# $Id$
# simple markdown to S5 translator (in progress)

import site # for exemaker

import os, re, sys
import htmlentitydefs
import markdown
import zipfile
import shutil

try:
    import xml.etree.cElementTree as ET
except ImportError:
    try:
        import cElementTree as ET
    except ImportError:
        import elementtree.ElementTree as ET

def gettext(elem):
    text = elem.text or ""
    for e in elem:
        text += gettext(e)
        if e.tail:
            text += e.tail
    return text

##
# Unpacks a ZIP file.

def zip_explode(filename, filter):

    zip = zipfile.ZipFile(filename)

    for name in zip.namelist():
        outfile = filter(name)
        if not outfile:
            continue
        dirname = os.path.dirname(outfile)
        if not os.path.isdir(dirname):
            os.makedirs(dirname)
        data = zip.read(name)
        if not data:
            continue # ?
        # print name, "->", outfile
        f = open(outfile, "wb")
        f.write(data)
        f.close()

##
# Slideshow wrapper.

class SlideShow:

    def __init__(self, directory):
        self.directory = directory

    ##
    # Prepares a slideshow directory, using data from an S5 template ZIP.

    def init(self, template_zip="s5-blank.zip", theme_zip=None):

        def filter(zipname):
            # get rid of mac crud
            if zipname[:8] == "__MACOSX":
                return None
            if os.path.basename(zipname) == ".DS_Store":
                return None
            parts = zipname.split("s5-blank/", 1)
            assert len(parts) == 2 and not parts[0], "unexpected zip contents"
            return os.path.join(self.directory, parts[1])

        zip_explode(template_zip, filter)

        def filter(zipname):
            # FIXME: it would of course be nice if the standard styles
            # could just be dropped on top of the ui directory, or be
            # entirely self-contained, but they seem to be somewhere in
            # between.  for now, fake things by stuffing all the files
            # inside ui/default
            if zipname.endswith(".js"):
                return None
            basename = os.path.basename(zipname)
            return os.path.join(self.directory, "ui", "default", basename)

        if theme_zip:
            zip_explode(theme_zip, filter)

    ##
    # Loads and prepares the S5 slideshow template.

    def load_template(self):

        filename = os.path.join(self.directory, "s5-blank.html")

        self.slideshow = ET.parse(filename)

        # clean up namespace
        NS_XHTML = "{http://www.w3.org/1999/xhtml}"
        for elem in self.slideshow.getiterator():
            if elem.tag.startswith(NS_XHTML):
                elem.tag = elem.tag[len(NS_XHTML):]

        # locate template elements
        self.elem_title = self.elem_footer = self.elem_presentation = None
        for elem in self.slideshow.getiterator():
            if elem.tag == "title":
                self.elem_title = elem
                elem.text = self.directory
            if elem.tag == "div":
                if elem.get("id") == "footer":
                    self.elem_footer = elem
                    del elem[:]
                elif elem.get("class") == "presentation":
                    self.elem_presentation = elem
                    del elem[:]

    ##
    # Sets the slideshow title.

    def set_title(self, title):
        self.elem_title.text = title

    ##
    # Sets the slideshow footer.

    def set_footer(self, location, title):
        del self.elem_footer[:]
        ET.SubElement(self.elem_footer, "h1").text = location
        ET.SubElement(self.elem_footer, "h2").text = title

    ##
    # Adds a slide to the slideshow.

    def add_slide(self, slide, handout=None):
        div = ET.SubElement(self.elem_presentation, "div")
        div.set("class", "slide")
        div[:] = slide[:]
        if handout:
            div.append(handout)

    ##
    # Writes the slideshow to disk.

    def save(self, filename=None):

        # fixup the element tree

        special_div_ids = ("controls", "currentSlide", "header")

        for elem in self.slideshow.getiterator():

            # handle INCREMENTAL lists
            if elem.tag in ("ul", "ol") and len(elem):
                if gettext(elem[0]).strip() == "INCREMENTAL":
                    del elem[0]
                    elem.set("class", "incremental")

            # some empty elements cannot use <tag /> syntax.  force
            # <tag></tag> syntax by adding a space to those elements.
            if (elem.tag == "div" and elem.get("id") in special_div_ids
                or elem.tag == "script"):
                elem.text = elem.text or " "

        # generate the output file

        # there's no simple way to make ET output the doctype header and
        # use a default namespace, so we have to do that outselves.

        filename = os.path.join(self.directory, filename or "index.html")

        out = open(filename, "w")

        print >>out, """
        <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
          "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
        <html xmlns="http://www.w3.org/1999/xhtml">
        """.strip()
        for elem in self.slideshow.getroot():
            ET.ElementTree(elem).write(out)
        print >>out, "</html>"

        out.close()

        print out.name, "ok"

def make_slideshow(filename, theme_zip=None):

    name = os.path.splitext(os.path.basename(filename))[0]

    text = open(filename).read()

    try:
        text = unicode(text, "utf-8")
    except UnicodeError:
        text = unicode(text, "iso-8859-1")

    def fixup(m):
        try:
            return unichr(htmlentitydefs.name2codepoint[m.group(1)])
        except KeyError:
            return "&amp;" + m.group(0) # ?

    text = re.sub("&(\w+);", fixup, text)

    text = text.encode("ascii", "xmlcharrefreplace")

    slideshow_dir = name + ".s5"

    slideshow = SlideShow(slideshow_dir)

    slideshow.init(theme_zip=theme_zip)

    # do we need tidy here?
    content = ET.XML("<body>" + markdown.markdown(text) + "</body>")

    slideshow.load_template()

    # look for variable block
    if content and content[0].tag == "p":
        elem = content[0]
        if elem.text and elem.text[0] == "$":
            del content[0]
            title = footer1 = footer2 = None
            for line in elem.text.split("\n"):
                line = line.strip()
                if line.startswith("$TITLE"):
                    title = line[6:]
                elif line.startswith("$FOOTER1"):
                    footer1 = line[8:]
                elif line.startswith("$FOOTER2"):
                    footer2 = line[8:]
                elif line:
                    print "---", "unknown variable:", line
            if title:
                slideshow.set_title(title)
            if footer1 or footer2:
                slideshow.set_footer(footer1, footer2)

    images = {}
    for elem in content.findall(".//img"):
        src = elem.get("src")
        if src:
            basename = os.path.basename(src)
            images[os.path.join(os.path.dirname(filename), src)] = basename
            elem.set("src", basename)

    # split on H1 tags
    slide = ET.Element("div")
    for elem in content:
        if elem.tag == "h1":
            if slide:
                slideshow.add_slide(slide)
                slide = ET.Element("div")
        slide.append(elem)
    if slide:
        slideshow.add_slide(slide)

    slideshow.save()

    # copy images
    for source, target in images.items():
        target = os.path.join(slideshow.directory, target)
        shutil.copy(source, target)
        print target, "ok"

if __name__ == "__main__":
    try:
        file = sys.argv[1]
    except IndexError:
        print "usage: s5make.py input.txt [theme.zip]"
        sys.exit()
    try:
        theme = sys.argv[2]
    except IndexError:
        theme = None
    make_slideshow(file, theme)