Merge pull request #842 from krinsman/master

Reorganize init_tornado_application, create some configurables

Author: parente

nbviewer/app.py

@@ -25,7 +25,7 @@
 
 from jinja2 import Environment, FileSystemLoader
 
-from traitlets import Unicode
+from traitlets import Unicode, Any, Set, default
 from traitlets.config import Application
 
 from .handlers import init_handlers
@@ -44,6 +44,11 @@
 from .log import log_request
 from .utils import git_info, jupyter_info, url_path_join
 
+try: # Python 3.8
+    from functools import cached_property
+except ImportError:
+    from .utils import cached_property
+
 #-----------------------------------------------------------------------------
 # Code
 #-----------------------------------------------------------------------------
@@ -74,83 +79,96 @@ def nrfoot():
 
 class NBViewer(Application):
 
-    config_file = Unicode('nbviewer_config.py', help="The config file to load").tag(config=True)
+    name = Unicode('nbviewer')
 
-    def init_tornado_application(self):
-        # NBConvert config
-        self.config.NbconvertApp.fileext = 'html'
-        self.config.CSSHTMLHeaderTransformer.enabled = False
+    config_file = Unicode('nbviewer_config.py', help="The config file to load").tag(config=True)
 
-        # DEBUG env implies both autoreload and log-level
-        if os.environ.get("DEBUG"):
-            options.debug = True
-            logging.getLogger().setLevel(logging.DEBUG)
-    
-        # setup memcache
-        mc_pool = ThreadPoolExecutor(options.mc_threads)
-    
-        # setup formats
-        formats = configure_formats(options, self.config, log.app_log)
-    
-        if options.processes:
-            pool = ProcessPoolExecutor(options.processes)
-        else:
-            pool = ThreadPoolExecutor(options.threads)
-    
-        memcache_urls = os.environ.get('MEMCACHIER_SERVERS',
-            os.environ.get('MEMCACHE_SERVERS')
-        )
-    
-        # Handle linked Docker containers
-        if(os.environ.get('NBCACHE_PORT')):
-            tcp_memcache = os.environ.get('NBCACHE_PORT')
-            memcache_urls = tcp_memcache.split('tcp://')[1]
-    
-        if(os.environ.get('NBINDEX_PORT')):
+    url_handler         = Unicode(default_value="nbviewer.providers.url.handlers.URLHandler",           help="The Tornado handler to use for viewing notebooks accessed via URL").tag(config=True)
+    local_handler       = Unicode(default_value="nbviewer.providers.local.handlers.LocalFileHandler",   help="The Tornado handler to use for viewing notebooks found on a local filesystem").tag(config=True)
+    github_blob_handler = Unicode(default_value="nbviewer.providers.github.handlers.GitHubBlobHandler", help="The Tornado handler to use for viewing notebooks stored as blobs on GitHub").tag(config=True)
+    github_tree_handler = Unicode(default_value="nbviewer.providers.github.handlers.GitHubTreeHandler", help="The Tornado handler to use for viewing directory trees on GitHub").tag(config=True)
+    gist_handler        = Unicode(default_value="nbviewer.providers.gist.handlers.GistHandler",         help="The Tornado handler to use for viewing notebooks stored as GitHub Gists").tag(config=True)
+    user_gists_handler  = Unicode(default_value="nbviewer.providers.gist.handlers.UserGistsHandler",    help="The Tornado handler to use for viewing directory containing all of a user's Gists").tag(config=True)
+
+    index = Any().tag(config=True)
+    @default('index')
+    def _load_index(self):
+        if os.environ.get('NBINDEX_PORT'):
             log.app_log.info("Indexing notebooks")
             tcp_index = os.environ.get('NBINDEX_PORT')
             index_url = tcp_index.split('tcp://')[1]
             index_host, index_port = index_url.split(":")
-            indexer = ElasticSearch(index_host, index_port)
         else:
             log.app_log.info("Not indexing notebooks")
             indexer = NoSearch()
-    
+        return indexer
+
+    # cache frontpage links for the maximum allowed time
+    max_cache_uris = Set().tag(config=True)
+    @default('max_cache_uris')
+    def _load_max_cache_uris(self):
+        max_cache_uris = {''}
+        for section in self.frontpage_setup['sections']:
+            for link in section['links']:
+                max_cache_uris.add('/' + link['target'])
+        return max_cache_uris
+
+    static_path = Unicode(default_value=pjoin(here, 'static')).tag(config=True)
+
+    static_url_prefix = Unicode().tag(config=True)
+    @default('static_url_prefix')
+    def _load_static_url_prefix(self):
+        return url_path_join(self.base_url, '/static/')
+
+    @cached_property
+    def base_url(self):
+        # prefer the JupyterHub defined service prefix over the CLI
+        base_url = os.getenv("JUPYTERHUB_SERVICE_PREFIX", options.base_url)
+        return base_url
+
+    @cached_property
+    def cache(self):
+        memcache_urls = os.environ.get('MEMCACHIER_SERVERS', os.environ.get('MEMCACHE_SERVERS'))
+        # Handle linked Docker containers
+        if os.environ.get('NBCACHE_PORT'):
+            tcp_memcache = os.environ.get('NBCACHE_PORT')
+            memcache_urls = tcp_memcache.split('tcp://')[1]
         if options.no_cache:
             log.app_log.info("Not using cache")
             cache = MockCache()
         elif pylibmc and memcache_urls:
+            # setup memcache
+            mc_pool = ThreadPoolExecutor(options.mc_threads)
             kwargs = dict(pool=mc_pool)
-            username = os.environ.get('MEMCACHIER_USERNAME', '')
-            password = os.environ.get('MEMCACHIER_PASSWORD', '')
+            username = os.environ.get("MEMCACHIER_USERNAME", "")
+            password = os.environ.get("MEMCACHIER_PASSWORD", "")
             if username and password:
                 kwargs['binary'] = True
                 kwargs['username'] = username
                 kwargs['password'] = password
                 log.app_log.info("Using SASL memcache")
             else:
-                log.app_log.info("Using plain memecache")
-    
-            cache = AsyncMultipartMemcache(memcache_urls.split(','), **kwargs)
+                log.app_log.info("Using plain memcache")
+
+            cache = AsyncMultiPartMemcache(memcache_urls.split(','), **kwargs)
         else:
             log.app_log.info("Using in-memory cache")
             cache = DummyAsyncCache()
-    
-        # setup tornado handlers and settings
-    
-        template_paths = pjoin(here, 'templates')
-    
-        if options.template_path is not None:
-            log.app_log.info("Using custom template path {}".format(
-                options.template_path)
-            )
-            template_paths = [options.template_path, template_paths]
-    
-        static_path = pjoin(here, 'static')
-        env = Environment(
-            loader=FileSystemLoader(template_paths),
-            autoescape=True
-        )
+
+        return cache
+
+    # for some reason this needs to be a computed property,
+    # and not a traitlets Any(), otherwise nbviewer won't run
+    @cached_property
+    def client(self):
+        AsyncHTTPClient.configure(HTTPClientClass)
+        client = AsyncHTTPClient()
+        client.cache = self.cache
+        return client
+
+    @cached_property
+    def env(self):
+        env = Environment(loader=FileSystemLoader(self.template_paths), autoescape=True)
         env.filters['markdown'] = markdown.markdown
         try:
             git_data = git_info(here)
@@ -159,106 +177,137 @@ def init_tornado_application(self):
             git_data = {}
         else:
             git_data['msg'] = escape(git_data['msg'])
-    
-    
+
         if options.no_cache:
-            # force jinja to recompile template every time
+            # force Jinja2 to recompile template every time
             env.globals.update(cache_size=0)
-        env.globals.update(nrhead=nrhead, nrfoot=nrfoot, git_data=git_data,
-            jupyter_info=jupyter_info(), len=len,
-        )
-        AsyncHTTPClient.configure(HTTPClientClass)
-        client = AsyncHTTPClient()
-        client.cache = cache
-    
-        # load frontpage sections
-        with io.open(options.frontpage, 'r') as f:
-            frontpage_setup = json.load(f)
-        # check if the json has a 'sections' field, otherwise assume it is
-        # just a list of sessions, and provide the defaults for the other
-        # fields
-        if 'sections' not in frontpage_setup:
-            frontpage_setup = {'title': 'nbviewer',
-                               'subtitle':
-                               'A simple way to share Jupyter Notebooks',
-                               'show_input': True,
-                               'sections': frontpage_setup}
-    
-        # cache frontpage links for the maximum allowed time
-        max_cache_uris = {''}
-        for section in frontpage_setup['sections']:
-            for link in section['links']:
-                max_cache_uris.add('/' + link['target'])
-    
+        env.globals.update(nrhead=nrhead, nrfoot=nrfoot, git_data=git_data, jupyter_info=jupyter_info(), len=len)
+
+        return env
+
+    @cached_property
+    def fetch_kwargs(self):
         fetch_kwargs = dict(connect_timeout=10,)
         if options.proxy_host:
-            fetch_kwargs.update(dict(proxy_host=options.proxy_host,
-                                     proxy_port=options.proxy_port))
-    
+            fetch_kwargs.update(proxy_host=options.proxy_host, proxy_port=options.proxy_port)
             log.app_log.info("Using web proxy {proxy_host}:{proxy_port}."
                              "".format(**fetch_kwargs))
-    
+        
         if options.no_check_certificate:
-            fetch_kwargs.update(dict(validate_cert=False))
-    
+            fetch_kwargs.update(validate_cert=False)
             log.app_log.info("Not validating SSL certificates")
-    
-        # prefer the jhub defined service prefix over the CLI
-        base_url = os.getenv('JUPYTERHUB_SERVICE_PREFIX', options.base_url)
+
+        return fetch_kwargs
+
+    @cached_property
+    def formats(self):
+        formats = configure_formats(options, self.config, log.app_log)
+        return formats
+
+    # load frontpage sections
+    @cached_property
+    def frontpage_setup(self):
+        with io.open(options.frontpage, 'r') as f:
+            frontpage_setup = json.load(f)
+        # check if the JSON has a 'sections' field, otherwise assume it is just a list of sessions,
+        # and provide the defaults of the other fields
+        if 'sections' not in frontpage_setup:
+            frontpage_setup = {
+                              'title':'nbviewer', 'subtitle':'A simple way to share Jupyter notebooks',
+                              'show_input':True, 'sections':frontpage_setup
+                              }
+        return frontpage_setup
+
+    @cached_property
+    def pool(self):
+        if options.processes:
+            pool = ProcessPoolExecutor(options.processes)
+        else:
+            pool = ThreadPoolExecutor(options.threads)
+        return pool
+
+    @cached_property
+    def rate_limiter(self):
+        rate_limiter = RateLimiter(limit=options.rate_limit, interval=options.rate_limit_interval, cache=self.cache)
+        return rate_limiter
+
+    @cached_property
+    def template_paths(self):
+        template_paths = pjoin(here, 'templates')
+        if options.template_path is not None:
+            log.app_log.info("Using custom template path {}".format(options.template_path))
+            template_paths = [options.template_path, template_paths]
+
+        return template_paths
+
+
+    def init_tornado_application(self):
+        # handle handlers
+        handlers = init_handlers(self.formats, options.providers, self.base_url, options.localfiles)
 
-        rate_limiter = RateLimiter(
-            limit=options.rate_limit,
-            interval=options.rate_limit_interval,
-            cache=cache,
-        )
-    
+        # NBConvert config
+        self.config.NbconvertApp.fileext = 'html'
+        self.config.CSSHTMLHeaderTransformer.enabled = False
+
+        # DEBUG env implies both autoreload and log-level
+        if os.environ.get("DEBUG"):
+            options.debug = True
+            logging.getLogger().setLevel(logging.DEBUG)
+   
+        # input traitlets to settings
         settings = dict(
-            log_function=log_request,
-            jinja2_env=env,
-            static_path=static_path,
-            static_url_prefix=url_path_join(base_url, '/static/'),
-            client=client,
-            formats=formats,
-            default_format=options.default_format,
-            providers=options.providers,
-            provider_rewrites=options.provider_rewrites,
-            config=self.config,
-            index=indexer,
-            cache=cache,
-            cache_expiry_min=options.cache_expiry_min,
-            cache_expiry_max=options.cache_expiry_max,
-            max_cache_uris=max_cache_uris,
-            frontpage_setup=frontpage_setup,
-            pool=pool,
-            gzip=True,
-            render_timeout=options.render_timeout,
-            localfile_path=os.path.abspath(options.localfiles),
-            localfile_follow_symlinks=options.localfile_follow_symlinks,
-            localfile_any_user=options.localfile_any_user,
-            fetch_kwargs=fetch_kwargs,
-            mathjax_url=options.mathjax_url,
-            rate_limiter=rate_limiter,
-            statsd_host=options.statsd_host,
-            statsd_port=options.statsd_port,
-            statsd_prefix=options.statsd_prefix,
-            base_url=base_url,
-            google_analytics_id=os.getenv('GOOGLE_ANALYTICS_ID'),
-            hub_api_token=os.getenv('JUPYTERHUB_API_TOKEN'),
-            hub_api_url=os.getenv('JUPYTERHUB_API_URL'),
-            hub_base_url=os.getenv('JUPYTERHUB_BASE_URL'),
-            ipywidgets_base_url=options.ipywidgets_base_url,
-            jupyter_widgets_html_manager_version=options.jupyter_widgets_html_manager_version,
-            jupyter_js_widgets_version=options.jupyter_js_widgets_version,
-            content_security_policy=options.content_security_policy,
-            binder_base_url=options.binder_base_url,
+                  config=self.config,
+                  index=self.index,
+                  max_cache_uris=self.max_cache_uris,
+                  static_path=self.static_path,
+                  static_url_prefix=self.static_url_prefix,
         )
-    
+        # input computed properties to settings
+        settings.update(
+                  base_url=self.base_url,
+                  cache=self.cache,
+                  client=self.client,
+                  fetch_kwargs=self.fetch_kwargs,
+                  formats=self.formats,
+                  frontpage_setup=self.frontpage_setup,
+                  jinja2_env=self.env,
+                  pool=self.pool,
+                  rate_limiter=self.rate_limiter,
+        )
+        # input settings from CLI options
+        settings.update(
+                  binder_base_url=options.binder_base_url,
+                  cache_expiry_max=options.cache_expiry_max,
+                  cache_expiry_min=options.cache_expiry_min,
+                  content_security_policy=options.content_security_policy,
+                  default_format=options.default_format,
+                  ipywidgets_base_url=options.ipywidgets_base_url,
+                  jupyter_js_widgets_version=options.jupyter_js_widgets_version,
+                  jupyter_widgets_html_manager_version=options.jupyter_widgets_html_manager_version,
+                  localfile_any_user=options.localfile_any_user,
+                  localfile_follow_symlinks=options.localfile_follow_symlinks,
+                  localfile_path=os.path.abspath(options.localfiles),
+                  mathjax_url=options.mathjax_url,
+                  provider_rewrites=options.provider_rewrites,
+                  providers=options.providers,
+                  render_timeout=options.render_timeout,
+                  statsd_host=options.statsd_host,
+                  statsd_port=options.statsd_port,
+                  statsd_prefix=options.statsd_prefix,
+        )
+        # additional settings
+        settings.update(
+                  google_analytics_id=os.getenv('GOOGLE_ANALYTICS_ID'),
+                  gzip=True,
+                  hub_api_token=os.getenv('JUPYTERHUB_API_TOKEN'),
+                  hub_api_url=os.getenv('JUPYTERHUB_API_URL'),
+                  hub_base_url=os.getenv('JUPYTERHUB_BASE_URL'),
+                  log_function=log_request,
+        )
+
         if options.localfiles:
             log.app_log.warning("Serving local notebooks in %s, this can be a security risk", options.localfiles)
 
-        # handle handlers
-        handlers = init_handlers(formats, options.providers, base_url, options.localfiles)
-    
         # create the app
         self.tornado_application = web.Application(handlers, debug=options.debug, **settings)