                        WordPerfect File Information
              Copyright 1984,1985,1986 WordPerfect Corporation
                                 Version 4.2

File Structure

WordPerfect files are ASCII files.  The text is neither encrypted nor
compacted.  WordPerfect files do not contain a ^Z as an end of file
character.  If a program places a ^Z and pads to the end of the block with
garbage, then WordPerfect may crash (padding with nulls or ^Z's is OK).

Formatting or function codes are embedded as they occur in the text.  No
code is used for beginning of file, or end of file.  When creating
WordPerfect files from other programs, keep in mind that 1) the initial WP
margin settings are usually 10 and 74 and it is best to keep the line length
65 characters or less unless you change the margin settings and 2) it is
best not to pad with spaces.

Function Codes

All function codes listed below are represented as base-8 (octal) numbers. 
Angle brackets are not actually a part of the function codes.  They are
simply used to indicate that the enclosed number is a single-byte code
unless clearly stated otherwise.  Square brackets are similarly not part of
the code but are simply used to clarify the notation.

Where multiple byte data is indicated within angle brackets, the data is
treated as a binary string.  Wherever an "old value" is indicated, a binary
zero can be used.  WordPerfect will automatically take care of updating the
"old value" locations.  On output, the "old value" can be ignored.

Multiple-Byte Functions

The code for multiple-byte functions (those above 300) always appears
twice--the first occurrence is the "open gate," and a second occurrence is
the "closing gate."  These functions present a special problem, because some
of the functions have a variable length.  The length of each function is
listed at the left margin.  Those functions whose length is indicated by a
-1 have a variable length.  In the case of variable-length, multiple-byte
function, a program should scan for the "closing gate."  For other multiple-
byte functions, a program should simply skip over the number of bytes indi-
cated.

Additional Notes

There is no beginning of field or beginning of record code in WordPerfect
secondary merge files.  The end of field separator is a ^R followed by a
Hard Return (Line Feed), and the end of record separator is a  followed by
a Hard Return (Line Feed).

For a spelling or grammar check, special care needs to be taken to allow for
the various types of hyphenation function codes (251-256), and for the hard
space (240).1     011 09   Tab
1     012 0a   Hard new line
1     013 0b   Soft new page
1     014 0c   Hard new page
1     015 0d   Soft new line

1     200 80   No-op (always deleted)
1     201 81   Turn right justification on
1     202 82   Turn right justification off
1     203 83   End of centered text
1     204 84   End of aligned or flushed text
1     205 85   Used as a temp starting pt for math calc
1     206 86   Center page from top to bottom
1     207 87   Begin col mode
1     210 88   End col mode (at top of page)
1     211 89   Tab after the right margin
1     212 8a   Widow/orphan on
1     213 8b   Widow/orphan off
1     214 8c   Hard end of line & soft end of page
1     215 8d   Footnote # (appears only inside of footnotes)
1     216 8e   Reserved
1     217 8f   Reserved
1     220 90   Red line on
1     221 91   Red line off
1     222 92   Strike out on
1     223 93   Strike out off
1     224 94   Underline on
1     225 95   Underline off
1     226 96   Reverse video on (reserved)
1     227 97   Reverse video off (reserved)
1     230 98   Table of contents placeholder
1     231 99   Overstrike
1     232 9a   Cancel hyphenation of following word
1     233 9b   End of generated text
1     234 9c   Bold off
1     235 9d   Bold on
1     236 9e   Hyphenation off
1     237 9f   Hyphenation on
1     240 a0   Hard space
1     241 a1   Do subtotal
1     242 a2   Subtotal entry
1     243 a3   Do total
1     244 a4   Total entry
1     245 a5   Do grand total
1     246 a6   Math calculation column
1     247 a7   Begin math mode
1     250 a8   End math mode
1     251 a9   Hard hyphen in line
1     252 aa   Hard hyphen at end of line
1     253 ab   Hard hyphen at end of page
1     254 ac   Soft hyphen
1     255 ad   Soft hyphen at end of line
1     256 ae   Soft hyphen at end of page
1     257 af   End of text cols & end of line
1     260 b0   End of text cols & end of page
1     261 b1   negative number (N) for math mode
1     262 b2   italics on (not implemented in IBM PC version)
1     263 b3   italics off (not implemented in IBM PC version)
1     264 b4   shadow on (not implemented in IBM PC version)
1     265 b5   shadow off (not implemented in IBM PC version)
1     266 b6   outline on (not implemented in IBM PC version)
1     267 b7   outline off (not implemented in IBM PC version)
1     274 bc   Superscript
1     275 bd   Subscript
1     276 be   Advance 1/2 line up
1     277 bf   Advance 1/2 line down

6     300 c0   Margin reset
          <300><old left mar><old right mar>
               <new left mar><new right mar><300>
4     301 c1   Spacing reset - values stored in half line values
          <301><old spacing><new spacing><301>
3     302 c2   Left margin release
          <302><# of spaces to go left><302>
5     303 c3   Center the following text
          <303><type of center><center col #><starting col #><303>
               -text- <203>
          Type = 0 for centering between margins
               = 1 for centering around current column
5     304 c4   Align or flush right
          <304><align char><align col #><starting col #><304> -text- <204>
          If align char = 12 (new line) then this is a flush right and the
               align col # is the right margin; otherwise, the align col #
               is the next tab stop.
          If the high bit of the align character is set, then this is dot
               leader align or dot leader flush right.
6     305 c5   Reset hyphenation zone (Hotzone)
          <305><old left hzone><old right hzone>
                <new left hzone><new right hzone><305>
4     306 c6   Set page number position
          <306><old pos code><new pos code><306>
          Code:    0 = None         1 = Top Left    2 = Top Center
                   3 = Top Right    4 = Top L & R   5 = Bottom Left
                   6 = Bot Center   7 = Bot Right   8 = Bottom L & R
6     307 c7   Set page number
          <307><old # high order><old # low order>
                <new # high order><new # low order><307>
          Only low order 15 bits determine the page number.
          If high bit is set, style is Roman, otherwise Arabic.
8     310 c8   Set page # col positions
          <310><old left><old cen><old right>
                <new left><new cen><new right><310>
42    311 c9   Set tabs (used in 2.2-4.1, 4.2 and later uses 361)
          <311><old tab table (20. Bytes)><new tab table (20. Bytes)><311>
          Each bit represents one position counting from bit 0 to bit 159
3     312 ca   Conditional end of page
          <312><number of single spaced lines not to be broken><312>
6     313 cb   Set pitch and/or font
          <313><old pitch><old font><new pitch><new font><313>
          If pitch is negative, then it is proportional.
4     314 cc   Set temporary margin (indent)
          <314><old tmar><new tmar><314>
3     315 cd   End of temporary margin (used in versions before 2.2)
          <315><temp marg><315>
4     316 ce   Set top margin (values are in half lines)
          <316><old top marg><new top marg><316>
3     317 cf   Suppress page characteristics
          <317><supress code><317>
          Code: (any or all bits may be inclusive or'ed together)
             1 = all suppressed
             2 = pg #'s suppressed
             4 = pg # moved to bottom
            10 = all headers suppressed
            20 = header a suppressed
            40 = header b suppressed
           100 = footer a suppressed
           200 = footer b suppressed
6     320 d0   Set form length
          <320><old form length><old # of text lines>
                <new form length><new # of text lines><320>
          Form length is stored as a number of 6 lpi lines
          Number of text lines is stored in half-lines
-1    321 d1   Header/footer
          <321><old def byte><# 1/2 lines used by old header/footer><377>
                <377><lmar><rmar> -text- <377><# of 1/2 lines><def><321>
          Type (low-order 2 bits)       Occurrence (high-order 6 bits)
               0 = header a                 0 = never
               1 = header b                 1 = all pages
               2 = footer a                 2 = odd pages
               3 = footer b                 4 = even pages
          Note that the low-order two bits of the old def byte must be   
correct
-1    322 d2   Footnote (used in 2.2-3.0, 4.0 and later uses 342)
          <322><fn #><# of half lines><377><lmar><rmar> -text- <322>
4     323 d3   Set footnote # (used in 2.2-3.0, 4.0 and later uses 344)
          <323><old #><new #><323>
4     324 d4   Advance to half-line # (stored in half-line values)
          <324><old line #><adv to 1/2 line #><324>
4     325 d5   Set lines per inch (only 6 or 8 lpi is valid)
          <325><old lpi code><new lpi code><325>
6     326 d6   Set extended tabs
          <326><old start><old increment><new start><new increment><326>
          Starting column position must be at least 160
-1    327 d7   Define math columns
          <327><Old Col. Def (24 Bytes)>
                   [<Old Calc 0>]<0>[<Old Calc 1>]<0>
                   [<Old Calc 2>]<0>[<Old Calc 3>]<0><377>
               <New Col. Def (24 Bytes)>
                   [<New Calc 0>]<0>[<New Calc 1>]<0>
                   [<New Calc 2>]<0>[<New Calc 3>]<0><327>
4     330 d8   Set alignment character
          <330><old char><new char><330>
4     331 d9   Set left margin release (# of columns to go left)
          <331><old #><new #><331> (used in 2.2-3.0, not used in 4.0)
4     332 da   Set underline mode
          <332><old mode><new mode><332>
               0 = normal underlining
               1 = double underlining
               2 = single underlining continuous
               3 = double continuous
4     333 db   Set sheet feeder bin number
          <333><old #><new #><333>
          Number is stored as one less than the bin number
               (bin #1 = 0, etc.)
-1    334 dc   End of page function (WordPerfect inserts this--changes each
               version)
          <334><# of 1/2 lines at end of page, low 7 bits><high 7 bits>
               <# of 1/2 lines used for footnotes>
               <# of pages used for footnotes>
               <# footnotes on this page><ceop flag><suppress code><334>
          If eop for last col on page then after <suppress>
               comes 23 more bytes:
               <# of 1/2 lines for col 1><# of 1/2 lines for col 2>
               <# of 1/2 lines for col 3>...<# of 1/2 lines for col 23>
               <line # of col on (0 if none on this page)>
24    335 dd   Define columns (used in 2.2-4.1, 4.2 and later uses 363)
          <335><old#cols><l1><r1><l2><r2><l3><r3><l4><r4><l5><r5>
                <new#cols><l1><r1><l2><r2><l3><r3><l4><r4><l5><r5><335>
          # cols:  low-order 7 bits = number
                   high-order 1 bit = 1 if parallel columns (for 4.1)
4     336 de   End of temp marg
          <336><old left tmar><old right tmar><336>
-1    337 df   Invisible characters (embedded printer command)
          <337><7-bit text><337>
          If character is >=277, then it is represented as <277><char-277>
4     340 e0   L/R temporary margin
          Old format (pre 4.0):
               <340><New right tmar><new left tmar><340>
          New format (4.0):
               <240><0><Difference between old and new left margin><340>
3     341 e1   Extended character
          <341><character><341>
-1    342 e2   New footnote/endnote (4.0 and later)
          <342><def><a><b><c><d><old ftnln><# ln pg 1><# ln pg 2>...
               <# Ln pg n><# pgs><377><lmar><rmar> - text - <342>
          Def: Bit 0: 0 = use numbers, 1 = use chars
               Bit 1: 0 = footnote, 1 = endnote
          a,b: If def bit 0 = 0, then a,b = footnote/endnote #
               If def bit 0 = 1, then a = # of chars, b = char
          c,d: Number of lines in foot/end note
          Note:  a,b and c,d are 14 bit numbers split into 7-bit bytes, high
                   order byte first
          Note:  For endnotes, there is just a null between <d> & <377>
150   343 e3   Footnote information (options) function
          <343><old values 74 bytes><new values 74 bytes><343>
          Byte Meaning
           1       spacing in footnotes
           2       spacing between footnotes
           3       number of lines to keep together
           4       flag byte (bits: b ln en ft n)
                        n: 1 if numbering starts on each page
                   en & ft:    0 = use numbers
                           1 = use characters
                           2 = use letters
                       ln: 0 = no line separator
                           1 = 2" line
                           2 = line from left to right margin
                           3 = 2" line & "Footnote continued" message
                        b: 0 = footnotes after text
                           1 = footnotes at bottom of page
          5        # of chars used in place of ftn #'s
          6-10 chars used in place of ftn #'s (null terminated if < 5)
          11       # of displayable chars in string for footnote-text
          12-26    string for footnote-text
          27       # of displayable chars in string for endnote-text
          28-42    string for endnote-text
          43       # of displayable chars in string for footnote-note
          44-58    string for footnote-note
          59       # of displayable chars in string for endnote-note
          60-74    string for endnote-note
6     344 e4   New set footnote # (4.0 and later)
          <344><old # high><old # low><new # high><new # low><344>
          Footnote numbers are 14 bit numbers split into 7-bit bytes, high
               order byte first
23    345 e5   Paragraph number definition
               (used in 2.2-4.1, 4.2 and later uses 356)
          <345><old 7 level numbers><old 7 def bytes>
               <new 7 def bytes><345>
          Def byte = two nibbles:
               Numbering style (low nibble) Punctuation (high nibble)
               0 = caps Roman               0 = nothing
               1 = l.c. Roman               1 = "." after number
               2 = caps letter          2 = ")" after number
               3 = l.c. letter          3 = "(" before, ")" after
               4 = Arabic
               5 = Arabic with previous levels separated by "."
11    346 e6   Paragraph number (used in 2.2-4.1, 4.2 and later uses 357)
          <346><new level #><def byte><old 7 numbers><346>
          Level # is 0 for first level, 1 for 2nd, etc.
3     347 e7   Begin marked text
          <347><def,info><347> - text - <350><def,info><350>
               Definition (high nibble)     Information (low nibble)
               0 = table of contents        level (0-4)
               2 = list                     list # (0-4)
3     350 e8   End marked text
          <350><def,info><350>
          Def, info same as for mark (347)
-1    351 e9   Define marked text
          <351><def,info><5 bytes definition><concordance file name><351>
          Def, info byte:
               Definition (high nibble)     Information (low nibble)
               0 = table of contents        level (0-4)
               1 = index                    concordance file flag
               2 = list                     list # (0-4)
               3 = ToA                      ToA section # (0-15)

          For Index the low nibble of the def, info byte is:
                   0 = no concordance file
                   1 = concordance file name included in function
          For ToC, 5 Definition bytes represent the five levels
          For index and lists only first definition byte is significant
               Definition bytes for ToC, index, and lists:
                   0 = no page numbers
                   1 = page # after text, preceded by two spaces
                   2 = page # after text, in parentheses, preceded by one
                       space
                   3 = page # flush right
                   4 = page # flush right with dot leader
          For ToA, just first def byte is significant
               Def byte=xxxBxxDU
                   B=1 if blank line should be inserted between authorities
                   D=1 if dot leader before page numbers
                   U=1 if underlining allowed
-1    352 ea   Define index mark
          <352><index heading><0><subheading><352>
32    353 eb   Date/Time function
          <353><30 byte, null terminated format string><353>
4     354 ec   Block Protect
          <354><Def><# of 1/2 lines in block><354>
          Def  = 0 for Block Protect On
               = 1 for Block Protect Off
-1    355 ed   Table of Authorities
          <355><Section Number><0>short form<0>full form<355>
               Section Number = 0-15 or 32
               0 thru 15 are valid section numbers
               a section number of 32 is used to indicate that the function
               contains only a short form and no full form
44    356 ee   Paragraph Number Definition (new)
          <356><old 7 level numbers (words)><old 7 def bytes>
               <new 7 def bytes><7 starting level #'s (words)><356>
          each def byte = two nibbles
          low nibble = numbering style      high nibble = punctuation    
          ----------                    -----------
          0 = caps roman                0 = nothing
          1 = l.c. roman                1 = "." After number
          2 = caps letter               2 = ")" after number
          3 = l.c. letter               3 = "(" before, ")" after
          4 = arabic
          5 = arabic with previous levels separated by "."
18    357 ef   Paragraph Number (new)
          <357><new level #><def byte><old 7 numbers (words)><357>
                level # is 0 for first level, 1 for 2nd, etc.
6     360 f0   Line Numbering
          <360><Old Interval><Old Position>
               <New Interval><New Position><360>
          interval = STR00000:
               bit 7 (S) = Status (1 = line numbering on)
               bit 6 (T) = Number only Text lines (1 = on)
               bit 5 (R) = Restart numbering on each page (1 = on)
               00000 = line numbering interval (0-30)
          position =   distance from left edge of paper to right edge of
                       line number in 10ths of an inch 
106   361 f1   Tab Reset
          <361><old tab table (32 bytes)><old type table (20 bytes)>
               <new tab table (32 bytes)><new type table (20 bytes)><361>
               Each bit in tab table represents one position counting bit 0
                   to 250.
                   Each nibble in type table is the type of the
                       corresponding tab, ie. the 5th nibble is the type for
                       the 5th bit set, up to 40 tabs, the rest are assumed
                       type 0.
               Types:  0 = Normal left justified tab
                       1 = Centered tab
                       2 = Right justified tab
                       3 = Decimal aligned tab
                       4 = Left justified tab with a dot leader
                       5 = not used
                       6 = Right justified with a dot leader
                       7 = Decimal aligned with a dot leader
-1    362 f2   Comment/Summary
          <362><Def><old ufcur><# Lines><0> - Text - <362>
          Def: Bit 0=0 if comment
                    =1 if document summary, all text in following format:
          Date Of Creation <HRt> Author <HRt> Typist <HRt> Comments
100   363 f3   Define columns (length is 24*2*2+4)
          <363><old#cols><l1><r1><l2><r2>...<l24><r24>
                <new#cols><l1><r1><l2><r2>...<l24><r24><363>
          # cols:  low-order 7 bits = number
                   high-order 1 bit = 1 if parallel columns

4     364 f4   Point size (not implemented in IBM PC version)
          <364><old size><new size><364>
          size is in points (1/72 inch)

-1    365 f5   Picture
          for Mac: 8 bytes for rectangle
                   4 bytes for type (PICT)
                   2 bytes for indentifier
          perhaps also def byte to indicate whether it takes up space or
          overlays existing text; perhaps size in half line (maybe a 0 size
          is sufficient to indicate it overlays existing text).

5     366 f6   Different display character when hyphenated
          <366><def byte><char when inline><char when hyphenated><366>
          bits in def byte:
               bit 0 = 0 if no hyphenation next to function
                     = 1 if hyphenation happens next to function
               bit 1 = 0 if this function is after the hyphenation
                     = 1 if this function preceeds the hyphenation
          if a char is a null, then nothing is displayed for it.
          examples:  papaatje = pa-pa-<366><1><"a"><0><366>tje   
                     dineetje = di-ne<366><3><"e"><"r"><366>-tje  

-1    367 f7   Leading (for Mac)

-1    370 f8   Kerning Adjustment (for Mac)