Add proper install and cli and refactor code

.gitignore

@@ -1,2 +1,5 @@
 config.py
-__pycache__/
+__pycache__/
+Pipfile.lock
+.mypy_cache
+cache.json

Pipfile

@@ -0,0 +1,20 @@
+[[source]]
+name = "pypi"
+url = "https://pypi.org/simple"
+verify_ssl = true
+
+[scripts]
+"list" = "invoke --list"
+"help" = "invoke --help"
+"login" = "invoke login"
+"list-files" = "invoke list-files"
+"search" = "invoke search"
+"links" = "invoke links"
+
+[dev-packages]
+
+[packages]
+clipboard = "*"
+tabulate = "*"
+python-levenshtein = "*"
+invoke = "*"

README.md

@@ -1,13 +1,68 @@
 # nextcloud-link-generator
+
 ## Installation
-Simply run ```python3 setup.py``` to install all needed libraries and create config.py. If you don't want your nextcloud password saved in config.py just skip the password question by hitting the enter key. 
-If you didn't save your passwort during setup you will be asked at running the script.
+
+Install dependencies with `pipenv install`.
+
+After installing you have to create a `config.py` file.
+This can be done by running `pipenv run login --host=<url of the nextcloud> --base-dir=<directory to search for in nextcloud>`.
+See also `config.py.example`.
 
 ## Usage
-Run ```python3 main.py <exam1> <exam2> <exam n>``` with the exams you want as commandline arguments. If there are spaces in the name, replace them with a minus (```-```)
+
+When using the cli the content of the directory is cached in the file `cache.json`.
+This should make searches very fast.
+The cache is refreshed if the `cache.json` is deleted or if the directory was changed or if the cache is not from the same date as today.
+The cache can also be completely bypassed (and will not be updated on disk) by using the `--no-cache` flag.
+
+List all exams:
+
+```bash
+pipenv run list-files
+# without on disk cache
+pipenv run list-files --no-cache
+```
+
+Search for exams:
+
+```bash
+# for fuzzy search
+pipenv run search --lectures="exam1,exam2,exam 3"
+# or for more exact matches
+pipenv run search --exact --lectures="exam1,exam2,exam 3"
+# to search without using the on disk cache
+pipenv run search --no-cache --lectures="exam1,exam2,exam 3"
+```
+
+Generate links for exams:
+
+```bash
+# for fuzzy search
+pipenv run links --lectures="exam1,exam2,exam 3"
+# or for more exact matches
+pipenv run links --exact --lectures="exam1,exam2,exam 3"
+# to generate links without using the on disk cache
+pipenv run links --no-cache --lectures="exam1,exam2,exam 3"
+```
+
+Get help for the commands:
+
+```bash
+# list all commands
+pipenv run list
+# get help for a specific command
+pipenv run help <command-name>
+# example:
+pipenv run help search
+```
+
+Note: `pipenv run` can be replaced by `invoke` if inside the python venv.
+
 
 ## Used libraries
+
   - https://github.com/owncloud/pyocclient at release 0.4
   - python-Levenshtein
   - tabulate
-  - clipboard
+  - clipboard
+  - invoke (only for cli)

tasks.py

@@ -0,0 +1,130 @@
+from invoke import task, context
+from pathlib import Path
+from urllib.request import urlopen, Request, urljoin, urlparse, urlunparse
+from urllib.error import HTTPError
+from json import loads
+from time import sleep, time
+import webbrowser
+
+
+@task(help={
+    'host': 'The url of the nextcloud server.',
+    'base-dir': 'The path to search for directories in the nextcloud server. (Can be changed later in config.py)',
+})
+def login(c, host, base_dir='/'):
+    '''Login with the login flow v2 of nextcloud to get an app password.
+
+    This will open a browser tab for you to login to.
+    The app password will be written in config.py.
+    '''
+    url = urlparse(host)
+    if (url.scheme is not None) and (url.scheme != 'https'):
+        print('The connection to {} does not use the secure https protocol. Aborting login.'.format(url))
+
+    # generate loginflow url
+    nextcloud_host_url = urlunparse(('https', url.netloc, url.path, None, None, None))
+    login_flow_url = urljoin(host, '/index.php/login/v2')
+
+    # send request
+    req = Request(url=login_flow_url, method='POST')
+    req.add_header('User-Agent', 'nextcloud-link-generator')
+    response = None
+    with urlopen(req) as res:
+        if res.status != 200:
+            print("Could not start login flow!")
+            return
+        response = loads('\n'.join(map(lambda s: s.decode(), res.readlines())))
+    print(response)
+
+    webbrowser.open_new_tab(response['login'])
+
+    poll_data = 'token=' + response['poll']['token']
+    poll_req = Request(url=response['poll']['endpoint'], data=poll_data.encode(), method='POST')
+
+    poll_response = None
+    start = time()
+    while (poll_response is None) and ((time() - start) < 300):
+        try:
+            with urlopen(poll_req) as res:
+                if res.status == 200:
+                    poll_response = loads('\n'.join(map(lambda s: s.decode(), res.readlines())))
+                    break
+                else:
+                    print('Something went wrong with the login flow.')
+                    return
+        except HTTPError as e:
+            status = e.getcode()
+            if status == 404:
+                sleep(1)
+            else:
+                print('Something went wrong with the login flow.')
+                return
+    else:
+        print('Login took too long. Try again and be faster next time.')
+        return
+
+    if poll_response is None:
+        print('Could not read app password!')
+        return
+
+    print(poll_response)
+
+    with open("config.py", 'w') as outfile:
+        outfile.write("user = '{}'\n".format(poll_response['loginName']))
+        outfile.write("password = '{}'\n\n".format(poll_response['appPassword']))
+        outfile.write("url = '{}'\n\n".format(poll_response['server']))
+        outfile.write("base_dir = '{}'\n\n".format(base_dir))
+
+
+@task(help={
+    'no-cache': 'Bypass the cache and load everything from the server.',
+})
+def list_files(c, no_cache=False):
+    '''List all directories in the basedir.'''
+    import config
+    from utilities import Nextcloud
+    nc = Nextcloud(config.url, config.user, config.password, config.base_dir)
+    for f in nc.get_filenames_with_path(no_cache=no_cache):
+        print(f['filename'], f['path'])
+
+
+@task(help={
+    'lectures': 'A list of lecture names to search for. Seperate lectures with ",".',
+    'exact': 'Only include exact substring matches (still ignores character case).',
+    'no-cache': 'Bypass the cache and load everything from the server.',
+})
+def search(c, lectures, exact=False, no_cache=False):
+    '''Search for directories in the basedir.'''
+    import config
+    from utilities import Nextcloud
+    nc = Nextcloud(config.url, config.user, config.password, config.base_dir)
+    for f in nc.get_files_for_lectures(lectures.split(','), exact_matches=exact, no_cache=no_cache):
+        print(f['searchableFilename'], f['filename'], f['path'], sep='\t')
+
+@task(help={
+    'lectures': 'A list of lecture names to search for. Seperate lectures with ",".',
+    'expiry': 'Nr. of days until links expire.',
+    'exact': 'Only include exact substring matches (still ignores character case).',
+    'no-cache': 'Bypass the cache and load everything from the server.',
+})
+def links(c, lectures, exact=False, no_cache=False, expiry=7):
+    '''Generate share links for directories in the basedir.'''
+    import config
+    from utilities import Helper
+    from tabulate import tabulate
+    from utilities import Nextcloud
+    nc = Nextcloud(config.url, config.user, config.password, config.base_dir)
+    files = nc.get_files_for_lectures(lectures.split(','), exact_matches=exact, no_cache=no_cache)
+    links = nc.get_links_for_files(files, link_expire_in_days=expiry)
+    clipboard = ''
+    table_content = []
+    for exam, link in links.items():
+        clipboard += '{exam}: {link}\n'.format(exam=exam, link=link)
+        table_content.append([exam, link])
+    table_content = sorted(table_content, key=lambda x: x[0])
+    try:
+        Helper.toClipboard(clipboard)
+    except Exception:
+        print(clipboard)
+    header = ["Exam", "Link"]
+    print(tabulate(table_content, headers=header))