document: read/restore all contents of a Word 2016 file

Adds support for serializing/deserializing of:

- webSettings
- fontTable
- endnotes
- footnotes
- themes
- optional numbering

common/schemas.go

@@ -27,10 +27,13 @@ const (
 	SharedStingsType         = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/sharedStrings"
 
 	// WML
-	HeaderType    = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/header"
-	FooterType    = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/footer"
-	NumberingType = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/numbering"
-	FontTableType = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/fontTable"
+	HeaderType      = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/header"
+	FooterType      = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/footer"
+	NumberingType   = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/numbering"
+	FontTableType   = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/fontTable"
+	WebSettingsType = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/webSettings"
+	FootNotesType   = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/footnotes"
+	EndNotesType    = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/endnotes"
 
 	// PML
 	SlideType                   = "http://schemas.openxmlformats.org/officeDocument/2006/relationships/slide"

creator.go

@@ -12,7 +12,6 @@ import (
 	"errors"
 	"fmt"
 	"reflect"
-	"sort"
 )
 
 // Any is the interface used for marshaling/unmarshaling xsd:any
@@ -21,73 +20,6 @@ type Any interface {
 	UnmarshalXML(d *xml.Decoder, start xml.StartElement) error
 }
 
-// Raw is used to unmarshal raw XML when we see an unknown tag
-type Raw struct {
-	XMLName xml.Name
-	Attrs   []xml.Attr `xml:",any,attr"`
-	Value   []byte     `xml:",innerxml"`
-}
-
-// MarshalXML allows raw to have the Any interface.
-func (r *Raw) MarshalXML(e *xml.Encoder, start xml.StartElement) error {
-	tmp := struct {
-		Attrs []xml.Attr `xml:",any,attr"`
-		Value []byte     `xml:",innerxml"`
-	}{}
-
-	tmp.Value = r.Value
-	start.Attr = nil
-	s := xml.StartElement{Name: r.XMLName}
-	tmpAttrs := make([]xml.Attr, len(r.Attrs))
-	copy(tmpAttrs, r.Attrs)
-
-	// fix namespaces in the element we're about to write
-	for i := 0; i < len(tmpAttrs); i++ {
-		attr := tmpAttrs[i]
-		// we unmarshaled an xmlns:foo="http:/foo.com" attribute for a <foo:bar/> element
-		if attr.Name.Space == "xmlns" && attr.Value == r.XMLName.Space {
-			// add xmlns:foo="http://foo.com" to the element
-			s.Attr = append(s.Attr, xml.Attr{Name: xml.Name{Local: "xmlns:" + attr.Name.Local}, Value: s.Name.Space})
-			s.Name.Local = attr.Name.Local + ":" + s.Name.Local
-			// rewrite <bar xmlns:foo="http://foo.com"/> to <foo:bar xmlns:foo="http://foo.com"/>
-			// nuke our namespace which would have been put as xmlns="http://foo.com"
-			s.Name.Space = ""
-		} else if attr.Name.Space == "xmlns" || attr.Name.Local == "xmlns" {
-		} else {
-			s.Attr = append(s.Attr, attr)
-		}
-	}
-
-	// ensure consistent output
-	sort.Slice(s.Attr, func(i, j int) bool {
-		if s.Attr[i].Name.Space != s.Attr[j].Name.Space {
-			return s.Attr[i].Name.Space < s.Attr[j].Name.Space
-		}
-		return s.Attr[i].Name.Local < s.Attr[j].Name.Local
-	})
-
-	if err := e.EncodeElement(tmp, s); err != nil {
-		return err
-	}
-	return nil
-}
-
-// UnmarshalXML allows raw to have the Any interface.
-func (r *Raw) UnmarshalXML(d *xml.Decoder, start xml.StartElement) error {
-	tmp := struct {
-		XMLName xml.Name
-		Attrs   []xml.Attr `xml:",any,attr"`
-		Value   []byte     `xml:",innerxml"`
-	}{}
-	if err := d.DecodeElement(&tmp, &start); err != nil {
-		return err
-	}
-	r.XMLName = tmp.XMLName
-	r.Attrs = tmp.Attrs
-	r.Value = tmp.Value
-	return nil
-}
-
 var creatorFns = map[string]interface{}{}
 
 // RegisterConstructor registers a constructor function used for unmarshaling
@@ -100,7 +32,7 @@ func RegisterConstructor(ns, name string, fn interface{}) {
 func CreateElement(start xml.StartElement) (Any, error) {
 	fn, ok := creatorFns[start.Name.Space+"/"+start.Name.Local]
 	if !ok {
-		r := &Raw{}
+		r := &XSDAny{}
 		return r, nil
 	}
 

creator_test.go

@@ -26,7 +26,10 @@ func TestRawEncode(t *testing.T) {
 	}
 	dec := xml.NewDecoder(f)
 	var got *bytes.Buffer
-	for i := 0; i < 2; i++ {
+
+	// should round trip multiple times with no changes after
+	// the first encoding
+	for i := 0; i < 5; i++ {
 		stng := wml.NewSettings()
 		if err := dec.Decode(stng); err != nil {
 			t.Errorf("error decoding settings: %s", err)
@@ -41,12 +44,14 @@ func TestRawEncode(t *testing.T) {
 		dec = xml.NewDecoder(bytes.NewReader(got.Bytes()))
 	}
 	xmlStr := got.String()
-	beg := strings.Index(xmlStr, "<w:hdrShapeDefaults>")
-	end := strings.Index(xmlStr, "</w:hdrShapeDefaults>")
+	beg := strings.LastIndex(xmlStr, "<w:hdrShapeDefaults>")
+	end := strings.LastIndex(xmlStr, "</w:hdrShapeDefaults>")
 
 	gotRaw := xmlStr[beg+20 : end]
-	exp := `<shapedefaults xmlns="urn:schemas-microsoft-com:office:office" spidmax="2049" xmlns:_="urn:schemas-microsoft-com:vml" _:ext="edit"/>`
+
+	exp := "<o:shapedefaults v:ext=\"edit\" spidmax=\"2049\" xmlns:o=\"urn:schemas-microsoft-com:office:office\" xmlns:v=\"urn:schemas-microsoft-com:vml\"><o:idmap v:ext=\"edit\" data=\"1\"/></o:shapedefaults>"
 	if gotRaw != exp {
 		t.Errorf("expected\n%q\ngot\n%q\n", exp, gotRaw)
 	}
+
 }

document/document.go

@@ -17,9 +17,9 @@ import (
 	"os"
 	"path/filepath"
 
-	"baliance.com/gooxml/schema/schemas.openxmlformats.org/officeDocument/2006/sharedTypes"
-
 	"baliance.com/gooxml/common"
+	dml "baliance.com/gooxml/schema/schemas.openxmlformats.org/drawingml"
+	st "baliance.com/gooxml/schema/schemas.openxmlformats.org/officeDocument/2006/sharedTypes"
 	wml "baliance.com/gooxml/schema/schemas.openxmlformats.org/wordprocessingml"
 
 	"baliance.com/gooxml/zippkg"
@@ -28,14 +28,19 @@ import (
 // Document is a text document that can be written out in the OOXML .docx format.
 type Document struct {
 	common.DocBase
-	x         *wml.Document
-	Settings  Settings
-	Numbering Numbering
-	Styles    Styles
-	headers   []*wml.Hdr
-	footers   []*wml.Ftr
-	docRels   common.Relationships
-	images    []*iref
+	x           *wml.Document
+	Settings    Settings
+	Numbering   Numbering
+	Styles      Styles
+	headers     []*wml.Hdr
+	footers     []*wml.Ftr
+	docRels     common.Relationships
+	images      []*iref
+	themes      []*dml.Theme
+	webSettings *wml.WebSettings
+	fontTable   *wml.Fonts
+	endNotes    *wml.Endnotes
+	footNotes   *wml.Footnotes
 }
 
 // New constructs an empty document that content can be added to.
@@ -43,7 +48,7 @@ func New() *Document {
 	d := &Document{x: wml.NewDocument()}
 	d.ContentTypes = common.NewContentTypes()
 	d.x.Body = wml.NewCT_Body()
-	d.x.ConformanceAttr = sharedTypes.ST_ConformanceClassTransitional
+	d.x.ConformanceAttr = st.ST_ConformanceClassTransitional
 	d.docRels = common.NewRelationships()
 
 	d.AppProperties = common.NewAppProperties()
@@ -138,8 +143,10 @@ func (d *Document) Save(w io.Writer) error {
 	if err := zippkg.MarshalXML(z, "word/document.xml", d.x); err != nil {
 		return err
 	}
-	if err := zippkg.MarshalXML(z, "word/numbering.xml", d.Numbering.X()); err != nil {
-		return err
+	if d.Numbering.X() != nil {
+		if err := zippkg.MarshalXML(z, "word/numbering.xml", d.Numbering.X()); err != nil {
+			return err
+		}
 	}
 	if err := zippkg.MarshalXML(z, "word/styles.xml", d.Styles.X()); err != nil {
 		return err
@@ -147,6 +154,32 @@ func (d *Document) Save(w io.Writer) error {
 	if err := zippkg.MarshalXML(z, "word/_rels/document.xml.rels", d.docRels.X()); err != nil {
 		return err
 	}
+	if d.webSettings != nil {
+		if err := zippkg.MarshalXML(z, "word/webSettings.xml", d.webSettings); err != nil {
+			return err
+		}
+	}
+	if d.fontTable != nil {
+		if err := zippkg.MarshalXML(z, "word/fontTable.xml", d.fontTable); err != nil {
+			return err
+		}
+	}
+	if d.endNotes != nil {
+		if err := zippkg.MarshalXML(z, "word/endnotes.xml", d.endNotes); err != nil {
+			return err
+		}
+	}
+	if d.footNotes != nil {
+		if err := zippkg.MarshalXML(z, "word/footnotes.xml", d.footNotes); err != nil {
+			return err
+		}
+	}
+	for i, thm := range d.themes {
+		if err := zippkg.MarshalXML(z, fmt.Sprintf("word/theme/theme%d.xml", i+1), thm); err != nil {
+			return err
+		}
+
+	}
 	for i, hdr := range d.headers {
 		fn := fmt.Sprintf("word/header%d.xml", i+1)
 		if err := zippkg.MarshalXML(z, fn, hdr); err != nil {
@@ -242,6 +275,9 @@ func Open(filename string) (*Document, error) {
 // Read reads a document from an io.Reader.
 func Read(r io.ReaderAt, size int64) (*Document, error) {
 	doc := New()
+	// numbering is not required
+	doc.Numbering.x = nil
+
 	td, err := ioutil.TempDir("", "gooxml-docx")
 	if err != nil {
 		return nil, err
@@ -301,7 +337,7 @@ func Read(r io.ReaderAt, size int64) (*Document, error) {
 		case common.SettingsType:
 			decMap[basePaths[doc.docRels]+r.Target()] = doc.Settings.X()
 		case common.NumberingType:
-			doc.Numbering.Clear()
+			doc.Numbering = NewNumbering()
 			decMap[basePaths[doc.docRels]+r.Target()] = doc.Numbering.X()
 		case common.StylesType:
 			doc.Styles.Clear()
@@ -314,6 +350,22 @@ func Read(r io.ReaderAt, size int64) (*Document, error) {
 			ftr := wml.NewFtr()
 			doc.footers = append(doc.footers, ftr)
 			decMap[basePaths[doc.docRels]+r.Target()] = ftr
+		case common.ThemeType:
+			thm := dml.NewTheme()
+			doc.themes = append(doc.themes, thm)
+			decMap[basePaths[doc.docRels]+r.Target()] = thm
+		case common.WebSettingsType:
+			doc.webSettings = wml.NewWebSettings()
+			decMap[basePaths[doc.docRels]+r.Target()] = doc.webSettings
+		case common.FontTableType:
+			doc.fontTable = wml.NewFonts()
+			decMap[basePaths[doc.docRels]+r.Target()] = doc.fontTable
+		case common.EndNotesType:
+			doc.endNotes = wml.NewEndnotes()
+			decMap[basePaths[doc.docRels]+r.Target()] = doc.endNotes
+		case common.FootNotesType:
+			doc.footNotes = wml.NewFootnotes()
+			decMap[basePaths[doc.docRels]+r.Target()] = doc.footNotes
 		case common.ImageType:
 			imgPath := basePaths[doc.docRels] + r.Target()
 			for i, f := range files {

document/document_test.go

@@ -39,7 +39,7 @@ func TestOpen(t *testing.T) {
 		t.Errorf("created an invalid document: %s", err)
 	}
 	wb.Save(&got)
-	testhelper.CompareZip(t, "simple-1.docx", got.Bytes())
+	testhelper.CompareZip(t, "simple-1.docx", got.Bytes(), true)
 }
 
 func TestOpenHeaderFooter(t *testing.T) {
@@ -70,3 +70,16 @@ func TestAddParagraph(t *testing.T) {
 		t.Errorf("expected 2 paragraphs, got %d", len(doc.Paragraphs()))
 	}
 }
+
+func TestOpenWord2016(t *testing.T) {
+	doc, err := document.Open("../testdata/Office2016/Word-Windows.docx")
+	if err != nil {
+		t.Errorf("error opening Windows Word 2016 document: %s", err)
+	}
+	got := bytes.Buffer{}
+	if err := doc.Save(&got); err != nil {
+		t.Errorf("error saving W216 file: %s", err)
+	}
+	testhelper.CompareGoldenZipFilesOnly(t, "../../testdata/Office2016/Word-Windows.docx", got.Bytes())
+	doc.SaveToFile("/tmp/gen/rewrote.docx")
+}

document/testdata/settings.xml

@@ -1,2 +1,2 @@
 <?xml version="1.0" encoding="UTF-8" standalone="yes"?>
-<w:settings xmlns="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:a="http://schemas.openxmlformats.org/drawingml/2006/main" xmlns:m="http://schemas.openxmlformats.org/officeDocument/2006/math" xmlns:ma="http://schemas.openxmlformats.org/schemaLibrary/2006/main" xmlns:pic="http://schemas.openxmlformats.org/drawingml/2006/picture" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships" xmlns:s="http://schemas.openxmlformats.org/officeDocument/2006/sharedTypes" xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:wp="http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing" xmlns:xml="http://www.w3.org/XML/1998/namespace"><w:zoom w:percent="100"/><w:removePersonalInformation/><w:removeDateAndTime/><w:defaultTabStop w:val="720"/><w:characterSpacingControl w:val="doNotCompress"/><w:hdrShapeDefaults><shapedefaults xmlns="urn:schemas-microsoft-com:office:office" spidmax="2049" xmlns:_="urn:schemas-microsoft-com:vml" _:ext="edit"/></w:hdrShapeDefaults><w:footnotePr><w:footnote w:id="-1"/><w:footnote w:id="0"/></w:footnotePr><w:endnotePr><w:endnote w:id="-1"/><w:endnote w:id="0"/></w:endnotePr><w:compat><w:compatSetting w:name="compatibilityMode" w:uri="http://schemas.microsoft.com/office/word" w:val="15"/><w:compatSetting w:name="overrideTableStyleFontSizeAndJustification" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/><w:compatSetting w:name="enableOpenTypeFeatures" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/><w:compatSetting w:name="doNotFlipMirrorIndents" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/><w:compatSetting w:name="differentiateMultirowTableHeaders" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/></w:compat><w:rsids><w:rsidRoot w:val="003C2D97"/><w:rsid w:val="002939C2"/><w:rsid w:val="003C2D97"/><w:rsid w:val="00A45D78"/></w:rsids><m:mathPr xmlns="http://schemas.openxmlformats.org/officeDocument/2006/math" xmlns:m="http://schemas.openxmlformats.org/officeDocument/2006/math" xmlns:s="http://schemas.openxmlformats.org/officeDocument/2006/sharedTypes" xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:xml="http://www.w3.org/XML/1998/namespace"><m:mathFont m:val="Cambria Math"/><m:brkBin m:val="before"/><m:brkBinSub m:val="--"/><m:smallFrac m:val=""/><m:dispDef/><m:lMargin m:val="0"/><m:rMargin m:val="0"/><m:defJc m:val="centerGroup"/><m:wrapIndent m:val="1440"/><m:intLim m:val="subSup"/><m:naryLim m:val="undOvr"/></m:mathPr><w:themeFontLang w:val="en-US"/><w:clrSchemeMapping w:bg1="light1" w:t1="dark1" w:bg2="light2" w:t2="dark2" w:accent1="accent1" w:accent2="accent2" w:accent3="accent3" w:accent4="accent4" w:accent5="accent5" w:accent6="accent6" w:hyperlink="hyperlink" w:followedHyperlink="followedHyperlink"/><w:shapeDefaults><shapedefaults xmlns="urn:schemas-microsoft-com:office:office" spidmax="2049" xmlns:_="urn:schemas-microsoft-com:vml" _:ext="edit"/><shapelayout xmlns="urn:schemas-microsoft-com:office:office" xmlns:_="urn:schemas-microsoft-com:vml" _:ext="edit"><o:idmap v:ext="edit" data="1"/></shapelayout></w:shapeDefaults><w:decimalSymbol w:val="."/><w:listSeparator w:val=","/></w:settings>
+<w:settings xmlns="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:a="http://schemas.openxmlformats.org/drawingml/2006/main" xmlns:m="http://schemas.openxmlformats.org/officeDocument/2006/math" xmlns:ma="http://schemas.openxmlformats.org/schemaLibrary/2006/main" xmlns:pic="http://schemas.openxmlformats.org/drawingml/2006/picture" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships" xmlns:s="http://schemas.openxmlformats.org/officeDocument/2006/sharedTypes" xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:wp="http://schemas.openxmlformats.org/drawingml/2006/wordprocessingDrawing" xmlns:xml="http://www.w3.org/XML/1998/namespace"><w:zoom w:percent="100"/><w:removePersonalInformation/><w:removeDateAndTime/><w:defaultTabStop w:val="720"/><w:characterSpacingControl w:val="doNotCompress"/><w:hdrShapeDefaults><o:shapedefaults v:ext="edit" spidmax="2049" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:v="urn:schemas-microsoft-com:vml"/></w:hdrShapeDefaults><w:footnotePr><w:footnote w:id="-1"/><w:footnote w:id="0"/></w:footnotePr><w:endnotePr><w:endnote w:id="-1"/><w:endnote w:id="0"/></w:endnotePr><w:compat><w:compatSetting w:name="compatibilityMode" w:uri="http://schemas.microsoft.com/office/word" w:val="15"/><w:compatSetting w:name="overrideTableStyleFontSizeAndJustification" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/><w:compatSetting w:name="enableOpenTypeFeatures" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/><w:compatSetting w:name="doNotFlipMirrorIndents" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/><w:compatSetting w:name="differentiateMultirowTableHeaders" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/></w:compat><w:rsids><w:rsidRoot w:val="003C2D97"/><w:rsid w:val="002939C2"/><w:rsid w:val="003C2D97"/><w:rsid w:val="00A45D78"/></w:rsids><m:mathPr xmlns="http://schemas.openxmlformats.org/officeDocument/2006/math" xmlns:m="http://schemas.openxmlformats.org/officeDocument/2006/math" xmlns:s="http://schemas.openxmlformats.org/officeDocument/2006/sharedTypes" xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:xml="http://www.w3.org/XML/1998/namespace"><m:mathFont m:val="Cambria Math"/><m:brkBin m:val="before"/><m:brkBinSub m:val="--"/><m:smallFrac m:val="false"/><m:dispDef/><m:lMargin m:val="0"/><m:rMargin m:val="0"/><m:defJc m:val="centerGroup"/><m:wrapIndent m:val="1440"/><m:intLim m:val="subSup"/><m:naryLim m:val="undOvr"/></m:mathPr><w:themeFontLang w:val="en-US"/><w:clrSchemeMapping w:bg1="light1" w:t1="dark1" w:bg2="light2" w:t2="dark2" w:accent1="accent1" w:accent2="accent2" w:accent3="accent3" w:accent4="accent4" w:accent5="accent5" w:accent6="accent6" w:hyperlink="hyperlink" w:followedHyperlink="followedHyperlink"/><w:shapeDefaults><o:shapedefaults v:ext="edit" spidmax="2049" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:v="urn:schemas-microsoft-com:vml"/><o:shapelayout v:ext="edit" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:v="urn:schemas-microsoft-com:vml"><o:idmap v:ext="edit" data="1"/></o:shapelayout></w:shapeDefaults><w:decimalSymbol w:val="."/><w:listSeparator w:val=","/></w:settings>

schema/schemas.openxmlformats.org/officeDocument/2006/math/common.go

@@ -17,7 +17,7 @@ import (
 )
 
 func ParseUnionST_OnOff(s string) (sharedTypes.ST_OnOff, error) {
-	return sharedTypes.ST_OnOff{}, nil
+	return sharedTypes.ParseUnionST_OnOff(s)
 }
 func ParseUnionST_TwipsMeasure(s string) (sharedTypes.ST_TwipsMeasure, error) {
 	ret := sharedTypes.ST_TwipsMeasure{}

schema/schemas.openxmlformats.org/officeDocument/2006/sharedTypes/common.go

@@ -43,11 +43,13 @@ var ST_PositiveFixedPercentagePatternRe = regexp.MustCompile(ST_PositiveFixedPer
 
 func ParseUnionST_OnOff(s string) (ST_OnOff, error) {
 	r := ST_OnOff{}
-	tru := true
 	switch s {
 	case "true", "1", "on":
+		tru := true
 		r.Bool = &tru
 	default:
+		fals := false
+		r.Bool = &fals
 	}
 	return r, nil
 }

spreadsheet/workbook_test.go

@@ -89,7 +89,7 @@ func TestOpen(t *testing.T) {
 		t.Errorf("created an invalid spreadsheet: %s", err)
 	}
 	wb.Save(&got)
-	testhelper.CompareZip(t, "simple-1.xlsx", got.Bytes())
+	testhelper.CompareZip(t, "simple-1.xlsx", got.Bytes(), true)
 }
 
 func TestOpenExcel2016(t *testing.T) {

testdata/settings.xml

@@ -1,2 +1,3 @@
-<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
-<w:settings xmlns:mc="http://schemas.openxmlformats.org/markup-compatibility/2006" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships" xmlns:m="http://schemas.openxmlformats.org/officeDocument/2006/math" xmlns:v="urn:schemas-microsoft-com:vml" xmlns:w10="urn:schemas-microsoft-com:office:word" xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:w14="http://schemas.microsoft.com/office/word/2010/wordml" xmlns:w15="http://schemas.microsoft.com/office/word/2012/wordml" xmlns:w16se="http://schemas.microsoft.com/office/word/2015/wordml/symex" xmlns:sl="http://schemas.openxmlformats.org/schemaLibrary/2006/main" mc:Ignorable="w14 w15 w16se"><w:zoom w:percent="100"/><w:removePersonalInformation/><w:removeDateAndTime/><w:defaultTabStop w:val="720"/><w:characterSpacingControl w:val="doNotCompress"/><w:hdrShapeDefaults><o:shapedefaults v:ext="edit" spidmax="2049"/></w:hdrShapeDefaults><w:footnotePr><w:footnote w:id="-1"/><w:footnote w:id="0"/></w:footnotePr><w:endnotePr><w:endnote w:id="-1"/><w:endnote w:id="0"/></w:endnotePr><w:compat><w:compatSetting w:name="compatibilityMode" w:uri="http://schemas.microsoft.com/office/word" w:val="15"/><w:compatSetting w:name="overrideTableStyleFontSizeAndJustification" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/><w:compatSetting w:name="enableOpenTypeFeatures" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/><w:compatSetting w:name="doNotFlipMirrorIndents" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/><w:compatSetting w:name="differentiateMultirowTableHeaders" w:uri="http://schemas.microsoft.com/office/word" w:val="1"/></w:compat><w:rsids><w:rsidRoot w:val="003C2D97"/><w:rsid w:val="002939C2"/><w:rsid w:val="003C2D97"/><w:rsid w:val="00A45D78"/></w:rsids><m:mathPr><m:mathFont m:val="Cambria Math"/><m:brkBin m:val="before"/><m:brkBinSub m:val="--"/><m:smallFrac m:val="0"/><m:dispDef/><m:lMargin m:val="0"/><m:rMargin m:val="0"/><m:defJc m:val="centerGroup"/><m:wrapIndent m:val="1440"/><m:intLim m:val="subSup"/><m:naryLim m:val="undOvr"/></m:mathPr><w:themeFontLang w:val="en-US"/><w:clrSchemeMapping w:bg1="light1" w:t1="dark1" w:bg2="light2" w:t2="dark2" w:accent1="accent1" w:accent2="accent2" w:accent3="accent3" w:accent4="accent4" w:accent5="accent5" w:accent6="accent6" w:hyperlink="hyperlink" w:followedHyperlink="followedHyperlink"/><w:shapeDefaults><o:shapedefaults v:ext="edit" spidmax="2049"/><o:shapelayout v:ext="edit"><o:idmap v:ext="edit" data="1"/></o:shapelayout></w:shapeDefaults><w:decimalSymbol w:val="."/><w:listSeparator w:val=","/><w15:chartTrackingRefBased/></w:settings>
+<?xml version="1.0" encoding="UTF-8" standalone="yes"?>
+<w:settings xmlns:mc="http://schemas.openxmlformats.org/markup-compatibility/2006" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:r="http://schemas.openxmlformats.org/officeDocument/2006/relationships" xmlns:m="http://schemas.openxmlformats.org/officeDocument/2006/math" xmlns:v="urn:schemas-microsoft-com:vml" xmlns:w10="urn:schemas-microsoft-com:office:word" xmlns:w="http://schemas.openxmlformats.org/wordprocessingml/2006/main" xmlns:w14="http://schemas.microsoft.com/office/word/2010/wordml" xmlns:w15="http://schemas.microsoft.com/office/word/2012/wordml" xmlns:w16se="http://schemas.microsoft.com/office/word/2015/wordml/symex" xmlns:sl="http://schemas.openxmlformats.org/schemaLibrary/2006/main" mc:Ignorable="w14 w15 w16se">
+<w:hdrShapeDefaults><o:shapedefaults v:ext="edit" spidmax="2049"><o:idmap v:ext="edit" data="1"/></o:shapedefaults></w:hdrShapeDefaults></w:settings>

testhelper/compare.go

@@ -36,7 +36,7 @@ func CompareGoldenXML(t *testing.T, expectedFn string, got []byte) {
 	dumpXmlDiff(t, exp, got)
 }
 
-func CompareZip(t *testing.T, expectedFn string, got []byte) {
+func CompareZip(t *testing.T, expectedFn string, got []byte, cmpFileContents bool) {
 	golden := filepath.Join("testdata", expectedFn)
 	zgot, err := zip.NewReader(bytes.NewReader(got), int64(len(got)))
 	if err != nil {
@@ -57,8 +57,19 @@ func CompareZip(t *testing.T, expectedFn string, got []byte) {
 	if err != nil {
 		t.Errorf("unable to read file: %s", err)
 	}
-	t.Run(expectedFn, compareZipContents(zexp, zgot))
+	t.Run(expectedFn, compareZipContents(zexp, zgot, cmpFileContents))
 }
+
+func CompareGoldenZipFilesOnly(t *testing.T, expectedFn string, got []byte) {
+	golden := filepath.Join("testdata", expectedFn)
+	if *update {
+		if err := ioutil.WriteFile(golden, got, 0644); err != nil {
+			t.Fatal(err)
+		}
+	}
+	CompareZip(t, expectedFn, got, false)
+}
+
 func CompareGoldenZip(t *testing.T, expectedFn string, got []byte) {
 	golden := filepath.Join("testdata", expectedFn)
 	if *update {
@@ -66,10 +77,10 @@ func CompareGoldenZip(t *testing.T, expectedFn string, got []byte) {
 			t.Fatal(err)
 		}
 	}
-	CompareZip(t, expectedFn, got)
+	CompareZip(t, expectedFn, got, true)
 }
 
-func compareZipContents(exp, got *zip.Reader) func(t *testing.T) {
+func compareZipContents(exp, got *zip.Reader, cmpFileContents bool) func(t *testing.T) {
 	return func(t *testing.T) {
 		expFiles := make([]*zip.File, len(exp.File))
 		copy(expFiles, exp.File)
@@ -85,8 +96,10 @@ func compareZipContents(exp, got *zip.Reader) func(t *testing.T) {
 					continue
 				}
 				if f.Name == g.Name {
-					// comparing ones that have the same name
-					t.Run(f.Name, compareFiles(f, g))
+					if cmpFileContents {
+						// comparing contents that have the same name
+						t.Run(f.Name, compareFiles(f, g))
+					}
 					expFiles[i] = nil
 					gotFiles[j] = nil
 				}

xsdany.go

@@ -0,0 +1,168 @@
+package gooxml
+
+import (
+	"encoding/xml"
+	"log"
+	"strings"
+)
+
+// XSDAny  is used to marshal/unmarshal xsd:any types in the OOXML schema.
+type XSDAny struct {
+	Attrs  []xml.Attr
+	Tokens []xml.Token
+}
+
+func cloneToken(tok xml.Token) xml.Token {
+	switch el := tok.(type) {
+	case xml.CharData:
+		cd := xml.CharData{}
+		cd = append(cd, el...)
+		return cd
+	case xml.StartElement, xml.EndElement:
+		return tok
+	default:
+		log.Fatalf("need to suppot %T", el)
+	}
+	return nil
+}
+
+// UnmarshalXML implements the xml.Unmarshaler interface.
+func (x *XSDAny) UnmarshalXML(d *xml.Decoder, start xml.StartElement) error {
+	x.Tokens = append(x.Tokens, cloneToken(start))
+lfor:
+	for {
+		tok, err := d.Token()
+		if err != nil {
+			return err
+		}
+		switch el := tok.(type) {
+		default:
+			x.Tokens = append(x.Tokens, cloneToken(tok))
+		case xml.EndElement:
+			x.Tokens = append(x.Tokens, cloneToken(tok))
+			if el.Name == start.Name {
+				break lfor
+			}
+		}
+	}
+	return nil
+}
+
+type nsSet struct {
+	urlToPrefix map[string]string
+	prefixToURL map[string]string
+	prefixes    []string //required for deterministic output
+}
+
+func (n *nsSet) getPrefix(ns string) string {
+	// do we have a prefix for this ns?
+	if sc, ok := n.urlToPrefix[ns]; ok {
+		return sc
+	}
+
+	// determine the last path portion of the namespace
+	// "urn:schemas-microsoft-com:office:office" = "office"
+	// "http://schemas.microsoft.com/office/word/2012/wordml" = "wordml"
+	split := strings.Split(ns, "/")
+	split = strings.Split(split[len(split)-1], ":")
+	// last segment of the namesapce
+	last := split[len(split)-1]
+	lng := 0
+	pfx := []byte{}
+	for {
+		if lng < len(last) {
+			pfx = append(pfx, last[lng])
+		} else {
+			pfx = append(pfx, '_')
+		}
+		lng++
+		// is this prefix unused?
+		if _, ok := n.prefixToURL[string(pfx)]; !ok {
+			n.prefixToURL[string(pfx)] = ns
+			n.urlToPrefix[ns] = string(pfx)
+			n.prefixes = append(n.prefixes, string(pfx))
+			return string(pfx)
+		}
+	}
+}
+
+func (n nsSet) applyToSE(se *xml.StartElement) {
+	if se.Name.Space == "" {
+		return
+	}
+
+	pfx := n.getPrefix(se.Name.Space)
+	se.Name.Space = ""
+	se.Name.Local = pfx + ":" + se.Name.Local
+	tmpAttr := se.Attr
+	se.Attr = nil
+	for _, attr := range tmpAttr {
+		// skip these as we create them later
+		if attr.Name.Space == "xmlns" {
+			continue
+		}
+		if attr.Name.Space != "" {
+			pfx := n.getPrefix(attr.Name.Space)
+			attr.Name.Space = ""
+			attr.Name.Local = pfx + ":" + attr.Name.Local
+		}
+		se.Attr = append(se.Attr, attr)
+	}
+}
+
+func (n nsSet) applyToEE(ee *xml.EndElement) {
+	if ee.Name.Space == "" {
+		return
+	}
+	pfx := n.getPrefix(ee.Name.Space)
+	ee.Name.Space = ""
+	ee.Name.Local = pfx + ":" + ee.Name.Local
+}
+
+// MarshalXML implements the xml.Marshaler interface.
+func (x *XSDAny) MarshalXML(e *xml.Encoder, start xml.StartElement) error {
+	if len(x.Tokens) == 0 {
+		return nil
+	}
+	ns := nsSet{urlToPrefix: make(map[string]string),
+		prefixToURL: make(map[string]string)}
+
+	// collect the namespaces
+	for _, tok := range x.Tokens {
+		if se, ok := tok.(xml.StartElement); ok {
+			if se.Name.Space != "" {
+				ns.getPrefix(se.Name.Space)
+			}
+			for _, attr := range se.Attr {
+				if attr.Name.Space != "" && attr.Name.Space != "xmlns" {
+					ns.getPrefix(attr.Name.Space)
+				}
+			}
+		}
+	}
+	// iniital element must be a StartElement
+	se := x.Tokens[0].(xml.StartElement)
+	ns.applyToSE(&se)
+	// add namespaces to first element
+	for _, pfx := range ns.prefixes {
+		ns := ns.prefixToURL[pfx]
+		se.Attr = append(se.Attr, xml.Attr{
+			Name:  xml.Name{Local: "xmlns:" + pfx},
+			Value: ns,
+		})
+	}
+	e.EncodeToken(se)
+
+	for _, tok := range x.Tokens[1:] {
+		if se, ok := tok.(xml.StartElement); ok {
+			ns.applyToSE(&se)
+			e.EncodeToken(se)
+		} else if ee, ok := tok.(xml.EndElement); ok {
+			ns.applyToEE(&ee)
+			e.EncodeToken(ee)
+		} else if err := e.EncodeToken(tok); err != nil {
+			return err
+		}
+	}
+	return nil
+}