Sync library documentation with reality.

Split mandoc_escape(3), mandoc_malloc(3), and mchars_alloc(3)
out of mandoc(3), adding lots of new information.

5 files changed, 882 insertions, 180 deletions

diff --git a/Makefile b/Makefile
index b18499cb..276eb59f 100644
--- a/Makefile
+++ b/Makefile
@@ -223,10 +223,13 @@ DISTFILES	 = LICENSE \
 		   mandoc.h \
 		   mandoc_aux.h \
 		   mandoc_char.7 \
+		   mandoc_escape.3 \
 		   mandoc_html.3 \
+		   mandoc_malloc.3 \
 		   manpath.h \
 		   mansearch.3 \
 		   mansearch.h \
+		   mchars_alloc.3 \
 		   mdoc.7 \
 		   mdoc.h \
 		   msec.in \
@@ -327,8 +330,11 @@ WWW_MANS	 = apropos.1.html \
 		   mandoc.1.html \
 		   preconv.1.html \
 		   mandoc.3.html \
+		   mandoc_escape.3.html \
 		   mandoc_html.3.html \
+		   mandoc_malloc.3.html \
 		   mansearch.3.html \
+		   mchars_alloc.3.html \
 		   tbl.3.html \
 		   mandoc.db.5.html \
 		   eqn.7.html \
@@ -393,7 +399,8 @@ base-install: base-build
 	$(INSTALL_LIB) man.h mandoc.h mandoc_aux.h mdoc.h \
 		$(DESTDIR)$(INCLUDEDIR)
 	$(INSTALL_MAN) mandoc.1 preconv.1 demandoc.1 $(DESTDIR)$(MANDIR)/man1
-	$(INSTALL_MAN) mandoc.3 tbl.3 $(DESTDIR)$(MANDIR)/man3
+	$(INSTALL_MAN) mandoc.3 mandoc_escape.3 mandoc_malloc.3 \
+		mchars_alloc.3 tbl.3 $(DESTDIR)$(MANDIR)/man3
 	$(INSTALL_MAN) man.7 mdoc.7 roff.7 eqn.7 tbl.7 mandoc_char.7 \
 		$(DESTDIR)$(MANDIR)/man7
 	$(INSTALL_DATA) example.style.css $(DESTDIR)$(EXAMPLEDIR)
diff --git a/mandoc.3 b/mandoc.3
index f110a805..9a99f9c2 100644
--- a/mandoc.3
+++ b/mandoc.3
@@ -20,21 +20,11 @@
 .Os
 .Sh NAME
 .Nm mandoc ,
-.Nm mandoc_calloc ,
-.Nm mandoc_escape ,
-.Nm mandoc_malloc ,
-.Nm mandoc_realloc ,
-.Nm mandoc_strdup ,
-.Nm mandoc_strndup ,
+.Nm man_deroff ,
 .Nm man_meta ,
 .Nm man_mparse ,
 .Nm man_node ,
-.Nm mchars_alloc ,
-.Nm mchars_free ,
-.Nm mchars_num2char ,
-.Nm mchars_num2uc ,
-.Nm mchars_spec2cp ,
-.Nm mchars_spec2str ,
+.Nm mdoc_deroff ,
 .Nm mdoc_meta ,
 .Nm mdoc_node ,
 .Nm mparse_alloc ,
@@ -50,57 +40,17 @@
 .Sh LIBRARY
 .Lb libmandoc
 .Sh SYNOPSIS
+.In sys/types.h
 .In mandoc.h
 .Fd "#define ASCII_NBRSP"
 .Fd "#define ASCII_HYPH"
 .Fd "#define ASCII_BREAK"
-.Ft "void *"
-.Fo mandoc_calloc
-.Fa "size_t nmemb"
-.Fa "size_t size"
-.Fc
-.Ft "enum mandoc_esc"
-.Fo mandoc_escape
-.Fa "const char **end"
-.Fa "const char **start"
-.Fa "int *sz"
-.Fc
-.Ft "void *"
-.Fn mandoc_malloc "size_t size"
-.Ft "struct mchars *"
-.Fo mandoc_realloc
-.Fa "void *ptr"
-.Fa "size_t size"
-.Fc
-.Ft "char *"
-.Fn mandoc_strdup
-.Fn mchars_alloc "void"
-.Ft void
-.Fn mchars_free "struct mchars *p"
-.Ft char
-.Fn mchars_num2char "const char *cp" "size_t sz"
-.Ft int
-.Fn mchars_num2uc "const char *cp" "size_t sz"
-.Ft "const char *"
-.Fo mchars_spec2str
-.Fa "const struct mchars *p"
-.Fa "const char *cp"
-.Fa "size_t sz"
-.Fa "size_t *rsz"
-.Fc
-.Ft int
-.Fo mchars_spec2cp
-.Fa "const struct mchars *p"
-.Fa "const char *cp"
-.Fa "size_t sz"
-.Fc
-.Ft void
+.Ft struct mparse *
 .Fo mparse_alloc
-.Fa "enum mparset inttype"
+.Fa "int options"
 .Fa "enum mandoclevel wlevel"
 .Fa "mandocmsg mmsg"
 .Fa "char *defos"
-.Fa "int quick"
 .Fc
 .Ft void
 .Fo (*mandocmsg)
@@ -138,6 +88,7 @@
 .Fa "struct mparse *parse"
 .Fa "struct mdoc **mdoc"
 .Fa "struct man **man"
+.Fa "char **sodest"
 .Fc
 .Ft "const char *"
 .Fo mparse_strerror
@@ -147,8 +98,14 @@
 .Fo mparse_strlevel
 .Fa "enum mandoclevel"
 .Fc
+.In sys/types.h
 .In mandoc.h
 .In mdoc.h
+.Ft void
+.Fo mdoc_deroff
+.Fa "char **dest"
+.Fa "const struct mdoc_node *node"
+.Fc
 .Ft "const struct mdoc_meta *"
 .Fo mdoc_meta
 .Fa "const struct mdoc *mdoc"
@@ -159,8 +116,14 @@
 .Fc
 .Vt extern const char * const * mdoc_argnames;
 .Vt extern const char * const * mdoc_macronames;
+.In sys/types.h
 .In mandoc.h
 .In man.h
+.Ft void
+.Fo man_deroff
+.Fa "char **dest"
+.Fa "const struct man_node *node"
+.Fc
 .Ft "const struct man_meta *"
 .Fo man_meta
 .Fa "const struct man *man"
@@ -215,37 +178,22 @@ or invoke
 .Fn mparse_reset
 and parse new files.
 .El
-.Pp
-The
-.Nm
-library also contains routines for translating character strings into glyphs
-.Pq see Fn mchars_alloc
-and parsing escape sequences from strings
-.Pq see Fn mandoc_escape .
 .Sh REFERENCE
 This section documents the functions, types, and variables available
 via
-.In mandoc.h .
+.In mandoc.h ,
+with the exception of those documented in
+.Xr mandoc_escape 3
+and
+.Xr mchars_alloc 3 .
 .Ss Types
 .Bl -ohang
-.It Vt "enum mandoc_esc"
-An escape sequence classification.
 .It Vt "enum mandocerr"
 A fatal error, error, or warning message during parsing.
 .It Vt "enum mandoclevel"
 A classification of an
 .Vt "enum mandocerr"
 as regards system operation.
-.It Vt "struct mchars"
-An opaque pointer to an object allowing for translation between
-character strings and glyphs.
-See
-.Fn mchars_alloc .
-.It Vt "enum mparset"
-The type of parser when reading input.
-This should usually be
-.Dv MPARSE_AUTO
-for auto-detection.
 .It Vt "struct mparse"
 An opaque pointer to a running parse sequence.
 Created with
@@ -261,38 +209,20 @@ messages emitted by the parser.
 .El
 .Ss Functions
 .Bl -ohang
-.It Fn mandoc_escape
-Scan an escape sequence, i.e., a character string beginning with
-.Sq \e .
-Pass a pointer to the character after the
-.Sq \e
-as
-.Va end ;
-it will be set to the supremum of the parsed escape sequence unless
-returning
-.Dv ESCAPE_ERROR ,
-in which case the string is bogus and should be
-thrown away.
-If not
-.Dv ESCAPE_ERROR
-or
-.Dv ESCAPE_IGNORE ,
-.Va start
-is set to the first relevant character of the substring (font, glyph,
-whatever) of length
-.Va sz .
-Both
-.Va start
-and
-.Va sz
-may be
-.Dv NULL .
-Declared in
-.In mandoc.h ,
-implemented in
-.Pa mandoc.c .
+.It Fn man_deroff
+Obtain a text-only representation of a
+.Vt struct man_node ,
+including text contained in its child nodes.
+To be used on children of the pointer returned from
+.Fn man_node .
+When it is no longer needed, the pointer returned from
+.Fn man_deroff
+can be passed to
+.Xr free 3 .
 .It Fn man_meta
-Obtain the meta-data of a successful parse.
+Obtain the meta-data of a successful
+.Xr man 7
+parse.
 This may only be used on a pointer returned by
 .Fn mparse_result .
 Declared in
@@ -306,67 +236,29 @@ Declared in
 implemented in
 .Pa man.c .
 .It Fn man_node
-Obtain the root node of a successful parse.
+Obtain the root node of a successful
+.Xr man 7
+parse.
 This may only be used on a pointer returned by
 .Fn mparse_result .
 Declared in
 .In man.h ,
 implemented in
 .Pa man.c .
-.It Fn mchars_alloc
-Allocate an
-.Vt "struct mchars *"
-object for translating special characters into glyphs.
-See
-.Xr mandoc_char 7
-for an overview of special characters.
-The object must be freed with
-.Fn mchars_free .
-Declared in
-.In mandoc.h ,
-implemented in
-.Pa chars.c .
-.It Fn mchars_free
-Free an object created with
-.Fn mchars_alloc .
-Declared in
-.In mandoc.h ,
-implemented in
-.Pa chars.c .
-.It Fn mchars_num2char
-Convert a character index (e.g., the \eN\(aq\(aq escape) into a
-printable ASCII character.
-Returns \e0 (the nil character) if the input sequence is malformed.
-Declared in
-.In mandoc.h ,
-implemented in
-.Pa chars.c .
-.It Fn mchars_num2uc
-Convert a hexadecimal character index (e.g., the \e[uNNNN] escape) into
-a Unicode codepoint.
-Returns \e0 (the nil character) if the input sequence is malformed.
-Declared in
-.In mandoc.h ,
-implemented in
-.Pa chars.c .
-.It Fn mchars_spec2cp
-Convert a special character into a valid Unicode codepoint.
-Returns \-1 on failure or a non-zero Unicode codepoint on success.
-Declared in
-.In mandoc.h ,
-implemented in
-.Pa chars.c .
-.It Fn mchars_spec2str
-Convert a special character into an ASCII string.
-Returns
-.Dv NULL
-on failure.
-Declared in
-.In mandoc.h ,
-implemented in
-.Pa chars.c .
+.It Fn mdoc_deroff
+Obtain a text-only representation of a
+.Vt struct mdoc_node ,
+including text contained in its child nodes.
+To be used on children of the pointer returned from
+.Fn mdoc_node .
+When it is no longer needed, the pointer returned from
+.Fn mdoc_deroff
+can be passed to
+.Xr free 3 .
 .It Fn mdoc_meta
-Obtain the meta-data of a successful parse.
+Obtain the meta-data of a successful
+.Xr mdoc
+parse.
 This may only be used on a pointer returned by
 .Fn mparse_result .
 Declared in
@@ -374,7 +266,9 @@ Declared in
 implemented in
 .Pa mdoc.c .
 .It Fn mdoc_node
-Obtain the root node of a successful parse.
+Obtain the root node of a successful
+.Xr mdoc
+parse.
 This may only be used on a pointer returned by
 .Fn mparse_result .
 Declared in
@@ -385,15 +279,33 @@ implemented in
 Allocate a parser.
 The arguments have the following effect:
 .Bl -tag -offset 5n -width inttype
-.It Ar inttype
-When set to
+.It Ar options
+When the
 .Dv MPARSE_MDOC
 or
-.Dv MPARSE_MAN ,
-only that parser will be used.
-With
-.Dv MPARSE_AUTO ,
-the document type will be automatically detected.
+.Dv MPARSE_MAN
+bit is set, only that parser is used.
+Otherwise, the document type is automatically detected.
+.Pp
+When the
+.Dv MPARSE_SO
+bit is set,
+.Xr roff 7
+.Ic \&so
+file inclusion requests are always honoured.
+Otherwise, if the request is the only content in an input file,
+only the file name is remembered, to be returned in the
+.Fa sodest
+argument of
+.Fn mparse_result .
+.Pp
+When the
+.Dv MPARSE_QUICK
+bit is set, parsing is aborted after the NAME section.
+This is for example useful in
+.Xr makewhatis 8
+.Fl Q
+to quickly build minimal databases.
 .It Ar wlevel
 Can be set to
 .Dv MANDOCLEVEL_FATAL ,
@@ -414,9 +326,6 @@ macro, overriding the
 .Dv OSNAME
 preprocessor definition and the results of
 .Xr uname 3 .
-.It Ar quick
-When set, parsing is aborted after the NAME section.
-This is for example useful to quickly build minimal databases.
 .El
 .Pp
 The same parser may be used for multiple files so long as
@@ -486,7 +395,7 @@ i.e., those where
 .Fn mparse_readfd
 returned less than MANDOCLEVEL_FATAL
 .Pc
-should invoke this function, in which case one of the two pointers will
+should invoke this function, in which case one of the three pointers will
 be filled in.
 Declared in
 .In mandoc.h ,
@@ -540,6 +449,8 @@ The following non-printing characters may be embedded in text strings:
 A non-breaking space character.
 .It Dv ASCII_HYPH
 A soft hyphen.
+.It Dv ASCII_BREAK
+A breakable zero-width space.
 .El
 .Pp
 Escape characters are also passed verbatim into text strings.
@@ -547,11 +458,9 @@ An escape character is a sequence of characters beginning with the
 backslash
 .Pq Sq \e .
 To construct human-readable text, these should be intercepted with
-.Fn mandoc_escape
-and converted with one of
-.Fn mchars_num2char ,
-.Fn mchars_spec2str ,
-and so on.
+.Xr mandoc_escape 3
+and converted with one the functions described in
+.Xr mchars_alloc 3 .
 .Ss Man Abstract Syntax Tree
 This AST is governed by the ontological rules dictated in
 .Xr man 7
@@ -596,7 +505,7 @@ where capitalised non-terminals represent nodes.
 .El
 .Pp
 The only elements capable of nesting other elements are those with
-next-lint scope as documented in
+next-line scope as documented in
 .Xr man 7 .
 .Ss Mdoc Abstract Syntax Tree
 This AST is governed by the ontological
@@ -732,10 +641,13 @@ front-ends to
 .Xr mandoc 1
 are unable to render them in any meaningful way.
 Furthermore, behaviour when encountering badly-nested blocks is not
-consistent across troff implementations, especially when using  multiple
+consistent across troff implementations, especially when using multiple
 levels of badly-nested blocks.
 .Sh SEE ALSO
 .Xr mandoc 1 ,
+.Xr mandoc_escape 3 ,
+.Xr mandoc_malloc 3 ,
+.Xr mchars_alloc 3 ,
 .Xr eqn 7 ,
 .Xr man 7 ,
 .Xr mandoc_char 7 ,
diff --git a/mandoc_escape.3 b/mandoc_escape.3
new file mode 100644
index 00000000..e76601a3
--- /dev/null
+++ b/mandoc_escape.3
@@ -0,0 +1,362 @@
+.\"	$Id$
+.\"
+.\" Copyright (c) 2014 Ingo Schwarze <schwarze@openbsd.org>
+.\"
+.\" Permission to use, copy, modify, and distribute this software for any
+.\" purpose with or without fee is hereby granted, provided that the above
+.\" copyright notice and this permission notice appear in all copies.
+.\"
+.\" THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+.\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+.\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+.\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+.\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+.\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+.\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+.\"
+.Dd $Mdocdate$
+.Dt MANDOC_ESCAPE 3
+.Os
+.Sh NAME
+.Nm mandoc_escape
+.Nd parse roff escape sequences
+.Sh LIBRARY
+.Lb libmandoc
+.Sh SYNOPSIS
+.In sys/types.h
+.In mandoc.h
+.Ft "enum mandoc_esc"
+.Fo mandoc_escape
+.Fa "const char **end"
+.Fa "const char **start"
+.Fa "int *sz"
+.Fc
+.Sh DESCRIPTION
+This function scans a
+.Xr roff 7
+escape sequence.
+.Pp
+An escape sequence consists of
+.Bl -dash -compact -width 2n
+.It
+an initial backslash character
+.Pq Sq \e ,
+.It
+a single ASCII character called the escape sequence identifier,
+.It
+and, with only a few exceptions, an argument.
+.El
+.Pp
+Arguments can be given in the following forms; some escape sequence
+identifiers only accept some of these forms as specified below.
+The first three forms are called the standard forms.
+.Bl -tag -width 2n
+.It \&In brackets: Ic \&[ Ns Ar argument Ns Ic \&]
+The argument starts after the initial
+.Sq \&[ ,
+ends before the final
+.Sq \&] ,
+and the escape sequence ends with the final
+.Sq \&] .
+.It Two-character argument short form: Ic \&( Ns Ar ar
+This form can only be used for arguments
+consisting of exactly two characters.
+It has the same effect as
+.Ic \&[ Ns Ar ar Ns Ic \&] .
+.It One-character argument short form: Ar a
+This form can only be used for arguments
+consisting of exactly one character.
+It has the same effect as
+.Ic \&[ Ns Ar a Ns Ic \&] .
+.It Delimited form: Ar C Ns Ar argument Ns Ar C
+The argument starts after the initial delimiter character
+.Ar C ,
+ends before the next occurrence of the delimiter character
+.Ar C ,
+and the escape sequence ends with that second
+.Ar C .
+Some escape sequences allow arbitrary characters
+.Ar C
+as quoting characters, some restrict the range of characters
+that can be used as quoting characters.
+.El
+.Pp
+Upon function entry,
+.Fa end
+is expected to point to the escape sequence identifier.
+The values passed in as
+.Fa start
+and
+.Fa sz
+are ignored and overwritten.
+.Pp
+By design, this function cannot handle those
+.Xr roff 7
+escape sequences that require in-place expansion, in particular
+user-defined strings
+.Ic \e* ,
+number registers
+.Ic \en ,
+width measurements
+.Ic \ew ,
+and numerical expression control
+.Ic \eB .
+These are handled by
+.Fn roff_res ,
+a private preprocessor function called from
+.Fn roff_parseln ,
+see the file
+.Pa roff.c .
+.Pp
+The function
+.Fn mandoc_escape
+is used
+.Bl -dash -compact -width 2n
+.It
+recursively by itself, because some escape sequence arguments can
+in turn contain other escape sequences,
+.It
+for error detection internally by the
+.Xr roff 7
+parser part of the
+.Lb libmandoc ,
+see the file
+.Pa roff.c ,
+.It
+above all externally by the
+.Xr mandoc
+formatting modules, in particular
+.Fl Tascii
+and
+.Fl Thtml ,
+for formatting purposes, see the files
+.Pa term.c
+and
+.Pa html.c ,
+.It
+and rarely externally by high-level utilities using the mandoc library,
+for example
+.Xr makewhatis 8 ,
+to purge escape sequences from text.
+.El
+.Sh RETURN VALUES
+Upon function return, the pointer
+.Fa end
+is set to the character after the end of the escape sequence,
+such that the calling higher-level parser can easily continue.
+.Pp
+For escape sequences taking an argument, the pointer
+.Fa start
+is set to the beginning of the argument and
+.Fa sz
+is set to the length of the argument.
+For escape sequences not taking an argument,
+.Fa start
+is set to the character after the end of the sequence and
+.Fa sz
+is set to 0.
+Both
+.Fa start
+and
+.Fa sz
+may be
+.Dv NULL ;
+in that case, the argument and the length are not returned.
+.Pp
+For sequences taking an argument, the function
+.Fn mandoc_escape
+returns one of the following values:
+.Bl -tag -width 2n
+.It Dv ESCAPE_FONT
+The escape sequence
+.Ic \ef
+taking an argument in standard form:
+.Ic \ef[ , \ef( , \ef Ns Ar a .
+Two-character arguments starting with the character
+.Sq C
+are reduced to one-character arguments by skipping the
+.Sq C .
+More specific values are returned for the most commonly used arguments:
+.Bl -column "argument" "ESCAPE_FONTITALIC"
+.It argument Ta return value
+.It Cm R No or Cm 1 Ta Dv ESCAPE_FONTROMAN
+.It Cm I No or Cm 2 Ta Dv ESCAPE_FONTITALIC
+.It Cm B No or Cm 3 Ta Dv ESCAPE_FONTBOLD
+.It Cm P Ta Dv ESCAPE_FONTPREV
+.It Cm BI Ta Dv ESCAPE_FONTBI
+.El
+.It Dv ESCAPE_SPECIAL
+The escape sequence
+.Ic \eC
+taking an argument delimited with the single quote character
+and, as a special exception, the escape sequences
+.Em not
+having an identifier, that is, those where the argument, in standard
+form, directly follows the initial backslash:
+.Ic \eC' , \e[ , \e( , \e Ns Ar a .
+Note that the one-character argument short form can only be used for
+argument characters that do not clash with escape sequence identifiers.
+.Pp
+If the argument consists of more than one character
+and starts with the character
+.Sq u ,
+.Dv ESCAPE_UNICODE
+is returned as described below.
+If the argument is just the single character
+.Sq u ,
+.Dv ESCAPE_ERROR
+is returned.
+.Pp
+The
+.Dv ESCAPE_SPECIAL
+special character escape sequences can be rendered using the functions
+.Fn mchars_spec2cp
+and
+.Fn mchars_spec2str
+described in the
+.Xr mchars_alloc 3
+manual.
+.It Dv ESCAPE_UNICODE
+Escape sequences of the same format as described above under
+.Dv ESCAPE_SPECIAL ,
+but with an argument starting with the character
+.Sq u :
+.Ic \eC'u , \e[u .
+As a special exception,
+.Fa start
+is set to the character after the
+.Sq u ,
+and the
+.Fa sz
+return value does not include the
+.Sq u
+either.
+.Pp
+Such Unicode character escape sequences can be rendered using the function
+.Fn mchars_num2uc
+described in the
+.Xr mchars_alloc 3
+manual.
+.It Dv ESCAPE_NUMBERED
+The escape sequence
+.Ic \eN
+followed by a delimited argument.
+The delimiter character is arbitrary except that digits cannot be used.
+If a digit is encountered instead of the opening delimiter, that
+digit is considered to be the argument and the end of the sequence, and
+.Dv ESCAPE_IGNORE
+is returned.
+.Pp
+Such ASCII character escape sequences can be rendered using the function
+.Fn mchars_num2char
+described in the
+.Xr mchars_alloc 3
+manual.
+.It Dv ESCAPE_IGNORE
+.Bl -bullet -width 2n
+.It
+The escape sequence
+.Ic \es
+followed by an argument in standard form or by an argument delimited
+by the single quote character:
+.Ic \es' , \es[ , \es( , \es Ns Ar a .
+As a special exception, an optional
+.Sq +
+or
+.Sq \-
+character is allowed after the
+.Sq s
+for all forms.
+.It
+The escape sequences
+.Ic \eF ,
+.Ic \eg ,
+.Ic \ek ,
+.Ic \eM ,
+.Ic \em ,
+.Ic \en ,
+.Ic \eV ,
+and
+.Ic \eY
+followed by an argument in standard form.
+.It
+The escape sequences
+.Ic \eA ,
+.Ic \eb ,
+.Ic \eD ,
+.Ic \eo ,
+.Ic \eR ,
+.Ic \eX ,
+and
+.Ic \eZ
+followed by an argument delimited by an arbitrary character.
+.It
+The escape sequences
+.Ic \eH ,
+.Ic \eh ,
+.Ic \eL ,
+.Ic \el ,
+.Ic \eS ,
+.Ic \ev ,
+and
+.Ic \ex
+followed by an argument delimited by a character that cannot occur
+in numerical expressions.
+However, if any character that can occur in numerical expressions
+is found instead of a delimiter, the sequence is considered to end
+with that character, and
+.Dv ESCAPE_ERROR
+is returned.
+.El
+.It Dv ESCAPE_ERROR
+Escape sequences taking an argument but not matching any of the above patterns.
+In particular, that happens if the end of the logical input line
+is reached before the end of the argument.
+.El
+.Pp
+For sequences that do not take an argument, the function
+.Fn mandoc_escape
+returns one of the following values:
+.Bl -tag -width 2n
+.It Dv ESCAPE_SKIPCHAR
+The escape sequence
+.Qq \ez .
+.It Dv ESCAPE_NOSPACE
+The escape sequence
+.Qq \ec .
+.It Dv ESCAPE_IGNORE
+The escape sequences
+.Qq \ed
+and
+.Qq \eu .
+.El
+.Sh FILES
+This function is implemented in
+.Pa mandoc.c .
+.Sh SEE ALSO
+.Xr mchars_alloc 3 ,
+.Xr mandoc_char 7 ,
+.Xr roff 7
+.Sh HISTORY
+This function has been available since mandoc 1.11.2.
+.Sh AUTHORS
+.An Kristaps Dzonsons Aq Mt kristaps@bsd.lv
+.An Ingo Schwarze Aq Mt schwarze@openbsd.org
+.Sh BUGS
+The function doesn't cleanly distinguish between sequences that are
+valid and supported, valid and ignored, valid and unsupported,
+syntactically invalid, or undefined.
+For sequences that are ignored or unsupported, it doesn't tell
+whether that deficiency is likely to cause major formatting problems
+and/or loss of document content.
+The function is already rather complicated and still parses some
+sequences incorrectly.
+.
+.ig
+For these sequences, the list given below specifies a starting string
+and either the length of the argument or an ending character.
+The argument starts after the starting string.
+In the former case, the sequence ends with the end of the argument.
+In the latter case, the argument ends before the ending character,
+and the sequence ends with the ending character.
+..
diff --git a/mandoc_malloc.3 b/mandoc_malloc.3
new file mode 100644
index 00000000..190432f7
--- /dev/null
+++ b/mandoc_malloc.3
@@ -0,0 +1,197 @@
+.\"	$Id$
+.\"
+.\" Copyright (c) 2014 Ingo Schwarze <schwarze@openbsd.org>
+.\"
+.\" Permission to use, copy, modify, and distribute this software for any
+.\" purpose with or without fee is hereby granted, provided that the above
+.\" copyright notice and this permission notice appear in all copies.
+.\"
+.\" THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+.\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+.\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+.\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+.\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+.\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+.\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+.\"
+.Dd $Mdocdate$
+.Dt MANDOC_MALLOC 3
+.Os
+.Sh NAME
+.Nm mandoc_malloc ,
+.Nm mandoc_realloc ,
+.Nm mandoc_reallocarray ,
+.Nm mandoc_calloc ,
+.Nm mandoc_strdup ,
+.Nm mandoc_strndup ,
+.Nm mandoc_asprintf
+.Nd memory allocation function wrappers used in the mandoc library
+.Sh LIBRARY
+.Lb libmandoc
+.Sh SYNOPSIS
+.In sys/types.h
+.In mandoc_aux.h
+.Ft "void *"
+.Fo mandoc_malloc
+.Fa "size_t size"
+.Fc
+.Ft "void *"
+.Fo mandoc_realloc
+.Fa "void *ptr"
+.Fa "size_t size"
+.Fc
+.Ft "void *"
+.Fo mandoc_reallocarray
+.Fa "void *ptr"
+.Fa "size_t nmemb"
+.Fa "size_t size"
+.Fc
+.Ft "void *"
+.Fo mandoc_calloc
+.Fa "size_t nmemb"
+.Fa "size_t size"
+.Fc
+.Ft "char *"
+.Fo mandoc_strdup
+.Fa "const char *s"
+.Fc
+.Ft "char *"
+.Fo mandoc_strndup
+.Fa "const char *s"
+.Fa "size_t maxlen"
+.Fc
+.Ft int
+.Fo mandoc_asprintf
+.Fa "char **ret"
+.Fa "const char *format"
+.Fa "..."
+.Fc
+.Sh DESCRIPTION
+These functions call the
+.Lb libc
+functions of the same names, passing through their return values when
+successful.
+In case of failure, they do not return, but instead call
+.Xr perror 3
+and
+.Xr exit 3 .
+They can be used both internally by any code in the
+.Lb libmandoc
+and externally by programs using that library, for example
+.Xr mandoc 1 ,
+.Xr apropos 1 ,
+and
+.Xr makewhatis 8 .
+.Pp
+The function
+.Fn mandoc_malloc
+allocates one new object, leaving the memory uninitialized.
+The functions
+.Fn mandoc_realloc
+and
+.Fn mandoc_reallocarray
+change the size of an existing object or array, possibly moving it.
+When shrinking the size, existing data is truncated; when growing,
+the additional memory is not initialized.
+The function
+.Fn mandoc_calloc
+allocates a new array, initializing it to zero.
+.Pp
+The argument
+.Fa size
+is the size of each object.
+The argument
+.Fa nmemb
+is the new number of objects in the array.
+The argument
+.Fa ptr
+is a pointer to the existing object or array to be resized; if it is
+.Dv NULL ,
+a new object or array is allocated.
+.Pp
+The functions
+.Fn mandoc_strdup
+and
+.Fn mandoc_strndup
+copy a string into newly allocated memory.
+For
+.Fn mandoc_strdup ,
+the string pointed to by
+.Fa s
+needs to be NUL-terminated.
+For
+.Fn mandoc_strndup ,
+at most
+.Fa maxlen
+bytes are copied.
+The function
+.Fn mandoc_asprintf
+writes output formatted according to
+.Fa format
+into newly allocated memory and returns a pointer to the result in
+.Fa ret .
+For all three string functions, the result is always NUL-terminated.
+.Pp
+When the objects and strings are no longer needed,
+the pointers returned by these functions can be passed to
+.Xr free 3 .
+.Sh RETURN VALUES
+The function
+.Fn mandoc_asprintf
+always returns the number of characters written, excluding the
+final NUL byte.
+It never returns -1.
+.Pp
+The other functions always return a valid pointer; they never return
+.Dv NULL .
+.Sh FILES
+These functions are implemented in
+.Pa mandoc_aux.c .
+.Sh SEE ALSO
+.Xr asprintf 3 ,
+.Xr exit 3 ,
+.Xr malloc 3 ,
+.Xr perror 3 ,
+.Xr strdup 3
+.Sh STANDARDS
+The functions
+.Fn malloc ,
+.Fn realloc ,
+and
+.Fn calloc
+are required by
+.St -ansiC .
+The functions
+.Fn strdup
+and
+.Fn strndup
+are required by
+.St -p1003.1-2008 .
+The function
+.Fn asprintf
+is a widespread extension that first appeared in the GNU C library.
+.Pp
+The function
+.Fn reallocarray
+is an extension that first appeared in
+.Ox 5.6 .
+If it is not provided by the operating system, the mandoc build system
+uses a bundled portable implementation.
+.Sh HISTORY
+The functions
+.Fn mandoc_malloc ,
+.Fn mandoc_realloc ,
+.Fn mandoc_calloc ,
+and
+.Fn mandoc_strdup
+have been available since mandoc 1.9.12,
+.Fn mandoc_strndup
+since 1.11.5,
+and
+.Fn mandoc_asprintf
+and
+.Fn mandoc_reallocarray
+since 1.12.4 and 1.13.0.
+.Sh AUTHORS
+.An Kristaps Dzonsons Aq Mt kristaps@bsd.lv
+.An Ingo Schwarze Aq Mt schwarze@openbsd.org
diff --git a/mchars_alloc.3 b/mchars_alloc.3
new file mode 100644
index 00000000..34c9fa3f
--- /dev/null
+++ b/mchars_alloc.3
@@ -0,0 +1,224 @@
+.\"	$Id$
+.\"
+.\" Copyright (c) 2014 Ingo Schwarze <schwarze@openbsd.org>
+.\"
+.\" Permission to use, copy, modify, and distribute this software for any
+.\" purpose with or without fee is hereby granted, provided that the above
+.\" copyright notice and this permission notice appear in all copies.
+.\"
+.\" THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+.\" WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+.\" MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+.\" ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+.\" WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+.\" ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+.\" OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+.\"
+.Dd $Mdocdate$
+.Dt MCHARS_ALLOC 3
+.Os
+.Sh NAME
+.Nm mchars_alloc ,
+.Nm mchars_free ,
+.Nm mchars_num2char ,
+.Nm mchars_num2uc ,
+.Nm mchars_spec2cp ,
+.Nm mchars_spec2str
+.Nd character table for mandoc
+.Sh LIBRARY
+.Lb libmandoc
+.Sh SYNOPSIS
+.In sys/types.h
+.In mandoc.h
+.Ft "struct mchars *"
+.Fn mchars_alloc "void"
+.Ft void
+.Fo mchars_free
+.Fa "struct mchars *table"
+.Fc
+.Ft char
+.Fo mchars_num2char
+.Fa "const char *decimal"
+.Fa "size_t sz"
+.Fc
+.Ft int
+.Fo mchars_num2uc
+.Fa "const char *hexadecimal"
+.Fa "size_t sz"
+.Fc
+.Ft int
+.Fo mchars_spec2cp
+.Fa "const struct mchars *table"
+.Fa "const char *name"
+.Fa "size_t sz"
+.Fc
+.Ft "const char *"
+.Fo mchars_spec2str
+.Fa "const struct mchars *table"
+.Fa "const char *name"
+.Fa "size_t sz"
+.Fa "size_t *rsz"
+.Fc
+.Sh DESCRIPTION
+These functions translate Unicode character numbers and
+.Xr roff 7
+character names into glyphs.
+See
+.Xr mandoc_char 7
+for a list of
+.Xr roff 7
+special characters.
+These functions are intended for external use by programs formatting
+.Xr mdoc 7
+and
+.Xr man 7
+pages for output, for example the
+.Xr mandoc 1
+output formatter modules and
+.Xr makewhatis 8 .
+The
+.Fa decimal ,
+.Fa hexadecimal ,
+.Fa name ,
+and
+.Fa size
+input arguments are usually obtained from the
+.Xr mandoc_escape 3
+parser function.
+.Pp
+The function
+.Fn mchars_num2char
+converts a
+.Fa decimal
+string representation of a character number consisting of
+.Fa sz
+digits into a printable ASCII character.
+If the input string is non-numeric or does not represent a printable
+ASCII character, the NUL character
+.Pq Sq \e0
+is returned.
+For example, the
+.Xr mandoc 1
+.Fl Tascii ,
+.Fl Tutf8 ,
+and
+.Fl Thtml
+output modules use this function to render
+.Xr roff 7
+.Ic \eN
+escape sequences.
+.Pp
+The function
+.Fn mchars_num2uc
+converts a
+.Fa hexadecimal
+string representation of a Unicode codepoint consisting of
+.Fa sz
+digits into an integer representation.
+If the input string is non-numeric or represents an ASCII character,
+the NUL character
+.Pq Sq \e0
+is returned.
+For example, the
+.Xr mandoc 1
+.Fl Tutf8
+and
+.Fl Thtml
+output modules use this function to render
+.Xr roff 7
+.Ic \e[u Ns Ar XXXX Ns Ic \&]
+and
+.Ic \eC\(aqu Ns Ar XXXX Ns Ic \(aq
+escape sequences.
+.Pp
+The function
+.Fn mchars_alloc
+allocates an opaque
+.Vt "struct mchars *"
+table object for subsequent use by the following two lookup functions.
+When no longer needed, this object can be destroyed with
+.Fn mchars_free .
+.Pp
+The function
+.Fn mchars_spec2cp
+looks up a
+.Xr roff 7
+special character
+.Fa name
+consisting of
+.Fa sz
+characters in the
+.Fa table
+and returns the corresponding Unicode codepoint.
+If the
+.Ar name
+is not recognized, \-1 is returned.
+For example, the
+.Xr mandoc 1
+.Fl Tutf8
+and
+.Fl Thtml
+output modules use this function to render
+.Xr roff 7
+.Ic \e[ Ns Ar name Ns Ic \&]
+and
+.Ic \eC\(aq Ns Ar name Ns Ic \(aq
+escape sequences.
+.Pp
+The function
+.Fn mchars_spec2str
+looks up a
+.Xr roff 7
+special character
+.Fa name
+consisting of
+.Fa sz
+characters in the
+.Fa table
+and returns an ASCII string representation.
+The length of the representation is returned in
+.Fa rsz .
+In many cases, the meaning of such ASCII representations
+is not quite obvious, so using
+.Xr roff 7
+special characters in documents intended for ASCII rendering
+is usually a bad idea.
+If the
+.Ar name
+is not recognized,
+.Dv NULL
+is returned.
+For example,
+.Xr makewhatis 8
+and the
+.Xr mandoc 1
+.Fl Tascii
+output module use this function to render
+.Xr roff 7
+.Ic \e[ Ns Ar name Ns Ic \&]
+and
+.Ic \eC\(aq Ns Ar name Ns Ic \(aq
+escape sequences.
+.Sh FILES
+These funtions are implemented in the file
+.Pa chars.c .
+.Sh SEE ALSO
+.Xr mandoc 1 ,
+.Xr mandoc_escape 3 ,
+.Xr mandoc_char 7 ,
+.Xr roff 7
+.Sh HISTORY
+These functions and their predecessors have been available since the
+following mandoc versions:
+.Bl -column "mchars_num2char()" "1.11.3" "chars_num2char()" "1.10.10"
+.It Sy function Ta since Ta Sy predecessor Ta since
+.It Fn mchars_alloc Ta 1.11.3 Ta Fn ascii2htab Ta 1.5.3
+.It Fn mchars_free Ta 1.11.2 Ta Fn asciifree Ta 1.6.0
+.It Fn mchars_num2char Ta 1.11.2 Ta Fn chars_num2char Ta 1.10.10
+.It Fn mchars_num2uc Ta 1.11.3 Ta \(em Ta \(em
+.It Fn mchars_spec2cp Ta 1.11.2 Ta Fn chars_spec2cp Ta 1.10.5
+.It Fn mchars_spec2str Ta 1.11.2 Ta Fn a2ascii Ta 1.5.3
+.El
+.Sh AUTHORS
+.An Kristaps Dzonsons Aq Mt kristaps@bsd.lv
+.An Ingo Schwarze Aq Mt schwarze@openbsd.org