hgh/hgh.py

#!/usr/bin/env python3

import argparse, os, time, re, subprocess, json, sys

config_json = os.path.expanduser("~/.config/hgh/hgh.json")

parser = argparse.ArgumentParser(description='Handles hugo stuff..')

group0 = parser.add_mutually_exclusive_group()
#group1 = parser.add_mutually_exclusive_group()
group2 = parser.add_mutually_exclusive_group()
group3 = parser.add_mutually_exclusive_group()

group0.add_argument("--site", help="Name of site to work with. Use --sites for a list of all available sites.", dest='site_name', action="store", type=str, default = None)
group0.add_argument("-p", "--printconfig", help="Print complete json config.", dest='printconfig', action="store_true")
group0.add_argument("--sites", help="Print all sites from from configuration", dest='sites', action="store_true")

group2.add_argument("--create-comments-file", help="Create a comments.json in current working directory with given amount of comments.", dest='create_comments_file', action="store_true")
group2.add_argument("-cpb", "--create-pagebundle", help="Create new page bundle. Add --template for specific template to use.", dest='create_pagebundle', action="store", type=str, default = None)
group2.add_argument("--time", help="Copy current date and time to clipboard.", dest='time', action="store_true")
group2.add_argument("-cdb", help="Change directory to base directory of site.", dest='cdb', action="store_true")
group2.add_argument("-cdc", help="Change directory to contentdir of site.", dest='cdc', action="store_true")
group2.add_argument("-cdl", help="Change directory to alternative contentdir of site.", dest='cdl', action="store_true")
group2.add_argument("--build", help="Build a site.", dest='site_build', action="store_true")
group2.add_argument("--upload", help="Build AND upload site to remote.", dest='site_upload', action="store_true")
group2.add_argument("--deploy", help="Deploy (build AND upload) a site.", dest='deploy', action="store_true")
group2.add_argument("--tmux", help="Start the tmux session for --site.", dest='tmux', action="store_true")
group2.add_argument("-hs", "--hugo-server", help="Serves site (full). Use --hugo-server-environment for different environments.", dest='hugo_server', action="store_true")
group2.add_argument("-hsl", "--hugo-server-light", help="Serves site (light).", dest='hugo_server_light', action="store_true")
group2.add_argument("-ctcc", "--create-tags-categories-caches", help="Serves site (light).", dest='create_tags_categories_caches', action="store_true")

#group3.add_argument("--deprecated_searchpost", help="Find pages/posts containing search term in 'path, slug, title, date or permalink'. Strings and Regular Expressions (regex) are allowed.", dest='searchpost', action="store", type=str, default = None)
group3.add_argument("-hse", "--hugo-server-env", help="Use this environment for hugo server. Default on hugo and here is 'development'. Full render is normally 'production'.", dest='hugo_server_environment', action="store", type=str, default = None)
group3.add_argument("-f", "--find", help="Find pages/posts containing search term in file content. Uses 'grep -r'. Strings and regex allowed.", dest='find_anything', action="store", type=str, default = None)
group3.add_argument("-ff", "--find-fuzzy", help="Fuzzy find pages/posts containing search term. Press F4 to edit a selected file with \"bin_editor\". Strings and regex allowed.", dest='find_anything_fuzzy', action="store", type=str, default = None)
group3.add_argument("-ft", "--find-tags", help="Find tags containing search term. Strings and regex are allowed.", dest='find_tags', action="store", type=str, default = None)
group3.add_argument("-fc", "--find-categories", help="Find categories containing search term. Strings and regex are allowed.", dest='find_categories', action="store", type=str, default = None)
group3.add_argument("--template", help="Use template for --create-pagebundle command.", dest='template', action="store", type=str, default = None)
group3.add_argument("--list-templates", help="List available templates.", dest='list_templates', action="store_true")

def get_config(filename):
    with open(filename, "r") as config_file:
        config = json.load(config_file)

    return(config)

def title_to_slug(title):
    slug_max_length=config["settings"]["slug_max_length"]

    slug=title.replace(" ", "-")
    slug=slug.lower()
    if len(slug) > slug_max_length:
        slug=slug[:slug_max_length].rstrip()

    special_char_map = {ord('ä'):'ae', ord('ü'):'ue', ord('ö'):'oe', ord('ß'):'ss'}
    slug = slug.translate(special_char_map)

    slug = re.sub('[^0-9a-zA-Z-]+', '', slug)

    return(slug)

def get_datetime():
    import datetime, pytz

    datetime = pytz.timezone('Europe/Berlin').localize(datetime.datetime.now()).strftime("%Y-%m-%dT%H:%M:%S%z")
    datetime = "{0}:{1}".format(
        datetime[:-2],
        datetime[-2:]
    ) # https://gist.github.com/mattstibbs/a283f55a124d2de1b1d732daaac1f7f8

    return(datetime)

def hugo_server(environment="development"):
    if args.hugo_server_environment:
        environment = args.hugo_server_environment

    print("Starting hugo-full for {site} (port:{port})\n".format(site=args.site_name, port=config["sites"][args.site_name]["serve_port"]))
    subprocess.run(["hugo", "server", "-e", environment, "--logLevel", "info", "--cleanDestinationDir", "--renderToDisk", "--disableFastRender", "-E", "-D", "-F", "-p", str(config["sites"][args.site_name]["serve_port"]) ], cwd=config["sites"][args.site_name]["basedir"])

    # ohne --disableFastRender gibt es bei meinem Theme im Blog das Problem, dass CSS komplett fehlt. Wer das nicht hat, kann das entfernen.

def hugo_server_light():
    print("Starting hugo-light for {site} (port:{port})\n".format(site=args.site_name, port=config["sites"][args.site_name]["serve_port"]))
    subprocess.run(["hugo", "server", "--logLevel",  "info", "--cleanDestinationDir", "--renderToDisk", "-E", "-D", "-F", "-c", config["sites"][args.site_name]["contentdir-light"], "-p", str(config["sites"][args.site_name]["serve_port"]) ], cwd=config["sites"][args.site_name]["basedir"])

def start_tmux():
    # Attach to a already running session named --site or create a new one.
    process = subprocess.run(["/usr/bin/tmux", "has-session", "-t", args.site_name], cwd=config["sites"][args.site_name]["basedir"])
    if process.returncode == 0:
        # tmux already exists, attach
        print("Attaching to tmux for {site}".format(site=args.site_name))
        subprocess.run(["/usr/bin/tmux", "attach", "-t", args.site_name], cwd=config["sites"][args.site_name]["basedir"])
    else:
        # Create a new session
        print("Starting tmux for {site}".format(site=args.site_name))
        os.chdir(config["sites"][args.site_name]["basedir"])
        os.system("/usr/bin/tmux new-session -s '{site_name}' -n 'Full' -c '{cwd}' \; send-keys '{editor} $(fzf)' \; split-window -v -p 5 \; new-window -n 'Light' -c '{cwd_light}' \; send-keys '{editor} $(fzf)' \; split-window -v -p 5 \; new-window -n 'Serve' \; send-keys 'hg.py --site {site_name} -sl' \; select-window -t 1 \; select-pane -t 0 \;".format(editor=config["settings"]["bin_editor"], site_name=args.site_name, cwd=config["sites"][args.site_name]["basedir"], cwd_light=config["sites"][args.site_name]["contentdir-light"]))
        # first window: basedir 
        # second window: contentdir-light
        # third window: serve with prefilled command prompt        

def site_build():
    print("Building {site}".format(site=args.site_name))
    subprocess.run(["hugo", "--minify", "--cleanDestinationDir" ], cwd=config["sites"][args.site_name]["basedir"])

# def searchpost(search): # hugo list all zeigt nur beiträge an, die auch im aktuellen zustand gebaut würden. Bei development bei mir also nur die Beiträge von 2023, findet also nicht alles.
#     process = subprocess.run(["hugo", "list", "all" ], cwd=config["sites"][args.site_name]["basedir"], capture_output=True, universal_newlines=True)
#
#     _counter = 0
#     for line in iter(process.stdout.split('\n')):
#         if re.search(search, line, re.IGNORECASE):
#             _counter = _counter + 1
#             print('\n' + line)
#
#
#     if _counter > 0:
#         print("\nFormat: path,slug,title,date,expiryDate,publishDate,draft,permalink")
#
#     print("Results for '{}': {}".format(search, _counter))

def interactive_tags(search):
    os.system("/usr/bin/cat {filename} | fzf --delimiter=':' -n 2 -q '{search}'".format(filename = config["sites"][args.site_name]["list_tags"], search=search))

def interactive_categories(search):
    os.system("/usr/bin/cat {filename} | fzf --delimiter=':' -n 2 -q '{search}'".format(filename = config["sites"][args.site_name]["list_categories"], search=search))

def find_anything(search):
    os.chdir(config["sites"][args.site_name]["contentdir"])
    os.system("/usr/bin/rg -i --column --line-number --no-heading \'{search}\'".format(search=search))

def find_anything_fuzzy(search):
    os.chdir(config["sites"][args.site_name]["contentdir"])
    os.system("/usr/bin/rg -i --column --line-number --no-heading \'{search}\' . | fzf --bind 'f4:execute({editor} {{1}})' {bind_explorer} {preview} {preview_window} --with-nth=1.. -e --delimiter=':'".format(search=search, editor=config["settings"]["bin_editor"], preview="--preview 'batcat --style=full --color=always --line-range :500 {1}'", preview_window="--preview-window=top:75%:wrap", bind_explorer="--bind \'f3:execute(xdg-open \"$(dirname {})\")\'"))

def site_upload():
    remote_user = config["sites"][args.site_name]["remote_user"]
    remote_host = config["sites"][args.site_name]["remote_host"]
    remote_path = config["sites"][args.site_name]["remote_path"]
    build_dir = config["sites"][args.site_name]["build_dir"]

    subprocess.run(["/usr/bin/rsync", "-avz", "--delete-delay", build_dir, "{}@{}:{}".format(remote_user, remote_host, remote_path) ], cwd=config[args.site_name]["basedir"])
    
def get_year_month_date():
    from datetime import date
    todays_date = date.today()
    year = todays_date.strftime('%Y')
    month = todays_date.strftime('%m')
    day = todays_date.strftime('%d')

    return(year, month, day)

def process_template(template, title, slug, datetime):
    from jinja2 import Template

    filename = config["settings"]["template_path"] + template + ".jinja"

    with open(filename) as f:
        tmpl = Template(f.read())

    return(tmpl.render(title = title, slug = slug, datetime = datetime))

def count_and_unify_occurences_from_file(filename):
    """
    Liest eine Datei ein, die tausende gleiche Zeilen mit Tags enthält, die aus vielen Ausgaben von yq kamen und erstellt daraus ein dict, in dem jede Zeile nur einmal vorkommt und zusätzlich die Anzahl der Vorkommen enthält.
    Ausgabe.
    Code von https://www.geeksforgeeks.org/python-count-occurrences-of-each-word-in-given-text-file/
    """
    f = open(filename, "r")
    d = dict()

    for line in f:
        tag = line.strip("- ")
        tag = tag.strip()

        if tag in d:
            d[tag] = d[tag] + 1
        else:
            d[tag] = 1

    return(d)

def write_unified_occurences_to_cache_file(filename, occurences):
    f = open(filename, "w")

    for key in list(sorted(occurences.keys())):
        f.write("{count}:{tag}\n".format(count=occurences[key], tag=key))

    f.close()

def create_tags_categories_caches():
    """ Nutzt yq, um alle Blogbeiträge und Seiten nach genutzen Tags und Kategorien zu durchsuchen und eine Liste dieser und deren Häufigkeit zu erstellen, die dann mit "hgh.py -ft" genutzt werden kann. """
    from tempfile import mkstemp

    fd_tags, filename_tags = mkstemp()
    fd_categories, filename_categories = mkstemp()

    os.chdir(config["sites"][args.site_name]["contentdir"])

    command_tags = "find -iname \"*.md\" -exec yq --front-matter=extract \"{typeof}\" '{{}}' >> \"{tmpfile}\" \;".format(typeof = ".tags", tmpfile = filename_tags)
    command_categories = "find -iname \"*.md\" -exec yq --front-matter=extract \"{typeof}\" '{{}}' >> \"{tmpfile}\" \;".format(typeof = ".categories", tmpfile = filename_categories)

    os.system(command_tags)
    os.system(command_categories)

    # tags
    write_unified_occurences_to_cache_file(config["sites"][args.site_name]["list_tags"], count_and_unify_occurences_from_file(filename_tags))

    # categories
    write_unified_occurences_to_cache_file(config["sites"][args.site_name]["list_categories"], count_and_unify_occurences_from_file(filename_categories))

""" Start des Tools ab hier """

config = get_config(config_json)
args = parser.parse_args()

if not args.site_name and not (args.sites or args.time or args.printconfig or args.list_templates or args.create_comments_file):
    parser.error("--site is required for all arguments except (--time, --printconfig, --templates, --create-comments-file).")

if args.create_comments_file:
    # Creates a comment.json from template
    filename = config["settings"]["comments_filename"]

    comment = """[{
"id": 0,
"author": "",
"text": "",
"website": "",
"date": "",
"time": "",
"reply_to": 0
}]"""

    if not os.path.isfile(filename):
        with open(filename, "w") as w:
            w.write(comment)

        print("{filename} created.\n\nDon't forget to add \"jsoncomments: true\" to frontmatter.".format(filename=filename))
    else:
        print("Aborted. There is already a file named \"{}\"".format(filename))

if args.sites:
    for key in config["sites"].keys():
        print("{}: {}".format(key, config["sites"][key]["basedir"]))

if args.create_tags_categories_caches:
    create_tags_categories_caches()

if args.list_templates:
    import glob
    print("Templates available in {}".format(config["settings"]["template_path"]))
    for file in glob.glob(config["settings"]["template_path"] + "*.jinja"):
        print("- \"{}\"".format(file.lstrip(config["settings"]["template_path"]).rstrip(".jinja")))

if args.time:
    #get_datetime()
    os.system("date +'%Y-%m-%dT%H:%M:%S%:z' | xsel -i -b") # geht bei mir leider nicht mit pyperclip3
    print("Check your clipboard :)")

if args.printconfig:
    print("Config file: {path}\n{dump}".format(path = config_json, dump = json.dumps(config, indent=4)))

if args.hugo_server:
    hugo_server()

if args.hugo_server_light:
    hugo_server_light()

if args.tmux:
    start_tmux()

# if args.searchpost:
#     print("Deprecated")
#     searchpost(args.searchpost)

if args.find_tags:
    interactive_tags(args.find_tags)

if args.find_categories:
    interactive_categories(args.find_categories)

if args.find_anything:
    find_anything(args.find_anything)

if args.find_anything_fuzzy:
    find_anything_fuzzy(args.find_anything_fuzzy)

if args.site_build:
    site_build()

if args.site_upload:
    site_upload()

if args.deploy:
    site_build()
    site_upload()


# Wie gehtn das, dass man dort dann auch landet und das Python sich beendet? Gar nicht, da man auf die Shell, in der Python aufgerufen wurde, keinerlei Zugriff hat.
# Man kann nur mit chdir das cwd wechseln und dort eine Shell aufrufen.

if args.cdb:
    subprocess.run([config["settings"]["shell"]], cwd=config["sites"][args.site_name]["basedir"])

if args.cdc:
    subprocess.run([config["settings"]["shell"]], cwd=config["sites"][args.site_name]["contentdir"])

if args.cdl:
    subprocess.run([config["settings"]["shell"]], cwd=config["sites"][args.site_name]["contentdir-light"])

if args.create_pagebundle: # Erstellt ein Page Bundle mit Dateinamen aus slug, der automatisch aus dem angegebenen Titel generiert wird. --create-pagebundle "Dies ist ein neuer Blogbeitrag"
    # Das funktioniert nur angepasst für Natenoms Blog. Bitte im Code anpassen. Todo...
    title = args.create_pagebundle
    slug = title_to_slug(title)
    if args.template == None:
        template = config["settings"]["default_template"]
    else:
        template = args.template

    blog_directory_posts = config["sites"][args.site_name]["contentdir"] + '/posts/'

    # Page Bundle (nur Verzeichnis) verarbeiten
    year, month, day = get_year_month_date()

    directory_of_pagebundle = blog_directory_posts + year + "/" + month + "/" + year + "-" + month + "-" + day + "-" + slug

    if os.path.isdir(directory_of_pagebundle): # Page Bundle Verzeichnis existiert schon.
        directory_of_pagebundle_exists = True
        # Checking if the list is empty or not
        if len(os.listdir(directory_of_pagebundle)) == 0:
            print("Info: Directory exists, but is empty, weiter gehts: {}".format(directory_of_pagebundle))
    else: # Page Bundle Verzeichnis existiert noch nicht.
        from pathlib import Path
        #directory_of_pagebundle_exists = False
        Path(directory_of_pagebundle).mkdir(parents=True, exist_ok=True)

    # index.md verarbeiten
    if os.path.isfile(directory_of_pagebundle + "/index.md"):
        print("Es gibt schon eine index.md im Page Bundle")
        sys.exit(1)
    else:
        filename = directory_of_pagebundle + "/index.md"
        filecontent = process_template(template, title, slug, get_datetime()) # erweitern, dass eine Liste mit allen möglichen Dingen übergeben wird, die man in einem Template nutzen könnte.

        with open(filename, "w") as w:
            w.write(filecontent)

        if config["settings"]["open_new_pagebundle_with_vscode"]:
            # tut nicht bei flatpak und normal gleichzeichtig: subprocess.run([config["settings"]["bin_vscode"].split(" "), config["sites"][args.site_name]["basedir"], directory_of_pagebundle + "/index.md"])
            os.system("{bin_vscode} {basedir} {filename} &".format(bin_vscode = config["settings"]["bin_vscode"], basedir = config["sites"][args.site_name]["basedir"], filename = directory_of_pagebundle + "/index.md"))
        else:
            print("Page Bundle erfolgreich erstellt:\n\tTemplate: {}\n\tSlug: {}\n\tVerzeichnis: {}".format(template, slug, directory_of_pagebundle))
            os.system("echo '{path}' | xsel -i -b".format(path=directory_of_pagebundle + "/index.md"))
            print("Hinweise:\n\tPfad zur Datei in Zwischenablage eingefügt. In VSCodium öffnen mit 'Strg + o' und einfügen, Leerzeichen am Ende entfernen. (Das Leerzeichen am Ende kommt von VSCode.)")