c ====================================================================
c MDFLIST2.CCT - Format a gloss index (finderlist)
c For Shoebox version 5.0
c See the *** comments to learn how to modify this change table.
c --------------------------------------------------------------------
c
c Input:     Lexical data fields selected by MDFLIST1.CCT according to
c            the user's choices (e.g. Audience, Include part of speech).
c            Each reversal item from a \re, \ge, \rn, or \gn field
c            along with related fields \lc, \hm, \se, \ps, \pn, \sn
c            has become a reversal record. Shoebox has resorted these
c            records by reversal item and part of speech.
c
c Output:    A Standard Format file which will be converted to RTF
c            according to the "MDF Rich Text Format" database type's
c            marker properties (field names in the .typ file correspond
c            to style names in the Word document template .dot file).
c            Shoebox uses one of the MDF_*.CCT change tables (selected
c            in Audience Setup) to change labels and remove redundant
c            periods after parts of speech in this table's output.
c
c 1994-02    David F. Coward (david_coward@sil.org)
c            SIL, Kotak Pos 1050, Ambon, Maluku 97001 INDONESIA
c            MDFLIST.CCT change table in MDF version 1.0 for DOS
c 1996-08-10 Mark R. Pedrotti (mark_pedrotti@sil.org)
c            JAARS, P.O. Box 248, Waxhaw, NC 28173-0248 USA
c            Adapted from David Coward's MDFLIST.CCT and MDFMERG.CCT.
c 1997-12-04 MRP - Change notation for "special symbols" to |{emdash} etc.
c 1998-03-31 MRP - Output sense number immediately following the subentry,
c            not the entry, if a reversal item has both of them.
c            Simplified by changing some markers that MDFLIST1.CCT outputs
c            and removing unneeded logic for character styles.
c 1998-04-14 MRP - Generalize the hierarchy logic.
c            EmdashForEmptyPartOfSpeech option.
c 1998-07-25 MRP - DontRepeatPartOfSpeech option.
c            Acknowledgment: Prototyped by Bernis Eliason (SIL Colombia).
c            Note: Now MDFList1.cct outputs homonym numbers as part of
c            the \lc field using |hm{...} subfield notation.
c            Change other option to DontReplaceEmptyPartOfSpeech.
c 1998-12-09 MRP - Add a regional finderlist option and distinguish
c            the audience's language from the language of the index.
c            Added HR and LR styles, use fv for vernacular lexeme.
c 1999-03-18 MRP - Change options to \_EnglishGlossIndex, etc.
c 2000-05-25 MRP - Allow subfield formatting in reversals and glosses.
c            For example: game of fv:sungka, play a...
c            If there are subfields, the \re, \rn, or \rr field
c            output from MDFList1.cct consists of two lines:
c            1. gloss index for sorting (without the subfield notation)
c            2. gloss index for printing
c
c --------------------------------------------------------------------

c ====================================================================
c INITIALIZATION
c --------------------------------------------------------------------

begin >
    store(WS) ' ' nl d9 endstore  c White space: space, newline, tab
    store(Backslash) '\' endstore
    store(Newline) nl endstore
    store(total) '0' endstore  c Initialize total number of entries
    c Set default options
    set(TwoColumns)  c Page Setup, Section, Columns, Number: 2
    set(Total)  c Options, Formatting: Total number of entries
    set(sn_under_se)  c Standard hierarchy

    c *** To select the following option change clear to set:
    clear(DontRepeatPartOfSpeech)
    clear(DontReplaceEmptyPartOfSpeech)
    
    use(gOptionFields)

c ====================================================================
c DEFINITIONS FOR OUTPUT FORMATTING
c --------------------------------------------------------------------

c ====================================================================
define(dResetStoresAndSwitches) >
    store(rx) clear(rx)          c \re Reversal (E) or \rn or \rr
    store(lc) clear(lc)          c \lc Lexical citation
    store(hm) clear(hm)          c \hm Homonym number
    store(se) clear(se)          c \se Subentry
    store(ps) clear(ps,psEmpty)  c \ps Part of speech
    store(pn) clear(pn,pnEmpty)  c \pn Part of speech (n)
    store(sn) clear(sn)          c \sn Sense number
    endstore

c ====================================================================
define(dOutputReversalRecord) >
    endstore

    if(rxPrev) begin  c If there's a previous record
        ifn(AlreadyPunctuated) begin
            c The previous record's final punctuation depends on
            c this record's headword: is it the same or different?
            ifeq(rx) cont(rxPrev) begin '|fs{;}' end else '|fs{.}' endif
            nl  c End of the last field in the previous reversal record
        end endif
    end endif

    if(rx) begin
        set(OutputParagraph)  c 1998-07-25 MRP
        ifneq(rx) cont(rxPrev) begin  c Only output headword when it changes
            incr(total)  c Increment total number of entries
            '\FP' nl  c Finderlist Paragraph
            clear(OutputParagraph)  c Has now been output
            c Why mark the tab as |fs? If the last headword on a page has
            c a homonym number, then a tab with the headword's style
            c would be used in the header STYLEREF instead of the headword.
            if(e) '\HE ' endif  c 1998-12-09 MRP
			if(n) '\HN ' endif
			if(r) '\HR ' endif
			out(rx) '|fs{ ' d9 '}' nl
            set(rxPrev) store(rxPrev) outs(rx) endstore
            clear(pxPrev) store(pxPrev) endstore  c 1998-07-25 MRP
        end endif  c ifneq(rx) cont(rxPrev)

        if(ps) begin 
            set(OutputPartOfSpeech)  c 1998-07-25 MRP
            if(pxPrev) begin
                ifeq(ps) cont(pxPrev) begin
                    if(DontRepeatPartOfSpeech)
                        clear(OutputParagraph)
                        clear(OutputPartOfSpeech)
                    endif  c if(DontRepeatPartOfSpeech)
                end endif  c ifeq(ps)
            end endif  c if(pxPrev)
            set(pxPrev) store(pxPrev) outs(ps) endstore

            c Use the same paragraph style (with a tab indent) for lines
            c under the same headword. Why mark the tab as |fs?
            c It avoids the need for the No Word Wrap marker property.
            c NOTE: If the user unchecks Include part of speech,
            c identical reversal items occur together in one paragraph.
            if(OutputParagraph) '\FP |fs{' d9 '}' nl endif
            
            if(OutputPartOfSpeech) begin
                if(psEmpty) begin
                    ifn(DontReplaceEmptyPartOfSpeech) '\ps |{emdash}' nl endif
                end else
                    '\ps ' out(ps) '.' nl
                endif  c if(psEmpty)
            end endif  c if(OutputPartOfSpeech)
        end endif  c if(ps)

        if(pn) begin 
            set(OutputPartOfSpeech)  c 1998-07-25 MRP
            if(pxPrev) begin
                ifeq(pn) cont(pxPrev) begin
                    if(DontRepeatPartOfSpeech)
                        clear(OutputParagraph)
                        clear(OutputPartOfSpeech)
                    endif  c if(DontRepeatPartOfSpeech)
                end endif  c ifeq(pn)
            end endif  c if(pxPrev)
            set(pxPrev) store(pxPrev) outs(pn) endstore

            c Use the same paragraph style (with a tab indent) for lines
            c under the same headword. Why mark the tab as |fs?
            c It avoids the need for the No Word Wrap marker property.
            c NOTE: If the user unchecks Include part of speech,
            c identical reversal items occur together in one paragraph.
            if(OutputParagraph) '\FP |fs{' d9 '}' nl endif

            if(OutputPartOfSpeech) begin
                if(pnEmpty) begin
                    ifn(DontReplaceEmptyPartOfSpeech) '\ps |{emdash}' nl endif
                end else
                    '\pn ' out(pn) '.' nl
                endif  c if(pnEmpty)
            end endif  c if(OutputPartOfSpeech)
        end endif  c if(pn)

        '\fv '  c 1998-12-09 MRP

        if(se) begin
            out(se)
            if(sn) begin
                if(sn_under_se) ' |sn{(' out(sn) ')}' clear(sn) endif
            end endif c if(sn)
        end endif  c if(se)

        if(lc) begin
            if(se) '|fs{,} |fl{see:} ' endif
            out(lc)
            if(hm) '|hm{' out(hm) '}' endif
            if(sn) ' |sn{(' out(sn) ')}' endif
        end endif  c if(lc)

        do(dResetStoresAndSwitches)
    end endif  c if(rx)
    
    clear(AlreadyPunctuated)
    if(LP) begin
        c Section heading for new letter (or syllable).
        c Unless this is the first section heading, the last record
        c in the previous section needs its final punctuation.
        if(rxPrev) '|fs{.}' nl set(AlreadyPunctuated) endif
        c In Shoebox RTF conversion, the \LS section's single-column
        c format appies to the text that FOLLOWS the break.
        c The \LP1 marker's style has space after and keep with next.
        if(TwoColumns) '\LS' nl '\LP ' else '\LP1 ' endif
        if(e) '|LE{' endif  c 1998-12-09 MRP
		if(n) '|LN{' endif
		if(r) '|LR{' endif
		out(LP) '}' nl
        if(TwoColumns) '\DS' nl endif  c Section break
        store(LP) endstore clear(LP)
    end endif  c if(LP)

c ====================================================================
define(dEndfile) >
    do(dOutputReversalRecord)
    if(rxPrev) '|fs{.}' nl endif  c Punctuate the final paragraph
    if(TwoColumns) '\SS' nl endif  c A single-column section FOLLOWS
    if(Total) '\CP Total number of entries: ' out(total) nl endif
    endfile

c ====================================================================
c RULES FOR INPUT PROCESSING
c --------------------------------------------------------------------

c ====================================================================
group(gOptionFields)  c Shoebox inserts these fields preceding the data

'\_National' nl > set(National)  c 1998-12-09 MRP
'\_EnglishGlossIndex'  nl > set(e) set(GlossOption)
'\_NationalGlossIndex' nl > set(n) set(GlossOption)
'\_RegionalGlossIndex' nl > set(r) set(GlossOption)

'\_SingleColumn' nl > clear(TwoColumns)
'\_NoTotal' nl > clear(Total)

'\_se_ps_sn' nl > set(sn_under_se)
'\_se_sn_ps' nl > set(sn_under_se)
'\_AlternateHierarchy' nl > next
'\_sn_se_ps' nl > clear(sn_under_se)
'\_sn_ps_se' nl > clear(sn_under_se)
'\_ps_se_sn' nl > set(sn_under_se)
'\_ps_sn_se' nl > clear(sn_under_se)

'\_DontRepeatPartOfSpeech' > set(DontRepeatPartOfSpeech)  c 1998-07-25 MRP
'\_DontReplaceEmptyPartOfSpeech' > set(DontReplaceEmptyPartOfSpeech)

'\_' > store(omit) use(gUnknownOptionField)

nl > ''

'' >
	ifn(GlossOption) begin  c 1998-12-09 MRP
		if(National) set(n) else set(e) endif
	end endif  c ifn(GlossOption)
	c set(r) clear(e,n)  c *** Test the regional option in Shoebox 4
	store(omit) use(gMain)

c --------------------------------------------------------------------
group(gUnknownOptionField)  c From an older or newer version of Shoebox

nl fol(Backslash) > endstore use(gOptionFields)

c ====================================================================
group(gMain)  c Read the reversal records

'\LP ' > store(LP) set(LP) use(gField)
          
'\lx' nl > do(dOutputReversalRecord)  c Output previous record

'\lc ' > store(lc) set(lc) use(gField) set(NonBreakingHyphen)
'\hm ' > store(hm) set(hm) use(gField)

'\se ' > store(se) set(se) use(gField)

'\ps ' > store(ps) set(ps) use(gField)
'\ps' fol(Newline) > store(omit) use(gField) set(ps,psEmpty)

'\pn ' > store(pn) set(pn) use(gField)
'\pn' fol(Newline) > store(omit) use(gField) set(pn,pnEmpty)

'\sn ' > store(sn) set(sn) use(gField)                        

'\rr ' > next  c 1998-12-09 MRP
'\rn ' > next
'\re ' > store(rx) set(rx) use(gField,gReversal) set(NonBreakingHyphen)

'\' > store(omit) use(gField)

c ====================================================================
group(gField)  c Read the contents of a data field

c NOTE: MDFLIST1.CCT has already removed extra white space from fields
c and converted underline to space in gloss fields used for reversal.

'-' > if(NonBreakingHyphen) '|{_}' else dup endif
'-' prec(WS) > '|{_}'

nl fol(Newline) > ''  c Shoebox inserts a newline between records

nl fol(Backslash) > use(gMain) clear(NonBreakingHyphen)

endfile > next
nl endfile > do(dEndfile)

c --------------------------------------------------------------------
group(gReversal)

c 2000-05-25 Allow subfield formatting in reversals and glosses.
nl > store(rx)

c END ===============================================================