ *sfnt.Font, _ []byte, cutset map[rune]struct{}) ([]byte, error) {
-		u := make([]byte, 0, 512)
-		for key := range cutset {
-			u = strconv.AppendInt(u, int64(key), 16)
-			u = append(u, ',')
-		}
-		if len(u) < 1 {
-			return nil, fmt.Errorf("cutset is too small")
-		}
-		cmd := exec.Command("hb-subset",
-			"--font-file="+path, // must be passed explicitly as an arg
-			"-u", string(u[:len(u)-1]),
-			"--retain-gids",
-			"-o", "/dev/stdout", // ditto for stdout
-		)
-		return cmd.Output()
+// HBSubset returns the source bytes of a font containing only the characters included in the cutset.
+// The path parameter represents the path of the source font. This function has the same limitations as HBSubset.
+func HBSubsetPath(path string, cutset map[rune]struct{}) ([]byte, error) {
+	u := make([]byte, 0, 512)
+	for key := range cutset {
+		u = strconv.AppendInt(u, int64(key), 16)
+		u = append(u, ',')
+	}
+	if len(u) < 1 {
+		return nil, fmt.Errorf("cutset is too small")
 	}
+	cmd := exec.Command("hb-subset",
+		"--font-file="+path, // must be passed explicitly as an arg
+		"-u", string(u[:len(u)-1]),
+		"--retain-gids",
+		"-o", "/dev/stdout", // ditto for stdout
+	)
+	return cmd.Output()
 }
 
-// HBSubset can be used as a gdf.FontSubsetFunc. For this function to work, the HarfBuzz hb-subset
-// tool must be installed. The HBSubset func may handle edge cases that the TTFSubset func does not. hb-subset
+// HBSubset returns the source bytes of a font containing only the characters included in the cutset.
+// For this function to work, the HarfBuzz hb-subsettool must be installed.
+// The HBSubset func may handle edge cases that the TTFSubset func does not. hb-subset
 // has a mature, well-tested API and is capable of handling more font formats than TTFSubset.
 // However, this approach requires os/exec, so it might not be suitable for all environments.
-func HBSubset(_ *sfnt.Font, src []byte, cutset map[rune]struct{}) ([]byte, error) {
+func HBSubset(src []byte, cutset map[rune]struct{}) ([]byte, error) {
 	u := make([]byte, 0, 512)
 	for key := range cutset {
 		u = strconv.AppendInt(u, int64(key), 16)

font/hb_windows.go

@@ -7,60 +7,56 @@ import (
 	"os"
 	"os/exec"
 	"strconv"
-
-	"golang.org/x/image/font/sfnt"
 )
 
-// HBSubsetPath returns a func that can be used as a gdf.FontSubsetFunc. The returned func
-// is similar to HBSubset and has the same limitations. The path parameter represents the
-// path of the source font.
-func HBSubsetPath(path string) func(*sfnt.Font, []byte, map[rune]struct{}) ([]byte, error) {
-	return func(_ *sfnt.Font, _ []byte, cutset map[rune]struct{}) ([]byte, error) {
-		var out *os.File
-		var err error
-		outPath := "tmp-font-out"
+// HBSubset returns the source bytes of a font containing only the characters included in the cutset.
+// The path parameter represents the path of the source font. This function has the same limitations as HBSubset.
+func HBSubsetPath(path string, cutset map[rune]struct{}) ([]byte, error) {
+	var out *os.File
+	var err error
+	outPath := "tmp-font-out"
 
-		// make sure we're not overwriting anything
-		var i int
-		for _, err = os.Stat(outPath); err == nil; i++ {
-			outPath += strconv.Itoa(i)
-		}
-		// set up output file
-		if out, err = os.Create(outPath); err != nil {
-			return nil, err
-		}
-		defer os.Remove(outPath)
-		if err = out.Close(); err != nil {
-			return nil, err
-		}
+	// make sure we're not overwriting anything
+	var i int
+	for _, err = os.Stat(outPath); err == nil; i++ {
+		outPath += strconv.Itoa(i)
+	}
+	// set up output file
+	if out, err = os.Create(outPath); err != nil {
+		return nil, err
+	}
+	defer os.Remove(outPath)
+	if err = out.Close(); err != nil {
+		return nil, err
+	}
 
-		u := make([]byte, 0, 512)
-		for key := range cutset {
-			u = strconv.AppendInt(u, int64(key), 16)
-			u = append(u, ',')
-		}
-		if len(u) < 1 {
-			return nil, fmt.Errorf("cutset is too small")
-		}
-		cmd := exec.Command("hb-subset",
-			"--font-file="+path, // must be passed explicitly as an arg
-			"-u", string(u[:len(u)-1]),
-			"--retain-gids",
-			"-o", outPath, // ditto for output
-		)
-		if err = cmd.Run(); err != nil {
-			return nil, err
-		}
-		return os.ReadFile(outPath)
+	u := make([]byte, 0, 512)
+	for key := range cutset {
+		u = strconv.AppendInt(u, int64(key), 16)
+		u = append(u, ',')
+	}
+	if len(u) < 1 {
+		return nil, fmt.Errorf("cutset is too small")
+	}
+	cmd := exec.Command("hb-subset",
+		"--font-file="+path, // must be passed explicitly as an arg
+		"-u", string(u[:len(u)-1]),
+		"--retain-gids",
+		"-o", outPath, // ditto for output
+	)
+	if err = cmd.Run(); err != nil {
+		return nil, err
 	}
+	return os.ReadFile(outPath)
+
 }
 
-// HBSubset can be used as a gdf.FontSubsetFunc.
-// For this function to work, the HarfBuzz hb-subset tool must be installed.
+// HBSubset returns the source bytes of a font containing only the characters included in the cutset.
+// For this function to work, the HarfBuzz hb-subsettool must be installed.
 // The HBSubset func may handle edge cases that the TTFSubset func does not. hb-subset
 // has a mature, well-tested API and is capable of handling more font formats than TTFSubset.
-// However, this approach requires os/exec, so it might not be suitable for all environments.
-func HBSubset(_ *sfnt.Font, src []byte, cutset map[rune]struct{}) ([]byte, error) {
+// However, this approach requires os/exec, so it might not be suitable for all environments..
+func HBSubset(src []byte, cutset map[rune]struct{}) ([]byte, error) {
 	// Instead of using /dev/stdin and /dev/stdout, on Windows, this function creates temp files.
 	// It's unclear if this is necessary, but the hb-subset tool is finnicky.
 	var in, out *os.File

font/hbc.go

@@ -2,14 +2,11 @@
 
 package font
 
-import (
-	"golang.org/x/image/font/sfnt"
-)
+import "fmt"
 
-// HBSubsetC can be used as a gdf.FontSubsetFunc. It calls functions in libharfbuzz and libharfbuzz-subset via CGo. In order
-// for this function to work, CGo must be enabled, HarfBuzz v>=2.9.0 must be installed on your system, and `hbsubsetc` must be passed
-// as a build tag to the Go compiler. If these conditions are not met, the function instead calls HBSubset (which may fail if hb-subset)
-// has not been installed.
-func HBSubsetC(_ *sfnt.Font, src []byte, charset map[rune]struct{}) ([]byte, error) {
-	return HBSubset(nil, src, charset)
+// HBSubsetC calls functions in libharfbuzz and libharfbuzz-subset via CGo and returns the source bytes of a font containing only the
+// characters included in the cutset. In order for this function to work, CGo must be enabled, HarfBuzz v>=2.9.0 must be installed on
+// your system, and `hbsubsetc` must be passed to the Go compiler as a build tag.
+func HBSubsetC(src []byte, charset map[rune]struct{}) ([]byte, error) {
+	return nil, fmt.Errorf("HBSubetC is not enabled; use the build tag hbsubsetc and consult github.com/cdillond/gdf/font for further information")
 }

font/hbc_cgo.go

@@ -48,18 +48,15 @@ import "C"
 import (
 	"fmt"
 	"unsafe"
-
-	"golang.org/x/image/font/sfnt"
 )
 
-// HBSubsetC can be used as a gdf.FontSubsetFunc. It calls functions in libharfbuzz and libharfbuzz-subset via CGo. In order
-// for this function to work, CGo must be enabled, HarfBuzz v>=2.9.0 must be installed on your system, and `hbsubsetc` must be passed
-// as a build tag to the Go compiler. If these conditions are not met, the function instead calls HBSubset (which may fail if hb-subset)
-// has not been installed.
-func HBSubsetC(_ *sfnt.Font, src []byte, charset map[rune]struct{}) ([]byte, error) {
+// HBSubsetC calls functions in libharfbuzz and libharfbuzz-subset via CGo and returns the source bytes of a font containing only the
+// characters included in the cutset. In order for this function to work, CGo must be enabled, HarfBuzz v>=2.9.0 must be installed on
+// your system, and `hbsubsetc` must be passed to the Go compiler as a build tag.
+func HBSubsetC(src []byte, cutset map[rune]struct{}) ([]byte, error) {
 	// convert runes to uint32_t chars readable by hb-subset
-	charset_u32 := make([]uint32, len(charset))
-	for char := range charset {
+	charset_u32 := make([]uint32, len(cutset))
+	for char := range cutset {
 		charset_u32 = append(charset_u32, uint32(char))
 	}
 	// allocate at least as much as the current file size

font/subsetter.go

@@ -0,0 +1,52 @@
+package font
+
+import (
+	"fmt"
+
+	"golang.org/x/image/font/sfnt"
+)
+
+type BasicSubsetter struct {
+	sFNT *sfnt.Font
+	src  []byte
+}
+
+func (b *BasicSubsetter) Subset(cutset map[rune]struct{}) ([]byte, error) {
+	return TTFSubset(b.sFNT, b.src, cutset)
+}
+func (b *BasicSubsetter) Init(SFNT *sfnt.Font, src []byte, _ string) {
+	b.sFNT = SFNT
+	b.src = src
+}
+
+type HarfBuzzSubsetter struct {
+	path string // Path to the font file; may be empty if Src is non-nil
+	src  []byte // Font source bytes; may be nil if Path is not empty
+}
+
+func (h *HarfBuzzSubsetter) Subset(cutset map[rune]struct{}) ([]byte, error) {
+	if h.path != "" {
+		return HBSubsetPath(h.path, cutset)
+	}
+	if h.src != nil {
+		return HBSubset(h.src, cutset)
+	}
+	return nil, fmt.Errorf("no font source data provided")
+}
+
+func (h *HarfBuzzSubsetter) Init(_ *sfnt.Font, src []byte, path string) {
+	h.path = path
+	h.src = src
+}
+
+type HarfBuzzCGoSubsetter struct {
+	src []byte // Font source bytes
+}
+
+func (h *HarfBuzzCGoSubsetter) Subset(cutset map[rune]struct{}) ([]byte, error) {
+	return HBSubsetC(h.src, cutset)
+}
+
+func (h *HarfBuzzCGoSubsetter) Init(_ *sfnt.Font, src []byte, _ string) {
+	h.src = src
+}

font/subset.go renamed to font/ttf_subset.go

@@ -36,10 +36,6 @@ var pdfTables = [...]TableTag{
 	//1886352244, // post
 }
 
-// NoSubset can be used as a gdf.FontSubsetFunc when you wish to avoid subsetting a given
-// font. Beware: this can negatively impact the output PDF file size.
-func NoSubset(_ *sfnt.Font, src []byte, _ map[rune]struct{}) ([]byte, error) { return src, nil }
-
 // TTFSubset is something of a poor man's subsetting function. It works - for TrueType fonts with 'glyf' tables only - by zeroing out
 // the outlines of all glyphs not corresponding to or directly referenced by f's glyphs for the runes in cutset,
 // truncating f's glyf and loca tables, and then writing only the required tables to the returned byte slice. The final subset font

gdf.go

@@ -2,8 +2,6 @@ package gdf
 
 import (
 	"io"
-
-	"github.com/cdillond/gdf/font"
 )
 
 type PDF struct {
@@ -69,17 +67,17 @@ func includeChildren(pdf *PDF, o obj) error {
 			for key := range f.charset {
 				tmp[key] = struct{}{}
 			}
-
-			if f.FontSubsetFunc == nil {
-				f.FontSubsetFunc = font.TTFSubset
-			}
-			b, err := f.FontSubsetFunc(f.SFNT, f.srcb, tmp)
-			if err != nil {
-				return err // log.Println(err.Error())
+			if f.Subsetter == nil {
+				f.source.buf = f.srcb
 			} else {
-				f.source.buf = b
+				f.Subsetter.Init(f.SFNT, f.srcb, f.srcPath)
+				b, err := f.Subsetter.Subset(tmp)
+				if err != nil {
+					return err
+				} else {
+					f.source.buf = b
+				}
 			}
-
 		}
 		includeObj(pdf, child)
 		if err := includeChildren(pdf, child); err != nil {