jbofihe.1

.TH "jbofihe" 1L "January 2001"
.SH NAME
jbofihe \- Lojban parser / grammar checker / word glosser
.SH SYNOPSIS
.B jbofihe
[-v] [ -h | --help | -? ] [ -k ] [ -t ] [-tf ] [ -l ] [ -b ] [ -x ] [ -H ] [ -m ] [ -w width ] [-ie] [-re] [-se] [-sev] [-cr] [-bt] [filename]
.SH DESCRIPTION
.I jbofihe
is a program that reads a piece of Lojban text and checks it for
grammatical correctness.  If it passes, the output will show the
nesting of the major grammatical constructs by means of bracketing.
Also, words will have their approximate English translations shown.
Some of these translations are context-dependent.
.SH OPTIONS
.TP
.B -v
Show the program version and exit.
.TP
.B -h
Show help page
.TP
.B --help
Show help page
.TP
.B -?
Show help page
.TP
.B -k
Display lists of tokens after parse and after pseudo-token insertion
(mainly useful for debugging the lexer and parser).
.TP
.B -t
Produce a syntax tree on stdout.  Nodes with only one child will be
filtered out.
.TP
.B -tf
Produce a full syntax tree on stdout, including nodes with only one
child.
.TP
.B -l
Produce output for typesetting with LaTeX.
.TP
.B -x
Produce plain text output.
.TP
.B  -w width
Set the width of the screen (in characters) for the -x option.
.TP
.B -H
Produce HTML output
.TP
.B -b
Produce output in 'block' mode, where each block shows the Lojban, the
English gloss and a term tag where applicable. The default is
'running' mode, with term tags, Lojban and the English gloss all on
the same line.  Only applicable to -l and -x (not to -h).
.TP
.B -m
Show the amount of memory used by the processing.
.TP
.B -ie
Display elided separators and terminators in the output.  They are shown in
upper case, to distinguish them from those originally present in the input
text.
.TP
.B -re
Require the input text to have all elidable cmavo actually present.  (May be
useful for forcing the user to think about the grammatical structure of the
input text more carefully when debugging parse errors.)
.TP
.B -se
Show any elidable cmavo that were present unnecessarily in the input text.
Where a sequence of elidable cmavo occur, but not all could have been elided, a
list of possible 'minimal' sequences is shown.  Each 'minimal' sequence is such
that : removing any words from it would cause the meaning of the text to change
(or make it ungrammatical); yet preserving any more of the original words is
unnecessary to preserve the meaning and grammatical correctness of the text.
Note, -se and -re are mutually incompatible.  If both switches are specified,
-re is assumed.
.TP
.B -sev
For each sequence of elidable cmavo, show where later ones could be valid
before earlier ones, and where the non-elidable word at the end of the sequence
could have been valid earlier in the sequence.  [This information is used to
determine the output of -se, so this is a verbose mode for -se].
.TP
.B -cr
Allow lujvo to contain the cultural rafsi forms CCVVCy (CCVVCV at end of word)
described in section 4.16 of the Reference Grammar.  This affects recognition
of valid lujvo and fu'ivla forms (the latter since it affects the recognition
of slinku'i failures).
.TP
.B -bt
Show backtrace through the pending parser states when a syntax error occurs.
This may give some insight into what the parser was expecting to see next, to
help in fixing the error.
.IP
The backtrace comes in three parts.  The first part is a list of rules which
are automatically reduced when the erroneous word is encountered.  In the
extreme case, when everything prior to the erroneous word constitutes a
complete text in its own right, the reduction sequence will go right the way to
the rule which matches a complete text.  Generally though, this will not be the
case.
.IP
The second part of the report shows the conditions where the grammar requires
the next word to be of a type other than that which actually appeared.  All the
active rules are shown (with periods showing the active points), followed by a
list of word classes (i.e. selma'o or BRIVLA/CMENE etc) that would have been
valid at this point.
.IP
The third part of the report shows the pending states on the parser stack.
This may give some information on whereabouts at a wider level the problem
situation has occurred.
.TP
.B filename
The name of the file containing text to be parsed.  If omitted, the text is
read from standard input.
.SH EXIT STATUS
.TP
.B 0
The text was scanned and parsed successfully - the text was grammatically valid
Lojban.
.TP
.B "1/2"
The text had grammatical errors that were detected by the main parsing
algorithm.
.TP
.B 3
The text contained words whose structure was not valid Lojban.  No grammatical
checking of the text was performed; the grammatical function of the erroneous
word(s) could not be determined.
.SH SEE ALSO
.PP
.TP
.B cmafihe
A program which does English word glossing without the grammar
parsing and checking stage.  This is useful to get words quickly
looked up when studying someone else's text which contains many
grammatical errors.
.TP
.B smujajgau
A program for compiling binary dictionaries for use by jbofihe
.SH FILES
.TP
.B /usr/local/lib/jbofihe/smujmaji.dat
This file contains all the dictionary data compiled (using
.I smujajgau
) into a binary format to facilitate lookup and search.
.TP
.B /usr/local/bin/jbofihe
The executable.
.SH ENVIRONMENT
.TP
.B JBOFIHE_DICTIONARY
This should point to the name of the file containing the dictionary.
The default is /usr/local/lib/jbofihe/smujmaji.dat
\".SH CAVEATS
.SH BUGS
ju'oru'e so'imei (Surely there are many).
.SH REFERENCES
.TP
.B http://www.rc0.org.uk/jbofihe
Home page for the jbofihe project
.TP
.B http://www.lojban.org/
Home page of the Lojban community
.SH AUTHOR
Richard Curnow
.B <rc@rc0.org.uk>