preconv is now on encoding-recognition parity with groff. This last

commit adds parsing of "File Variables" in the first two lines in order
to grok the encoding.  This completes groff's recognition sequence (-e,
BOM, File variables, -D, default).  I've also cleaned up the manual to
indicate this and for some general readability.

preconv is now compiled by default in the Makefile.

1 files changed, 46 insertions, 27 deletions

diff --git a/preconv.1 b/preconv.1
index 266cbc7c..12ea6f80 100644
--- a/preconv.1
+++ b/preconv.1
@@ -1,4 +1,4 @@
-.\"	$Id: preconv.1,v 1.2 2011/05/26 12:14:46 kristaps Exp $
+.\"	$Id: preconv.1,v 1.3 2011/05/26 14:43:07 kristaps Exp $
 .\"
 .\" Copyright (c) 2011 Kristaps Dzonsons <kristaps@bsd.lv>
 .\"
@@ -42,18 +42,8 @@ Its arguments are as follows:
 .Bl -tag -width Ds
 .It Fl D Ar enc
 The default encoding.
-This is case-insensitive.
-See
-.Sx Algorithm
-and
-.Sx Encodings .
 .It Fl e Ar enc
 The document's encoding.
-This is case-insensitive.
-See
-.Sx Algorithm
-and
-.Sx Encodings .
 .It Ar file
 The input file.
 .El
@@ -63,27 +53,23 @@ If
 is not provided,
 .Nm
 accepts standard input.
-Output is written to standard output.
-Unicode characters in the ASCII range are printed as regular ASCII
-characters; those above this range are printed using the
+See
+.Sx Algorithm
+for encoding choice.
+.Pp
+The recoded input is written to standard output: Unicode characters in
+the ASCII range are printed as regular ASCII characters, while those
+above this range are printed using the
 .Sq \e[uNNNN]
 format documented in
 .Xr mandoc_char 7 .
 .Pp
 If input bytes are improperly formed in the current encoding, they're
 passed unmodified to standard output.
-.Ss Encodings
-The
+For some encodings, such as UTF-8, unrecoverable input sequences will
+cause
 .Nm
-utility accepts the
-.Ar utf\-8 ,
-.Ar us\-ascii ,
-and
-.Ar latin\-1
-encodings as arguments to
-.Fl D Ar enc
-or
-.Fl e Ar enc .
+to stop processing and exit.
 .Ss Algorithm
 An encoding is chosen according to the following steps:
 .Bl -enum
@@ -91,13 +77,41 @@ An encoding is chosen according to the following steps:
 From the argument passed to
 .Fl e Ar enc .
 .It
-If a BOM exists, utf\-8 encoding is selected.
+If a BOM exists, UTF\-8 encoding is selected.
+.It
+From the coding tags parsed from
+.Qq File Variables
+on the first two lines of input.
+A file variable is an input line of the form
+.Pp
+.Dl \%.\e\(dq -*- key: val [; key: val ]* -*-
+.Pp
+where
+.Cm key
+is
+.Qq coding
+and
+.Cm val
+is the name of the encoding.
+A typical usage may be
+.Pp
+.Dl \%.\e\(dq -*- mode: troff; coding: utf-8 -*-
 .It
 From the argument passed to
 .Fl D Ar enc .
 .It
 If all else fails, Latin\-1 is used.
 .El
+.Pp
+The
+.Nm
+utility recognises the UTF\-8, us\-ascii, and latin\-1 encodings as
+passed to the
+.Fl e
+and
+.Fl D
+arguments, or as coding tags.
+Encodings are matched case-insensitively.
 .\" .Sh IMPLEMENTATION NOTES
 .\" Not used in OpenBSD.
 .\" .Sh RETURN VALUES
@@ -107,7 +121,12 @@ If all else fails, Latin\-1 is used.
 .\" .Sh FILES
 .Sh EXIT STATUS
 .Ex -std
-.\" .Sh EXAMPLES
+.Sh EXAMPLES
+Explicitly page a UTF\-8 manual
+.Pa foo.1
+in the current locale:
+.Pp
+.Dl $ preconv \-e utf\-8 foo.1 | mandoc -Tlocale | less
 .\" .Sh DIAGNOSTICS
 .\" For sections 1, 4, 6, 7, & 8 only.
 .\" .Sh ERRORS