generate.py

#!/usr/bin/env python3
# -*- coding: UTF-8 -*-
import re
import sys
import os
import shutil
import argparse
import click
import jinja2
from jinja2 import Environment, FileSystemLoader
import http
import http.server
import socketserver
import tempfile
from PIL import Image
from resizeimage import resizeimage
import functools
import io
import traceback

root_path = os.path.join(sys.path[0], '..')
sys.path.insert(1, root_path)
import config

languages = config.languages

def readInfoFile(path):
    d = {}
    reg = re.compile(r'^(.*)=(.*)$')
    with open(path, 'r', encoding='utf-8') as infof:
        for line in infof.readlines():
           res = reg.match(line)
           if res is not None:
               d[res.groups()[0]] = res.groups()[1]
    return d


class Structure(object):
    def __init__(self):
        self.root_page = Page()
        def recurse(path, page):
            page.path = path
            page.readFromInfo(readInfoFile(os.path.join(path, 'info')))
            for x in os.listdir(path):
                newPath = os.path.join(path, x)
                if os.path.isdir(newPath):
                    newPage = Page()
                    page.addChild(newPage)
                    recurse(newPath, newPage)
            page.sortChildren()
        recurse(os.path.join(root_path, 'content/pages/'), self.root_page)

    def getPageByUrl(self, url, lang):
        page = self.root_page
        while len(url) > 0:
            url_part = url[0]
            url = url[1:]
            foundPage = False
            for i in page.children:
                if i.urlPart(lang) == url_part:
                    page = i
                    foundPage = True
                    break
            if not foundPage:
                return None
        return page

    def getPageByPath(self, path_arr, relative_to=None):
        page = relative_to if relative_to is not None else self.root_page 
        while len(path_arr) > 0:
            path_part = path_arr[0]
            path_arr = path_arr[1:]

            if path_part == ".":
                continue

            if path_part == "..":
                page = page.parent
                if page is None:
                    return None
            else:
                foundPage = False
                for i in page.children:
                    if i.pathPart() == path_part:
                        page = i
                        foundPage = True
                        break
                if not foundPage:
                    return None
        return page

    def printMenu(self, lang, filed, current_page=None):
        def visible_policy(page):
            if not page.visible():
                return False

            # Page is child of active
            if page.parent == current_page:
                return True

            # Page is active or indirect parent of active
            current_parent = current_page
            while current_parent is not None:
                if current_parent == page:
                    return True
                current_parent = current_parent.parent

            # Parent of page is indirect parent of active <=> Page is sibling of "active"
            current_parent = current_page
            while current_parent is not None:
                if current_parent == page.parent:
                    return True
                current_parent = current_parent.parent

            return False

        def recurse(page, depth, doRecursion=True):
            if page == current_page:
                print('<li class="active">', file=filed)
            else:
                print("<li>", file=filed)
            page.printLink(filed, lang)
            if doRecursion:
                visible_children = [x for x in page.children if visible_policy(x)]
                if len(visible_children) > 0:
                    print("<ul>", file=filed)
                    for item in visible_children:
                        recurse(item, depth + 1)
                    print("</ul>", file=filed)
            print("</li>", file=filed)
        print('<ul class="menu">', file=filed)
        # Show children of root page on level 0
        recurse(self.root_page, 0, False)
        root_visible_children = [x for x in self.root_page.children if visible_policy(x)]
        for p in root_visible_children:
            recurse(p, 0, True)
        print("</ul>", file=filed)


class Page(object):
    def __init__(self):
        self.parent = None
        self.children = []
        self.info = {}
        self.path = None

    def visible(self):
        return not 'invisible' in self.info or not self.info['invisible']

    def readFromInfo(self, info):
        self.info = info

    def addChild(self, child):
        self.children.append(child)
        child.parent = self

    def sortChildren(self):
        def getKey(child):
            return child.pathPartNum()
        self.children = sorted(self.children, key=getKey)

    def title(self, lang):
        return self.info['title-' + lang]

    def menuTitle(self, lang):
        if 'menutitle-' + lang in self.info:
            return self.info['menutitle-' + lang]
        else:
            return self.info['title-' + lang]

    def urlPart(self, lang):
        return self.info['url-' + lang]

    def pathPart(self):
        splitName = os.path.basename(self.path).split('+', maxsplit=1)
        return splitName[1] if len(splitName) >= 2 else splitName[0]

    def pathPartNum(self):
        splitName = os.path.basename(self.path).split('+', maxsplit=1)
        return int(splitName[0]) if len(splitName) >= 2 else 9999999

    def _url_no_lang(self, lang):
        url = ''
        if self.parent is not None:
            url += self.parent._url_no_lang(lang)
            url += '/'
        url += self.urlPart(lang)
        return url

    def url(self, lang):
        return '/' + lang + self._url_no_lang(lang) + '.html'

    def printLink(self, filed, lang, text=None):
        if text is None:
            text = self.menuTitle(lang)
        print('<a href="' + self.url(lang) + '">' + text + '</a>', file=filed)

    def printLanguages(self, filed):
        global languages
        print("<ul>", file=filed)
        for k, v in languages:
            print('<li>', file=filed)
            linkContent = '<img src="/images/languages/' + k + '.gif" title="' + v + '" alt="' + v + '"/>'
            self.printLink(filed, k, linkContent)
            print('</li>', file=filed)
        print("</ul>", file=filed)


class DirectLoader(jinja2.BaseLoader):
    def get_source(self, environment, template):
        path = template
        if not os.path.exists(path):
            raise jinja2.TemplateNotFound(template)
        mtime = os.path.getmtime(path)
        with open(path, 'r', encoding='utf-8') as f:
            source = "{% extends 'base.html' %}{% block content %}" + f.read() + "{% endblock %}"
        return source, path, lambda: mtime == os.path.getmtime(path)

structure = Structure()


def generate_image(path, width, max_width, out_dir):
    '''
    Generates an image with the specified width and returns its path.
    If max_width is specified and the width of the input image
    is smaller than max_width the image is not modified.
    '''
    print(f"generate_image {path} {width}")
    new_name = os.path.splitext(path)[0] + "-" + str(width) + os.path.splitext(path)[1]
    with open(out_dir + path, 'r+b') as f:
        with Image.open(f) as image:
            if max_width is not None and image.width < max_width:
                return path
            img = resizeimage.resize_width(image, width, validate=False)
            img.save(out_dir + new_name, image.format)
    return new_name


def generate_link(path, lang, page):
    print(f"generate_link {path} (lang={lang}) from {page.url(lang)}")
    params = path.split("/")
    relative_to = None
    if params[0] == ".":
        relative_to = page
    if params[0] == "..":
        relative_to = page
    if params[0] == "":  # path starts with /
        params = params[1:]
    href_page = structure.getPageByPath(params, relative_to)
    if href_page is None:
        raise RuntimeError(f"Link undefined - path {path} does not exist")
        #return jinja2.Undefined(name="href")
    else:
        return href_page.url(lang)


def check_exists(path, out_dir):
    print(f"check_exists {path}")
    if not os.path.exists(out_dir + path):
        raise RuntimeError(f"Path {path} does not exist")
    else:
        return path

def enumerate_folder(path, out_dir):
    print(f"enumerate_folder {path}")
    return sorted(os.listdir("content/static" + path))


def generate_page(page, lang, structure, out_dir, url=None):
    template_path = os.path.join(page.path, lang+'.html')
    page_path = page.url(lang) if url is None else url

    filename = out_dir + page_path

    print(f"generate_page {template_path} to {filename}")

    env = Environment(loader=FileSystemLoader(['templates']))
    loader = DirectLoader()

    env.globals = {
        "href": functools.partial(generate_link, lang=lang, page=page),
        "generate_image": functools.partial(generate_image, out_dir=out_dir),
        "check_exists": functools.partial(check_exists, out_dir=out_dir),
        "enumerate_folder": functools.partial(enumerate_folder, out_dir=out_dir),
        "language_code": lang,
    }
    template = loader.load(env, template_path, env.globals)

    dirname = os.path.dirname(filename)
    if not os.path.exists(dirname):
        os.makedirs(dirname)

    menu = io.StringIO()
    structure.printMenu(lang, menu, page)

    languages = io.StringIO()
    page.printLanguages(languages)

    args = {
        "title": page.title(lang),
        "menu": menu.getvalue(),
        "languages": languages.getvalue()
    }

    with open(filename, "w") as fh:
        fh.write(template.render(**args))


@click.command()
@click.option("--port", default=5000)
def serve(port):

    with tempfile.TemporaryDirectory() as out_dir:
        print(f"Serving pages from {out_dir}")

        # Copy static content to out_dir
        src = 'content/static'
        src_files = os.listdir(src)
        for file_name in src_files:
            full_file_name = os.path.join(src, file_name)
            full_dest_name = os.path.join(out_dir, file_name)
            if (os.path.isfile(full_file_name)):
                shutil.copy(full_file_name, out_dir)
            else:
                shutil.copytree(full_file_name, full_dest_name)


        class Handler(http.server.SimpleHTTPRequestHandler):
            '''
            Handler checks if a request is for a known page.
            If it is it regenerates the page before serving.
            Afterwards it continues serving from the directory as usual
            '''
            def __init__(self, *args, **kwargs):
                super().__init__(*args, directory=out_dir, **kwargs)

            def send_head(self):
                print(self.path)
                if self.path.endswith(".html"):
                    path = self.path[0:-5]
                    print(f"Path without html '{path}'")
                    values = path.split("/")[1:]
                    lang = values[0]
                    url = values[1:]

                    page = structure.getPageByUrl(url, lang)
                    if page is not None:
                        try:
                            #TODO Handle errors and send 500
                            generate_page(page, lang, structure, out_dir)
                        except Exception as e:
                            self.send_error(http.HTTPStatus.INTERNAL_SERVER_ERROR, "Failed to generate page")
                            print(traceback.format_exc())
                            return None
                return super().send_head()


        socketserver.TCPServer.allow_reuse_address = True
        with socketserver.TCPServer(("", port), Handler) as httpd:
            print("serving at port", port)
            httpd.serve_forever()

@click.command()
@click.option("--out_dir", required=True)
def generate(out_dir):
    if out_dir.endswith("/"):
        out_dir = out_dir[0:-1]

    structure = Structure()

    #Delete existing output
    if os.path.exists(out_dir):
        for the_file in os.listdir(out_dir):
            file_path = os.path.join(out_dir, the_file)
            try:
                if os.path.isfile(file_path):
                    os.unlink(file_path)
                elif os.path.isdir(file_path):
                    shutil.rmtree(file_path)
            except Exception as e:
                print(e)

    #Copy static content
    src = os.path.join(root_path, 'content/static')
    src_files = os.listdir(src)
    for file_name in src_files:
        full_file_name = os.path.join(src, file_name)
        full_dest_name = os.path.join(out_dir, file_name)
        if (os.path.isfile(full_file_name)):
            shutil.copy(full_file_name, out_dir)
        else:
            shutil.copytree(full_file_name, full_dest_name)

    #Generate pages
    def recurse_pages(page):
        for lang, langname in languages:
            generate_page(page, lang, structure, out_dir)
        for c in page.children:
            recurse_pages(c)

    recurse_pages(structure.root_page)
    #Index with default language
    generate_page(structure.root_page, config.default_language, structure, out_dir, url='/index.html')

@click.group()
def cli():
    pass

cli.add_command(serve)
cli.add_command(generate)

if __name__ == '__main__':
    cli()