WordPerfect 6 and later documents

WordPerfect 6.0 introduced the format that every later version still writes, marked by major version 2 in the WPC header. The prefix is an index of packets (fonts, styles, summary, graphics), and the document area is a stream of text and function codes like that of 5.x, but with different code ranges and group layouts. This page follows libwpd; numbers are little-endian. The bytes inside each group and packet are on the 6.x function and packet data page.

Index header and prefix packets

The index header starts at the offset in bytes 14–15 of the WPC header (at least 16).

Offset	Size	Field	Meaning
`0`	1	Flags	2.
`1`	1	Reserved	0.
`2`	2	Number of indexes	How many 14-byte entries follow.
`4`	10	Reserved	–
`14`	14 each	Index entries	See below.

Offset	Size	Field	Meaning
`0`	1	Flags	Bit 0: the packet has child packets.
`1`	1	Packet type	See packet types.
`2`	2	Use count	How many codes in the document refer to the packet.
`4`	2	Hide count	How many of those references are hidden.
`6`	4	Data size	Length of the packet’s data.
`10`	4	Data offset	Where the data is in the file.

Function codes in the document refer to packets by their position in this index (their “prefix ID”).

Packet types

Type	Packet
`0x02`	Document settings (descriptor pool)
`0x07`	Hyperlink
`0x08`	General WordPerfect text: text kept outside the main stream, such as headers, footnotes and box contents
`0x10`	Next total page count
`0x12`	Extended document summary (see below)
`0x1B`	Comment (annotation)
`0x23`	Printer selection
`0x25`	Initial font
`0x30`	Style data
`0x31`	Outline style (numbering: 0 Arabic, 1 lower-case letters, 2 upper-case letters, 3 lower-case Roman, 4 upper-case Roman)
`0x34`	Shared and personal file dates
`0x40`	Graphics file name
`0x41`	Graphics box style
`0x43`	Fill style
`0x4E`	Check-as-you-go (spelling) data
`0x55`	Desired font
`0x5E`	Prefix time stamp
`0x69`	Table style
`0x6D`	World Wide Web data
`0x6F`	Cached graphics file data (the picture itself)
`0x77`	Undo information (WordPerfect 6.1)

Many other packet types exist; libwpd names only these, and their full list is not public.

Document summary fields

Field numbers in the extended document summary packet:

N	Field	N	Field	N	Field
1	Abstract	2	Account	3	Address
4	Attachments	5	Author	6	Authorisation
7	Bill to	8	Blind copy	9	Carbon copy
10	Category	11	Checked by	12	Client
13	Comments	14	Creation date	15	Date completed
16	Department	17	Descriptive name	18	Descriptive type
19	Destination	20	Disposition	21	Division
22	Document number	23	Editor	24	Forward to
25	Group	26	Keywords	27	Language
28	Mail stop	29	Matter	30	Office
31	Owner	32	Project	33	Publisher
34	Purpose	35	Received from	36	Recorded by
37	Recorded date	38	Reference	39	Revision date (unused)
40	Revision notes	41	Revision number	42	Section
43	Security	44	Source	45	Status
46	Subject	47	Telephone number	48	Typist
49	Version date	50	Version notes	51	Version number

The document area

Bytes	Meaning
`0x00`	Not used; ignored.
`0x01`–`0x20`	The 32 “default extended international characters”: see below. Note that `0x20` is not a space.
`0x21`–`0x7F`	ASCII text.
`0x80`–`0xCF`	Single-byte functions, including the space.
`0xD0`–`0xEF`	Variable-length function groups.
`0xF0`–`0xFE`	Fixed-length function groups.
`0xFF`	Reserved; not a valid code.

Default extended international characters

Byte	Character	Byte	Character	Byte	Character	Byte	Character
`01`	å	`09`	â	`11`	è	`19`	Õ
`02`	Å	`0A`	ã	`12`	ê	`1A`	ö
`03`	æ	`0B`	Ã	`13`	í	`1B`	Ö
`04`	Æ	`0C`	ç	`14`	ñ	`1C`	ü
`05`	ä	`0D`	Ç	`15`	Ñ	`1D`	Ü
`06`	Ä	`0E`	ë	`16`	ø	`1E`	ú
`07`	á	`0F`	é	`17`	Ø	`1F`	ù
`08`	à	`10`	É	`18`	õ	`20`	ß

Single-byte functions

“EOL” is end of line, “EOC” end of column, “EOP” end of page. Hard ones were typed; soft ones are where WordPerfect’s layout broke the line; deletable ones are hard codes the program may remove.

Code	Function	Code	Function
`0x80`	Soft space (an ordinary space)	`0xBF`	Table off
`0x81`	Hard (non-breaking) space	`0xC0`	Table row at hard EOP
`0x82`	Soft hyphen in a line	`0xC1`	Table row at hard EOC at hard EOP
`0x83`	Soft hyphen at EOL	`0xC2`	Table row at hard EOC
`0x84`	Hard hyphen	`0xC3`	Table row at EOC at EOP
`0x85`	Automatic hyphen	`0xC4`	Table row at EOC
`0x87`	Dormant hard return	`0xC5`	Table row
`0xB4`	Deletable hard EOP	`0xC6`	Table cell
`0xB5`	Deletable hard EOC at EOP	`0xC7`	Hard EOP (page break)
`0xB6`	Deletable hard EOC	`0xC8`	Hard EOC at EOP
`0xB7`	Deletable hard EOL at EOC at EOP	`0xC9`	Hard EOC (column break)
`0xB8`	Deletable hard EOL at EOC	`0xCA`	Hard EOL at EOC at EOP
`0xB9`	Deletable hard EOL	`0xCB`	Hard EOL at EOC
`0xBA`	Deletable soft EOL at EOC at EOP	`0xCC`	Hard EOL (hard return, end of paragraph)
`0xBB`	Deletable soft EOL at EOC	`0xCD`	Soft EOL at EOC at EOP
`0xBC`	Deletable soft EOL	`0xCE`	Soft EOL at EOC
`0xBD`	Table off at EOC at EOP	`0xCF`	Soft EOL (wrapped line)
`0xBE`	Table off at EOC

0x86 and 0x88 to 0xB3 are single-byte functions that libwpd does not name.

Variable-length function groups

Layout of a group, as libwpd reads it:

Size	Field	Meaning
1	Code	`0xD0`–`0xEF`.
1	Sub-code	Which function of the group.
2	Size	Length of the whole group, both code bytes included.
1	Flags	Bit 7: prefix IDs follow.
1 + 2n	Prefix IDs	If bit 7: a count, then that many 16-bit packet numbers.
2	Non-deletable size	Length of the data that must be kept; what follows it, up to the end, may be dropped by programs that do not understand it.
…	Data	Depends on the group and sub-code.
2	Size	The size again.
1	Code	The code again.

Code	Group	Sub-codes named by libwpd
`0xD0`	End of line	1 soft EOL; 2 soft EOC; 3 soft EOC at EOP; 4 hard EOL; 5 hard EOL at EOC; 6 hard EOL at EOP; 7 hard EOC; 8 hard EOC at EOP; 9 hard EOP; 10 table cell; 11 table row and cell; 12 table row at EOC; 13 table row at EOP; 14 table row at hard EOC; 15 table row at hard EOC at hard EOP; 16 table row at hard EOP; 17 table off; 18 table off at EOC; 19 table off at EOC at EOP; 20–28 the deletable forms (soft EOL, soft EOC, soft EOC at EOP, hard EOL, hard EOL at EOC, hard EOL at EOP, hard EOC, hard EOC at EOP, hard EOP). Data items inside: 128 row information (5 bytes), 129 cell formula, 130 top gutter (4), 131 bottom gutter (4), 132 cell information (9), 133 cell spanning (4), 134 cell fill colours (10), 135 cell line colour (6), 136 cell number type (6), 137 cell floating-point number (11), 139 cell prefix flag (3), 140 recalculation error number (3), 141 do not end a paragraph style here (1)
`0xD1`	Page	0 top margin; 1 bottom margin; 2 suppress page features (flags 1 page number, 2 page number at bottom centre, 4 header A, 8 header B, 16 footer A, 32 footer B, 64 watermark A, 128 watermark B); 3 page number position; 17 paper size and type (form)
`0xD2`	Column	0 left margin; 1 right margin; column definitions with type 0 newspaper, 1 newspaper balanced, 2 parallel, 3 parallel with block protect
`0xD3`	Paragraph	1 line spacing; 4 tab set; 5 justification (0 left, 1 full, 2 centre, 3 right, 4 full including the last line, 5 reserved); 10 spacing after paragraph; 11 first-line indent; 12 left margin adjustment; 13 right margin adjustment; 14 outline definition
`0xD4`	Character	0 alignment character; 24 colour; 25 shading; 26 font face; 27 font size; 29 comment; 30 dot leader characters; 42 table definition on; 43 table definition off; 44 table column; 50 paragraph number on; 51 paragraph number off
`0xD5`	Cross-reference	–
`0xD6`	Header, footer, watermark	0 header A; 1 header B; 2 footer A; 3 footer B; 4 watermark A; 5 watermark B (pages: bit 1 odd, 2 even); the text is in a general text packet
`0xD7`	Footnote, endnote	0 footnote on; 1 footnote off; 2 endnote on; 3 endnote off
`0xD8`	Set number	0 user counter; 1 box; 2 page; 3 secondary page; 4 chapter; 5 volume; 6 paragraph; 7 footnote; 8 endnote; 9 line
`0xD9`	Numbering method	The same ten counters; page methods 0 Arabic, 1 lower-case letters, 2 upper-case letters, 3 lower-case Roman, 4 upper-case Roman
`0xDA`	Display number	Even sub-codes start, odd ones end, the display of: 0/1 user counter, 2/3 box, 4/5 page, 6/7 secondary page, 8/9 chapter, 10/11 volume, 12/13 paragraph, 14/15 footnote, 16/17 endnote, 18/19 line, 20/21 total pages
`0xDB`	Increment number	The same ten counters as set number
`0xDC`	Decrement number	The same ten counters
`0xDD`	Style	4 paragraph style begin on (part 1); 5 begin off (part 1); 6 begin on (part 2); 7 begin off (part 2); 8 end on; 9 end off; 10 global on; 11 global off
`0xDE`	Merge	–
`0xDF`	Box	0 box anchored to a character; 1 to a paragraph; 2 to the page; 3 graphics line. Flags say which data follows: 0x8000 counter, 0x4000 position, 0x2000 content, 0x1000 caption, 0x0800 border, 0x0400 fill, 0x0200 wrapping, 0x0100 hypertext, 0x0080 HTML, 0x0040 grouping, 0x0020 drawing object
`0xE0`	Tab	The tab type, which libwpd describes as bits 3 to 8 of the tab group
`0xE1`	Platform	0/1 DDE start/end; 2/3 publish start/end; 4/5 subscribe start/end; 6 subtitle mark; 7 subtitle text; 8 subtitle option; 9 subtitle text options; 10 suppress subtitle; 11 third party; 12/13 pen foreground colour (32/64-bit); 14/15 pen background colour; 16 pen style; 17 pen pattern; 18/19 pen size (16/32-bit); 20/21 brush gradient; 22/23 brush foreground colour; 24/25 brush background colour; 26 brush pattern; 27 miscellaneous attributes; 28 Presentations slide options; 29 Presentations organisation chart node; 30 hide for conversions; 31 hide off; 32 SGML start tag; 33 start tag end; 34 end tag; 35 end tag end; 36 character reference; 37 end character reference; 38 entity reference; 39 end entity reference; 40/41 ignore on/off; 42/43 marked section on/off; 44/45 processing instruction on/off
`0xE2`	Formatter	–

“–” means the group exists but libwpd names none of its sub-codes. Groups 0xE3 to 0xEF are not described. The layout of each sub-code’s data is only partly known from libwpd; what is known is on the 6.x data page.

Fixed-length function groups

A fixed-length group starts and ends with its code; the sizes below include both.

Code	Size	Function
`0xF0`	4	Extended character: character number, character set
`0xF1`	5	Undo: 0 invalid text start, 1 invalid text end, 2 valid text start, 3 valid text end
`0xF2`	3	Attribute on: one attribute
`0xF3`	3	Attribute off
`0xF4`	3	Unknown
`0xF5`	3	Unknown
`0xF6`	4	Unknown
`0xF7`	4	Unknown
`0xF8`	4	Unknown
`0xF9`	5	Unknown
`0xFA`	5	Unknown
`0xFB`	6	Highlight on
`0xFC`	6	Highlight off
`0xFD`	8	Unknown
`0xFE`	8	Unknown

Attributes

0 extra large, 1 very large, 2 large, 3 small, 4 fine, 5 superscript, 6 subscript, 7 outline, 8 italic, 9 shadow, 10 redline, 11 double underline, 12 bold, 13 strikeout, 14 underline, 15 small caps, 16 blink.

Tab types

Value	Meaning	Value	Meaning
`0x00`	Back tab	`0x08`	Centre between margins
`0x01`	Table tab	`0x09`	Centre on current position
`0x02`	Left tab	`0x0A`	Centre tab
`0x04`	Bar tab	`0x10`	Flush right
`0x06`	Left indent	`0x12`	Right tab
`0x07`	Left and right indent	`0x1A`	Decimal tab

Character sets

Extended characters (F0 character set F0) use WordPerfect’s character sets: 0 ASCII, 1 multinational, 2 phonetic symbols, 3 box drawing, 4 typographic symbols, 5 iconic symbols, 6 maths and science, 7 maths and science extended, 8 Greek, 9 Hebrew, 10 Cyrillic, 11 Japanese, 12 Tibetan, 13 Arabic, 14 Arabic script. libwpd maps each character to Unicode.

Units

Lengths such as margins and tab positions are in WordPerfect units, 1,200 to the inch.

48 65 6C 6C 6F 80  F2 0C F2  77 6F 72 6C 64  F3 0C F3  CC

“Hello”, a space (0x80), bold on, “world”, bold off, hard return.

.WPD File Format – WordPerfect 6 and later documents