the-screenless-office/screenless/bureau/publications/publications.py

# publications office for reading and writing
from datetime import datetime
import json
import os
import random
import string
import urllib.request, urllib.parse, urllib.error

import code128
import lxml.html
import PIL
from readability import readability

from bureau import Bureau, add_command, add_api


class Publications(Bureau):
    """
    The Publications Office serves as a kind of screenless content management
    system.  Create, update and organize your sites while doing most of the work
    on paper or anything you can photograph.
    """

    name = "Publications Office"
    prefix = "PB"
    version = 0

    def __init__(self):
        Bureau.__init__(self)

        # set up db for published sites
        # TODO: rename this to something less ambiguous
        self.db = os.path.expanduser("~/.screenless/PB.data")
        if not os.path.exists(self.db):
            os.mkdir(self.db)

        # set up urldb for short-codes
        self.urldb = self.dbenv.open_db(b"urldb")
        

    @add_command("new", "Create a new Publication/Site")
    def new_site(self):
        """
        Create a new Publication/Site, set up config and tace a picture from
        the document camera as the index page.  Finally, it will print out
        the main page with commands for working with the site.
        """
        site_dir = os.path.join(self.db, "1")
        site_id = 1
        while os.path.exists(site_dir):
            site_id += 1
            site_dir = os.path.join(self.db, str(site_id))
        os.mkdir(site_dir)

        root_d = {"template": "default", "id": site_id}
        with open(os.path.join(site_dir, "root.json", "w")) as root_json:
            root_json.write(json.dumps(root_d))

        photo = self.send("PX", "photo")["photo"]

        # TODO: come up with a generic set of img form operations for Bureau
        #       should map regions defined with percentages to names
        form_img = PIL.Image.open(photo)
        fx, fy = form_img.size
        title_region = (0, 0, 0.5 * fx, 0.125 * fy)
        title_img = form_img.crop(title_region)
        content_region = (0, 0.125 * fy, fx, fy)
        content_img = form_img.crop(content_region)

    def _update_page(self, site, page):
        pass

    @add_command("r", "Print a web page for reading")
    def print_url(self, data):
        """
        Print out a web page for reading.  The command requires a short-code,
        typically referenced via barcode.  Short-codes refer to full resource
        URIs recorded in the Publications office 'urldb' database.
        """
        shortcode, _ = data.split(".")
        with self.dbenv.begin(db=self.urldb) as txn:
            print("looking up short-code:", shortcode)
            url = txn.get(shortcode.encode('utf-8')).decode()

        if not url:
            print("ERROR: no valid URL in db for short code: ", shortcode)
            return

        # download 
        headers = {'User-Agent': 'Mozilla/5.0'}
        req = urllib.request.Request(url, None, headers)
        urldata = urllib.request.urlopen(req)

        # re-render with readability
        doc = readability.Document(urldata.read(),
                                   url=url)
        timestamp = datetime.now().strftime("Sourced %d %B, %Y at %H:%M")
       
        html = lxml.html.document_fromstring(doc.summary())

        notecount = 0
        # store links then make corresponding svg barcodes
        for link in html.findall(".//a"):
            notecount += 1
            tmpcode = ''.join(random.choice(string.ascii_letters +\
                                            string.digits)\
                              for _ in range(5))

            with self.dbenv.begin(write=True, db=self.urldb) as txn:
                if "href" in link.attrib:
                    txn.put(tmpcode.encode(), link.attrib["href"].encode())

            svg = code128.svg("PBr." + tmpcode)

            footnote = html.makeelement("div")
            footnote.attrib["class"] = "footnote"
            notetext = html.makeelement("div")
            notetext.text = str(notecount) + ". " + link.attrib["href"]
            footnote.append(notetext)
            footnote.append(lxml.html.fromstring(svg.encode()))
            html.append(footnote)

        self.print_full("article.html", title=doc.title(),
                        article=lxml.html.tostring(html).decode("utf-8"),
                        url=url, date=timestamp)


def main():
    pb = Publications()
    pb.run()


if __name__ == "__main__":
    main()