[RFC][PATCH] Make manpages via DocBook XML
Jon TURNEY
jon.turney@dronecode.org.uk
Tue Jul 21 13:42:00 GMT 2015
As I wrote a while ago, I've been looking for a better way to generate
the Cygwin package containing manpages for newlib.
Attached is a patch implementing a replacement for the makedoc tool,
which generates DocBook XML rather than .texinfo, which can then be
processed into manpages and other formats.
I would appreciate any comments you have on the approach and the
implementation.
There are at least the following issues with this patch:
* This only converts to DocBook the minimum necessary to generate manpages.
Converting the full content of libc.texinfo, libm.texinfo, sys.tex,
reent.tex, iconv.tex, etc. is hard to do automatically, so would need to
be done by hand.
Since maintaining the documentation in two formats seems unlikely to be
a good idea, this would mean fully converting to DocBook, removing all
the current texinfo source and .info generation and generating the .info
from DocBook instead.
* There's a lot of boilerplate in all the Makefiles to generate
documentation, and this adds more. I don't know why this isn't in
Makefile.shared
* This fails to fully achieve the objective of generating good manpages.
A few functions (e.g. sccanf, ssprintf) use nested tables in a way
which can't be represented by troff markup (see [1]). So the markup of
the documentation for these functions would need to be improved,
presumably by writing it directly in DocBook.
* I haven't tested that documentation is correctly included/excluded
depending on the newlib configuration.
* build avoidance needs doing correctly
* The documentation license contains a clause specifically permitting
printing after processing with Tex. Presumably the copyright holder, Red
Hat, would need to approve a change permitting printing after processing
with other tools.
[1] https://sourceforge.net/p/docbook/bugs/1362/
-------------- next part --------------
From dc69f69ee7db72052e1f13a9099c91812a771a70 Mon Sep 17 00:00:00 2001
From: Jon TURNEY <jon.turney@dronecode.org.uk>
Date: Sat, 18 Jul 2015 19:12:45 +0100
Subject: [PATCH] Make newlib manpages via DocBook XML
Add makedocbook, a tool to process makedoc markup and output DocBook XML
refentries.
Add chapter-texi2docbook, to help automatically convert chapter .texi files to
DocBook XML. (This does more work than is strictly necessary: For generating man
pages all we care about is the content of the refentries, so all we need to
convert are @include of the makedoc generated .def files to xi:include
makedocbook generated .xml. Fully converting the documentation would also need
to convert the handwritten contents of lib[cm].texinfo, where only a skeleton
lib[cm].xml is currently used.)
All the source files which are processed with makedoc, process with makedocbook
as well
Generate and install man pages, HTML and PDF
v2:
makedoc COURIER lines are used for preformatted content over multiple lines, so
must be shown verbatim. Convert to <literallayout class="monospace"> not <code>
also generate PDF
add an index, insert index entries for functions and texinfo @[cf]index
Remove unneeded searchpaths since we xinclude everything during xsltproc processing
Structure using chapters
Mark optional includes with an empty fallback
Signed-off-by: Jon TURNEY <jon.turney@dronecode.org.uk>
---
newlib/doc/.gitignore | 3 +
newlib/doc/chapter-texi2docbook.py | 142 +++++++
newlib/doc/makedocbook.py | 792 +++++++++++++++++++++++++++++++++++++
newlib/libc/Makefile.am | 20 +-
newlib/libc/ctype/Makefile.am | 24 +-
newlib/libc/libc.in.xml | 43 ++
newlib/libc/locale/Makefile.am | 21 +-
newlib/libc/misc/Makefile.am | 22 +-
newlib/libc/posix/Makefile.am | 22 +-
newlib/libc/search/Makefile.am | 19 +-
newlib/libc/signal/Makefile.am | 24 +-
newlib/libc/stdio/Makefile.am | 24 +-
newlib/libc/stdio64/Makefile.am | 23 +-
newlib/libc/stdlib/Makefile.am | 24 +-
newlib/libc/string/Makefile.am | 21 +-
newlib/libc/time/Makefile.am | 24 +-
newlib/libm/Makefile.am | 19 +-
newlib/libm/common/Makefile.am | 17 +-
newlib/libm/complex/Makefile.am | 24 +-
newlib/libm/libm.in.xml | 11 +
newlib/libm/math/Makefile.am | 22 +-
newlib/libm/mathfp/Makefile.am | 22 +-
newlib/man.xsl | 13 +
newlib/refcontainers.xslt | 14 +
24 files changed, 1322 insertions(+), 68 deletions(-)
create mode 100644 newlib/doc/.gitignore
create mode 100755 newlib/doc/chapter-texi2docbook.py
create mode 100755 newlib/doc/makedocbook.py
create mode 100644 newlib/libc/libc.in.xml
create mode 100644 newlib/libm/libm.in.xml
create mode 100644 newlib/man.xsl
create mode 100644 newlib/refcontainers.xslt
diff --git a/newlib/doc/.gitignore b/newlib/doc/.gitignore
new file mode 100644
index 0000000..b6fcb73
--- /dev/null
+++ b/newlib/doc/.gitignore
@@ -0,0 +1,3 @@
+# PLY artefacts
+parser.out
+parsetab.py
diff --git a/newlib/doc/chapter-texi2docbook.py b/newlib/doc/chapter-texi2docbook.py
new file mode 100755
index 0000000..1cd1e2e
--- /dev/null
+++ b/newlib/doc/chapter-texi2docbook.py
@@ -0,0 +1,142 @@
+#!/usr/bin/env python
+#
+# python script to convert the handwritten chapter .texi files which include
+# the generated files for each function to equivalent DocBook XML
+#
+# This has very a simple, fixed idea about how those chapter .texi files will be
+# structured
+#
+
+from __future__ import print_function
+from enum import Enum
+import sys
+import re
+
+class State(Enum):
+ normal = 0
+ para = 1
+ table = 2
+ enumerate = 3
+ skip = 4
+
+def main():
+ state = State.normal
+ first_row = True
+ first_node = True
+
+ print ('<?xml version="1.0" encoding="UTF-8"?>')
+ print ('<!DOCTYPE chapter PUBLIC "-//OASIS//DTD DocBook V4.5//EN" "http://www.oasis-open.org/docbook/xml/4.5/docbookx.dtd">')
+
+ for l in sys.stdin.readlines():
+ l = l.rstrip()
+
+ # transform < and >
+ l = l.replace("<", "<")
+ l = l.replace(">", ">")
+
+ # transform @file{foo} to <filename>foo</filename>
+ l = re.sub("@file{(.*?)}", "<filename>\\1</filename>", l)
+ # ditto code
+ l = re.sub("@code{(.*?)}", "<code>\\1</code>", l)
+ # ditto dfn
+ l = re.sub("@dfn{(.*?)}", "<firstterm>\\1</firstterm>", l)
+ # ditto emph
+ l = re.sub("@emph{(.*?)}", "<emphasis>\\1</emphasis>", l)
+ # ditto samp
+ l = re.sub("@samp{(.*?)}", "<literal>\\1</literal>", l)
+ # ditto var
+ l = re.sub("@var{(.*?)}", "<varname>\\1</varname>", l)
+
+ # handle @*
+ l = l.replace("@*", "</para><para>", 1)
+
+ # remove @exdent
+ l = l.replace("@exdent", "", 1)
+ # remove @noindent
+ l = l.replace("@noindent", "", 1)
+
+ if l.startswith("@node"):
+ l = l.replace("@node", "", 1)
+ l = l.strip()
+ l = l.lower()
+ if first_node:
+ print ('<chapter id="%s" xmlns:xi="http://www.w3.org/2001/XInclude">' % l.replace(' ', '_'))
+ first_node = False
+ continue
+ elif l.startswith("@chapter "):
+ l = l.replace("@chapter ", "", 1)
+ print ('<title>%s</title>' % l)
+ print ('<para>')
+ state = State.para
+ continue
+ elif not l:
+ if state == State.para:
+ print ('</para>')
+ print ('<para>')
+ continue
+ elif l.startswith("@table"):
+ # table introduces a 2-column table
+ l = l.replace("@table @code", "<informaltable><tgroup cols='2'><tbody>")
+ state = State.table
+ first_row = True
+ elif l.startswith("@item"):
+ if state == State.table:
+ if not first_row:
+ l = "</entry></row>" + l
+ first_row = False
+ l = l.replace("@item", "<row><entry><code>")
+ l = l + "</code></entry><entry>"
+ elif state == State.enumerate:
+ if not first_row:
+ l = "</para></listitem>" + l
+ first_row = False
+ l = l.replace("@item", "<listitem><para>")
+ elif l.startswith("@end table"):
+ l = l.replace("@end table", "</entry></row></tbody></tgroup></informaltable>")
+ state = State.para
+ elif l.startswith("@enumerate"):
+ if state == State.para:
+ l = "</para>" + l
+ state = State.enumerate
+ l = l.replace("@enumerate", "<orderedlist>")
+ first_row = True
+ elif l.startswith("@end enumerate"):
+ if state == State.enumerate:
+ l = l.replace("@end enumerate", "</para></listitem></orderedlist><para>")
+ state = State.para
+ elif l.startswith("@menu"):
+ if state == State.para:
+ print("</para>")
+ state = State.skip
+ continue
+ elif l.startswith("@endmenu"):
+ state = State.normal
+ continue
+ elif l.startswith("@cindex") or l.startswith("@findex"):
+ l = l.split(None, 1)[1]
+ l = "<indexterm><primary>" + l + "</primary></indexterm>"
+ elif (l.startswith("@page") or l.startswith("@example") or
+ l.startswith("@end example") or l.startswith("@ifset") or
+ l.startswith("@end ifset")):
+ continue
+ elif l.startswith("@include "):
+ l = l.replace("@include ", "", 1)
+ l = l.replace(".def", ".xml", 1)
+ print ('<xi:include href="%s"/>' % l.strip())
+
+ # we can skip over everything between @menu and @endmenu
+ if state == State.skip:
+ continue
+
+ match = re.match('@[a-z]+', l)
+ if match:
+ print ("texinfo command %s ?" % match.group(), file=sys.stderr)
+
+ print (l)
+
+ if state == State.para:
+ print ('</para>')
+ print ('</chapter>')
+
+if __name__ == "__main__" :
+ main()
diff --git a/newlib/doc/makedocbook.py b/newlib/doc/makedocbook.py
new file mode 100755
index 0000000..81d9bd9
--- /dev/null
+++ b/newlib/doc/makedocbook.py
@@ -0,0 +1,792 @@
+#!/usr/bin/env python
+#
+# python script to process makedoc instructions in a source file and produce
+# DocBook XML output
+#
+
+#
+# This performs 3 stages of processing on it's input, in a similar fashion
+# to makedoc:
+#
+# 1. Discard everything outside of /* */ comments
+# 2. Identify lines which contains commands (a single uppercase word)
+# 3. Apply each command to the text of the following lines (up to the next
+# command or the end of the comment block), to produce some output
+#
+# The resulting output contains one or more DocBook XML refentry elements.
+#
+# To make the output a valid XML document which can be xincluded, those refentry
+# elements are contained by a refcontainer element. refcontainer is not part of
+# the DocBook DTD and should be removed by a suitable XSLT.
+#
+
+from __future__ import print_function
+
+import sys
+import re
+from optparse import OptionParser
+import lxml.etree
+import ply.lex as lex
+import ply.yacc as yacc
+
+rootelement = None # root element of the XML tree
+refentry = None # the current refentry
+verbose = 0
+
+def dump(s, stage, threshold = 1):
+ if verbose > threshold:
+ print('*' * 40, file=sys.stderr)
+ print(stage, file=sys.stderr)
+ print('*' * 40, file=sys.stderr)
+ print('%s' % s, file=sys.stderr)
+ print('*' * 40, file=sys.stderr)
+
+#
+# Stage 1
+#
+
+def skip_whitespace_and_stars(i, src):
+
+ while i < len(src) and (src[i].isspace() or (src[i] == '*' and src[i+1] != '/')):
+ i += 1
+
+ return i
+
+# Discard everything not inside '/* */' style-comments which start at column 0
+# Discard any leading blank space or '*'
+# Discard a single leading '.'
+# Discard blank lines after a blank line
+def comment_contents_generator(src):
+ i = 0
+
+ while i < len(src) - 2:
+ if src[i] == '\n' and src[i+1] == '/' and src[i+2] == '*':
+ i = i + 3
+
+ i = skip_whitespace_and_stars(i, src)
+
+ if src[i] == '.':
+ i += 1
+
+ while i < len(src):
+ if src[i] == '\n':
+ yield '\n'
+ i += 1
+
+ # allow a single blank line
+ if i < len(src) and src[i] == '\n':
+ yield '\n'
+ i += 1
+
+ i = skip_whitespace_and_stars(i, src)
+
+ elif src[i] == '*' and src[i+1] == '/':
+ i = i + 2
+ # If we have just output \n\n, this adds another blank line.
+ # This is the only way a double blank line can occur.
+ yield '\nEND\n'
+ break
+ else:
+ yield src[i]
+ i += 1
+ else:
+ i += 1
+
+def remove_noncomments(src):
+ src = '\n' + src
+ dst = ''.join(comment_contents_generator(src))
+ dump(dst, 'extracted from comments')
+
+ return dst
+
+#
+# Stage 2
+#
+
+# A command is a single word, all uppercase, and alone on a line
+#
+# Some 'QUICKREF' lines have trailing text. This is probably a markup error,
+# but accept it.
+def iscommand(l):
+ if re.match('^[A-Z_]+\s*$', l) or l.startswith('QUICKREF'):
+ return True
+ return False
+
+def command_block_generator(content):
+ command = 'START'
+ text = ''
+
+ for l in content.splitlines():
+ if iscommand(l):
+ yield (command, text)
+ command = l.rstrip()
+ text = ''
+ else:
+ text = text + l + '\n'
+ yield (command, text)
+
+# Look for commands, which give instructions how to process the following input
+def process(content):
+ content = content.lstrip()
+
+ dump(content, 'about to process for commands')
+
+ # process into a list of tuples of commands and the associated following text
+ # it is important to maintain the order of the sections the commands generate
+ processed = list(command_block_generator(content))
+
+ return processed
+
+#
+# Stage 3
+#
+
+# invoke each command on it's text
+def perform(processed):
+ for i in processed:
+ c = i[0].rstrip()
+ t = i[1].strip() + '\n'
+
+ if verbose:
+ print("performing command '%s'" % c, file=sys.stderr)
+
+ if c in command_dispatch_dict:
+ command_dispatch_dict[c](c, t)
+ else:
+ print("command '%s' is not recognized" % c, file=sys.stderr)
+ # the text following an unrecognized command is discarded
+
+# FUNCTION (aka TYPEDEF)
+#
+def function(c, l):
+ global refentry
+ global rootelement
+
+ l = l.strip()
+ if verbose:
+ print('FUNCTION %s' % l, file=sys.stderr)
+
+ # Some pages use an en-dash '--' rather than an em-dash '---', so that
+ # should be handled as well, but that is probably a markup error
+ if '---' in l:
+ separator = '---'
+ elif '--' in l:
+ separator = '--'
+
+ if ';' in l:
+ # fpclassify has an unusual format we also need to handle
+ spliton = ';'
+ l = l.splitlines()[0]
+ elif len(l.splitlines()) > 1:
+ # a few pages like mktemp have two '---' lines
+ spliton = ';'
+ o = ''
+ for i in l.splitlines():
+ if separator in i:
+ o += i + ';'
+ else:
+ o += i
+ l = o[:-1]
+ else:
+ spliton = '\n'
+
+ namelist = []
+ descrlist = []
+ for a in l.split(spliton):
+ (n, d) = a.split(separator, 1)
+ namelist = namelist + n.split(',')
+ descrlist = descrlist + [d]
+
+ # only copysign and log1p use <[ ]> markup in descr,
+ # only gets() uses << >> markup
+ # but we should handle it correctly
+ descr = line_markup_convert(', '.join(descrlist))
+
+ # fpclassify includes an 'and' we need to discard
+ namelist = map(lambda v: re.sub('^and ', '', v.strip(), 1), namelist)
+ # strip off << >> surrounding name
+ namelist = map(lambda v: v.strip().lstrip('<').rstrip('>'), namelist)
+
+ if verbose:
+ print(namelist, file=sys.stderr)
+ # additional alternate names may also appear in INDEX commands
+
+ # create the root element if needed
+ if rootelement is None:
+ rootelement = lxml.etree.Element('refentrycontainer')
+
+ # FUNCTION implies starting a new refentry
+ if refentry is not None:
+ print("FUNCTION without NEWPAGE", file=sys.stderr)
+ exit(1)
+
+ # create the refentry
+ refentry = lxml.etree.SubElement(rootelement, 'refentry')
+ refentry.append(lxml.etree.Comment(' Generated by makedocbook.py '))
+ refentry.set('id', namelist[0].lstrip('_'))
+
+ refmeta = lxml.etree.SubElement(refentry, 'refmeta')
+ # refentrytitle will be same as refdescriptor, the primary name
+ refentrytitle = lxml.etree.SubElement(refmeta, 'refentrytitle')
+ refentrytitle.text = namelist[0]
+ manvolnum = lxml.etree.SubElement(refmeta, 'manvolnum')
+ manvolnum.text = '3'
+
+ refnamediv = lxml.etree.SubElement(refentry, 'refnamediv')
+ # refdescriptor is the primary name, assume we should use the one which
+ # appears first in the list
+ refdescriptor = lxml.etree.SubElement(refnamediv, 'refdescriptor')
+ refdescriptor.text = namelist[0]
+ # refname elements exist for all alternate names
+ for n in namelist:
+ refname = lxml.etree.SubElement(refnamediv, 'refname')
+ refname.text = n
+ refpurpose = lxml.etree.SubElement(refnamediv, 'refpurpose')
+ refnamediv.replace(refpurpose, lxml.etree.fromstring('<refpurpose>' + descr + '</refpurpose>'))
+
+ # Only FUNCTION currently exists, which implies that the SYNOPSIS should be
+ # a funcsynopsis. If TYPEDEF was to be added, SYNOPSIS should be processed
+ # in a different way, probably producing a refsynopsis.
+
+# INDEX
+# may occur more than once for each FUNCTION giving alternate names this
+# function should be indexed under
+#
+def index(c, l):
+ l = l.strip()
+
+ if verbose:
+ print('INDEX %s' % l, file=sys.stderr)
+
+ # discard anything after the first word
+ l = l.split()[0]
+
+ # add indexterm
+ # (we could just index under all the refnames, but we control the indexing
+ # separately as that is what makedoc does)
+ indexterm = lxml.etree.SubElement(refentry, 'indexterm')
+ primary = lxml.etree.SubElement(indexterm, 'primary')
+ primary.text = l
+
+ # to validate, it seems we need to maintain refentry elements in a certain order
+ refentry[:] = sorted(refentry, key = lambda x: x.tag)
+
+ # adds another alternate refname
+ refnamediv = refentry.find('refnamediv')
+
+ # as long as it doesn't already exist
+ if not refnamediv.xpath(('refname[.="%s"]') % l):
+ refname = lxml.etree.SubElement(refnamediv, 'refname')
+ refname.text = l
+ if verbose > 1:
+ print('added refname %s' % l, file=sys.stderr)
+ else:
+ if verbose > 1:
+ print('duplicate refname %s discarded' % l, file=sys.stderr)
+
+ # to validate, it seems we need to maintain refnamediv elements in a certain order
+ refnamediv[:] = sorted(refnamediv, key = lambda x: x.tag)
+
+
+# SYNOPSIS aka ANSI_SYNOPSIS
+# ANSI-style synopsis
+#
+# Note that makedoc would also process <<code>> markup here, but there are no
+# such uses.
+#
+def synopsis(c, t):
+ refsynopsisdiv = lxml.etree.SubElement(refentry, 'refsynopsisdiv')
+ funcsynopsis = lxml.etree.SubElement(refsynopsisdiv, 'funcsynopsis')
+
+ s = ''
+ for l in t.splitlines():
+ if re.match('\s*[#[]', l):
+ # a #include, #define etc.
+ # fpclassify contains some comments in [ ] brackets
+ funcsynopsisinfo = lxml.etree.SubElement(funcsynopsis, 'funcsynopsisinfo')
+ funcsynopsisinfo.text = l.strip() + '\n'
+ else:
+ s = s + l
+
+ # some prototypes are missing the terminating ';', so add it if needed
+ if s.endswith(')'):
+ s = s + ';'
+
+ if ';' in s:
+ synopsis_for_prototype(funcsynopsis, s)
+ s = ''
+
+ if s.strip():
+ print("surplus synopsis '%s'" % s, file=sys.stderr)
+ raise
+
+def synopsis_for_prototype(funcsynopsis, s):
+ s = s.strip()
+
+ # funcsynopsis has a very detailed content model, so we need to massage the
+ # bare prototype into it. Fortunately, since the parameter names are marked
+ # up, we have enough information to do this.
+ for fp in s.split(';'):
+ fp = fp.strip()
+ if fp:
+
+ if verbose:
+ print("'%s'" % fp, file=sys.stderr)
+
+ match = re.match(r'(.*?)([\w\d]*) ?\((.*)\)', fp)
+
+ if verbose:
+ print(match.groups(), file=sys.stderr)
+
+ funcprototype = lxml.etree.SubElement(funcsynopsis, 'funcprototype')
+ funcdef = lxml.etree.SubElement(funcprototype, 'funcdef')
+ funcdef.text = match.group(1)
+ function = lxml.etree.SubElement(funcdef, 'function')
+ function.text = match.group(2)
+
+ if match.group(3).strip() == 'void':
+ void = lxml.etree.SubElement(funcprototype, 'void')
+ else:
+ # Split parameters on ',' except if it is inside ()
+ for p in re.split(',(?![^()]*\))', match.group(3)):
+ p = p.strip()
+
+ if verbose:
+ print(p, file=sys.stderr)
+
+ if p == '...':
+ varargs = lxml.etree.SubElement(funcprototype, 'varargs')
+ else:
+ paramdef = lxml.etree.SubElement(funcprototype, 'paramdef')
+ parameter = lxml.etree.SubElement(paramdef, 'parameter')
+
+ # <[ ]> enclose the parameter name
+ match2 = re.match('(.*)<\[(.*)\]>(.*)', p)
+
+ if verbose:
+ print(match2.groups(), file=sys.stderr)
+
+ paramdef.text = match2.group(1)
+ parameter.text = match2.group(2)
+ parameter.tail = match2.group(3)
+
+
+# DESCRIPTION
+# (RETURNS, ERRORS, PORTABILITY, BUGS, WARNINGS, SEEALSO, NOTES are handled the same)
+#
+# Create a refsect with a title corresponding to the command
+#
+# Nearly all the the existing DESCRIPTION contents could be transformed into
+# DocBook with a few regex substitutions. Unfortunately, pages like sprintf and
+# sscanf, have very complex layout using nested tables and itemized lists, which
+# it is best to parse in order to transform correctly.
+#
+
+def refsect(t, s):
+ refsect = lxml.etree.SubElement(refentry, 'refsect1')
+ title = lxml.etree.SubElement(refsect, 'title')
+ title.text = t.title()
+
+ if verbose:
+ print('%s has %d paragraphs' % (t, len(s.split('\n\n'))) , file=sys.stderr)
+
+ if verbose > 1:
+ dump(s, 'before lexing')
+
+ # dump out lexer token sequence
+ lex.input(s)
+ for tok in lexer:
+ print(tok, file=sys.stderr)
+
+ # parse the section text for makedoc markup and the few pieces of texinfo
+ # markup we understand, and output an XML marked-up string
+ xml = parser.parse(s, tracking=True, debug=(verbose > 2))
+
+ dump(xml, 'after parsing')
+
+ xml = '<refsect1>' + xml + '</refsect1>'
+
+ refsect.extend(lxml.etree.fromstring(xml))
+
+def seealso(c, t):
+ refsect('SEE ALSO', t)
+
+# NEWPAGE
+#
+# start a new refentry
+
+def newpage(c, t):
+ global refentry
+ refentry = None
+
+# command dispatch table
+
+def discarded(c, t):
+ return
+
+command_dispatch_dict = {
+ 'FUNCTION' : function,
+ 'TYPEDEF' : function, # TYPEDEF is not currently used, but described in doc.str
+ 'INDEX' : index,
+ 'TRAD_SYNOPSIS' : discarded, # K&R-style synopsis, obsolete and discarded
+ 'ANSI_SYNOPSIS' : synopsis,
+ 'SYNOPSIS' : synopsis,
+ 'DESCRIPTION' : refsect,
+ 'RETURNS' : refsect,
+ 'ERRORS' : refsect,
+ 'PORTABILITY' : refsect,
+ 'BUGS' : refsect,
+ 'WARNINGS' : refsect,
+ 'SEEALSO' : seealso,
+ 'NOTES' : refsect, # NOTES is not described in doc.str, so is currently discarded by makedoc, but that doesn't seem right
+ 'QUICKREF' : discarded, # The intent of QUICKREF and MATHREF is not obvious, but they don't generate any output currently
+ 'MATHREF' : discarded,
+ 'START' : discarded, # a START command is inserted to contain the text before the first command
+ 'END' : discarded, # an END command is inserted merely to terminate the text for the last command in a comment block
+ 'NEWPAGE' : newpage,
+}
+
+#
+# Utility functions
+#
+
+# apply transformations which are easy to do in-place
+def line_markup_convert(p):
+ s = p;
+
+ # process the texinfo escape for an @
+ s = s.replace('@@', '@')
+
+ # escape characters not allowed in XML
+ s = s.replace('&','&')
+ s = s.replace('<','<')
+ s = s.replace('>','>')
+
+ # convert <<somecode>> to <code>somecode</code> and <[var]> to
+ # <varname>var</varname>
+ # also handle nested << <[ ]> >> correctly
+ s = s.replace('<<','<code>')
+ s = s.replace('<[','<varname>')
+ s = s.replace(']>','</varname>')
+ s = s.replace('>>','</code>')
+
+ # also convert some simple texinfo markup
+ # convert @emph{foo} to <emphasis>foo</emphasis>
+ s = re.sub('@emph{(.*?)}', '<emphasis>\\1</emphasis>', s)
+ # convert @minus{} to U+2212 MINUS SIGN
+ s = s.replace('@minus{}', '−')
+ # convert @dots{} to U+2026 HORIZONTAL ELLIPSIS
+ s = s.replace('@dots{}', '…')
+
+ # very hacky way of dealing with @* to force a newline
+ s = s.replace('@*', '</para><para>')
+
+ if (verbose > 3) and (s != p):
+ print('%s-> line_markup_convert ->\n%s' % (p, s), file=sys.stderr)
+
+ return s
+
+#
+# lexer
+#
+
+texinfo_commands = {
+ 'ifnottex' : 'IFNOTTEX',
+ 'end ifnottex' : 'ENDIFNOTTEX',
+ 'tex' : 'IFTEX',
+ 'end tex' : 'ENDIFTEX',
+ 'comment' : 'COMMENT',
+ 'c ' : 'COMMENT',
+ }
+
+# token names
+tokens = [
+ 'BLANKLINE',
+ 'BULLETEND',
+ 'BULLETSTART',
+ 'COURIER',
+ 'EOF',
+ 'ITEM',
+ 'TABLEEND',
+ 'TABLESTART',
+ 'TEXINFO',
+ 'TEXT',
+] + list(set(texinfo_commands.values()))
+
+# regular expression rules for tokens, in priority order
+# (all these expressions should match a whole line)
+def t_TEXINFO(t):
+ # this matches any @command. but not @command{} which just happens to be at
+ # the start of a line
+ r'@\w+[^{]*?\n'
+
+ # if the line starts with a known texinfo command, change t.type to the
+ # token for that command
+ for k in texinfo_commands.keys():
+ if t.value[1:].startswith(k):
+ t.type = texinfo_commands[k]
+ break
+
+ return t
+
+def t_COURIER(t):
+ r'[.|].*\n'
+ t.value = line_markup_convert(t.value[1:])
+ return t
+
+def t_BULLETSTART(t):
+ r'O\+\n'
+ return t
+
+def t_BULLETEND(t):
+ r'O-\n'
+ return t
+
+def t_TABLESTART(t):
+ r'o\+\n'
+ return t
+
+def t_TABLEEND(t):
+ r'o-\n'
+ return t
+
+def t_ITEM(t):
+ r'o\s.*\n'
+ t.value = re.sub('o\s', '', lexer.lexmatch.group(0), 1)
+ t.value = line_markup_convert(t.value)
+ return t
+
+def t_TEXT(t):
+ r'.+\n'
+ t.value = line_markup_convert(t.value)
+ t.lexer.lineno += 1
+ return t
+
+def t_BLANKLINE(t):
+ r'\n'
+ t.lexer.lineno += 1
+ return t
+
+def t_eof(t):
+ if hasattr(t.lexer,'at_eof'):
+ # remove eof flag ready for lexing next input
+ delattr(t.lexer,'at_eof')
+ t.lexer.lineno = 0
+ return None
+
+ t.type = 'EOF'
+ t.lexer.at_eof = True;
+
+ return t
+
+# Error handling rule
+def t_error(t):
+ print("tokenization error, remaining text '%s'" % t.value, file=sys.stderr)
+ exit(1)
+
+lexer = lex.lex()
+
+#
+# parser
+#
+
+def parser_verbose(p):
+ if verbose > 2:
+ print(p[0], file=sys.stderr)
+
+def p_input(p):
+ '''input : paragraph
+ | input paragraph'''
+ if len(p) == 3:
+ p[0] = p[1] + '\n' + p[2]
+ else:
+ p[0] = p[1]
+ parser_verbose(p)
+
+# Strictly, text at top level should be paragraphs (i.e terminated by a
+# BLANKLINE), while text contained in rows or bullets may not be, but this
+# grammar doesn't enforce that for simplicity's sake.
+def p_paragraph(p):
+ '''paragraph : paragraph_content maybe_eof_or_blankline'''
+ p[0] = '<para>\n' + p[1] + '</para>'
+ parser_verbose(p)
+
+def p_paragraph_content(p):
+ '''paragraph_content : paragraph_line
+ | paragraph_line paragraph_content'''
+ if len(p) == 3:
+ p[0] = p[1] + p[2]
+ else:
+ p[0] = p[1]
+ parser_verbose(p)
+
+def p_paragraph_line(p):
+ '''paragraph_line : TEXT
+ | texinfocmd
+ | courierblock
+ | table
+ | bulletlist'''
+ p[0] = p[1]
+
+def p_empty(p):
+ 'empty :'
+ p[0] = ''
+
+def p_maybe_eof_or_blankline(p):
+ '''maybe_eof_or_blankline : empty
+ | EOF
+ | BLANKLINE
+ | BLANKLINE EOF'''
+ p[0] = ''
+
+def p_maybe_lines(p):
+ '''maybe_lines : empty
+ | paragraph maybe_lines'''
+ if len(p) == 3:
+ p[0] = p[1] + p[2]
+ else:
+ p[0] = p[1]
+ parser_verbose(p)
+
+def p_maybe_blankline(p):
+ '''maybe_blankline : empty
+ | BLANKLINE'''
+ p[0] = ''
+
+def p_courierblock(p):
+ '''courierblock : courier'''
+ p[0] = '<literallayout class="monospaced">' + p[1] + '</literallayout>'
+ parser_verbose(p)
+
+def p_courier(p):
+ '''courier : COURIER
+ | COURIER courier'''
+ if len(p) == 3:
+ p[0] = p[1] + p[2]
+ else:
+ p[0] = p[1]
+ parser_verbose(p)
+
+def p_bullet(p):
+ '''bullet : ITEM maybe_lines
+ | ITEM BLANKLINE maybe_lines'''
+ if len(p) == 3:
+ # Glue any text in ITEM into the first para of maybe_lines
+ # (This is an unfortunate consequence of the line-based tokenization we do)
+ if p[2].startswith('<para>'):
+ p[0] = '<listitem><para>' + p[1] + p[2][len('<para>'):] + '</listitem>'
+ else:
+ p[0] = '<listitem><para>' + p[1] + '</para>' + p[2] + '</listitem>'
+ else:
+ p[0] = '<listitem><para>' + p[1] + '</para>' + p[3] + '</listitem>'
+ parser_verbose(p)
+
+def p_bullets(p):
+ '''bullets : bullet
+ | bullet bullets'''
+ if len(p) == 3:
+ p[0] = p[1] + '\n' + p[2]
+ else:
+ p[0] = p[1]
+ parser_verbose(p)
+
+def p_bulletlist(p):
+ '''bulletlist : BULLETSTART bullets BULLETEND maybe_blankline'''
+ p[0] = '<itemizedlist>' + p[2] + '</itemizedlist>'
+ parser_verbose(p)
+
+def p_row(p):
+ '''row : ITEM maybe_lines
+ | ITEM BLANKLINE maybe_lines'''
+ if len(p) == 3:
+ p[0] = '<row><entry><code>' + p[1] + '</code></entry><entry>' + p[2] + '</entry></row>'
+ else:
+ p[0] = '<row><entry><code>' + p[1] + '</code></entry><entry>' + p[3] + '</entry></row>'
+ parser_verbose(p)
+
+def p_rows(p):
+ '''rows : row
+ | row rows'''
+ if len(p) == 3:
+ p[0] = p[1] + '\n' + p[2]
+ else:
+ p[0] = p[1]
+ parser_verbose(p)
+
+def p_table(p):
+ '''table : TABLESTART rows TABLEEND maybe_blankline'''
+ p[0] = '<informaltable><tgroup cols="2"><tbody>' + p[2] + '</tbody></tgroup></informaltable>'
+ parser_verbose(p)
+
+def p_texinfocmd(p):
+ '''texinfocmd : unknown_texinfocmd
+ | comment
+ | nottex
+ | tex'''
+ p[0] = p[1]
+
+def p_unknown_texinfocmd(p):
+ '''unknown_texinfocmd : TEXINFO'''
+ print("unknown texinfo command '%s'" % p[1].strip(), file=sys.stderr)
+ p[0] = p[1]
+ parser_verbose(p)
+
+def p_nottex(p):
+ '''nottex : IFNOTTEX paragraph_content ENDIFNOTTEX'''
+ p[0] = p[2]
+
+def p_tex(p):
+ '''tex : IFTEX paragraph_content ENDIFTEX'''
+ # text for TeX formatter inside @iftex is discarded
+ p[0] = ''
+
+def p_comment(p):
+ '''comment : COMMENT'''
+ # comment text is discarded
+ p[0] = ''
+
+def p_error(t):
+ print('parse error at line %d, token %s, next token %s' % (t.lineno, t, parser.token()), file=sys.stderr)
+ exit(1)
+
+parser = yacc.yacc(start='input')
+
+#
+#
+#
+
+def main(file):
+ content = file.read()
+ content = remove_noncomments(content)
+ processed = process(content)
+ perform(processed)
+
+ # output the XML tree
+ s = lxml.etree.tostring(rootelement, pretty_print=True)
+
+ if not s:
+ print('No output produced (perhaps the input has no makedoc markup?)', file=sys.stderr)
+ exit(1)
+
+ print(s)
+
+ # warn about texinfo commands which didn't get processed
+ match = re.search('@[a-z*]+', s)
+ if match:
+ print('texinfo command %s remains in output' % match.group(), file=sys.stderr)
+
+#
+#
+#
+
+if __name__ == '__main__' :
+ options = OptionParser()
+ options.add_option('-v', '--verbose', action='count', dest = 'verbose')
+ (opts, args) = options.parse_args()
+
+ verbose = opts.verbose
+
+ if len(args) > 0:
+ main(open(args[0], 'rb'))
+ else:
+ main(sys.stdin)
diff --git a/newlib/libc/Makefile.am b/newlib/libc/Makefile.am
index 21a74fe..86c77f4 100644
--- a/newlib/libc/Makefile.am
+++ b/newlib/libc/Makefile.am
@@ -223,6 +223,23 @@ info_TEXINFOS = libc.texinfo
libc_TEXINFOS = sigset.texi extra.texi posix.texi stdio64.texi iconvset.texi \
targetdep.tex $(SUBDEFS)
+docbook: $(SUBDEFS) libc.in.xml
+ xsltproc --xinclude --path ${builddir} --nonet ${srcdir}/../refcontainers.xslt ${srcdir}/libc.in.xml >libc.xml
+ -xmllint --postvalid --noent --nonet --noout libc.xml
+ xmlto --skip-validation html-nochunks --stringparam funcsynopsis.style=ansi libc.xml
+ xmlto --skip-validation pdf --with-dblatex --stringparam funcsynopsis.style=ansi libc.xml
+ xmlto --skip-validation man -m ${srcdir}/../man.xsl libc.xml
+
+doc: docbook
+
+install-data-local:
+ mkdir -p $(DESTDIR)$(htmldir)
+ $(INSTALL_DATA) libc.html $(DESTDIR)$(htmldir)/
+ mkdir -p $(DESTDIR)$(docdir)
+ $(INSTALL_DATA) libc.pdf $(DESTDIR)$(docdir)/
+ mkdir -p $(DESTDIR)$(mandir)/man3
+ $(INSTALL_DATA) *.3 $(DESTDIR)$(mandir)/man3
+
.PHONY: force
force:
@@ -230,7 +247,8 @@ CLEANFILES = $(CRT0) \
sigset.texi stmp-sigset extra.texi stmp-extra \
stdio64.texi stmp-stdio64 targetdep.tex stmp-targetdep \
tmp-sigset.texi tmp-iconvset.texi tmp-extra.texi \
- tmp-stdio64.texi tmp-posix.texi tmp-targetdep.texi
+ tmp-stdio64.texi tmp-posix.texi tmp-targetdep.texi \
+ *.xml libc.hml libc.pdf *.3
ACLOCAL_AMFLAGS = -I .. -I ../..
CONFIG_STATUS_DEPENDENCIES = $(newlib_basedir)/configure.host
diff --git a/newlib/libc/ctype/Makefile.am b/newlib/libc/ctype/Makefile.am
index 785bf7a..03fbe4b 100644
--- a/newlib/libc/ctype/Makefile.am
+++ b/newlib/libc/ctype/Makefile.am
@@ -98,19 +98,33 @@ CHEWOUT_FILES= \
wctrans.def \
wctype.def
-SUFFIXES = .def
+SUFFIXES = .def .xml
CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = ctype.tex
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
- cat $(srcdir)/ctype.tex >> $(TARGETDOC)
+doc-texinfo: $(CHEWOUT_FILES)
+ cat $(srcdir)/$(CHAPTER) >> $(TARGETDOC)
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/$(CHAPTER) >../$(DOCBOOK_CHAPTER)
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
$(lpfx)ctype_.$(oext): ctype_.c ctype_iso.h ctype_cp.h
diff --git a/newlib/libc/libc.in.xml b/newlib/libc/libc.in.xml
new file mode 100644
index 0000000..50d4460
--- /dev/null
+++ b/newlib/libc/libc.in.xml
@@ -0,0 +1,43 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook V4.5//EN" "http://www.oasis-open.org/docbook/xml/4.5/docbookx.dtd">
+<book id="libc" xmlns:xi="http://www.w3.org/2001/XInclude">
+ <bookinfo>
+ <productname>newlib</productname>
+ </bookinfo>
+
+ <!-- XXX: also should put the rest of the content from libc.texinfo here, but I don't need it for just generating manpages -->
+ <xi:include href="stdlib.xml"/>
+ <xi:include href="ctype.xml"/>
+ <xi:include href="stdio.xml"/>
+ <!-- stdio64 is optional -->
+ <xi:include href="stdio64.xml">
+ <xi:fallback/>
+ </xi:include>
+
+ <xi:include href="strings.xml"/>
+ <xi:include href="wcstrings.xml"/>
+ <!-- signals is optional -->
+ <xi:include href="signal.xml">
+ <xi:fallback/>
+ </xi:include>
+
+ <xi:include href="time.xml"/>
+ <xi:include href="locale.xml"/>
+ <!-- reent.tex contains fixed content and nothing seems to include the .def made in reent/ -->
+
+ <xi:include href="misc.xml"/>
+ <!-- posix is optional -->
+ <xi:include href="posix.xml">
+ <xi:fallback/>
+ </xi:include>
+ <!-- XXX: stdarg.h and vararg.h are directly described in libc.texinfo -->
+
+ <!-- iconv is optional -->
+ <xi:include href="iconv.xml">
+ <xi:fallback/>
+ </xi:include>
+
+ <!-- processing should insert index here -->
+ <index/>
+
+</book>
diff --git a/newlib/libc/locale/Makefile.am b/newlib/libc/locale/Makefile.am
index 66cc7ba..545a072 100644
--- a/newlib/libc/locale/Makefile.am
+++ b/newlib/libc/locale/Makefile.am
@@ -44,12 +44,25 @@ CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = locale.tex
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
- cat $(srcdir)/locale.tex >> $(TARGETDOC)
+doc-texinfo: $(CHEWOUT_FILES)
+ cat $(srcdir)/$(CHAPTER) >> $(TARGETDOC)
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
diff --git a/newlib/libc/misc/Makefile.am b/newlib/libc/misc/Makefile.am
index 8ac62a2..6774a13 100644
--- a/newlib/libc/misc/Makefile.am
+++ b/newlib/libc/misc/Makefile.am
@@ -29,11 +29,25 @@ CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = misc.tex
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
- cat $(srcdir)/misc.tex >> $(TARGETDOC)
+doc-texinfo: $(CHEWOUT_FILES)
+ cat $(srcdir)/${CHAPTER} >> $(TARGETDOC)
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
diff --git a/newlib/libc/posix/Makefile.am b/newlib/libc/posix/Makefile.am
index 0056cc1..7dae470 100644
--- a/newlib/libc/posix/Makefile.am
+++ b/newlib/libc/posix/Makefile.am
@@ -61,13 +61,27 @@ CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = posix.tex
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
- cat $(srcdir)/posix.tex >> $(TARGETDOC)
+doc-texinfo: $(CHEWOUT_FILES)
+ cat $(srcdir)/${CHAPTER} >> $(TARGETDOC)
+
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
AM_CFLAGS = -D_GNU_SOURCE
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
diff --git a/newlib/libc/search/Makefile.am b/newlib/libc/search/Makefile.am
index defb6bb..34248ee 100644
--- a/newlib/libc/search/Makefile.am
+++ b/newlib/libc/search/Makefile.am
@@ -62,7 +62,7 @@ lib_a_CFLAGS = $(AM_CFLAGS)
noinst_DATA =
endif # USE_LIBTOOL
-SUFFIXES = .def
+SUFFIXES = .def .xml
CHEWOUT_FILES = \
bsearch.def \
@@ -73,12 +73,23 @@ CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
+doc-texinfo: $(CHEWOUT_FILES)
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
include $(srcdir)/../../Makefile.shared
diff --git a/newlib/libc/signal/Makefile.am b/newlib/libc/signal/Makefile.am
index 7a7e64c..04b2ccc 100644
--- a/newlib/libc/signal/Makefile.am
+++ b/newlib/libc/signal/Makefile.am
@@ -23,17 +23,31 @@ include $(srcdir)/../../Makefile.shared
CHEWOUT_FILES = psignal.def raise.def signal.def
-SUFFIXES = .def
+SUFFIXES = .def .xml
CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = signal.tex
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
- cat $(srcdir)/signal.tex >> $(TARGETDOC)
+doc-texinfo: $(CHEWOUT_FILES)
+ cat $(srcdir)/$(CHAPTER) >> $(TARGETDOC)
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
diff --git a/newlib/libc/stdio/Makefile.am b/newlib/libc/stdio/Makefile.am
index 0a086e7..c179272 100644
--- a/newlib/libc/stdio/Makefile.am
+++ b/newlib/libc/stdio/Makefile.am
@@ -421,20 +421,34 @@ CHEWOUT_FILES = \
vfwprintf.def \
vfwscanf.def
-SUFFIXES = .def
+SUFFIXES = .def .xml
CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = stdio.tex
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
- cat $(srcdir)/stdio.tex >> $(TARGETDOC)
+doc-texinfo: $(CHEWOUT_FILES)
+ cat $(srcdir)/$(CHAPTER) >> $(TARGETDOC)
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
$(lpfx)clearerr.$(oext): local.h
$(lpfx)clearerr_u.$(oext): local.h
diff --git a/newlib/libc/stdio64/Makefile.am b/newlib/libc/stdio64/Makefile.am
index 914cbca..b7cacfd 100644
--- a/newlib/libc/stdio64/Makefile.am
+++ b/newlib/libc/stdio64/Makefile.am
@@ -52,18 +52,31 @@ CHEWOUT_FILES = \
ftello64.def \
tmpfile64.def
-SUFFIXES = .def
+SUFFIXES = .def .xml
CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = stdio64.tex
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
- cat $(srcdir)/stdio64.tex >> $(TARGETDOC)
+doc-texinfo: $(CHEWOUT_FILES)
+ cat $(srcdir)/$(CHAPTER) >> $(TARGETDOC)
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
diff --git a/newlib/libc/stdlib/Makefile.am b/newlib/libc/stdlib/Makefile.am
index c86010a..97ec905 100644
--- a/newlib/libc/stdlib/Makefile.am
+++ b/newlib/libc/stdlib/Makefile.am
@@ -290,20 +290,34 @@ CHEWOUT_FILES= \
wcstombs.def \
wctomb.def
-SUFFIXES = .def
+SUFFIXES = .def .xml
CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = stdlib.tex
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
- cat $(srcdir)/stdlib.tex >> $(TARGETDOC)
+doc-texinfo: $(CHEWOUT_FILES)
+ cat $(srcdir)/$(CHAPTER) >> $(TARGETDOC)
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
$(lpfx)dtoa.$(oext): dtoa.c mprec.h
$(lpfx)ldtoa.$(oext): ldtoa.c mprec.h
diff --git a/newlib/libc/string/Makefile.am b/newlib/libc/string/Makefile.am
index ba49af8..fc4602f 100644
--- a/newlib/libc/string/Makefile.am
+++ b/newlib/libc/string/Makefile.am
@@ -145,18 +145,31 @@ wcswidth.def wcsxfrm.def wcwidth.def wmemchr.def \
wmemcmp.def wmemcpy.def wmemmove.def wmemset.def \
memmem.def memrchr.def rawmemchr.def strchrnul.def
-SUFFIXES = .def
+SUFFIXES = .def .xml
CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
+doc-texinfo: $(CHEWOUT_FILES)
cat $(srcdir)/strings.tex >> $(TARGETDOC)
cat $(srcdir)/wcstrings.tex >> $(TARGETDOC)
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/strings.tex >../strings.xml
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/wcstrings.tex >../wcstrings.xml
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
diff --git a/newlib/libc/time/Makefile.am b/newlib/libc/time/Makefile.am
index 95f088a..39d47bf 100644
--- a/newlib/libc/time/Makefile.am
+++ b/newlib/libc/time/Makefile.am
@@ -60,17 +60,31 @@ CHEWOUT_FILES = \
tzset.def \
wcsftime.def
-SUFFIXES = .def
+SUFFIXES = .def .xml
CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = time.tex
TARGETDOC = ../tmp.texi
-doc: $(CHEWOUT_FILES)
- cat $(srcdir)/time.tex >> $(TARGETDOC)
+doc-texinfo: $(CHEWOUT_FILES)
+ cat $(srcdir)/$(CHAPTER) >> $(TARGETDOC)
-CLEANFILES = $(CHEWOUT_FILES) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(CHEWOUT_FILES:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(CHEWOUT_FILES) *.ref $(DOCBOOK_OUT_FILES)
diff --git a/newlib/libm/Makefile.am b/newlib/libm/Makefile.am
index bc45816..9d03dc2 100644
--- a/newlib/libm/Makefile.am
+++ b/newlib/libm/Makefile.am
@@ -58,10 +58,27 @@ math/stmp-def: stmp-targetdep ; @true
complex/stmp-def: stmp-targetdep ; @true
+docbook: math/stmp-def complex/stmp-def libm.in.xml
+ xsltproc --xinclude --path ${builddir} --nonet ${srcdir}/../refcontainers.xslt ${srcdir}/libm.in.xml >libm.xml
+ -xmllint --xinclude -path ${builddir} --postvalid --noent --nonet --noout libm.xml
+ xmlto --skip-validation --searchpath ${builddir} html-nochunks --stringparam funcsynopsis.style=ansi libm.xml
+ xmlto --skip-validation --searchpath ${builddir} pdf --with-dblatex --stringparam funcsynopsis.style=ansi libm.xml
+ xmlto --skip-validation --searchpath ${builddir} man -m ${srcdir}/../man.xsl libm.xml
+
+doc: docbook
+
+install-data-local:
+ mkdir -p $(DESTDIR)$(htmldir)
+ $(INSTALL_DATA) libm.html $(DESTDIR)$(htmldir)/
+ mkdir -p $(DESTDIR)$(docdir)
+ $(INSTALL_DATA) libm.pdf $(DESTDIR)$(docdir)/
+ mkdir -p $(DESTDIR)$(mandir)/man3
+ $(INSTALL_DATA) *.3 $(DESTDIR)$(mandir)/man3/
+
.PHONY: force
force:
-CLEANFILES = tmp.texi targetdep.tex stmp-targetdep
+CLEANFILES = tmp.texi targetdep.tex stmp-targetdep *.xml libm.hml libm.pdf *.3
ACLOCAL_AMFLAGS = -I .. -I ../..
CONFIG_STATUS_DEPENDENCIES = $(newlib_basedir)/configure.host
diff --git a/newlib/libm/common/Makefile.am b/newlib/libm/common/Makefile.am
index d5e0ef9..6c868d5 100644
--- a/newlib/libm/common/Makefile.am
+++ b/newlib/libm/common/Makefile.am
@@ -69,13 +69,24 @@ CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
TARGETDOC = ../tmp.texi
-doc: $(chobj)
+doc-texinfo: $(chobj)
-CLEANFILES = $(chobj) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(chobj:.def=.xml)
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(chobj) *.ref $(DOCBOOK_OUT_FILES)
# A partial dependency list.
diff --git a/newlib/libm/complex/Makefile.am b/newlib/libm/complex/Makefile.am
index 1c4a502..f66e45e 100644
--- a/newlib/libm/complex/Makefile.am
+++ b/newlib/libm/complex/Makefile.am
@@ -40,20 +40,34 @@ chobj = cabs.def cacos.def cacosh.def carg.def \
csin.def csinh.def csqrt.def ctan.def ctanh.def
-SUFFIXES = .def
+SUFFIXES = .def .xml
CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = complex.tex
TARGETDOC = ../tmp.texi
-doc: $(chobj)
- cat $(srcdir)/complex.tex >> $(TARGETDOC)
+doc-texinfo: $(chobj)
+ cat $(srcdir)/$(CHAPTER) >> $(TARGETDOC)
-CLEANFILES = $(chobj) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(chobj:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(chobj) *.ref $(DOCBOOK_OUT_FILES)
# A partial dependency list.
diff --git a/newlib/libm/libm.in.xml b/newlib/libm/libm.in.xml
new file mode 100644
index 0000000..c0e0506
--- /dev/null
+++ b/newlib/libm/libm.in.xml
@@ -0,0 +1,11 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE book PUBLIC "-//OASIS//DTD DocBook V4.5//EN" "http://www.oasis-open.org/docbook/xml/4.5/docbookx.dtd">
+<book id="libm" xmlns:xi="http://www.w3.org/2001/XInclude">
+ <bookinfo>
+ <productname>newlib</productname>
+ </bookinfo>
+
+ <!-- XXX: also should put the rest of the content from libm.texinfo here, but I don't need it for just generating manpages -->
+ <xi:include href="complex.xml"/>
+ <xi:include href="math.xml"/>
+</book>
diff --git a/newlib/libm/math/Makefile.am b/newlib/libm/math/Makefile.am
index 2616ada..c8fb416 100644
--- a/newlib/libm/math/Makefile.am
+++ b/newlib/libm/math/Makefile.am
@@ -82,14 +82,28 @@ CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = math.tex
TARGETDOC = ../tmp.texi
-doc: $(chobj)
- cat $(srcdir)/math.tex >> $(TARGETDOC)
+doc-texinfo: $(chobj)
+ cat $(srcdir)/$(CHAPTER) >> $(TARGETDOC)
-CLEANFILES = $(chobj) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(chobj:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(chobj) *.ref $(DOCBOOK_OUT_FILES)
# A partial dependency list.
diff --git a/newlib/libm/mathfp/Makefile.am b/newlib/libm/mathfp/Makefile.am
index 0caab07..65a5e27 100644
--- a/newlib/libm/mathfp/Makefile.am
+++ b/newlib/libm/mathfp/Makefile.am
@@ -93,14 +93,28 @@ CHEW = ../../doc/makedoc -f $(srcdir)/../../doc/doc.str
.c.def:
$(CHEW) < $< > $*.def 2> $*.ref
- touch stmp-def
+ @touch stmp-def
+CHAPTER = mathfp.tex
TARGETDOC = ../tmp.texi
-doc: $(chobj)
- cat $(srcdir)/mathfp.tex >> $(TARGETDOC)
+doc-texinfo: $(chobj)
+ cat $(srcdir)/$(CHAPTER) >> $(TARGETDOC)
-CLEANFILES = $(chobj) *.ref
+DOCBOOK_CHEW = ${srcdir}/../../doc/makedocbook.py
+DOCBOOK_OUT_FILES = $(chobj:.def=.xml)
+DOCBOOK_CHAPTER = ${CHAPTER:.tex=.xml}
+
+.c.xml:
+ $(DOCBOOK_CHEW) < $< > $*.xml || ( rm $*.xml && false )
+ @touch stmp-xml
+
+doc-docbook: $(DOCBOOK_OUT_FILES)
+ ${srcdir}/../../doc/chapter-texi2docbook.py <$(srcdir)/${CHAPTER} >../${DOCBOOK_CHAPTER}
+
+doc: doc-docbook doc-texinfo
+
+CLEANFILES = $(chobj) *.ref $(DOCBOOK_OUT_FILES)
# A partial dependency list.
diff --git a/newlib/man.xsl b/newlib/man.xsl
new file mode 100644
index 0000000..a0e2736
--- /dev/null
+++ b/newlib/man.xsl
@@ -0,0 +1,13 @@
+<?xml version='1.0'?>
+<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform" version='1.0'>
+
+<!-- don't truncate long manual names -->
+<xsl:param name="man.th.extra3.max.length" select="45" />
+
+<!-- don't moan about missing metadata -->
+<xsl:param name="refentry.meta.get.quietly" select="1" />
+
+<!-- generate ansi rather than k&r style function synopses -->
+<xsl:param name="funcsynopsis.style" select="ansi" />
+
+</xsl:stylesheet>
diff --git a/newlib/refcontainers.xslt b/newlib/refcontainers.xslt
new file mode 100644
index 0000000..85396f6
--- /dev/null
+++ b/newlib/refcontainers.xslt
@@ -0,0 +1,14 @@
+<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
+ <!-- trivial XSLT which removes the refentrycontainer layer -->
+ <xsl:output method="xml" encoding="UTF-8" indent="yes" doctype-public="-//OASIS//DTD DocBook V4.5//EN" doctype-system="http://www.oasis-open.org/docbook/xml/4.5/docbookx.dtd" />
+ <xsl:strip-space elements="*"/>
+
+ <!-- Whenever you match any node but refentrycontainer or any attribute -->
+ <xsl:template match="node()[not(self::refentrycontainer)]|@*">
+ <!-- Copy the current node -->
+ <xsl:copy>
+ <!-- Including any attributes it has and any child nodes -->
+ <xsl:apply-templates select="node()|@*"/>
+ </xsl:copy>
+ </xsl:template>
+</xsl:stylesheet>
--
2.4.5
More information about the Newlib
mailing list