• ClasX-LBF Command -- Braille Result
• (\\lineline) -- {$l$l}
• (\\parapara) -- {$p$p}
• (\\linepara) -- {$l$p}
• (\\paraline) -- {$p$l}

When you want {$l$l} in your final braille chapters, you must code it as (\\lineline). If you just typed ($l$l), CLEANUP would think it's redundant, and collapse it to ($p). From our discussions with linear transcribers, we believe the (\\paraline) command is appropriate to mark stanzas in poetry--more details in Part 12.

Data entry

You must mark new paragraphs with ($p). Typing (<CR><space><space>) for a new paragraph won't work: the result would be {$l}, not {$p}. For new braille lines, you can use hard <CR> and ($l) absolutely interchangeably.

Generally, you should mark all skipped lines with ($s), not <CR><CR (You will get better paper braille results, too!) You can use multiple <CR>s in title pages--more on this in Part 5. In consultation with your ever-present experienced transcriber, place "do-it-yourself" commands like (\\paraline) where appropriate; they are deleted by MAKEPAPER.

Part 2: Page Numbers and Print Page Indicators

Unlike paper braille, linear braille devices don't have output pages. A significant number of TranscriBEX \\ commands are concerned with braille page numbering and conditionally moving to a new braille page, so LBF$ automatically eliminates these commands.

The LBF print page indicator follows the general pattern of space, dot 5, dots 2-5, the print page number, dots 2-5, dot 2: {"3#a31}. TranscriBEX has three different print page indicator commands: (\\pph[number]) for print page indicators that appear immediately before a heading; (\\ppc[number]) for continuing print page indicators; and (\\pp[number]) for all others. ClasX-LBF eliminates (\\ppc[number]), since its only function is changing the continued print page indicator on output braille pages. The LBF$ transformation changes the other two commands to three characters: "3<control-T>. As you send your final, formatted chapters to the linear braille device, BEX adds the closing {31} characters to the print page indicator (and deletes the <control-T)

Continuing print page indicators on VB pages

The VB stores data on a cassette tape, and it internally divides the text into "pages" which cannot exceed 1,000 characters. The VB user can only search one VB page at a time. Responding to requests from transcribers and users, BEX creates a "continued print page indicator" at the beginning of each VB page. This "continuing print page indicator" is just like the Code of Braille Textbook Formats and Technique's requirement for paper braille. LBF's print page indicator uses the form {"3#a31}, where {#a} is any page number. When the first VB page has print page 60, then the second VB page for print page 60 begins with {"3a#fj31}. Since option T - ToVB controls how text is divided into VB pages, it creates the "continued print page indicator" automatically as it sends text to the VB.

Data Entry

The ClasX-LBF process takes care of all braille page numbering commands and print page indicators automatically. As long as you follow the TranscriBEX data entry guidelines, you're sitting in clover.

Part 3: Running Heads

Since linear braille devices don't have output pages, there's no place to put a running head. Most braille documents have a title page, and the [TEXT] in the running head summarizes the full title of the document. The ClasX-LBF process can delete the (\\runninghead[TEXT]<CR>) command and data from your \\ chapters.

However, "casual" documents may lack a title page. In this situation, the text of the runninghead on the first braille page serves as the title of the document. When this is true, deleting the [TEXT] would eliminate important information! Instead, you should change the (\\runninghead) to a major heading before you run the data through LBF$. When your original looks like:

\\(bs \\runninghead May Meeting Minutes <CR>Called to order at 4 P.M. .... \\)xs

Part 4: Headings

Paper and linear braille handle headings quite differently. The Code of Braille Textbook Formats and Techniques describes three levels of headings for paper braille transcriptions: major headings, minor headings, and paragraph headings. Which level to show is left to the transcriber. The Code does not require the transcription to show the exact hierarchy of headings. On the other hand, the Provisional Guidelines for Literary Linear Braille Format marks the most major heading with {$h}, the next level of heading with {$h#a}, the next level as {$h#b}, and so on. At the same time, LBF asks that the fewest possible number of heading levels be assigned.

For example, a textbook may have four heading levels. In the paper braille, the first and second level are both shown as major headings, while the third and fourth level are transcribed as minor headings. Some intervention is required to modify this data for a linear transcription showing four headings. ClasX-LBF provides you with some flexibility in your transcription of headings. The transformation chapter LBF$ does the work of changing \\ commands into {$} "format symbols." The variety of rules in LBF$ give you two basic approaches.

Approach 1: Only show two heading levels

The first approach requires dramatically less effort on your part, but it could mean that your linear transcription does not totally conform to LBF. LBF$ turns every (\\hd) into {$h}, and changes (\\mh) to {$h#a}. When your inkprint only has two levels of headings to begin with, then your linear transcription conforms perfectly. When your inkprint has four levels of headings, this approach means the braille reader does not get all the hierarchical information prescribed by LBF.

Approach 2: The (\\hd#) and (\\mh#) commands

The LBF$ transformation chapter respects a digit appended to TranscriBEX's (\\hd) and (\\mh) commands. LBF$ does not alter the digit in transformation: when you use the digit 1, then the final $2 chapters have the braille digit {#a}. When you enter (\\hd1) then you get {$h#a}. And when you enter (\\mh1) you also get {$h#a}.

In other words, when you use ClasX-LBF's (\\hd#) and (\\mh#) commands, it's the value of the number that determines where the heading comes in the hierarchy. The ClasX-LBF process ignores the ranking implicit in the paper braille distinction between "minor" and "major" headings: both (\\hd1) and (\\mh1) become {$h#a}.

Keep in mind that LBF says the highest level of heading is unmarked; that is, {$h} is the most important. To signal the highest level of heading, use (\\hd) without any number at all. {$h#a} is the first level of heading below "major" headings. Type this as (\\hd1) or (\\mh1). For the next level of heading, use (\\hd2) or (\\mh2), the next level is (\\hd3) or (\\mh3) and so forth.

The Code and LBF do agree on "paragraph headings," where the text of the heading is marked by italics signs at the start of the paragraph. After the heading text, prose continues without moving to a new paragraph. You use TranscriBEX's (\\ib) italics begin and (\\if) italics finish commands when transcribing paragraph headings for paper or linear braille.

Dual-purpose data entry

Since the LBF$ transformation chapter changes both (\\hd1) and (\\mh1) to the same {$h#a} format symbol, you have some leeway in your data entry. Mark your major headings with (\\hd#) and your minor headings with (\\mh#). Before you can make a paper braille version, you must use the MAKEPAPER transformation, which gets rid of all LBF-specific commands. MAKEPAPER deletes any digit after a (\\hd) or (\\mh) command.

For example, your document has four levels of headings. For paper braille, the first two levels are transcribed as "major headings," while the next two levels are "minor headings." In your dual-purpose chapters, use (\\hd), (\\hd1), (\\mh2), and (\\mh3). LBF$ changes these to the "format symbol" with the appropriate digit: {$h}, {$h#a}, {$h#b}, and {$h#c} respectively. MAKEPAPER changes these to plain (\\hd) and (\\mh) commands.

Modifying existing paper \\ chapters

When your existing paper braille chapters show more than two levels of headings, you must somehow add the hierarchical digits to the existing (\\hd) and (\\mh) commands. Section 3, Part 2 discusses the TRACER transformation chapter. TRACER places a unique character after the (\\hd) and (\\mh) commands, making it easy to find them all with a single locate command in the Editor.

However, you may be able to add the heading level numbers automatically with Replace characters. For many inkprint documents, the words present in a heading follow a regular pattern. You can find (\\mhSection) and change it to (\\mh2Section). An extensive sample of this approach appears in Section 3, Part 2.

Data entry for long headings

On TranscriBEX pages 6:9-10, we introduce the (\\hd) command for major headings. We advise you to manually divide long headings to make the most pleasing line breaks. Unfortunately, the commands we suggest cause problems in a linear braille transcription. Our sample shows:

\\(bs $s\\hd[Start of heading that's more]

<CR>\\hd[than 33 characters in braille] \\)xs

Because the second line of the heading also begins with (\\hd), you end up with two {$h} "format symbols." To prevent this, use the centering command (\\c) when manually dividing headings.

\\(bs $s\\hd[Start of heading that's more]

<CR>\\c[than 33 characters in braille] \\)xs

LBF$ simply deletes all (\\c) commands, so each heading begins with just one {$h}.

Part 5: Title Pages

Both English Braille and the Code of Braille Textbook Formats and Techniques categorize the text appearing on a braille title page into "defined areas of information." LBF wants each of these defined areas to begin with a {$p} "format symbol."

For example, a title page in textbook format consists of

• the title
• the subtitle
• the author(s)
• the publisher's permission
• the list of copyright dates
• the year of braille transcription
• the name of transcriber
• information about the contents of one volume.

These eight items are each one "defined area of information."

In the TranscriBEX Manual, we said to start each line of the title page with (<CR>\\c). An alternative command for centering title page lines, (<CR>\\ct), was described in the June 1987 RDC Newsletter. The LBF$ transformation chapter eliminates both of these constructs.

Dual-purpose data entry

When you type the title page, use ($p\\c) or ($p\\ct) to introduce each defined area of information. If you need to skip a line to center your title page information, you can use multiple <CR>s: ClasX-LBF is smart enough to change them to ($p) indicators. Inside each area of information, begin the lines with (<CR>\\c) or (<CR>\\ct). Since LBF$ strips out the centering command, after transformation, the "$" chapters have the ($p) in the right places.

Modifying existing paper \\ chapters

Edit the chapter with the title page, and check that ($p) or two <CR>s appear at the start of each defined are of information. Don't bother to delete the other (<CR>\\c) commands, as LBF$ does it for you.

Part 6: Indexes

For entries in an index (or glossary), LBF uses a system similar to headings. The main level entries begin with {$5}. (The screen braille 5 is dots 2-6, the Grade 2 en contraction.) The first level down is marked with {$5#a}, the second level gets {$5#b}, and so forth. The {$5#n} "format symbol" implies a new braille line or new paragraph.

The braille codes divide indexes into two types: "simple" and "complex." The entries in "simple" indexes are all at the same level, while the more common "complex" index has two or more levels of nesting. Modifying paper data entry for simple indexes requires more effort than modifying paper data entry for complex indexes.

A "complex" index begins each item with a hierarchical \\ command, while a "simple" index starts each item with one <CR Since LBF requires a "format symbol" at the start of each line, you must use the S-NDX transformation to improve your \\ chapters that contain a simple index.

When you create a "complex" index that follows all the rules in TranscriBEX Section 11, then the LBF$ transformation takes care of all the modification for you. You don't need to manually change a thing. However, in TranscriBEX Section 18, we tell you about a data entry shortcut for indexes. On pages 18:5-6, we say you don't have to repeat the hierarchical commands for second and subsequent entries at the same level. This shortcut works wonderfully for paper braille, but fails in linear braille.

Modifying existing paper \\ chapters

For a simple index, check to make sure that the (\\simpleindex) and (\\rt) commands bracket the data. Then use Replace characters with the S-NDX transformation chapter to improve your \\ chapters. S-NDX places the (\\ind) command at the start of each entry. LBF$ transforms (\\ind) to the appropriate "format symbol," and MAKE$ creates the same format from (\\ind) as from a <CR> in a simple index.

For a complex index, check to see if you've used the data entry shortcut. If you have, insert the missing (\\mi), (\\si), (\\ssi), etc. commands.

Dual-purpose data entry

For complex indexes, do not use the data entry shortcut. Type the relevant (\\mi), (\\si), (\\ssi), etc. command at the start of each and every index entry. For a simple index, you have two choices. You can begin each entry in a simple index with (\\ind). For example:

\\(bs \\simpleindex \\ind Apples 3

<CR>\\ind Artichokes 12

<CR>\\ind Bananas 7

<CR>\\ind Broccoli 2 \\rt ... \\)xs

If you don't like typing a lot, follow the existing guidelines in the TranscriBEX Manual. Then use the S-NDX transformation mentioned above to place the (\\ind) commands automatically; the MAKE$ or ALL$ transformation chapters treat (\\ind) just like <CR>

Part 7: Outlines and Other Hierarchies

LBF marks the level of every item in an outline. The most important level is unmarked; the first level down is labelled "1", the next level is "2", and so on. The LBF "format symbols" for outlines are {$\}, {$\#a}, {$\#b} and so forth. You may be wondering what the backslash is doing there. {\} is dots 1-2-5-6, the Grade 2 ou contraction. So this symbol is mnemonic; it's just not immediately evident in screen braille.

Just as with headings, the {$\#n} "format symbol" implies the start of a new braille line. You don't want {$p$\#a[entry]}, so LBF$ collapses this to {$\#a[entry]}.

TranscriBEX gives you two ways to enter outlines: "short outlines" use the index command family: (\\mi), (\\si), (\\ssi), and so forth up to (\\6si). "Long outlines," on the other hand, use the outline command family: (\\mo), (\\so), (\\sso) up to (\\6so). LBF treats "short" and "long" outlines identically; each item is preceded with the {$\#n} "format symbol" appropriate to its position in the outline. However, LBF treats outlines differently than indexes, which are marked with {$5#n} "format symbols."

Since ClasX-LBF and TranscriBEX need different kinds of information, we've had to dream up a new set of commands for "short outlines." Our store of mnemonics is getting low: we shortened the word outline by lopping off the first letter and came up with the (\\mu) family.

Dual Purpose Data Entry for Outlines

Begin a main level short outline entry with (<CR>\\mu), the first sub-level with (<CR>\\su), the second sub-level with (<CR>\\ssu), the third sub-level with (<CR>\\sssu), the fourth sub-level with (<CR>\\4su), the fifth sub-level with (<CR>\\5su), and the final and sixth with (<CR>\\6su). The MAKEPAPER transformation chapter changes the "u" family to TranscriBEX's (\\mi), (\\si), etc. index family of commands.

Modifying existing paper \\ chapters

Following the TranscriBEX instructions, your paper braille chapters containing short outlines use the index family of commands. The SHORT-OUT transformation chapter on the ClasX-LBF disk changes these index commands to their equivalent short outline commands.

Part 8: Tables of Contents

Paper braille uses indent and runover variation to show the nesting of headings in a table of contents. LBF uses explicit "format symbols": the same {$h#n} symbols used for headings. The main entry in a table of contents starts with {$h}; the first level down starts with {$h#a}, the second level down uses {$h#b}, and so on. LBF does not use guide dots to separate the table of contents entry from its page number: there's simply one space between the entry and the digits.

Tables of contents can be either "simple" or "complex." "Complex" tables of contents require a TranscriBEX \\ command at the start of each line, but TranscriBEX data entry for "simple" tables of contents requires just a <CR> at the start of each line. Since LBF wants the explicit {$h} "format symbol," you must use the S-TOC transformation chapter to improve your "L" chapters with simple tables of contents.

When you create a "complex" table of contents that follows all the rules in TranscriBEX Section 12, then the LBF$ transformation takes care of all the modification for you. You don't need to manually change a thing. However, in TranscriBEX Section 18, we tell you about a data entry shortcut for tables of contents. On pages 18:5-6, we say you don't have to repeat the hierarchical commands for second and subsequent entries at the same level. This shortcut works great for paper braille, but fails for linear braille.

Modifying existing paper \\ chapters

For simple tables of contents, check to make sure that the (\\contents) and (\\endcontents) commands bracket the data. Make sure that you have that (\\endcontents) at the end: (\\rt) is not an acceptable substitute! Then use Replace characters with the S-TOC transformation chapter to improve your \\ chapters.

For complex tables of contents, check to see if you've used the data entry shortcut. If you have, insert the missing (\\con), (\\mc), (\\sc), etc. commands. (To make insertion faster, copy the command to the Clipboard. Find and insert all the (\\sc) commands at one swoop, then copy (\\ssc) to the Clipboard and locate again for that command.)

Dual-purpose data entry

For complex tables of contents, do not use the data entry shortcut. Type the relevant (\\con) or (\\mc), (\\sc), etc. command at the start of each and every contents entry.

For simple tables of contents, begin each line with the (\\con) command. For example:

\\(bs \\contents \\hd CONTENTS \\left Chapter \\right Page

<CR>\\con 1 It All Begins in the Soil \\gd 1

<CR>\\con 2 Composting and Mulching \\gd 21

<CR>\\con 3 What to do About the Weather \\gd 41 ... \\)xs

If you don't feel like typing all the (\\con) commands, you can do your data entry as you always did, then use the S-TOC transformation described above.

Part 9: Notes

Like the Code and English Braille, LBF defines two types of note: the "transcriber's note," which adds text not present in inkprint; and the plain "note," used for existing notes. The LBF start of transcriber's note symbol is {$tn}; the end of the transcriber's note is marked with {$tn=}. All other notes begin with {$n} and end with {$n=}. For transcriber's notes, TranscriBEX uses (\\tn) and (\\endtn), so LBF$ simply changes these commands to {$tn} and {$tn=}.

Just placing the LBF "format symbol" may not be sufficient, however. Transcriber's notes often explain how the braille is representing a particular inkprint structure. Since the format strategies for linear and paper braille are so different, the text of the transcriber's note may need to be changed. You should manually check each transcriber's note to make sure it still applies.

The Code sets out a variety of paper braille note formats. LBF begins all notes with {$n} and ends them with {$n=}. LBF$ changes TranscriBEX (\\note) and (\\specialnote) commands to {$n}, and changes (\\endnote) and (\\endspecialnote) to {$n=}.

Part 10: Questions

The Provisional Guidelines for Literary Linear Braille Format don't mention the transcriptions of questions at all. But since TranscriBEX fully supports questions, it seemed a pity to throw all that information away when you modified a paper braille \\ chapter for linear braille output. All the commands relating to questions are RDC's "friendly extension" of the LBF rules. None of the "format symbols" introduced here are sanctioned by BANA. Please discuss the propriety of using these "friendly extensions" with a certified transcriber.

As with indexes and tables of contents, questions can have either a "simple" or "complex" structure. RDC's "friendly extensions" only address complex questions. The TranscriBEX commands for questions establish nesting indents and runover that show the nesting of questions and their answer choices.

We borrowed LBF's "format symbol" for outlines, {$\#n}, to show where each question fits in a hierarchy. In RDC's system, the main level of questions are flagged with {$\}. The first level of sub-question is flagged with {$\#a}, the second level down gets {$\#b}, and so on down to the fourth level of sub-question, marked by {$\#d}. The {\#n} "format symbol" implies a new braille line or new paragraph.

Data Entry

LBF$ handles all of TranscriBEX's \\ commands for complex questions automatically; no effort is required on your part.

Part 11: Tables

Columnar material is inherently two-dimensional, while a linear braille display is inherently one-dimensional. LBF provides comprehensive guidelines for presenting tables to make them readable on a single line of braille cells.

Three types of tables

The Code and TranscriBEX support three kinds of tables: paragraph, stair-step, and line-for-line. ClasX-LBF transforms all three types into the single format prescribed by LBF. Here's a quick review of the three TranscriBEX table types.

The only \\ commands for paragraph tables are (\\items) to start and (\\enditems) to finish. The column headings are shown within a transcriber's note. Inside the table, you use to illuminate the structure of the data. The general pattern of TranscriBEX data entry for paragraph tables is:

\\(bs \\tn The print column form has been changed to the following form in braille:$p[Row label]:[First Column Label]; ... [Last Column Label]. \\endtn\\items

$p[Row label]: [first entry in row]; ... [last entry in row].

...

$p[Row label]: [first entry in row]; ... [last entry in row].\\enditems$p \\)xs

In a stair-step table, each row entry appears on a separate braille line. The horizontal position of the entry moves right to follow its column position. Again, the column headings are shown within a transcriber's note. The TranscriBEX data entry looks like:

\\tn The print columns have been changed to the following format in braille:

• <CR>\\fc [First Column Label]
• <CR>\\nc [Column Label]
• <CR>\\nc [Last Column Label] \\endtn
• <CR>\\fc [First Column Entry Row 1]
• <CR>\\nc [Column Entry Row 1]
• <CR>\\nc [Last Column Entry Row 1]
• ...
• <CR>\\fc [First Column Entry Last Row]
• <CR>\\nc [Column Entry Last Row]
• <CR>\\nc [Last Column Entry Last Row] \rt$p \\)xs

Finally, a line-for-line table uses the most TranscriBEX commands. Horizontal rules in inkprint become (\\heavyline), (\\lightline), or (\\lineof*) commands. Each tabular line begins with one <CR Every column heading begins with a (\\hw#) command; every table entry begins with a (\\w#) or (\\w#r) command. The value of # depends on how wide the entry is. If you have created a line-for-line table, you know what's involved; we won't attempt to summarize its structure here.

Three paper table types collapse to one linear format

One reason the Code provides three types of tables is to maximize use of space on the braille page. Since a linear braille device never allows the braille reader to compare two lines vertically, LBF uses a different strategy altogether for tables.

Every table must begin with {$tb} and terminate with {$tb=}; you must manually type these "format symbols" in your "L" chapters. In addition to alerting the braille reader to the presence of a table, the ($tb) and ($tb=) characters are crucial to the ClasX-LBF process for tables. They serve as the "on string" and "off string" for the PRINT-TABLE and BRL-TABLE transformation chapters, explained in detail in Section 3, Part 4.

LBF requires that "every column be searchable and retrievable." To this end, LBF asks the transcriber to "key out" every column and row. When you type the ($tb) and ($tb=) characters in the right spots, the ClasX-LBF process automatically letters columns for you, and also repeats these letters for each row entry.

In addition to these column key letters, LBF wants all tables to use a regular pattern of punctuation. Each row begins with a paragraph symbol; the row label ends with a colon; each entry ends with a semicolon; and the final entry in each row ends with a period. Again, the ClasX-LBF process places this punctuation automatically as long as you type the ($tb) and ($tb=) in the right spots.

TranscriBEX Section 15 provides data entry guidelines for all three types of tables. They all require extensive modification to conform to LBF. Luckily, the PRINT-TABLE and BRL-TABLE transformation chapters do 99% of the work for you: full details in Section 3, Part 4.

Part 12: Poetry

LBF advises you to minimize use of the {$l} new line "format symbol." But one place it is appropriate is for showing poetic lines. For numbered poetic lines, LBF attaches the line number to the {$l} "format symbol"; the line of poetry numbered "10" would start as {$l#aj[poetic line]}

For paper braille, a skipped line and change in paragraph format signal the transition between prose and poetry; each stanza begins with a skipped line. LBF uses {$p$l} to mark the start of each poem and the start of each stanza. To signal the end of poetry and return to prose, LBF uses {$l=}. (Don't use {$l=} at the end of every line, just at the end of the poem.)

The LBF$ transformation chapter deletes all the TranscriBEX \\ commands that establish indent, runover, and a "line number zone." LBF$ changes the <CR>s to new-line ($l) indicators, and it also respects the [number] in (<CR>\\ln[number]), attaching it to the ($l).

TranscriBEX doesn't have an "end of poem" command: at the end of poetry, you use the generic (\\rt) "regular text" command to restore standard braille paragraphs. This means you must place the end-of-poetry "format symbol" manually.

Modifying existing paper \\ chapters

Poems must begin with (\\poem$s), not ($s\\poem). Type {$l=} before the (\\rt) command that signals the return to prose. Use the POEM transformation chapter to change all the ($s) skip-line indicators marking stanzas to ($p\\paraline).

Dual-purpose Data Entry

Type (\\paraline) at the start of each stanza; MAKEPAPER strips this out. Type ($l=\\rt) to end every poem.

Part 13: Captions

LBF uses two "format symbols" to enclose captions. {$cp} marks the start of a caption, while {$cp=} finishes it off. TranscriBEX has no command that's specifically designed for captions. You generally enclose captions within (\\note) and (\\endnote) commands, but you can use other commands as well.

For dual-purpose data entry, enter the ($cp) and ($cp=) "format symbols" whenever you transcribe a caption. (The MAKEPAPER transformation chapter throws these away.) When you're improving paper braille \\ chapters, you must add the ($cp) and ($cp=) "format symbols" manually.

Part 14: Boxes

LBF discourages your transcribing boxed braille material for linear braille devices, but it does provide you with "format symbols," just in case. Boxed material begins with {$b} and ends up with {$b=}.

TranscriBEX discourages you from boxes altogether; on page 1:6 we flatly state that boxes are not supported. However, you could find your own way to create boxes with TranscriBEX. At any rate, this is one case where the LBF format is a snap. Simply type ($b) at the start of the box, and ($b=) at its end.

If you've developed systematic commands for transcribing boxes, then you could change those commands to the ($b) and ($b=) symbols automatically. Write the appropriate transformation rules and add them to TRACER--see Section 3, Part 2 for details.

Part 15: Computer Braille Code

Transcribing with the new Computer Braille Code ("CBC") is not discussed in the Provisional Guidelines for Literary Linear Braille Format. Yet the other half of your ClasX package gives you all the tools to create CBC transcriptions. Since linear braille devices are generally computer terminals, their owners are the braille readers most likely to use computers.

As a "friendly extension" to the LBF system, RDC supplies you with LBF-like "format symbols" to use for CBC materials. Keep in mind that these "format symbols" are not approved by BANA. Please discuss the propriety of using these "friendly extensions" with a certified transcriber.

In this Part we explain our LBF-like symbols, but we do not provide an explanation of the ClasX-CBC commands. We assume you've read the ClasX-CBC Manual, and you understand what CBC is and how ClasX-CBC creates it.

For some computer notation, CBC uses a skipped line to signal the change in braille codes. Our system shows the transition between displayed CBC and literary text with {$l$l}. The CBC Indicators that signal the switch between embedded Grade 2 and CBC don't require any special handling.

Likewise, the Countable Spaces Indicator is comprehensible in linear braille without intervention. CBC uses a unique indent and runover for computer program listings. The ClasX-CBC commands follow the pattern of (\\#c); the lines that begin at the inkprint left margin use (\\0c), the first level of indent is (\\1c), and so forth up to (\\7c). This type of hierarchy is similar to information in outlines, so we use LBF's {$\#n} symbol. The main level is unnumbered: (\\0c) becomes {$\}; (\\1c) becomes {$\#a}; (\\2c) becomes {$\#b}; and so on up to (\\7c), which is turned into {$\#g}.

When your \\ chapters use ClasX-CBC commands, run Replace characters with the LBF$C transformation chapter instead of LBF$. If you don't, you'll get <beep> ***error*** messages. As mentioned in ClasX-CBC Section 2, Part 6, MAKE$C does not delete multiple spaces. LBF$ does delete multiple spaces, and the majority of transcriptions won't include CBC material. Also, LBF$C contains many more transformation rules than plain LBF$ so it's slower. The moral is, only use LBF$C when your chapters contain CBC material.

Data Entry

In ClasX-CBC Section 2, Part 7, we talk about a data entry shortcut for program listings. We say there that the first program line at a given level requires a (\\#c) command, but subsequent lines at the same level may begin with just <CR As is the case with indexes and tables of contents, this data entry shortcut won't work for linear braille. You must insert the "missing" (\\#c) commands. LBF$C handles all other CBC issues automatically.

Part 16: BEX $$ Commands

Nowhere in the TranscriBEX Manual do we tell you to enter BEX's native $$ commands in your original \\ chapters. The whole point of the TranscriBEX system is the mnemonic \\ commands. However, we're not software police, and you may have found a good reason to use BEX's format commands to get your transcribing done. Anything that works is great!

But when you do use $$ commands in your \\ chapters, you must delete them before you send your \\ chapters through the ClasX-LBF process. When you send your final $2 chapters to the linear braille device, it gets almost every character in the BEX chapter. (See Section 3, Part 10 for why it's not completely character-for-character.) You don't want your lovely linear braille transcription spoiled by random dollar signs!

The CLEANUP transformation deletes theand format commands that make the translator do braille italics. It also deletes the format commands that the translator places for CBC material. We considered trying to delete "stray" $$ commands with the CLEANUP transformation, but rejected the idea.

Since only you understand the function the $$ commands are serving, you need to take care of these by hand. If you use BEX's tab commands to format a table, just deleting the commands does not create a correctly-formatted LBF table. (We've provided some shortcuts for just this situation--see Section 3, Part 4 for details.)

Part 17: Checklist for Dual-Purpose Data Entry

To minimize the effort of creating linear and paper braille transcriptions from a single set of data, keep in mind the unique requirements of each medium. Your dual-purpose chapters will contain more commands than needed for either medium alone. ClasX-LBF knows how to eliminate those that are irrelevant to linear braille. The MAKEPAPER transformation chapter deletes commands that only pertain to linear braille. Follow these guidelines to create dual-purpose chapters:

• Always enter skipped lines in text as ($s), not <CR><CR
• Enter LBF-specific commands while typing poetry. Begin your poem with (\\poem$s\\paraline). At each stanza, type ($s\\paraline). Finish the poem with ($l=\\rt). This obviates the need for the POEM transformation.
• Enter LBF "format symbols" while typing captions. Begin the caption with ($cp), and finish it with ($cp=).
• Enter LBF "format symbols" while typing tables. Begin the table with ($tb), and finish it with ($tb=).
• Enter LBF "format symbols" while creating boxed material. Begin the box with ($b), and finish it with ($b=).
• Enter heading level numbers for documents requiring them. When the linear transcription has more than two levels of headings, use (\\hd) for the most important level, then (\\hd1) or (\\mh1) for the next level, (\\hd2) or (\\mh2) for the next level, and so forth. When breaking long headings into two or more lines, begin second and subsequent lines with (<CR>\\c), not (<CR>\\hd).
• Begin each defined area of information in a title page with ($p) or <CR><CR>, not a single <CR
• Label every entry in simple indexes and tables of contents. Place (<CR>\\ind) at the start of each index entry, and (<CR>\\con) at the start of each contents entry. This obviates the need for the S-NDX and S-TOC transformations.
• Use the "short outline" family of commands. Mark the main level with (<CR>\\mu), the first sub-level with (<CR>\\su), the second sub-level with (<CR>\\ssu), the third sub-level with (<CR>\\sssu), the fourth sub-level with (<CR>\\4su), the fifth sub-level with (<CR>\\5su), and the final and sixth sub-level with (<CR>\\6su). This obviates the need for the SHORT-OUT transformation.
• Label every entry in hierarchical formats. For tables of contents, indexes, short outlines, and indented computer program listings, enter the appropriate TranscriBEX command at the start of each and every line.

Use a naming system that reminds you that you've created dual purpose chapters: we recommend ending your chapter names with D for dual purpose. Before beginning a braille braille transcription of this data, use Replace Chapters with the MAKEPAPER transformation on you ClasX-LBF disk. As you create these modified paper-braille chapters, use the #D naming method, which deletes the last character.

Section 3: Processing Data with ClasX-LBF

The entire ClasX-LBF process is similar to the TranscriBEX process you already know. Here's a conceptual overview:

• Stage 1: Creating "L" chapters. In addition to inkprint or braille data entry, you run Replace characters several times. The transformations required depend on what kind of formats appear in the "\\ command" chapters that form the basis for your linear transcriptions. In the worst case scenario, you need to transform your data seven times before this stage is finished. In the best case scenario, only two transformations are needed.
• Stage 2: Creating the $2 chapters. Use Replace characters and the LBF$ transformation to change the \\ commands to LBF "format symbols," creating the "$" chapters. The LBF$ transformation also deletes a significant number of \\ commands that are irrelevant to linear braille transcriptions. For inkprint data entry, use Grade 2 translator to change the "$" chapters to $2 chapters. Optionally, use Replace characters with the FINETUNE transformation chapter to fix rare translator errors. For both inkprint and braille data entry, use Replace characters with the CLEANUP transformation chapter to remove $$ commands the translator uses, as well as the rest of the \\ commands. Finally, proofread these $2 chapters to check for formatting problems, using either a Preview Paperless Brailler or any screen mode.
• Stage 3: Sending data to linear braille device. For tape-based VersaBraille, use option T - ToVB on the Main Menu. For any other linear braille device (including the disk-based VersaBraille II), use option P - Print on the Main Menu, specifying the printer number you've configured as a class P - Paperless Brailler.

The bulk of this Section is devoted to Stage 1. This Part 1 has a master chart of all the Stage 1 steps. Parts 2 through 6 explore each step in detail. The procedures for Stage 2 begin with Part 7, while Stage 3 is covered in Parts 10 and 11.

Part 1: Overview of Stage 1

As we outlined in Section 1, Part 2, data you move through the ClasX-LBF process could be created with two different goals in mind. You may start out with existing \\ command chapters. When you did data entry, you did not think about linear braille. We'll call this type of data "paper \\ chapters." Some of the seven transformations cited above add commands that ClasX-LBF needs to make correct linear transcriptions.

Now that you have ClasX-LBF, you can also create chapters that contain all the information LBF requires, as well as all the TranscriBEX commands. These are "dual-purpose \\ chapters." To make paper braille from dual-purpose \\ chapters, you must first use the MAKEPAPER transformation chapter; this strips out the commands that are only relevant to ClasX-LBF. The significant advantage of dual-purpose data entry is that Stage 1 then requires fewer transformations.

You must begin with clean \\ chapters

Whether you do dual-purpose data entry or use paper \\ chapters, it's crucial that you have faithfully corrected these original chapters. Once you're familiar with TranscriBEX, it's tempting to only fix minor problems in the final formatted, translated chapters. But in order for ClasX-LBF to work correctly, the starting point chapters must contain all the appropriate \\ commands.

For example, suppose you neglected to enter the (\\rt) command at the end of a simple index. You've edited enough $2 chapters to know that MAKE$ changes (\\rt) to ( $$ml0 $$ms0 $$mr0 $$i2), so you simply type those commands in your final chapters. For a simple index, ClasX-LBF needs that (\\rt); without it, your data would get scrambled.

Stage 1 Summary

The charts on the next two pages summarize all the possible steps in Stage 1. An abbreviated version also appears on the ClasX-LBF Reference Card. \\7coltable

• 1 Which chapters
• 2 BEX Option
• 3 Source Chapter
• 4 Naming Method
• 5 Target Chapter
• 6 Transformation Chapter
• 7 Comments
• 1 All
• 2 Replace characters
• 3 \\
• 4 #AL
• 5 L
• 6 TRACER
• 7 Creates the first version of the "L" chapters. Copy up to RAM drive if possible.
• 1 All
• 2 Editor
• 3 L
• 4 --
• 5 --
• 6 --
• 7 Locate for <Del> character flag placed by TRACER; insert some LBF "format symbols" manually; compare data with checklist (page LS3:9); note which chapters have tables, short outlines, simple contents, simple indexes, and poems. Use control-V to proofread each page in the Editor, checking for data entry errors. See Part 2 for details.
• 1 All
• 2 Replace characters
• 3 L
• 4 #AB
• 5 LB
• 6 CONFIRM
• 7 Creates backup chapters, copy down from RAM to floppy. If any replacements occur, you need to find the <Del>s you neglected to handle, then repeat this step.
• 1 Only those containing paragraph, stair-step, or line-for-line tables
• 2 Replace characters
• 3 L
• 4 0S
• 5 L
• 6 PRINT-TABLE or BRL-TABLE
• 7 Between ($tb) and ($tb=), places column keys and punctuation. See Part 4 for full details.
• 1 Only those containing short outlines
• 2 Replace characters
• 3 L
• 4 0S
• 5 L
• 6 SHORT-OUT
• 7 Changes index-family commands to short-outline commands. Unneeded when you use (\\mu), (\\su), (\\ssu) commands in dual-purpose data entry. See Part 5 for details.
• 1 Only those containing simple tables of contents
• 2 Replace characters
• 3 L
• 4 0S
• 5 L
• 6 S-TOC
• 7 Between (\\contents) and (\\endcontents), inserts (\\con) after each <CR Unneeded when you use (\\con) in dual-purpose data entry. See Part 5 for details.
• 1 Only those containing simple indexes
• 2 Replace characters
• 3 L
• 4 0S
• 5 L
• 6 S-NDX
• 7 Between (\\simpleindex) and (\\rt), inserts (\\ind) after each <CR Unneeded when you use (\\ind) in dual-purpose data entry. See Part 5 for details.
• 1 Only those containing poems
• 2 Replace characters
• 3 L
• 4 0S
• 5 L
• 6 POEM
• 7 Between (\\poem) and ($l=), inserts (\\paraline) after each ($s). Unneeded when you use (\\paraline) in dual-purpose data entry. See Part 5 for details.
• 1 All
• 2 Control-V in Editor or Print to screen
• 3 L
• 4 --
• 5 --
• 6 --
• 7 Confirm all is well; fix errors in these chapters and proof until satisfied. See Part 6 for details.
• 1 All
• 2 Copy chapters
• 3 L
• 4 #AB
• 5 LB
• 6 --
• 7 Updating backup chapters, copying down from RAM to floppy. \\end7coltable

As you can see from this chart, only two transformations are mandatory: TRACER and CONFIRM. The other transformations are only needed when you have tables, short outlines, simple tables of contents, simple indexes, or poems in your \\ chapters. Now that you see all the transformations that you might require, you see why ClasX-LBF cries out for RAM drives.

When you have more than 128K memory in your Apple, then you only wait for floppy disk access twice during all of Stage 1. When you Replace with TRACER you "copy up"; that is, your source dual-purpose or paper \\ chapters are on floppy disk and your target chapters are on RAM drive. When you use CONFIRM you "copy down"; your source "L" chapters are on RAM drive and your target chapters are on floppy disks.

Save Time with Smaller Chapters

Five of the Stage 1 transformations are very specific: you only have to modify a certain kind of data. When you transcribe an entire braille volume in one BEX chapter, then you end up reading and writing a lot of irrelevant data when fixing up simple tables of contents, indexes, short outlines, poems, and tables.

A better approach is to limit each BEX chapter to around 20 braille pages. Many users move to a new BEX page for each new inkprint page, which makes it easy to find your way around the data. Section 5, Part 5 demonstrates breaking up a massive BEX chapter into more comfortable working units.

Part 2: Effects of TRACER

Stage 1 always begins with Replace characters and the TRACER transformation chapter. TRACER accomplishes two goals: it forces you to make a copy of your original paper or dual-purpose \\ chapters, and it places a unique marker on most of the \\ commands you have to change manually. You must work with copies of your original \\ chapters, because at the end of Stage 1, the "L" chapters don't contain all the \\ commands needed for paper braille. We suggest you use the AL naming method to add the letter L to the end of your source chapter names. The initial number sign in the chart is a shorthand reminder that you're not just copying to and from your default disk drive. Instead, you're copying from one disk to a different disk (hopefully, from floppy to RAM drive). Replace the # with the disk drive number appropriate to your extended disk system.

Throughout Section 2, we told you about specific commands you had to fix manually. TRACER functions like the radioactive milkshake used for gastric x-rays. (Disgusting, but true.) TRACER inserts one <Del> character before the majority of commands that need your attention. After TRACER is finished, you review the data in the Editor. On the first BEX page, enter control-L <Del> control-A to locate for the <Del> character. After the first time, you can just enter control-L control-A. (BEX "remembers" the locate string definition until you define a new one.)

When you have devised your own TranscriBEX commands, or you use BEX's $$ commands to format your chapters, you should add rules to TRACER that flag these commands for your attention. Section 2, Part 4 is devoted to the issue of headings. When you are comfortable with Replace characters, you can add rules to the TRACER transformation to flag all your headings for you. After we describe the effects of the unmodified TRACER supplied on your ClasX-LBF disk, we address adding heading rules in detail.

Once you have dealt with the command flagged by <Del>, you can simply press the spacebar to overwrite the <Del> character with a space. That's a lot faster than using BEX's control-D <space> command to erase one character, since you don't have to wait for the screen to refresh. The LBF$ and LBF$C transformations get rid of a space before a \\ command, so you don't have to worry about extra spaces leaking in to your data.

Interpreting TRACER's Findings

The following list enumerates all the places TRACER puts a <Del> character, in the natural front-to-back order of a braille book. Each item describes the decisions you need to make or the characters you need to add or delete. Unless otherwise specified, simply <space> over the <Del> character after you've handled each flagged command.

• Flagged Item -- What To Do
• (\\bookprelim) & (\\textbookprelim) -- If it's a title page, make sure that each "defined areas of information" begins with one ($p) or two <CR>s.
• (\\contents) -- If it's a simple contents and you don't have (\\con) at the start of each entry, note chapter name for replacement with S-TOC. Make sure (\\endcontents) appears at end.
• (\\runninghead) -- If "casual" document without title page, replace the (\\runninghead) command with appropriate heading command. All other cases, space over the <Del> character and let LBF$ and CLEANUP delete the command for you.
• (\\hd) and (\\mh) -- Change the <Del> character to a digit if you're using the (\\hd#) and (\\mh#) approach to headings.
• (<CR>\\hd) -- Possibly the second line of a long heading you've manually divided. Change the (\\hd) to (\\c).
• (\\tn) -- Beginning of transcriber's note. Consider whether note needs rewording to reflect different format strategies of linear braille.
• (\\poem) -- If it's ($s\\poem), then transpose to (\\poem$s). Stanzas must be marked with ($s), not ($p) or <CR><CR Advance to the end of the poem and insert the "end of poetry" ($l=) format symbol, then note chapter name for replacing with POEM.
• (\\table) & (\\endtable) -- Beginning and ending of a line-for-line table. Type the start-of-table ($tb) "format symbol" before the (\\table) command. Type the end-of-table ($tb=) "format symbol" after the (\\endtable) command. Note the chapter name for replacing with PRINT-TABLE or BRL-TABLE.
• ( \\endtn \\items ) -- Possibly the transition between column headings and table rows in a paragraph-style table. Enter control-Z control-P to zoom cursor back by paragraphs to the start of the table. Type ($tb) then advance cursor to the (\\enditems) command and type ($tb=). Note the chapter name for replacing with PRINT-TABLE or BRL-TABLE.
• (\\endtn<CR>\\fc) -- The transition between column headings and table rows in a stair-step table. Zoom cursor back to the start of the table and type ($tb); advance cursor to the (\\rt) command at the table's end and type ($tb=). Note the chapter name for replacing with PRINT-TABLE or BRL-TABLE.
• (\\simpleindex) -- If you don't have (\\ind) at the start of each entry, note chapter name for replacing with S-NDX. Check to make sure that you have (\\rt) at the end.
• (\\i#r#) -- You have established some indent and runover format not supplied with TranscriBEX. Decide how to transcribe this format following the LBF guidelines, and enter appropriate LBF "format symbols," if required.
• Any $$ commands (except underlining) -- Delete the command with control-D control-W. You have used BEX format commands to create some format not handled by TranscriBEX. Decide how to transcribe this format following the LBF guidelines, and enter appropriate LBF "format symbols," if required. If you're using BEX tab commands for a table, see Part 4 for a shortcut to manually fixing this.

Manual Checklist during TRACER Review

Unfortunately, the TRACER transformation can't mark absolutely everything you need to change. For example, Section 2, Part 13 discusses the LBF "format symbols" {$cp} and {$cp=} that enclose captions. Since you can use a variety of TranscriBEX commands to transcribe a caption in paper braille, TRACER can't reliably find them all. Similarly, although we do make a "best guess" for tables, TRACER can't automatically find all of them. As you manually review each BEX page in the Editor, answer four questions:

• Are there paragraph, stair-step, or line-for-line tables? If yes, do three things: 1. Type the start-of-table ($tb) "format symbol" immediately before the heading that begins the table. 2. Type the end-of-table ($tb=) "format symbol" after the final table entry. 3. Note the chapter name for further handling with PRINT-TABLE or BRL-TABLE.
• Are there any captions? If yes, enclose within the start-of-caption ($cp) and end-of-caption ($cp=) "format symbols." Manually delete the TranscriBEX commands that formatted the caption.
• Is there any boxed material? If yes, enclose within the start-of-box ($b) and end-of-box ($b=) "format symbols." Manually delete the TranscriBEX or $$ commands that format the box.
• Are there "short outlines" formatted with TranscriBEX index-family commands? If yes, check whether real index material also appears within the same chapter. If so, you must split off the material into two BEX chapters--a sample appears in Section 5, Part 5.

Adding Heading Rules to TRACER

In Section 2, Part 4, we discussed two ways to handle headings for linear braille. One approach adds a number after TranscriBEX's (\\hd) and (\\mh) commands. We mention in passing that it's often possible to place all the digits automatically with Replace characters. The TRACER transformation is the perfect opportunity for these changes, since you always check your data manually in the Editor after you use TRACER. As you review the transformed data, you can confirm that the headings are as they should be, and manually fix any place where your rules didn't work.

TRACER's existing heading rules

As supplied on the ClasX-LBF disk, the last three rules in TRACER flag every heading in your document with the <Del> character. After TRACER is finished, (\\hd) becomes (\\hd<Del>) and (\\mh) becomes (\\mh<Del>). As you're locating for the <Del> character, you overwrite the <Del> with the appropriate number. The drawback to this approach is that it requires many decisions on your part. It's a lot faster to add the heading level numbers automatically, which we explore in a few paragraphs.

When your document only uses two or three levels of headings, then you don't need to add a digit to the (\\hd) and (\\mh) commands. In this case, the <Del> characters for headings that TRACER places are irrelevant, and you can easily prevent those rules from being executed. Here's how:

• 1. Use Copy chapters to make a MY TRACER chapter to modify.
• 2. Edit MY TRACER and enter control-A <space> to place the cursor at the end of the BEX page.
• 3. Enter control-L \\hd control-Z. Your cursor lands on the find string of the first heading-related rule.
• 4. Enter control-I to insert, then type ||| to end the list of rules, then enter control-N to finish the insert. The heading rules are still there, but you've ended the list of rules prematurely. If you decide you want to use the rules after all, you can just delete those three "early" vertical bars.
• 5. Quit the Editor with control-Q.

When you begin Stage 1, use this modified MY TRACER chapter instead of the supplied TRACER transformation chapter.

Adding heading numbers automatically

For many inkprint documents, the words in headings follow a regular pattern. You can find (\\mhChapter) and change it to (\\mh2Chapter). A good place to look for these patterns is the inkprint Table of Contents. The only difficulty would be if you misspelled one of the key words you're finding and changing. The use of a transformation chapter like this depends on the predictability of the text and the accuracy of the data entry.

Here's a truly convoluted example. The major inkprint division are called "chapters." So the most important heading always starts with the word "Chapter" followed by a digit. Then each chapter is divided into four parts: the first is always labeled "Overview," the next heading depends on the topic in that chapter; the third part is always "Questions" and the last is "For Further Study." In the paper version, these are marked with (\\hd).

The next level of heading begins with a number. Under "Questions," for example, there is "1.2: About this Chapter"; "1.3: About the Supplemental Reading"; and "1.4: About the Video Tape." These are also transcribed as "major headings" in the paper version, so they also begin with (\\hd).

Your linear transcription shows three levels of headings. It's worthwhile to consult BEX Master Level Section 6 in this situation, since contextual Replace lets you define a find string more generally than basic Replace. As you'll see in the following sample, a single rule can find (\\hd) followed by any two-digit number.