path: root/panfry/document.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
import os
import re
import pandoc
from tempfile import NamedTemporaryFile
from panfry.util import *
from panfry.page import TextFile, Page, TOCItem
from panfry.toc import TOCItem


class Document:

    def __init__(self, path):
        self.src_path = os.path.join(path, 'src')
        self.meta = self.get_meta(self.src_path)
        self.pages = self.get_pages(self.src_path)
        self.full_toc = True
        self.standalone = True
        self.json_toc = False

        self.css_file = 'css/style.css'
        self.assets_dir = 'assets'
        self.pandoc_options = []

    @property
    def set_simple_toc(self):
            self.full_toc = False

    @property
    def set_json_toc(self):
            self.json_toc = True
            self.standalone = False

    def set_pandoc_options(self, options):
            self.pandoc_options = options

    @property
    def set_full_toc(self):
            self.full_toc = True

    def set_templater(self, templater):
        self.templater = templater

    def expand_int_links(self, source):
        '''
        Converts all internal links to full url links
        so that links reference accross the split up html pages

        Arguments:
        source = original markdown source

        returns: markdown source with expanded links
        '''
        re_link1 = re.compile(r'(\[(?P<s>[^\]^]+)\])\((?P<l>#.*)\)', re.S)
        re_link2 = re.compile(r'(\[(?P<s>[^\]^]+)\][^(])', re.S)
        re_link3 = re.compile(r'(\[(?P<s>[^\]^]+)\])\[(?P<h>[^\]].*)\]', re.S)
        source = re_link1.sub(self._get_cross_link, source)
        source = re_link2.sub(self._get_cross_link, source)
        source = re_link3.sub(self._get_cross_link, source)

        return source

    def _get_cross_link(self, m):
        if 'h' in m.groupdict():
            heading = m.group('h')
        else:
            heading = m.group('s')

        for page in self.pages:
            for tocitem in page.toc:
                if tocitem.heading == heading.replace('\n', ' '):
                    return '[%s](%s%s)' % (heading,
                                           page.htmlfile,
                                           tocitem.header_id,
                                           )

        return '[%s]' % ref_txt

    @property
    def title(self):
        if self.title_block.firstline.startswith('%'):
            title = self.title_block.firstline.split('%')[1].strip()
            for line in self.title_block.lines[1:]:
                if line.startswith('%'):
                    break
                title = '%s %s' % (title, line.strip())
            return title
        else:
            return pages[0].title

    def get_pages(self, path):
        pages = []
        filelist = self.toc.lines
        for filename in filelist:
            print("Reading in %s..." % filename)
            if not filename:  # blank line in toc
                print("WARNING: %s contains blank lines" % self.toc.filename)
                continue
            source = read_file(os.path.join(path, filename))
            if source:
                pages.append(Page(filename, source))
            else:
                print("WARNING: Page: %s not found or is empty." % filename)

        return pages

    def get_meta(self, path):
        metafiles = ['TOC', 'TITLE_BLOCK']
        meta = {}
        for filename in metafiles:
            source = read_file(os.path.join(path, filename))
            if source:
                metafile = TextFile(filename, source)
                setattr(self, filename.lower(), metafile)
            else:
                print("!E: %s not found or is empty. Aborting..." % filename)
                exit(1)

        return meta

    def next_page(self, page):
        try:
            idx = self.pages.index(page)
        except:
            return ''
        if idx >= len(self.pages) - 1:
            return ''
        return self.pages[idx+1]

    def prev_page(self, page):
        try:
            idx = self.pages.index(page)
        except:
            return ''
        if idx == 0:
            return ''
        return self.pages[idx-1]

    @property
    def toc_links(self):
        '''
        Returns a list of dictionaries. Each dictionary element contains
        a page title and the html file name.
        '''
        links = []
        for page in self.pages:
            if self.full_toc:
                toc = page.toc
            else:
                toc = [page.toc[0]]

            for tocitem in toc:
                link = '%s%s' % (page.htmlfile, tocitem.header_id)
                links.append(dict(link=unicode(link, "utf8"),
                                  text=unicode(tocitem.heading, "utf8"),
                                  level=tocitem.level,
                                  ))

        return links

    @property
    def toc_json(self):
        lastchild = top = toc = TOCItem(self.title, 1, 'title_block.html')

        curlevel = 0
        for page in self.pages:
            heading = ''
            for line in page.markdown.split('\n'):
                line = line.strip()
                if heading and re.match('[=]{2}', line):
                    level = 1
                elif heading and re.match('[-]{2}', line):
                    level = 2
                elif re.match('#+.+[A-z|0-9]', line):
                    level = len(re.match('#+', line).group())
                    heading = line.split(' ', 1)[1].strip()
                else:
                    heading = line.strip()
                    continue

                if curlevel < level:
                    toc = lastchild
                    curlevel = toc.level
                while toc.parent and curlevel >= level:
                    toc = toc.parent
                    curlevel = toc.level

                lastchild = toc.add_child(heading, level, page.htmlfile)

        return top.json

    @property
    def filename_base(self):
        filename = self.title
        filename = re.sub(r'\s+', ' ', filename)
        filename = re.sub(r'[\n ]', '_', filename)
        filename = re.sub(r'[:,]', '-', filename)
        filename = re.sub(r'_+', '_', filename)
        filename = re.sub(r'-_', '-', filename)
        filename = re.sub(r'-+', '-', filename)

        return filename

    @property
    def pdf_filename(self):
        return self.filename_base + '.pdf'

    @property
    def epub_filename(self):
        return self.filename_base + '.epub'

    def publish_pdf(self, pub_path):
        pdf_path = os.path.join(pub_path, self.pdf_filename)
        src = self.title_block.source
        for page in self.pages:
            src += '\n%s' % page.source

        doc = pandoc.Document()
        doc.markdown = src

        pandoc.set_cwd(os.path.abspath(self.src_path))
        doc.add_argument('latex-engine=xelatex')
        for option in self.pandoc_options:
            doc.add_argument(option)
        doc.to_file(pdf_path)
        pandoc.set_cwd(None)

        return self.pdf_filename

    def publish_epub(self, pub_path):
        epub_path = os.path.join(pub_path, self.epub_filename)
        src = self.title_block.source
        for page in self.pages:
            src += '\n%s' % page.source

        doc = pandoc.Document()
        doc.markdown = src

        print("epub_path: %s" % epub_path)
        pandoc.set_cwd(os.path.abspath(self.src_path))
        for option in self.pandoc_options:
            doc.add_argument(option)
        doc.to_file(epub_path)
        pandoc.set_cwd(None)

        return self.epub_filename

    def publish_assets(self, pub_path):
        if not os.path.isdir(self.assets_dir):
            return
        for asset in os.listdir(self.assets_dir):
            src = os.path.join(self.assets_dir, asset)
            dst = os.path.join(pub_path, asset)
            if os.path.isdir(src):
                copy(src, dst)

    def publish_html(self, pub_path):
        pandoc.set_cwd(None)

        for page in self.pages:
            print("generating %s..." % page.htmlfile)
            template_file = NamedTemporaryFile(mode='w',
                                               suffix='pf.template',
                                               delete=False)
            template = self.templater.page_template(self, page)
            template_file.write(template)
            template_file.close()
            doc = pandoc.Document()
            doc.add_argument('toc')
            doc.add_argument('template=%s' % template_file.name)
            doc.add_argument('css=%s' % self.css_file)

            for option in self.pandoc_options:
                if option == 'number-sections':
                    continue
                doc.add_argument(option)

            if self.standalone:
                markdown = '%s\n' % self.title_block.source
            else:
                markdown = ''

            markdown += self.expand_int_links(page.markdown)
            doc.markdown = markdown
            content = doc.html5
            write_file(os.path.join(pub_path, page.htmlfile),
                       unicode(content, 'utf-8'))

        # If there is not explicit index.html, then link 'index.html'
        # to the toplevel page.
        if self.standalone and not 'index.md' in self.pages:
            src = os.path.join(self.pages[0].htmlfile)
            ref = os.path.join(pub_path, 'index.html')
            if os.path.exists(ref):
                os.remove(ref)
            os.symlink(src, ref)

        if self.json_toc:
            write_file(os.path.join(pub_path, 'toc.json'),
                       unicode(self.toc_json, 'utf-8'))

        ###### Copy any assets to publish directory
        self.publish_assets(pub_path)