Files
container.training/slides/markmaker.py
2017-11-05 11:52:27 -08:00

194 lines
6.3 KiB
Python
Executable File

#!/usr/bin/env python
# transforms a YAML manifest into a HTML workshop file
import glob
import logging
import os
import re
import string
import subprocess
import sys
import yaml
logging.basicConfig(level=os.environ.get("LOG_LEVEL", "INFO"))
class InvalidChapter(ValueError):
def __init__(self, chapter):
ValueError.__init__(self, "Invalid chapter: {!r}".format(chapter))
def anchor(title):
title = title.lower().replace(' ', '-')
title = ''.join(c for c in title if c in string.ascii_letters+'-')
return "toc-" + title
def insertslide(markdown, title):
title_position = markdown.find("\n# {}\n".format(title))
slide_position = markdown.rfind("\n---\n", 0, title_position+1)
logging.debug("Inserting title slide at position {}: {}".format(slide_position, title))
before = markdown[:slide_position]
extra_slide = """
---
name: {anchor}
class: title
{title}
.nav[[Back to table of contents](#{toclink})]
.debug[(automatically generated title slide)]
""".format(anchor=anchor(title), title=title, toclink=title2chapter[title])
after = markdown[slide_position:]
return before + extra_slide + after
def flatten(titles):
for title in titles:
if isinstance(title, list):
for t in flatten(title):
yield t
else:
yield title
def generatefromyaml(manifest, filename):
manifest = yaml.load(manifest)
markdown, titles = processchapter(manifest["chapters"], filename)
logging.debug("Found {} titles.".format(len(titles)))
toc = gentoc(titles)
markdown = markdown.replace("@@TOC@@", toc)
for title in flatten(titles):
markdown = insertslide(markdown, title)
exclude = manifest.get("exclude", [])
logging.debug("exclude={!r}".format(exclude))
if not exclude:
logging.warning("'exclude' is empty.")
exclude = ",".join('"{}"'.format(c) for c in exclude)
# Insert build info. This is super hackish.
markdown = markdown.replace(
".debug[",
".debug[{}\n\nThese sides have been built from commit: {}\n\n".format(dirtyfiles, commit),
1)
html = open("workshop.html").read()
html = html.replace("@@MARKDOWN@@", markdown)
html = html.replace("@@EXCLUDE@@", exclude)
html = html.replace("@@CHAT@@", manifest["chat"])
html = html.replace("@@TITLE@@", manifest["title"])
return html
title2chapter = {}
def gentoc(titles, depth=0, chapter=0):
if not titles:
return ""
if isinstance(titles, str):
title2chapter[titles] = "toc-chapter-1"
logging.debug("Chapter {} Title {}".format(chapter, titles))
return " "*(depth-2) + "- [{}](#{})\n".format(titles, anchor(titles))
if isinstance(titles, list):
if depth==0:
sep = "\n\n.debug[(auto-generated TOC)]\n---\n\n"
head = ""
tail = ""
elif depth==1:
sep = "\n"
head = "name: toc-chapter-{}\n\n## Chapter {}\n\n".format(chapter, chapter)
tail = ""
else:
sep = "\n"
head = ""
tail = ""
return head + sep.join(gentoc(t, depth+1, c+1) for (c,t) in enumerate(titles)) + tail
# Arguments:
# - `chapter` is a string; if it has multiple lines, it will be used as
# a markdown fragment; otherwise it will be considered as a file name
# to be recursively loaded and parsed
# - `filename` is the name of the file that we're currently processing
# (to generate inline comments to facilitate edition)
# Returns: (epxandedmarkdown,[list of titles])
# The list of titles can be nested.
def processchapter(chapter, filename):
if isinstance(chapter, unicode):
return processchapter(chapter.encode("utf-8"), filename)
if isinstance(chapter, str):
if "\n" in chapter:
titles = re.findall("^# (.*)", chapter, re.MULTILINE)
slidefooter = ".debug[{}]".format(makelink(filename))
chapter = chapter.replace("\n---\n", "\n{}\n---\n".format(slidefooter))
chapter += "\n" + slidefooter
return (chapter, titles)
if os.path.isfile(chapter):
return processchapter(open(chapter).read(), chapter)
if isinstance(chapter, list):
chapters = [processchapter(c, filename) for c in chapter]
markdown = "\n---\n".join(c[0] for c in chapters)
titles = [t for (m,t) in chapters if t]
return (markdown, titles)
raise InvalidChapter(chapter)
# Try to figure out the URL of the repo on GitHub.
# This is used to generate "edit me on GitHub"-style links.
try:
if "REPOSITORY_URL" in os.environ:
repo = os.environ["REPOSITORY_URL"]
else:
repo = subprocess.check_output(["git", "config", "remote.origin.url"])
repo = repo.strip().replace("git@github.com:", "https://github.com/")
if "BRANCH" in os.environ:
branch = os.environ["BRANCH"]
else:
branch = subprocess.check_output(["git", "rev-parse", "--abbrev-ref", "HEAD"])
base = subprocess.check_output(["git", "rev-parse", "--show-prefix"])
base = base.strip().strip("/")
urltemplate = ("{repo}/tree/{branch}/{base}/{filename}"
.format(repo=repo, branch=branch, base=base, filename="{}"))
except:
logging.exception("Could not generate repository URL; generating local URLs instead.")
urltemplate = "file://{pwd}/{filename}".format(pwd=os.environ["PWD"], filename="{}")
try:
commit = subprocess.check_output(["git", "rev-parse", "--short", "HEAD"])
except:
logging.exception("Could not figure out HEAD commit.")
commit = "??????"
try:
dirtyfiles = subprocess.check_output(["git", "status", "--porcelain"])
except:
logging.exception("Could not figure out repository cleanliness.")
dirtyfiles = "?? git status --porcelain failed"
def makelink(filename):
if os.path.isfile(filename):
url = urltemplate.format(filename)
return "[{}]({})".format(filename, url)
else:
return filename
if len(sys.argv) != 2:
logging.error("This program takes one and only one argument: the YAML file to process.")
else:
filename = sys.argv[1]
if filename == "-":
filename = "<stdin>"
manifest = sys.stdin
else:
manifest = open(filename)
logging.info("Processing {}...".format(filename))
sys.stdout.write(generatefromyaml(manifest, filename))
logging.info("Processed {}.".format(filename))