assemble.py

#!/usr/bin/env python3

import re
item_re = re.compile('^ *\(#([a-zA-z]\w*)\) *$')
# note that this ignores any anchor, the filename (sans suffix) is used as anchor instead:
title_re = re.compile('^### ([^<#]+)')
section_re = re.compile('^### ([a-zA-z]\w*.*)')

def item(name, newflag):
    """ returns tuple of (title, text) """
    text = ""
    with open('items/' + name+'.md', 'r') as itemfile:
        title = None
        blankcount = 0
        for line in itemfile.readlines():
            if title is None:
                match = title_re.match(line)
                if match:
                    title = match.group(1).strip()
                    #text += '\n### {0} <a name="{1}"/></a> ###\n'.format(title, name)
                    text += '\n### {2}{0} <a name="{1}"/></a> \n'.format(title, name, newflag)
                continue
            if line.strip() == "":
                blankcount += 1
            else:
                blankcount = 0
            if blankcount<2: text += line 
        while blankcount<1:
            text += '\n'
            blankcount += 1
    return (title, text)

import subprocess
import time
def is_new(path:str):
    cmd = 'git log --format=%at -1 {0}'.format(path)
    ts = subprocess.check_output(cmd.split())
    if not ts:
        return True
    now = int(time.time())
    if now-int(ts) < (5*3600*24): # if it was added less than 5 days ago, assume it is new
        return True
    return False

import sys
import re
from dateutil.parser import parse
if __name__ == '__main__':

    infilename = sys.argv[1]
    nameparts = re.match('(?P<label>\w+)-(?P<date>\d+).md', infilename).groupdict()

    if nameparts['label'] != 'outline':
        print("Warning: input filename doesn't follow expected pattern of 'outline-20190129.md'")

    #outfilename = 'weekly-email-{0}'.format(infilename.rsplit('-',1)[-1])
    outfilename = 'weekly-email-{0}.md'.format(nameparts['date'])

    #item_re = re.compile('^ *\(#([a-zA-z]\w*)\) *$')
    ## note that the anchor point in the item title will be replaced:
    #title_re = re.compile('^## ([^<#]+)')

    email_title_re = re.compile('^# NERSC Weekly Email, Week of (?P<date>\w+ \d+, \d+)')

    # track which item files were used, and report on unused items:
    items = set()

#
#git log --format=%at -1 laborday.md
    item_text = ""
    with open(infilename,'r') as infile, open(outfilename, 'w') as outfile:
        lines = infile.readlines()
        m = email_title_re.match(lines[0])
        if m and parse(nameparts['date']) != parse(m.group('date')):
            print("Warning: the title date {0} doesn't seem to match the outline date {1}".format(m.group('date'), nameparts['date']))
        sectioncount=0
        for line in lines:
            # look for section headings:
            match = section_re.match(line)
            if match:
                sectioncount+=1
                title = match.group(1)
                item_text += '([back to top](#top))\n\n'
                item_text += '---\n'
                item_text += '## {0} <a name="section{1}"/></a> ##\n'.format(title, sectioncount)
                outfile.write('## [{0}](#section{1}) ##\n'.format(title, sectioncount))
                continue
            match = item_re.match(line)
            if match:
                name = match.group(1)
                newflag = '(NEW) ' if is_new('items/{0}.md'.format(name)) else ''
                title, text = item(name, newflag)
                items.add(name+'.md')
                item_text += text
                outfile.write('- [{2}{0}](#{1})\n'.format(title, name, newflag))
            else:
                outfile.write(line)
        # now add the items:
        outfile.write(item_text)

    # check items/ dir for things we might have missed in the outline:
    from os import listdir
    from os.path import isfile, join
    allitems = set([f for f in listdir('items') if isfile(join('items', f))])
    unused = allitems - items
    if len(unused) != 0:
        print("Warning: the following items were found in items/ but not used:")
        for i in unused:
            print(i)