#!/usr/bin/env python3 """A simple blog generator with Pandoc as backend.""" # TODO: auto retouch: prompt for git commit amend after touching # (display commit message to avoid amending the wrong commit) # pylint: disable=too-many-lines import argparse from contextlib import contextmanager import copy import curses import datetime import email.utils import fileinput import io import http.client import http.server import multiprocessing import os import re import shutil import signal import string import subprocess import sys import tempfile import time import urllib.parse import blessed import bs4 import colorama import dateutil.parser import dateutil.tz import lxml.etree as ET from bs4 import UnicodeDammit from pprint import pprint import requests import toml from rss import * from utils import utils ROOTDIR = os.path.dirname(os.path.realpath(__file__)) config_blog = toml.load(os.path.join(ROOTDIR, "config/blog.toml")) config_generator = toml.load(os.path.join(ROOTDIR, "config/generator.toml")) config = {**config_blog, **config_generator} # Blog configuration BLOG_HOME = config["blog"]["BLOG_HOME"] CUSTOM_DOMAIN = config["blog"]["CUSTOM_DOMAIN"] BLOG_TITLE = config["blog"]["BLOG_TITLE"] BLOG_DESCRIPTION = config["blog"]["BLOG_DESCRIPTION"] LANGUAGE = config["blog"]["LANGUAGE"] AUTHOR = config["blog"]["AUTHOR"] AUTHOR_EMAIL = config["blog"]["AUTHOR_EMAIL"] ATOM_ICON_PATH = config["blog"]["ATOM_ICON_PATH"] RSS_ICON_PATH = config["blog"]["RSS_ICON_PATH"] RSS_ICON_WIDTH = config["blog"]["RSS_ICON_WIDTH"] RSS_ICON_HEIGHT = config["blog"]["RSS_ICON_HEIGHT"] #generator configurations GENERATOR_NAME = config["generator"]["GENERATOR_NAME"] GENERATOR_HOME_PAGE = config["generator"]["GENERATOR_HOME_PAGE"] SOURCEDIR = os.path.join(ROOTDIR, config["generator"]["SOURCEDIR"]) POSTSDIR = os.path.join(SOURCEDIR, config["generator"]["POSTSDIR"]) INDEXMD = os.path.join(SOURCEDIR, config["generator"]["INDEXMD"]) GENERATORSOURCE = os.path.join(ROOTDIR, config["generator"]["GENERATORSOURCE"]) HTMLTEMPLATE = os.path.join(SOURCEDIR, config["generator"]["HTMLTEMPLATE"]) BUILDDIR = os.path.join(ROOTDIR, config["generator"]["BUILDDIR"]) ATOM = os.path.join(BUILDDIR, config["generator"]["ATOM"]) RSS = os.path.join(BUILDDIR, config["generator"]["RSS"]) INDEXHTML = os.path.join(BUILDDIR, config["generator"]["INDEXHTML"]) EXCLUDELIST = os.path.join(SOURCEDIR, config["generator"]["EXCLUDELIST"]) FEED_MAX_ENTRIES = config["generator"]["FEED_MAX_ENTRIES"] CODE_LINE_HEIGHT = config["generator"]["CODE_LINE_HEIGHT"] # declare the global foreground ANSI codes BLACK = "" BLUE = "" CYAN = "" GREEN = "" MAGENTA = "" RED = "" WHITE = "" YELLOW = "" RESET = "" def generate_menu(): """Generate menu.""" sys.stderr.write("generating menu\n") fd, tmppath = tempfile.mkstemp() os.close(fd) # Put in a list the pages where the menu will be written html_fileList = [] for root, dirs, files in os.walk(BUILDDIR): for name in files: if name.endswith(".html"): try: html_fileList.append(os.path.join(root.split('build/')[1], name)) except IndexError: html_fileList.append(name) # Generate the string who contain the links of the menu htmly_website_page = "" # Writing the menu in all pages contained in the variable in place of the -- generate menu here -- for html_file in html_fileList: with open(tmppath, 'w', encoding='utf-8') as tmpfile: if os.path.exists("build/"+html_file): with open("build/"+html_file, 'r', encoding='utf-8') as indexmd: lines = indexmd.readlines() with open("build/"+html_file, 'w', encoding='utf-8') as indexmd: for line in lines: indexmd.write(re.sub(r'-- generate menu here --', htmly_website_page, line)) os.remove(tmppath) def generate_table(): """Generate table.""" first_comp = 1 first_pr = 1 tr_class = "odd" documents_fileList = [] documents_fileList.append("/website/bts-sio.html") fd, tmppath = tempfile.mkstemp() os.close(fd) htmly_website_page = "" if os.path.exists(BUILDDIR+"/website/bts-sio.html"): sys.stderr.write("generating table\n") # Put in a list the pages where the menu will be written #for root, dirs, files in os.walk(BUILDDIR+"/website/Documents/Situation2"): # for name in files: # if name.endswith(".html"): # try: # documents_fileList.append(os.path.join(root.split('build')[1], name)) # except IndexError: # documents_fileList.append(name) # Generate the string who contain the links of the menu #htmly_website_page = "" # Writing the menu in all pages contained in the variable in place of the -- generate submenu here -- for document_file in documents_fileList: with open(tmppath, 'w', encoding='utf-8') as tmpfile: if os.path.exists("build"+document_file): with open("build"+document_file, 'r', encoding='utf-8') as indexmd: lines = indexmd.readlines() with open("build"+document_file, 'w', encoding='utf-8') as indexmd: for line in lines: indexmd.write(re.sub(r'
-- table --', '', line))
                with open("build"+document_file, 'r', encoding='utf-8') as indexmd:
                    lines = indexmd.readlines()
                    with open("build"+document_file, 'w', encoding='utf-8') as indexmd:
                        for line in lines:
                            if (re.match('^\$.*', line) and first_pr == 1):
                                line_edited=''
                                indexmd.write(re.sub(r'^\$.*', line_edited, line))
                                first_pr = 0
                                first_comp = 1
                            elif (re.match('^\$.*', line)):
                                if (tr_class == "odd"):
                                    tr_class = "even"
                                else:
                                    tr_class = "odd"
                                line_edited=''
                                indexmd.write(re.sub(r'^\$.*', line_edited, line))
                            else:
                                indexmd.write(line)
                with open("build"+document_file, 'r', encoding='utf-8') as indexmd:
                    lines = indexmd.readlines()
                    with open("build"+document_file, 'w', encoding='utf-8') as indexmd:
                        for line in lines:

                            if (re.match('^    \$.*\$$', line)):
                                indexmd.write(re.sub(r'^    \$.*\$$', "
  • "+line.split("$")[1]+'
  • ', line)) first_comp = 1 elif (re.match('^ \$.*[^\$]$', line)): if first_comp == 1: indexmd.write(re.sub(r'^ \$.*[^\$]$', "", line)) else: indexmd.write(re.sub(r'^ \$.*', "", line)) with open("build"+document_file, 'r', encoding='utf-8') as indexmd: lines = indexmd.readlines() with open("build"+document_file, 'w', encoding='utf-8') as indexmd: for line in lines: indexmd.write(re.sub(r"-- end table --", "
    CompétenceActivitéJustification
    • "+line.split("$")[1]+'
    • ', line)) first_comp = 0 else: indexmd.write(re.sub(r'^ \$.*[^\$]$', "
    • "+line.split("$")[1]+'
    • ', line)) else: indexmd.write(line) with open("build"+document_file, 'r', encoding='utf-8') as indexmd: lines = indexmd.readlines() with open("build"+document_file, 'w', encoding='utf-8') as indexmd: for line in lines: if (re.match('^ \$.*', line)): indexmd.write(re.sub(r'^ \$.*', "
    • "+line.split("$")[1]+"
    • "+line+"
    ", line)) os.remove(tmppath) def generate_blog_list(feed): """"Generate blog list """ sys.stderr.write("generating blog list\n") html_fileList = [] for root, dirs, files in os.walk(BUILDDIR): for name in files: if re.search(r'blog',root): if name.endswith(".html"): try: html_fileList.append(os.path.join(root.split('blog/')[1], name)) except IndexError: html_fileList.append(name) # generate TOC for html_file in html_fileList: div_blog_list = u'
    \n\n' year = 10000 # will be larger than the latest year for quite a while # recall that entries are in reverse chronological order table_opened = False for entry in feed.entries: date = entry.updated_datetime if date.year < year: # close the previous table if there is one if table_opened: div_blog_list += u'\n' # write a new

    tag with the smaller year year = date.year div_blog_list += u'\n

    .:{0}:.

    \n\n'.format(year) div_blog_list += u'\n' table_opened = True # write a new table row entry in Markdown, in the format: # # # # # monthday = date.strftime("%b %d") div_blog_list += (u'' '\n' % (date.isoformat(), monthday, entry.relpath, entry.title_text)) if table_opened: div_blog_list += u'
    [Blah blah](/blog/2015-05-04-blah-blah.html)
    %s
    \n' div_blog_list += u'
    ' fd, tmppath = tempfile.mkstemp() os.close(fd) with open(tmppath, 'w', encoding='utf-8') as tmpfile: if os.path.exists("build/blog/index.html"): with open("build/blog/index.html", 'r', encoding='utf-8') as indexmd: lines = indexmd.readlines() with open("build/blog/index.html", 'w', encoding='utf-8') as indexmd: for line in lines: indexmd.write(re.sub(r'{% generate blog_list here %}', div_blog_list, line)) def generate_notes_list(): """"Generate notes list """ sys.stderr.write("generating notes list\n") html_fileList = [] for root, dirs, files in os.walk(BUILDDIR): for name in files: if re.search(r'notes',root): if name.endswith(".html"): try: html_fileList.append(os.path.join(root.split('notes/')[1], name)) except IndexError: html_fileList.append(name) div_notes_list = u'
    \n\n' year = 10000 # will be larger than the latest year for quite a while # recall that entries are in reverse chronological order table_opened = False for name in list(reversed(sorted(os.listdir(os.path.join(BUILDDIR, "notes"))))): if re.match(r"^[0-9]{4}-[0-9]{2}-[0-9]{2}.*\.html", name): htmlpath = os.path.join(BUILDDIR, "notes", name) #tentry = AtomEntry() #item = RssItem() try: with open(htmlpath, encoding="utf-8") as htmlfile: soup = bs4.BeautifulSoup(htmlfile.read(), "lxml") # generate atom entry #entry.author = copy.deepcopy(feed.author) # assume it's always the same author #entry_url = urllib.parse.urljoin(BLOG_HOME, "blog/%s" % name) #entry.id_text = entry_url #entry.id = ET.Element("id") #entry.id.text = entry_url relpath = "/notes/%s" % name #entry.link = ET.Element("link", href=entry_url) title_text = soup.title.text #entry.title = ET.Element("title", type="html") #entry.title.text = entry.title_text post_date = soup.find("meta", attrs={"name": "date"})["content"] updated_datetime = dateutil.parser.parse(post_date) date = updated_datetime if date.year < year: # close the previous table if there is one if table_opened: div_notes_list += u'\n' # write a new

    tag with the smaller year year = date.year div_notes_list += u'\n

    .:{0}:.

    \n\n'.format(year) div_notes_list += u'\n' table_opened = True # write a new table row entry in Markdown, in the format: # # # # # monthday = date.strftime("%b %d") div_notes_list += (u'' '\n' % (date.isoformat(), monthday, relpath, title_text)) except Exception: sys.stderr.write("error: failed to generate feed entry from %s\n" % name) with open(htmlpath, encoding="utf-8") as htmlfile: sys.stderr.write("dumping HTML:%s\n\n" % htmlfile.read()) raise if table_opened: div_notes_list += u'
    [Blah blah](/blog/2015-05-04-blah-blah.html)
    %s
    \n' div_notes_list += u'
    ' fd, tmppath = tempfile.mkstemp() os.close(fd) with open(tmppath, 'w', encoding='utf-8') as tmpfile: if os.path.exists("build/notes/index.html"): with open("build/notes/index.html", 'r', encoding='utf-8') as indexmd: lines = indexmd.readlines() with open("build/notes/index.html", 'w', encoding='utf-8') as indexmd: for line in lines: indexmd.write(re.sub(r'{% generate notes_list here %}', div_notes_list, line)) def generate_index(feed): """Generate index.html from index.md and a TOC.""" sys.stderr.write("generating index.html\n") # generate TOC tocbuff = io.StringIO() tocbuff.write('
    ') year = 10000 # will be larger than the latest year for quite a while # recall that entries are in reverse chronological order table_opened = False for entry in feed.entries: date = entry.updated_datetime if date.year < year: # close the previous table if there is one if table_opened: tocbuff.write(u'\n') # write a new

    tag with the smaller year year = date.year tocbuff.write(u'\n

    .:{0}:.

    \n\n'.format(year)) tocbuff.write(u'\n') table_opened = True # write a new table row entry in Markdown, in the format: # # # # # monthday = date.strftime("%b %d") tocbuff.write(u'' '\n' % (date.isoformat(), monthday, entry.title_text, entry.relpath)) if table_opened: tocbuff.write(u'
    [Blah blah](/blog/2015-05-04-blah-blah.html)
    [%s](%s)
    \n') tocbuff.write('
    ') # create tempfile with index.md and the TOC concatenated, and generate index.html from that # pylint: disable=invalid-name fd, tmppath = tempfile.mkstemp() os.close(fd) with open(tmppath, 'w', encoding='utf-8') as tmpfile: if os.path.exists(INDEXMD): with open(INDEXMD, 'r', encoding='utf-8') as indexmd: tmpfile.write(u"%s\n\n
    \n\n" % indexmd.read()) tmpfile.write("%s\n" % tocbuff.getvalue()) tocbuff.close() pandoc_args = [ "pandoc", tmppath, "--template", HTMLTEMPLATE, "--highlight-style=pygments", "-o", INDEXHTML, ] try: subprocess.check_call(pandoc_args) except subprocess.CalledProcessError: sys.stderr.write("error: failed to generate index.html\n") os.remove(tmppath) def generate_sitemap(feed): """Generate sitemap.xml.""" sitemap = ET.Element("urlset", xmlns="http://www.sitemaps.org/schemas/sitemap/0.9") # index sitemap.append(utils.make_sitemap_url_element(BLOG_HOME, feed.updated, "daily", 1.0)) # other top level pages for name in os.listdir(BUILDDIR): if (not name.endswith(".html") or name == "index.html" or re.match("google[a-z0-9]+\.html", name)): # exclude Google's site ownership verification file continue link = urllib.parse.urljoin(BLOG_HOME, name) fullpath = os.path.join(BUILDDIR, name) # try to extract updated time updated = None with open(fullpath, encoding="utf-8") as htmlobj: soup = bs4.BeautifulSoup(htmlobj.read(), "lxml") if soup.footer is not None: updated_tag = soup.footer.find(attrs={"class": "updated"}) if updated_tag is not None: updated = dateutil.parser.parse(updated_tag.text) sitemap.append(utils.make_sitemap_url_element(link, updated, "monthly", 0.9)) # blog entries for entry in feed.entries: sitemap.append(utils.make_sitemap_url_element(entry.link, entry.updated, "monthly", 0.9)) sitemappath = os.path.join(BUILDDIR, "sitemap.xml") with open(sitemappath, "w", encoding="utf-8") as sitemapfile: sitemapfile.write('\n%s\n' % ET.tostring(sitemap).decode('utf-8')) sys.stderr.write("wrote sitemap.xml\n") def rewrite_title(): """Override the title of some page for a better render""" sys.stderr.write("Overriding some titles\n") filenames =['build/index.html', 'build/blog/index.html', 'build/notes/index.html'] for root, dirs, files in os.walk(BUILDDIR): for name in files: if re.search(r'website($)',root): if name.endswith(".html"): try: filenames.append("build"+os.path.join(root.split(BUILDDIR)[1], name)) except IndexError: filenames.append(name) if re.search(r'Documents($)',root): if name.endswith(".html"): try: filenames.append("build"+os.path.join(root.split(BUILDDIR)[1], name)) except IndexError: filenames.append(name) if re.search(r'notes($)',root): if name.endswith(".html"): try: filenames.append("build"+os.path.join(root.split(BUILDDIR)[1], name)) except IndexError: filenames.append(name) if re.search(r'blog($)',root): if name.endswith(".html"): try: filenames.append("build"+os.path.join(root.split(BUILDDIR)[1], name)) except IndexError: filenames.append(name) h1_titles_list = [] h1_title = [] h2_titles_list = [] h2_title = [] fd, tmppath = tempfile.mkstemp() os.close(fd) for filename in filenames: soup = bs4.BeautifulSoup(open(filename), "lxml") for myh1 in soup.find_all("h1"): if re.match("^(?!.*article-title).*$", str(myh1)): h1_id = myh1['id'] h1_name = myh1.string h1_title.append(str(myh1)) myh1['class'] = "h1" myh1.string = "" h1_span_left = soup.new_tag("span") h1_span_left['class'] = "left-h1" h1_span_left.string = "█▓▒░" h1_span_title = soup.new_tag("span") h1_span_title['class'] = "title-h1" h1_span_title.string = "「"+h1_name+"」" h1_span_right = soup.new_tag("span") h1_span_right['class'] = "right-h1" h1_span_right.string = "░▒▓█" myh1.string.insert_before(h1_span_left) myh1.span.insert_after(h1_span_right) myh1.span.insert_after(h1_span_title) h1_title.append(myh1) h1_title.append(h1_name) h1_titles_list.append(h1_title) h1_title = [] for myh2 in soup.find_all("h2"): if re.match("^(?!.*blog-index-year-title).*$", str(myh2)): h2_id = myh2['id'] h2_name = myh2.string h2_title.append(str(myh2)) myh2['class'] = "h2" myh2.string = "" h2_span_left = soup.new_tag("span") h2_span_left['class'] = "left-h2" h2_span_left.string = ".: " h2_span_title = soup.new_tag("span") h2_span_title['class'] = "title-h2" h2_span_title.string = h2_name h2_span_right = soup.new_tag("span") h2_span_right['class'] = "right-h2" h2_span_right.string = " :." myh2.string.insert_before(h2_span_left) myh2.span.insert_after(h2_span_right) myh2.span.insert_after(h2_span_title) h2_title.append(myh2) h2_title.append(h2_name) h2_titles_list.append(h2_title) h2_title = [] tested_title_list = [] tested_title = [] for filename in filenames: soup = bs4.BeautifulSoup(open(filename), "lxml") if os.path.exists(filename): with open(filename, 'r', encoding='utf-8') as indexmd: lines = indexmd.readlines() with open(filename, 'w', encoding='utf-8') as indexmd: for line in lines: string = "" for title in h1_titles_list: if re.match(".*"+title[0]+".*", line): string = str(title[1]) for title in h2_titles_list: if re.match(".*"+title[0]+".*", line): string = str(title[1]) if string != "": indexmd.write(re.sub(line, string, line)) else: indexmd.write(line) def generate_index_and_feed(): """Generate index.html and feeds (atom and rss).""" # pylint: disable=too-many-statements,attribute-defined-outside-init,invalid-name sys.stderr.write("generating atom and rss feeds\n") # initialize atom feed feed = AtomFeed() feed.author = ET.fromstring( "" "{author}" "{home}" "{email}" "".format(author=AUTHOR, home=BLOG_HOME, email=AUTHOR_EMAIL)) feed.generator = ET.Element("generator", uri=GENERATOR_HOME_PAGE) feed.generator.text = GENERATOR_NAME if ATOM_ICON_PATH is not None: feed.icon = ET.Element("icon") feed.icon.text = urllib.parse.urljoin(BLOG_HOME, ATOM_ICON_PATH) feed.id_text = BLOG_HOME feed.id = ET.Element("id") feed.id.text = feed.id_text feed.links = [ ET.Element("link", href=urllib.parse.urljoin(BLOG_HOME, "atom.xml"), rel="self", type="application/atom+xml"), ET.Element("link", href=BLOG_HOME, rel="alternate", type="text/html"), ] feed.title_text = BLOG_TITLE feed.title = ET.fromstring("{title}".format(title=BLOG_TITLE)) feed.subtitle_text = BLOG_DESCRIPTION feed.subtitle = ET.fromstring("{subtitle}" .format(subtitle=BLOG_DESCRIPTION)) # initialize rss feed rss = RssFeed() rss.rssurl = urllib.parse.urljoin(BLOG_HOME, "rss.xml") rss.title = ET.Element("title") rss.title.text = BLOG_TITLE rss.link = ET.Element("link") rss.link.text = BLOG_HOME rss.description = ET.Element("description") rss.description.text = BLOG_DESCRIPTION rss.language = ET.Element("language") rss.language.text = LANGUAGE rss.author_text = "{email} ({name})".format(email=AUTHOR_EMAIL, name=AUTHOR) rss.managingEditor = ET.Element("managingEditor") rss.managingEditor.text = rss.author_text rss.webMaster = ET.Element("webMaster") rss.webMaster.text = rss.author_text rss.generator = ET.Element("generator") rss.generator.text = "{generator} ({url})".format(generator=GENERATOR_NAME, url=GENERATOR_HOME_PAGE) rss.image = ET.Element("image") if RSS_ICON_PATH is not None: ET.SubElement(rss.image, "url").text = urllib.parse.urljoin(BLOG_HOME, RSS_ICON_PATH) rss.image.append(copy.deepcopy(rss.title)) rss.image.append(copy.deepcopy(rss.link)) ET.SubElement(rss.image, "width").text = str(RSS_ICON_WIDTH) ET.SubElement(rss.image, "height").text = str(RSS_ICON_HEIGHT) # update times will be set after everthing finishes for name in os.listdir(os.path.join(BUILDDIR, "blog")): if re.match(r"^[0-9]{4}-[0-9]{2}-[0-9]{2}.*\.html", name): htmlpath = os.path.join(BUILDDIR, "blog", name) entry = AtomEntry() item = RssItem() try: with open(htmlpath, encoding="utf-8") as htmlfile: soup = bs4.BeautifulSoup(htmlfile.read(), "lxml") # generate atom entry entry.author = copy.deepcopy(feed.author) # assume it's always the same author entry_url = urllib.parse.urljoin(BLOG_HOME, "blog/%s" % name) entry.id_text = entry_url entry.id = ET.Element("id") entry.id.text = entry_url entry.relpath = "/blog/%s" % name entry.link = ET.Element("link", href=entry_url) entry.title_text = soup.title.text entry.title = ET.Element("title", type="html") entry.title.text = entry.title_text post_date = soup.find("meta", attrs={"name": "date"})["content"] entry.updated_datetime = dateutil.parser.parse(post_date) entry.updated = ET.Element("updated") # pylint: disable=no-member entry.updated.text = entry.updated_datetime.isoformat() # process content tags_to_remove = [] # mark header and footer for removal article = soup.article if article.header is not None: tags_to_remove.append(article.header) # mark line numbers for removal for line_number_span in article.find_all("span", attrs={"class": "line-number"}): tags_to_remove.append(line_number_span) # mark script tags for removal for script_tag in article.find_all("script"): tags_to_remove.append(script_tag) # make internal links absolute utils.absolutify_links(article, entry_url) # remove marked tags for tag in tags_to_remove: tag.extract() entry.content_html = ''.join([str(content) for content in article.contents]) entry.content = ET.Element("content", type="html") entry.content.text = ET.CDATA(entry.content_html) entry.assemble_entry() feed.entries.append(entry) # generate rss item item.title = ET.Element("title") item.title.text = entry.title_text item.link = ET.Element("link") item.link.text = entry_url item.description = ET.Element("description") item.description.text = ET.CDATA(entry.content_html) item.author = ET.Element("author") item.author.text = rss.author_text item.guid = ET.Element("guid", isPermaLink="true") item.guid.text = item.link.text item.timestamp = entry.updated_datetime.timestamp() item.pubDate = ET.Element("pubDate") item.pubDate.text = email.utils.formatdate(item.timestamp, usegmt=True) item.assemble_item() rss.items.append(item) except Exception: sys.stderr.write("error: failed to generate feed entry from %s\n" % name) with open(htmlpath, encoding="utf-8") as htmlfile: sys.stderr.write("dumping HTML:%s\n\n" % htmlfile.read()) raise # sort entries by reverse chronological order feed.entries.sort(key=lambda entry: entry.updated_datetime, reverse=True) rss.items.sort(key=lambda item: item.timestamp, reverse=True) generate_index(feed) generate_menu() generate_table() generate_blog_list(feed) generate_notes_list() rewrite_title() feed.updated_datetime = utils.current_datetime() feed.updated = ET.Element("updated") feed.updated.text = feed.updated_datetime.isoformat() rss.update_timestamp = time.time() rss.pubDate = ET.Element("pubDate") rss.pubDate.text = email.utils.formatdate(rss.update_timestamp, usegmt=True) rss.lastBuildDate = ET.Element("lastBuildDate") rss.lastBuildDate.text = email.utils.formatdate(rss.update_timestamp, usegmt=True) with open(ATOM, "w", encoding="utf-8") as atom: atom.write("%s\n" % feed.dump_feed(FEED_MAX_ENTRIES)) sys.stderr.write("wrote atom.xml\n") with open(RSS, "w", encoding="utf-8") as rssxml: rssxml.write("%s\n" % rss.dump_rss(FEED_MAX_ENTRIES)) sys.stderr.write("wrote rss.xml\n") generate_sitemap(feed) def postprocess_html_file(htmlfilepath): """Perform a series of postprocessing to an HTML file.""" with open(htmlfilepath, "r+", encoding="utf-8") as htmlfileobj: soup = bs4.BeautifulSoup(htmlfileobj.read(), "lxml") # a series of postprocessing (extensible) utils.process_image_sizes(soup) utils.link_img_tags(soup) utils.process_footnote_backlinks(soup) # write back htmlfileobj.seek(0) htmlfileobj.write(str(soup)) htmlfileobj.truncate() def static_vars(**kwargs): def decorate(func): for k in kwargs: setattr(func, k, kwargs[k]) return func return decorate # exclude_list is only inialized once to avoid constant disk IO @static_vars(exclude_list=None) def generate_blog(fresh=False, report_total_errors=True): """Generate the blog in BUILDDIR. Parameters ---------- fresh : bool If True, remove all existing build artifects and start afresh; otherwise, only copy or build new or modified files. Default is False. report_total_errors : bool If True, a line will be printed to stderr at the end of build (assuming the function doesn't raise early) reporting the total number of errors, e.g., "build finished with 0 errors". This is turned on by default, but pass False to turn it off, which will result in a completely silent session if nothing changed. This is useful for auto-regen, for instance. Returns ------- failed_builds : int Number of build failures. """ # pylint: disable=too-many-branches,too-many-locals,too-many-statements if not os.path.isdir(SOURCEDIR): raise OSError("source directory %s does not exist" % SOURCEDIR) if not os.path.exists(HTMLTEMPLATE): raise OSError("HTML template %s not found" % HTMLTEMPLATE) if not os.path.isdir(BUILDDIR): if os.path.exists(BUILDDIR): os.remove(BUILDDIR) os.mkdir(BUILDDIR, mode=0o755) if fresh: for name in os.listdir(BUILDDIR): if name == ".git": continue obj = os.path.join(BUILDDIR, name) if os.path.isdir(obj): shutil.rmtree(obj) else: os.remove(obj) # nojekyll: https://help.github.com/articles/files-that-start-with-an-underscore-are-missing/ if not os.path.exists(os.path.join(BUILDDIR, ".nojekyll")): with open(os.path.join(BUILDDIR, ".nojekyll"), "w") as fileobj: pass if CUSTOM_DOMAIN and not os.path.exists(os.path.join(BUILDDIR, "CNAME")): with open(os.path.join(BUILDDIR, "CNAME"), "w") as fileobj: fileobj.write(CUSTOM_DOMAIN) failed_builds = 0 generator_mtime = os.path.getmtime(GENERATORSOURCE) template_mtime = os.path.getmtime(HTMLTEMPLATE) fundamental_mtime = max(generator_mtime, template_mtime) anything_modified = False exclude_list = generate_blog.exclude_list # get value of static variable if exclude_list is None: try: with open(EXCLUDELIST) as fp: exclude_list = [os.path.abspath(os.path.join(SOURCEDIR, line.rstrip())) for line in list(fp) if line.strip() != "" and not line.startswith('#')] except OSError: exclude_list = [] generate_blog.exclude_list = exclude_list # assign to static variable for the future for root, dirs, files in os.walk(SOURCEDIR): # If root is in exclude list, skip all files and remove all subdirs from traversal list. if root in exclude_list: dirs[:] = [] continue relroot = os.path.relpath(root, start=SOURCEDIR) dstroot = os.path.join(BUILDDIR, relroot) if not os.path.isdir(dstroot): if os.path.exists(dstroot): os.remove(dstroot) os.mkdir(dstroot, mode=0o755) for name in files: if name.startswith('.') or os.path.join(root, name) in exclude_list: continue extension = name.split(".")[-1] if extension not in ["css", "js", "asc", "html", "jpg", "md", "png", "svg", "ico", "txt", "eot", "ttf", "woff", "woff2"]: continue relpath = os.path.join(relroot, name) srcpath = os.path.join(root, name) if extension == "md": dstpath = os.path.join(dstroot, re.sub(r'\.md$', '.html', name)) else: dstpath = os.path.join(dstroot, name) if ((not os.path.exists(dstpath) or os.path.getmtime(dstpath) <= max(fundamental_mtime, os.path.getmtime(srcpath)))): # new post or modified post anything_modified = True if srcpath == INDEXMD: continue # index will be processed separately if extension in ["css", "js", "asc", "html", "jpg", "png", "svg", "ico", "txt", "eot", "ttf", "woff", "woff2"]: sys.stderr.write("copying %s\n" % relpath) shutil.copy(srcpath, dstpath) elif extension == "md": sys.stderr.write("compiling %s\n" % relpath) pandoc_args = [ "pandoc", srcpath, "--template", HTMLTEMPLATE, "--highlight-style=pygments", "-o", dstpath, ] try: subprocess.check_call(pandoc_args) except subprocess.CalledProcessError: failed_builds += 1 sys.stderr.write("error: failed to generate %s" % relpath) # postprocess generated HTML file postprocess_html_file(dstpath) if anything_modified: generate_index_and_feed() sys.stderr.write("done\n") if report_total_errors: sys.stderr.write("build finished with %d errors\n" % failed_builds) return failed_builds def generate(args): """Wrapper for generate_blog(fresh=False).""" # pylint: disable=unused-argument exit(generate_blog(fresh=False)) def regenerate(args): """Wrapper for generate_blog(fresh=True).""" # pylint: disable=unused-argument exit(generate_blog(fresh=True)) def sanitize(string): """Sanitize string (title) for URI consumption.""" if isinstance(string, bytes): string = string.decode('utf-8') # to lowercase string = string.lower() # strip all non-word, non-hyphen and non-whitespace characters string = re.sub(r"[^\w\s-]", "", string) # replace consecutive whitespaces with a single hyphen string = re.sub(r"\s+", "-", string) # percent encode the result return urllib.parse.quote(string) def edit_post_with_editor(path): """Launch text editor to edit post at a given path. Text editor is $VISUAL, then if empty, $EDITOR, then if still empty, vi. """ if "VISUAL" in os.environ: editor = os.environ["VISUAL"] elif "EDITOR" in os.environ: editor = os.environ["EDITOR"] else: editor = "vi" subprocess.call([editor, path]) def new_post(title): """Create a new post with metadata pre-filled. The path to the new post is printed to stdout. Returns ------- 0 On success. """ date = utils.current_datetime() filename_date = date.strftime("%Y-%m-%d") iso_date = date.isoformat() display_date = "%s %d, %d" % (date.strftime("%B"), date.day, date.year) title_sanitized = sanitize(title) filename = "%s-%s.md" % (filename_date, title_sanitized) fullpath = os.path.join(POSTSDIR, filename) if not os.path.isdir(POSTSDIR): if os.path.exists(POSTSDIR): os.remove(POSTSDIR) os.mkdir(POSTSDIR, mode=0o755) if os.path.exists(fullpath): sys.stderr.write("%serror: '%s' already exists, please pick a different title%s\n" % (RED, fullpath, RESET)) return 1 with open(fullpath, 'w', encoding='utf-8') as newpost: newpost.write("---\n") newpost.write('title: "%s"\n' % title) newpost.write("date: %s\n" % iso_date) newpost.write("date_display: %s\n" % display_date) newpost.write("---\n\n") sys.stderr.write("New post created in:\n") print(fullpath) edit_post_with_editor(fullpath) return 0 def new_post_cli(args): """CLI wrapper around new_post.""" new_post(args.title) def touch(filename): """Update the timestamp of a post to the current time.""" filename = os.path.basename(filename) fullpath = os.path.join(POSTSDIR, filename) if not os.path.exists(fullpath): sys.stderr.write("%serror: post %s not found %s\n" % (RED, fullpath, RESET)) return 1 filename_prefix_re = re.compile(r"^[0-9]{4}-[0-9]{2}-[0-9]{2}") if not filename_prefix_re.match(filename): sys.stderr.write(RED) sys.stderr.write("error: post %s is not a valid post\n" % filename) sys.stderr.write("error: the filename of a valid post begins with " "a date in the form xxxx-xx-xx\n") sys.stderr.write(RESET) return 1 # update timestamp in the metadata section of the post whatchanged = io.StringIO() date = utils.current_datetime() iso_date = date.isoformat() display_date = "%s %d, %d" % (date.strftime("%B"), date.day, date.year) filename_date = date.strftime("%Y-%m-%d") with fileinput.input(files=(fullpath), inplace=True) as lines: meta_fences = 0 for line in lines: if line.startswith("---"): meta_fences += 1 sys.stdout.write(line) continue if meta_fences >= 2: # already went past the metadata section sys.stdout.write(line) continue if line.startswith("date: "): updated_line = "date: %s\n" % iso_date sys.stdout.write(updated_line) whatchanged.write("-%s+%s\n" % (line, updated_line)) continue if line.startswith("date_display: "): updated_line = "date_display: %s\n" % display_date sys.stdout.write(updated_line) whatchanged.write("-%s+%s\n" % (line, updated_line)) continue sys.stdout.write(line) sys.stderr.write("\n%schangeset:%s\n\n%s" % (YELLOW, RESET, whatchanged.getvalue())) whatchanged.close() # check if the file needs to be renamed new_filename = filename_prefix_re.sub(filename_date, filename) if new_filename != filename: new_fullpath = os.path.join(POSTSDIR, new_filename) os.rename(fullpath, new_fullpath) sys.stderr.write("%srenamed to %s%s\n" % (YELLOW, new_filename, RESET)) return 0 def touch_cli(args): """CLI wrapper around touch.""" touch(args.filename) def deploy(args): """Deploys build directory to origin/master without regenerating. Returns ------- 0 On success. Exit early with nonzero status otherwise. """ # pylint: disable=unused-argument,too-many-statements # check whether root is dirty os.chdir(ROOTDIR) dirty = subprocess.check_output(["git", "status", "--porcelain"]) if dirty: sys.stderr.write(YELLOW) sys.stderr.write("Project root is dirty.\n") sys.stderr.write("You may want to commit in your changes " "to the source branch, since the SHA and title " "of the latest commit on the source branch will be " "incorporated into the commit message on " "the deployment branch. Type s[hell] on the " "next prompt to open an interactive shell.\n") sys.stderr.write(RESET) while True: sys.stderr.write("Continue? [yNs] ") answer = input() if not answer: # default abort = True break elif answer.startswith(('y', 'Y')): abort = False break elif answer.startswith(('n', 'N')): abort = True break elif answer.startswith(('s', 'S')): shell = (os.environ['SHELL'] if 'SHELL' in os.environ and os.environ['SHELL'] else 'zsh') subprocess.call(shell) stilldirty = subprocess.check_output(["git", "status", "--porcelain"]) if stilldirty: sys.stderr.write(YELLOW) sys.stderr.write("Project root is still dirty.\n") sys.stderr.write(RESET) else: sys.stderr.write("Please answer yes or no.\n") if abort: sys.stderr.write("%saborting deployment%s\n" % (RED, RESET)) return 1 # extract latest commit on the source branch source_commit = subprocess.check_output( ["git", "log", "-1", "--pretty=oneline", "source", "--"]).decode('utf-8').strip() # cd into BUILDDIR and assemble commit message sys.stderr.write("%scommand: cd '%s'%s\n" % (BLUE, BUILDDIR, RESET)) os.chdir(BUILDDIR) # extract updated time from atom.xml if not os.path.exists("atom.xml"): sys.stderr.write("atom.xml not found, cannot deploy\naborting\n") return 1 atomxml = ET.parse("atom.xml").getroot() updated = atomxml.find('{http://www.w3.org/2005/Atom}updated').text commit_message = ("Site updated at %s\n\nsource branch was at:\n%s\n" % (updated, source_commit)) # commit changes in BUILDDIR sys.stderr.write("%scommand: git add --all%s\n" % (BLUE, RESET)) subprocess.check_call(["git", "add", "--all"]) sys.stderr.write("%scommand: git commit --no-verify --gpg-sign --message='%s'%s\n" % (BLUE, commit_message, RESET)) try: subprocess.check_call(["git", "commit", "--gpg-sign", "--message=%s" % commit_message]) except subprocess.CalledProcessError: sys.stderr.write("\n%serror: git commit failed%s\n" % (RED, RESET)) return 1 # check dirty status dirty = subprocess.check_output(["git", "status", "--porcelain"]) if dirty: sys.stderr.write(RED) sys.stderr.write("error: failed to commit all changes; " "build directory still dirty\n") sys.stderr.write("error: please manually inspect what was left out\n") sys.stderr.write(RESET) return 1 # push to origin/master sys.stderr.write("%scommand: git push origin master%s\n" % (BLUE, RESET)) try: subprocess.check_call(["git", "push", "origin", "master"]) except subprocess.CalledProcessError: sys.stderr.write("\n%serror: git push failed%s\n" % (RED, RESET)) return 1 return 0 def gen_deploy(args): """Regenerate and deploy.""" # pylint: disable=unused-argument,too-many-branches # try to smartly determine the latest post, and prompt to touch it current_time = time.time() latest_post = None latest_postdate = 0 latest_mtime = 0 for name in os.listdir(POSTSDIR): matchobj = re.match(r"^([0-9]{4})-([0-9]{2})-([0-9]{2})-.*\.md", name) if not matchobj: continue fullpath = os.path.join(POSTSDIR, name) mtime = os.path.getmtime(fullpath) # get post date from the date metadata field of the post postdate = 0 with open(fullpath) as postobj: for line in postobj: dateregex = r"^date: (\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}-\d{2}:?\d{2})" datematch = re.match(dateregex, line.rstrip()) if datematch: postdate = dateutil.parser.parse(datematch.group(1)).timestamp() break # skip the post if it is dated more than three days ago if current_time - postdate > 3 * 24 * 3600: continue if mtime > latest_mtime: latest_post = name latest_postdate = postdate latest_mtime = mtime # prompt for touching if the latest post determined above was # modified within the last hour but the date registered in the post # isn't within the last ten minutes if ((latest_post is not None and current_time - latest_mtime < 3600 and current_time - latest_postdate > 600)): sys.stderr.write("%sIt appears that %s might be a new post.\n" "Do you want to touch its timestamp?%s\n" % (GREEN, latest_post, RESET)) while True: yesnoquit = input("[ynq]: ") if yesnoquit.startswith(("Y", "y")): yesno = True break elif yesnoquit.startswith(("N", "n")): yesno = False break elif yesnoquit.startswith(("Q", "q")): sys.stderr.write("%saborting gen_deploy%s\n" % (RED, RESET)) return 1 else: sys.stderr.write("Please answer yes, no, or quit.\n") if yesno: sys.stderr.write("%stouching %s%s\n" % (BLUE, latest_post, RESET)) touch(latest_post) sys.stderr.write("\n") generate_blog(fresh=True) deploy(None) class HTTPServerProcess(multiprocessing.Process): """This class can be used to run an HTTP server.""" def __init__(self, rootdir): """Initialize the HTTPServerProcess class. Parameters ---------- rootdir : str The root directory to serve from. """ super().__init__() self.rootdir = rootdir def run(self): """Create an HTTP server and serve forever. Runs on localhost. The default port is 8000; if it is not available, a random port is used instead. """ os.chdir(self.rootdir) # pylint: disable=invalid-name HandlerClass = http.server.SimpleHTTPRequestHandler try: httpd = http.server.HTTPServer(("", 8001), HandlerClass) except OSError: httpd = http.server.HTTPServer(("", 0), HandlerClass) _, portnumber = httpd.socket.getsockname() sys.stderr.write("server serving on http://localhost:%d\n" % portnumber) try: httpd.serve_forever() except KeyboardInterrupt: httpd.shutdown() def preview(args): """Serve the blog and auto regenerate upon changes.""" # pylint: disable=unused-argument server_process = HTTPServerProcess(BUILDDIR) server_process.start() sys.stderr.write("watching for changes\n") sys.stderr.write("send SIGINT to stop\n") # install a SIGINT handler only for this process sigint_raised = False def sigint_mitigator(signum, frame): """Translate SIGINT to setting the sigint_raised flag.""" nonlocal sigint_raised sigint_raised = True signal.signal(signal.SIGINT, sigint_mitigator) # Watch and auto-regen. # No need to actually implement watch separately, since # generate_blog(fresh=False, report_total_errors=False) already # watches for modifications and only regens upon changes, and it is # completely silent when there's no change. while not sigint_raised: generate_blog(fresh=False, report_total_errors=False) time.sleep(0.5) sys.stderr.write("\nSIGINT received, cleaning up...\n") server_process.join() return 0 def list_posts(): """List all posts, with date, title, and path to source file. This function only lists posts that has been built (since it reads metadata from HTML rather than Markdown). Returns ------- posts : list A list of posts, in reverse chronological order, where each element is a tuple of (date, title, path to source file). """ posts = [] for name in os.listdir(os.path.join(BUILDDIR, "blog")): if not re.match(r"^[0-9]{4}-[0-9]{2}-[0-9]{2}.*\.html", name): continue htmlpath = os.path.join(BUILDDIR, "blog", name) entry = AtomEntry() item = RssItem() try: with open(htmlpath, encoding="utf-8") as htmlfile: soup = bs4.BeautifulSoup(htmlfile.read(), "lxml") title = soup.title.text date = dateutil.parser.parse(soup.find("meta", attrs={"name": "date"})["content"]) source_path = os.path.join(POSTSDIR, re.sub(r'.html$', '.md', name)) posts.append((date, title, source_path)) except Exception: sys.stderr.write("error: failed to read metadata from HTML file %s\n" % name) with open(htmlpath, encoding="utf-8") as htmlfile: sys.stderr.write("dumping HTML:%s\n\n" % htmlfile.read()) raise posts.sort(key=lambda post: post[0], reverse=True) return posts class PostSelector: def __init__(self, term, posts): self._term = term self.posts_per_page = term.height - 2 self.pages = [posts[i:i+self.posts_per_page] for i in range(0, len(posts), self.posts_per_page)] self.num_pages = len(self.pages) self.pagepos = 0 self.postpos = 0 self.inserting = False # True if in the middle of inserting a post #, False otherwise term.enter_fullscreen() print(term.clear(), end="") sys.stdout.flush() self.selection = "" self.quit = False self.display_page() def _clear_to_eol(self): term = self._term print(term.clear_eol, end="") sys.stdout.flush() def _print_line(self, line, linenum, highlight=False): term = self._term width = term.width with term.location(0, linenum): if highlight: print(term.reverse(line[:width]), end="") else: print(line[:width], end="") self._clear_to_eol() def _print_post(self, page, pos, highlight=False): if pos >= len(page): # if position out of range, just clear the line self._print_line("", pos + 1, highlight) else: date, title, path = page[pos] line = "%3d: %s %s" % (pos, date.strftime("%m/%d/%y"), title) self._print_line(line, pos + 1, highlight) def display_page(self): term = self._term page = self.pages[self.pagepos] with term.hidden_cursor(): topline = " PAGE %d/%d POST %d" % (self.pagepos + 1, self.num_pages, self.postpos) if self.inserting: topline += term.blink("_") self._print_line(topline, 0, highlight=True) for i in range(self.posts_per_page): self._print_post(page, i) # highlight selected post self._print_post(page, self.postpos, highlight=True) bottomline = " Press h for help." self._print_line(bottomline, term.height - 1, highlight=True) def dispatch(self, key): term = self._term if key in string.digits: # insert if self.inserting: newpostpos = 10 * self.postpos + int(key) if newpostpos < len(self.pages[self.pagepos]): self.postpos = newpostpos else: self.postpos = int(key) self.inserting = True elif key.name == "KEY_DELETE": self.postpos //= 10 self.inserting = True else: self.inserting = False if key.name == "KEY_ENTER": self.selection = self.pages[self.pagepos][self.postpos][2] if key in {"q", "Q"}: self.quit = True elif key.name == "KEY_DOWN" or key in {"n", "N"}: if self.postpos + 1 < len(self.pages[self.pagepos]): self.postpos += 1 elif key.name == "KEY_UP" or key in {"p", "P"}: if self.postpos > 0: self.postpos -= 1 elif key.name == "KEY_RIGHT" or key in {".", ">"}: if self.pagepos + 1 < self.num_pages: self.pagepos += 1 self.postpos = 0 elif key.name == "KEY_LEFT" or key in {",", "<"}: if self.pagepos > 0: self.pagepos -= 1 self.postpos = 0 elif key in {"h", "H"}: print(term.clear_eol, end="") sys.stdout.flush() help_text_lines = [ "Next post: n or ", "Previous post: p or ", "Next page: . or > or ", "Previous page: , or < or ", "Select post: or ", "Select by number: type number as shown (delete or backspace to edit)", "Get help: h", "Quit program: q", ] for i in range(term.height - 1): self._print_line(help_text_lines[i] if i < len(help_text_lines) else "", i) bottomline = " Press any key to continue." self._print_line(bottomline, term.height - 1, highlight=True) with term.raw(): term.inkey() def restore(self): term = self._term term.exit_fullscreen() print(term.clear(), end="") sys.stdout.flush() def select(self): term = self._term try: while True: with term.raw(): self.dispatch(term.inkey()) if self.selection or self.quit: break self.display_page() except Exception: raise finally: self.restore() return self.selection def edit_existing_post(args): selector = PostSelector(blessed.Terminal(), list_posts()) selection = selector.select() if selection: print(selection) edit_post_with_editor(selection) else: return 1 def main(): """CLI interface.""" description = "Simple blog generator in Python with Pandoc as backend." parser = argparse.ArgumentParser(description=description) subparsers = parser.add_subparsers(dest="action") subparsers.required = True parser_new_post = subparsers.add_parser( "new_post", aliases=["n", "new"], description="Create a new post with metadata pre-filled.") parser_new_post.add_argument("title", help="title of the new post") parser_new_post.set_defaults(func=new_post_cli) parser_new_post = subparsers.add_parser( "touch", aliases=["t", "tou"], description="""Touch an existing post, i.e., update its timestamp to current time. Why is this ever useful? Well, the timestamp filled in by new_post is the time of creation, but one might spend several hours after the creation of the file to finish the post. Sometimes the post is even created on one day and finished on another (say created at 11pm and finished at 1am). Therefore, one may want to retouch the timestamp before publishing.""") parser_new_post.add_argument("filename", help="path or basename of the source file, " "e.g., 2015-05-05-new-blog-new-start.md") parser_new_post.set_defaults(func=touch_cli) parser_generate = subparsers.add_parser( "generate", aliases=["g", "gen"], description="Generate new or changed objects.") parser_generate.set_defaults(func=generate) parser_regenerate = subparsers.add_parser( "regenerate", aliases=["r", "regen"], description="Regenerate the entire blog afresh.") parser_regenerate.set_defaults(func=regenerate) parser_new_post = subparsers.add_parser( "preview", aliases=["p", "pre"], description="Serve the blog locally and auto regenerate upon changes.") parser_new_post.set_defaults(func=preview) parser_new_post = subparsers.add_parser( "deploy", aliases=["d", "dep"], description="Deploy build/ to origin/master without regenerating.") parser_new_post.set_defaults(func=deploy) parser_new_post = subparsers.add_parser( "gen_deploy", aliases=["gd", "gendep"], description="Rebuild entire blog and deploy build/ to origin/master.") parser_new_post.set_defaults(func=gen_deploy) parser_new_post = subparsers.add_parser( "edit", aliases=["e", "ed"], description="Bring up post selector to select post for editing.") parser_new_post.set_defaults(func=edit_existing_post) with utils.init_colorama(): args = parser.parse_args() returncode = args.func(args) exit(returncode) if __name__ == '__main__': main()