diff --git a/go.mod b/go.mod index 6fc54b2..8a00306 100644 --- a/go.mod +++ b/go.mod @@ -8,9 +8,9 @@ require ( github.com/go-text/typesetting v0.3.5 github.com/tdewolff/canvas v0.0.0-20260923214215-09804640d00c github.com/tdewolff/font v0.0.0-20260913163313-54f98bb59ee6 - github.com/xiaoqidun/j2kgo v0.0.0-20261009050019-d6d8c42b70bd + github.com/xiaoqidun/j2kgo v0.0.0-20261009063910-4865a21ef893 github.com/xiaoqidun/jbig2 v1.1.3 - github.com/xiaoqidun/pdfgo v0.0.0-20261009054335-bd8a5e721992 + github.com/xiaoqidun/pdfgo v0.0.0-20261009064143-9a945dbdcbb6 golang.org/x/crypto v0.57.0 golang.org/x/image v0.46.0 golang.org/x/text v0.42.0 diff --git a/go.sum b/go.sum index 7a7d43d..8422d02 100644 --- a/go.sum +++ b/go.sum @@ -58,12 +58,12 @@ github.com/tdewolff/parse/v2 v2.8.16 h1:bLk5svUOQRkW/Y2SJ+DeENSIkZBcTIkq+Atyv5D8 github.com/tdewolff/parse/v2 v2.8.16/go.mod h1:XdsoSFThlVIRIajAuqz1evNY7bagZS8LBOPA3aVopwQ= github.com/tdewolff/test v1.0.12 h1:7F21DqIajswxuche0geHdrUZRCWE4oko4b7bcmkkrxk= github.com/tdewolff/test v1.0.12/go.mod h1:XPuWBzvdUzhCuxWO1ojpXsyzsA5bFoS3tO/Q3kFuTG8= -github.com/xiaoqidun/j2kgo v0.0.0-20261009050019-d6d8c42b70bd h1:8aJ5cOOwA68CKp0FwcFgS27zSdcSJnYbmu8TJSn/HVQ= -github.com/xiaoqidun/j2kgo v0.0.0-20261009050019-d6d8c42b70bd/go.mod h1:DCwAfuDzCDYqUiAlZ7BDwigTjaXw7o9L1FtEma/D1aA= +github.com/xiaoqidun/j2kgo v0.0.0-20261009063910-4865a21ef893 h1:jmQlBO7R+bcimOMs9jjyDWR1FqiPNGtZqA/V1WiYfPA= +github.com/xiaoqidun/j2kgo v0.0.0-20261009063910-4865a21ef893/go.mod h1:DCwAfuDzCDYqUiAlZ7BDwigTjaXw7o9L1FtEma/D1aA= github.com/xiaoqidun/jbig2 v1.1.3 h1:p0d8FMLiLfzQuX278T4VTfTz8tIJEw87u4v63lH4CBc= github.com/xiaoqidun/jbig2 v1.1.3/go.mod h1:8QonVWtENmh2kc6l+leqeGl7psb79DWlHDkCpdChjfQ= -github.com/xiaoqidun/pdfgo v0.0.0-20261009054335-bd8a5e721992 h1:W5ukWZe5ru3A2Mlq1sqXw1hnW9e2ViK1LYvxG63baz0= -github.com/xiaoqidun/pdfgo v0.0.0-20261009054335-bd8a5e721992/go.mod h1:n0s/KSvZRp3IGc2fGCkrg0shh2KPj++HYCzNDJj/ydg= +github.com/xiaoqidun/pdfgo v0.0.0-20261009064143-9a945dbdcbb6 h1:r1G1wTBO32uPvfT3iMl1S96+yj4294bNuFiHHGA+vM0= +github.com/xiaoqidun/pdfgo v0.0.0-20261009064143-9a945dbdcbb6/go.mod h1:swzoqA5+wSHsRPesJS1VOMLhSyWkLCSQ+IgHNYXbcKo= github.com/xyproto/randomstring v1.0.5 h1:YtlWPoRdgMu3NZtP45drfy1GKoojuR7hmRcnhZqKjWU= github.com/xyproto/randomstring v1.0.5/go.mod h1:rgmS5DeNXLivK7YprL0pY+lTuhNQW3iGxZ18UQApw/E= github.com/yuin/goldmark v1.8.5 h1:r6N5afV5qj/5S4UTch8agZHJ8UxNCMwX7WjkkJam2NA= diff --git a/ofdgo_pdf.go b/ofdgo_pdf.go index baf10c6..a320d61 100644 --- a/ofdgo_pdf.go +++ b/ofdgo_pdf.go @@ -81,6 +81,7 @@ type pdfImporter struct { pageActionPage int fontIDs map[*pdfgo.Font]string fonts map[*pdfgo.Font]*pdfImportedFont + symbolFonts map[*pdfgo.Font]*pdfImportedFont fontWarnings map[string]bool clipTexts map[*pdfgo.TextClip][]TextObject clipPaths *renderCache[[32]byte, pdfClipPath] diff --git a/ofdgo_pdf_font.go b/ofdgo_pdf_font.go index 53d4d42..2f0a3b8 100644 --- a/ofdgo_pdf_font.go +++ b/ofdgo_pdf_font.go @@ -143,6 +143,120 @@ func (p *pdfImporter) importedFont(source *pdfgo.Font) (*pdfImportedFont, error) return result, nil } +// symbolFont 将标准符号字体的旧式字符表转换为Unicode映射,保留原字形编号 +// 入参: source 未嵌入的PDF字体 +// 返回: *pdfImportedFont 可用的符号字体,无匹配时为空, error 字体或取消错误 +func (p *pdfImporter) symbolFont(source *pdfgo.Font) (*pdfImportedFont, error) { + name := pdfEmbeddedFontName(source.Name) + if source.Subtype != "Type1" || name != "Symbol" && name != "ZapfDingbats" { + return nil, nil + } + if imported, ok := p.symbolFonts[source]; ok { + return imported, nil + } + if p.symbolFonts == nil { + p.symbolFonts = make(map[*pdfgo.Font]*pdfImportedFont) + } + if err := p.ctx.Err(); err != nil { + return nil, err + } + definition := &Font{ID: name, FontName: name, FamilyName: name} + reader := &Reader{doc: &Document{}, fontResourcesRead: true, fontCache: map[string]*Font{name: definition}} + resolved, err := p.editor.newRenderer(reader).ResolveFont(name, true) + if err != nil || len(resolved.Data) == 0 || p.editor.backends.FontResources == nil { + p.symbolFonts[source] = nil + return nil, nil + } + tables, err := fontFileTables(resolved.Data, 0) + if err != nil || !pdfSymbolCmap(tables["cmap"]) { + p.symbolFonts[source] = nil + return nil, nil + } + standard, err := p.reader.ReadFontContext(p.ctx, pdfgo.Dictionary{"Type": pdfgo.Name("Font"), "Subtype": pdfgo.Name("Type1"), "BaseFont": pdfgo.Name(name)}) + if err != nil { + return nil, err + } + codes := make([]byte, 256) + for index := range codes { + codes[index] = byte(index) + } + glyphs, err := standard.DecodeContext(p.ctx, codes) + if err != nil { + return nil, err + } + original := parseCmapMappings(tables["cmap"]) + mapping := make(map[rune]uint16) + names := make(map[string]uint16) + for _, glyph := range glyphs { + if glyph.Name == ".notdef" || glyph.Name == "" { + continue + } + id := original[rune(glyph.Code)] + for _, base := range [...]rune{0xf000, 0xf100, 0xf200} { + if candidate := original[base+rune(glyph.Code)]; candidate != 0 { + if id != 0 && id != candidate { + return nil, fmt.Errorf("ambiguous PDF symbol font mapping") + } + id = candidate + } + } + if id == 0 { + continue + } + names[glyph.Name] = id + if char, size := utf8.DecodeRuneInString(glyph.Text); size != 0 && size == len(glyph.Text) { + mapping[char] = id + } + } + if len(mapping) == 0 { + p.symbolFonts[source] = nil + return nil, nil + } + tables["cmap"] = buildCmapTable(0, mapping) + program, err := serializeOTF(tables) + if err != nil { + return nil, err + } + resource, err := p.editor.backends.FontResources.OpenFontResource(FontFile{Name: name + ".ttf", Data: program}, 0) + if err != nil { + return nil, err + } + metrics, err := p.editor.backends.Fonts.OpenFont(resource.Data) + if err != nil { + return nil, err + } + if _, ok := metrics.(FontOutlines); !ok { + p.symbolFonts[source] = nil + return nil, nil + } + result := &pdfImportedFont{resource: resource, checksum: sha256.Sum256(resource.Data), metrics: metrics, type1Glyphs: names} + p.symbolFonts[source] = result + return result, nil +} + +// pdfSymbolCmap 识别不含Unicode字符表的Windows符号字体 +// 入参: data 字符映射表 +// 返回: bool 是否为旧式符号字体 +func pdfSymbolCmap(data []byte) bool { + if len(data) < 4 { + return false + } + count := int(binary.BigEndian.Uint16(data[2:])) + if count > (len(data)-4)/8 { + return false + } + symbol := false + for i := range count { + pos := 4 + i*8 + platform, encoding := binary.BigEndian.Uint16(data[pos:]), binary.BigEndian.Uint16(data[pos+2:]) + if platform == 0 || platform == 3 && (encoding == 1 || encoding == 10) { + return false + } + symbol = symbol || platform == 3 && encoding == 0 + } + return symbol +} + // pdfFontProgram 为PDF子集字体补齐封装表,保留字形轮廓和编号 // 无效字符表按PDF已解码字形重建,不改变字符对应的字形编号 // 入参: source PDF字体, type1 已解析的Type1程序,nil时按需解析 diff --git a/ofdgo_pdf_group.go b/ofdgo_pdf_group.go index a73c36a..17f87b3 100644 --- a/ofdgo_pdf_group.go +++ b/ofdgo_pdf_group.go @@ -370,10 +370,13 @@ func (p *pdfImporter) compileGroup(build func(*pdfImporter) error) (*RasterPage, if p.fonts == nil { p.fonts = make(map[*pdfgo.Font]*pdfImportedFont) } + if p.symbolFonts == nil { + p.symbolFonts = make(map[*pdfgo.Font]*pdfImportedFont) + } if p.clipPaths == nil { p.clipPaths = &renderCache[[32]byte, pdfClipPath]{limit: 8 << 20} } - local := pdfImporter{ctx: p.ctx, reader: p.reader, editor: editor, renderer: p.renderer, matrix: p.matrix, pageBox: p.pageBox, pageWidth: p.pageWidth, pageHeight: p.pageHeight, rasterDPI: p.rasterDPI, warning: p.warning, rasterWarned: true, fontIDs: map[*pdfgo.Font]string{}, fonts: p.fonts, clipPaths: p.clipPaths} + local := pdfImporter{ctx: p.ctx, reader: p.reader, editor: editor, renderer: p.renderer, matrix: p.matrix, pageBox: p.pageBox, pageWidth: p.pageWidth, pageHeight: p.pageHeight, rasterDPI: p.rasterDPI, warning: p.warning, rasterWarned: true, fontIDs: map[*pdfgo.Font]string{}, fonts: p.fonts, symbolFonts: p.symbolFonts, clipPaths: p.clipPaths} if err := build(&local); err != nil { return nil, box, err } diff --git a/ofdgo_pdf_object.go b/ofdgo_pdf_object.go index f23b3a5..9d59781 100644 --- a/ofdgo_pdf_object.go +++ b/ofdgo_pdf_object.go @@ -326,13 +326,29 @@ func (p *pdfImporter) text(mark pdfgo.TextMark) error { if err != nil { return err } - } else if imported = p.embeddedTextFont(font, mark.Glyphs); imported != nil { - reused, embedded = true, true - mark.Glyphs = slices.Clone(mark.Glyphs) - for index := range mark.Glyphs { - char, _ := utf8.DecodeRuneInString(mark.Glyphs[index].Text) - mark.Glyphs[index].ID = imported.metrics.GlyphIndex(char) - mark.Glyphs[index].HasID = true + } else { + imported = p.embeddedTextFont(font, mark.Glyphs) + symbol := false + if imported == nil { + imported, err = p.symbolFont(font) + if err != nil { + return err + } + symbol = imported != nil + } + if imported != nil { + reused, embedded = true, true + mark.Glyphs = slices.Clone(mark.Glyphs) + for index := range mark.Glyphs { + glyph := &mark.Glyphs[index] + if symbol { + glyph.ID = imported.type1Glyphs[glyph.Name] + } else { + char, _ := utf8.DecodeRuneInString(glyph.Text) + glyph.ID = imported.metrics.GlyphIndex(char) + } + glyph.HasID = true + } } } id := p.fontIDs[font]