mhfixmsg.man: Replace UTF-8 `§' with `\(sc' troff escape.

man page source is ASCII.  GNU groff's troff's source is ISO 8859-1.
The `§' in UTF-8 looks like `§' in ISO 8859-1.  If those two bytes make
it to a UTF-8 terminal then it appears as `§', but target something
else, e.g. PDF, and the flaw shows.

Also adjust test-manpages to grep for bytes other than HT, and SPACE to
tilde, in the built man pages.  A test of the source could be also be
done for the Makefile's `all' target, but this one should remain to spot
the build process breaking things, e.g. non-ASCII in $prefix.

diff --git a/man/mhfixmsg.man b/man/mhfixmsg.man
index c418184..dd6c340 100644 (file)
--- a/man/mhfixmsg.man
+++ b/man/mhfixmsg.man
@@ -74,7 +74,7 @@ then the part is not decoded (and a message will be displayed if
 .B \-verbose
 is enabled).  Note that
 .B \-decodetext
 .B \-verbose
 is enabled).  Note that
 .B \-decodetext

-binary can produce messages that are not compliant with RFC 5322, §2.1.1.
+binary can produce messages that are not compliant with RFC 5322, \(sc2.1.1.

 .PP
 When the
 .B \-decodetext
 .PP
 When the
 .B \-decodetext


@@ -108,7 +108,7 @@ with its mandatory
 argument.  The
 .B \-nodecodeheaderfieldbodies
 inhibits this transformation.  The transformation can produce a message
 argument.  The
 .B \-nodecodeheaderfieldbodies
 inhibits this transformation.  The transformation can produce a message

-that does not conform with RFC 2047, §1, paragraph 6, because the decoded
+that does not conform with RFC 2047, \(sc1, paragraph 6, because the decoded

 header field body could contain unencoded non-ASCII characters.  It is
 therefore not enabled by default.
 .PP
 header field body could contain unencoded non-ASCII characters.  It is
 therefore not enabled by default.
 .PP


@@ -120,7 +120,7 @@ switch selects this behavior and is enabled by default.  The
 switch causes carriage return characters to be stripped from, and not
 inserted in, text content when it is decoded and encoded.  Note that
 its use can cause the generation of MIME messages that do not conform
 switch causes carriage return characters to be stripped from, and not
 inserted in, text content when it is decoded and encoded.  Note that
 its use can cause the generation of MIME messages that do not conform

-with RFC 2046, §4.1.1, paragraph 1.
+with RFC 2046, \(sc4.1.1, paragraph 1.

 .PP
 The
 .B \-textcharset
 .PP
 The
 .B \-textcharset


@@ -206,7 +206,7 @@ The
 switch enables a transformation to change the
 Content-Transfer-Encoding from an invalid value to 8-bit in message
 parts with a Content-Type of multipart and message, as required by
 switch enables a transformation to change the
 Content-Transfer-Encoding from an invalid value to 8-bit in message
 parts with a Content-Type of multipart and message, as required by

-RFC 2045, §6.4.  That condition is indicated by a \*(lqmust be
+RFC 2045, \(sc6.4.  That condition is indicated by a \*(lqmust be

 encoded in 7bit, 8bit, or binary\*(rq error message from
 .B mhlist
 and other
 encoded in 7bit, 8bit, or binary\*(rq error message from
 .B mhlist
 and other

diff --git a/test/manpages/test-manpages b/test/manpages/test-manpages
index 6b2ee95..768c8f2 100755 (executable)
--- a/test/manpages/test-manpages
+++ b/test/manpages/test-manpages
@@ -48,8 +48,9 @@ fi
 rm -f "$actual"
 touch "$actual"
 
 rm -f "$actual"
 touch "$actual"
 

-

 for manpage in man?/*.?; do
 for manpage in man?/*.?; do

+  LC_ALL=C grep -n '[^  -~]' /dev/null "$manpage" 2>&1 |
+  sed 's/^/non-ascii:/' >>"$actual"

   groff -Tutf8 -mandoc $warnings "$manpage" >/dev/null 2>>"$actual"
 done
 
   groff -Tutf8 -mandoc $warnings "$manpage" >/dev/null 2>>"$actual"
 done