#!/usr/bin/env python3
"""A simple blog generator with Pandoc as backend."""
import argparse
from contextlib import contextmanager
import datetime
import io
import http.client
import http.server
import multiprocessing
import os
import re
import shutil
import signal
import subprocess
import sys
import tempfile
import time
import urllib.parse
import xml.etree.ElementTree as ET
import bs4
import colorama
import dateutil.parser
import dateutil.tz
ROOTDIR = os.path.dirname(os.path.realpath(__file__))
SOURCEDIR = os.path.join(ROOTDIR, "source")
INDEXMD = os.path.join(SOURCEDIR, "index.md")
TEMPLATEDIR = os.path.join(ROOTDIR, "templates")
HTMLTEMPLATE = os.path.join(TEMPLATEDIR, "template.html")
BUILDDIR = os.path.join(ROOTDIR, "build")
ATOM = os.path.join(BUILDDIR, "atom.xml")
INDEXHTML = os.path.join(BUILDDIR, "index.html")
FEED_MAX_ENTRIES = 20
# Hack ET to support CDATA.
# I know _escape_cdata pops out of nowhere but I won't investigate until
# it breaks.
# XML suuuuuucks.
# http://stackoverflow.com/a/30019607/1944784
def cdata(text=None):
"""Generate an XML CDATA element (ET.Element)."""
element = ET.Element('![CDATA[')
element.text = text
return element
# pylint: disable=protected-access,undefined-variable
ET._original_serialize_xml = ET._serialize_xml
def _serialize_xml(write, elem, qnames, namespaces, short_empty_elements,
**kwargs):
"""Hacked _serialize_xml, tested to work in Python 3.4.3."""
if elem.tag == '![CDATA[':
write("\n<{}{}]]>\n".format(elem.tag, elem.text))
if elem.tail:
write(_escape_cdata(elem.tail))
else:
return ET._original_serialize_xml(write, elem, qnames, namespaces,
short_empty_elements, **kwargs)
ET._serialize_xml = ET._serialize['xml'] = _serialize_xml
# declare the global foreground ANSI codes
BLACK = ""
BLUE = ""
CYAN = ""
GREEN = ""
MAGENTA = ""
RED = ""
WHITE = ""
YELLOW = ""
RESET = ""
@contextmanager
def init_colorama():
"""Set global foreground modifying ANSI codes.
BLACK, BLUE, CYAN, GREEN, MAGENTA, RED, WHITE, YELLOW, and RESET.
"""
# pylint: disable=exec-used,invalid-name
colorama.init()
for color, ansi in colorama.Fore.__dict__.items():
exec("global {0}; {0} = '{1}'".format(color, ansi))
yield
colorama.deinit()
class AtomFeed(object):
"""Class for storing atom:feed date and metadata."""
# pylint: disable=invalid-name,too-many-instance-attributes
def __init__(self):
"""Define available attributes."""
self.author = None # atom:author
self.generator = None # atom:generator, optional
self.icon = None # atom:icon, optional
self.logo = None # atom:logo, optional
self.id_text = None # atom:id, just use URI
self.id = None # atom:id
self.links = [] # list of atom:link
self.title_text = None # the text of atom:title
self.title = None # atom:title
self.updated_datetime = None # update time as a datetime object
self.updated = None # atom:updated
self.entries = [] # list of atom:entry, in reverse time order
self.feed = None # atom:feed, assembled
def assemble_feed(self):
"""Assemble atom:feed."""
self.feed = ET.Element("feed", xmlns="http://www.w3.org/2005/Atom")
self.feed.append(self.title)
for link in self.links:
self.feed.append(link)
self.feed.append(self.updated)
self.feed.append(self.id)
self.feed.append(self.author)
if self.icon is not None:
self.feed.append(self.icon)
if self.logo is not None:
self.feed.append(self.icon)
if self.generator is not None:
self.feed.append(self.generator)
# include at most FEED_MAX_ENTRIES entries in the feed
for entry in self.entries[:FEED_MAX_ENTRIES]:
self.feed.append(entry.entry)
def dump_feed(self):
"""Dump atom:feed XML."""
if self.feed is None:
self.assemble_feed()
return ET.tostring(self.feed).decode('utf-8')
class AtomEntry(object):
"""Class for storing atom:entry data and metadata."""
# pylint: disable=invalid-name,too-many-instance-attributes
def __init__(self):
"""Define available attributes."""
self.author = None # atom:author
self.id_text = None # atom:id, just use URI
self.id = None # atom:id
self.relpath = None # HTML page path relative to home
self.link = None # atom:link
self.title_text = None # plain text title
self.title = None # atom:title
self.updated_datetime = None # update time as a datetime object
self.updated = None # atom:updated
self.content_html = None # content as HTML markup
self.content = None # atom:content
self.entry = None # atom:entry, assembled
def assemble_entry(self):
"""Assemble atom:entry."""
self.entry = ET.Element("entry")
self.entry.append(self.title)
self.entry.append(self.link)
self.entry.append(self.updated)
self.entry.append(self.id)
self.entry.append(self.author)
self.entry.append(self.content)
def dump_entry(self):
"""Dump atom:entry XML."""
if self.entry is None:
self.assemble_entry()
return ET.tostring(self.entry).decode('utf-8')
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
for entry in feed.entries:
date = entry.updated_datetime
if date.year < year:
# write a new
tag with the smaller year
year = date.year
tocbuff.write(u'\n
{0}
\n\n'.format(year))
# write a new
entry (
) in Markdown, in the format:
# *
# [Blah blah](/blog/2015-05-04-blah-blah.html)
monthday = date.strftime("%b %d")
tocbuff.write(u'* [%s](%s)\n' %
(date.isoformat(), monthday, entry.title_text, entry.relpath))
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_index_and_feed():
"""Generate index.html and atom feed."""
sys.stderr.write("generating atom feed\n")
# initialize feed
feed = AtomFeed()
# TODO: Put hard-coded values in a config file
feed.author = ET.fromstring(""
"Zhiming Wang"
"https://github.com/zmwangx"
"zmwangx@gmail.com"
"")
feed.generator = ET.Element("generator", uri="https://github.com/zmwangx/zmwangx.github.io")
feed.generator.text = "pyblog"
feed.icon = ET.Element("icon")
feed.icon.text = "http://zmwangx.github.io/img/icon-400.png"
feed.id_text = "http://zmwangx.github.io"
feed.id = ET.Element("id")
feed.id.text = feed.id_text
feed.links = [
ET.Element("link", href="http://zmwangx.github.io/atom.xml", rel="self"),
ET.Element("link", href="http://zmwangx.github.io/"),
]
feed.title_text = "dl? cmplnts?"
feed.title = ET.fromstring("%s" % feed.title_text)
# update time will be set after everthing finishes
postspath = os.path.join(BUILDDIR, "blog")
# traverse all posts in reverse time order
for name in sorted(os.listdir(postspath), reverse=True):
if re.match(r"^(\d{4})-(\d{2})-(\d{2}).*\.html", name):
htmlpath = os.path.join(postspath, name)
entry = AtomEntry()
with open(htmlpath, encoding="utf-8") as htmlfile:
soup = bs4.BeautifulSoup(htmlfile.read())
entry.author = feed.author # assume it's always the same author
entry.id_text = "%s/blog/%s" % (feed.id_text, name)
entry.id = ET.Element("id")
entry.id.text = entry.id_text
entry.relpath = "/blog/%s" % name
entry.link = ET.Element("link", href=entry.id_text)
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()
# extract the article content without header and footer
article = soup.article
article.header.extract()
article.footer.extract()
entry.content_html = ''.join([str(content)
for content in article.contents])
entry.content = ET.Element("content", type="html")
entry.content.append(cdata(entry.content_html))
entry.assemble_entry()
feed.entries.append(entry)
generate_index(feed)
feed.updated_datetime = datetime.datetime.fromtimestamp(round(time.time()),
dateutil.tz.tzlocal())
feed.updated = ET.Element("updated")
feed.updated.text = feed.updated_datetime.isoformat()
with open(ATOM, 'w', encoding='utf-8') as atom:
atom.write("%s\n" % feed.dump_feed())
sys.stderr.write("wrote atom.xml\n")
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)
failed_builds = 0
template_mtime = os.path.getmtime(HTMLTEMPLATE)
anything_modified = False
for root, _, files in os.walk(SOURCEDIR):
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('.'):
continue
extension = name.split(".")[-1]
if extension not in ["css", "jpg", "md", "png", "svg", "ico"]:
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(template_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", "jpg", "png", "svg", "ico"]:
sys.stderr.write("copying %s\n" % relpath)
shutil.copy(srcpath, dstpath)
elif extension == "md":
sys.stderr.write("generating %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)
if anything_modified:
generate_index_and_feed()
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 new_post(args):
"""Create a new post with metadata pre-filled.
The path to the new post is printed to stdout.
Returns
-------
0
On success.
"""
title = args.title
date = datetime.datetime.fromtimestamp(round(time.time()),
dateutil.tz.tzlocal())
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)
postdir = os.path.join(SOURCEDIR, "blog")
fullpath = os.path.join(postdir, filename)
if not os.path.isdir(postdir):
if os.path.exists(postdir):
os.remove(postdir)
os.mkdir(postdir, mode=0o755)
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")
sys.stderr.write("New post created in:\n")
print(fullpath)
return 0
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.\n")
sys.stderr.write(RESET)
while True:
sys.stderr.write("Continue? [yN] ")
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
else:
sys.stderr.write("Please answer yes or no.\n")
if abort:
sys.stderr.write("%saborting deployment%s\n" % (RED, RESET))
exit(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")
exit(1)
atomxml = ET.parse("atom.xml").getroot()
updated = atomxml.find('{http://www.w3.org/2005/Atom}updated').text
commit_message = ("site updated %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 --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))
exit(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)
exit(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))
exit(1)
return 0
def gen_deploy(args):
"""Regenerate and deploy."""
# pylint: disable=unused-argument
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(("", 8000), 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):
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 main():
"""CLI interface."""
description = "Simple blog generator in Python with Pandoc as backend."
parser = argparse.ArgumentParser(description=description)
subparsers = parser.add_subparsers()
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(
"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)
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(
"preview", aliases=["p", "pre"],
description="Serve the blog locally and auto regenerate upon changes.")
parser_new_post.set_defaults(func=preview)
with init_colorama():
args = parser.parse_args()
returncode = args.func(args)
exit(returncode)
if __name__ == '__main__':
main()