💫 Improve model downloads: check for existing install, customise pip and use requests library again (#2346)

* Go back to using requests instead of urllib (closes #2320)

Fewer dependencies are good, but this one was simply causing too many other problems around SSL verification and Python 2/3 compatibility. requests is a popular enough package that it's okay for spaCy to depend on it – and this will hopefully make model downloads less flakey.

* Only download model if not installed (see #1456)

Use #egg=model==version to allow pip to check for existing installations. The download is only started if no installation matching the package/version is found. Fixes a long-standing inconvenience.

* Pass additional options to pip when installing model (resolves #1456)

Treat all additional arguments passed to the download command as pip options to allow user to customise the command. For example:

python -m spacy download en --user

* Add CLI option to enable installing model package dependencies

* Revert "Add CLI option to enable installing model package dependencies"

This reverts commit 9336ffe695.

* Update documentation

requirements.txt

@@ -9,5 +9,6 @@ plac<1.0.0,>=0.9.6
 ujson>=1.35
 dill>=0.2,<0.3
 regex==2017.4.5
+requests>=2.13.0,<3.0.0
 pytest>=3.0.6,<4.0.0
 mock>=2.0.0,<3.0.0

setup.py

@@ -195,7 +195,8 @@ def setup_package():
                 'pathlib',
                 'ujson>=1.35',
                 'dill>=0.2,<0.3',
-                'regex==2017.4.5'],
+                'regex==2017.4.5',
+                'requests>=2.13.0,<3.0.0'],
             classifiers=[
                 'Development Status :: 5 - Production/Stable',
                 'Environment :: Console',

spacy/cli/_messages.py

@@ -7,7 +7,7 @@ class Messages(object):
     M002 = ("Creating a shortcut link for 'en' didn't work (maybe you "
             "don't have admin permissions?), but you can still load the "
             "model via its full package name: nlp = spacy.load('{name}')")
-    M003 = ("Server error ({code}: {desc})")
+    M003 = ("Server error ({code})")
     M004 = ("Couldn't fetch {desc}. Please find a model for your spaCy "
             "installation (v{version}), and download it manually. For more "
             "details, see the documentation: https://spacy.io/usage/models")

spacy/cli/download.py

@@ -2,15 +2,14 @@
 from __future__ import unicode_literals
 
 import plac
+import requests
 import os
 import subprocess
 import sys
-import ujson
 
-from .link import link
 from ._messages import Messages
+from .link import link
 from ..util import prints, get_package_path
-from ..compat import url_read, HTTPError
 from .. import about
 
 
@@ -18,41 +17,23 @@ from .. import about
     model=("model to download, shortcut or name)", "positional", None, str),
     direct=("force direct download. Needs model name with version and won't "
             "perform compatibility check", "flag", "d", bool),
-    insecure=("insecure mode - disables the verification of certificates",
-              "flag", "i", bool),
-    ca_file=("specify a certificate authority file to use for certificates "
-             "validation. Ignored if --insecure is used", "option", "c"))
-def download(model, direct=False, insecure=False, ca_file=None):
+    pip_args=("additional arguments to be passed to `pip install` when "
+              "installing the model"))
+def download(model, direct=False, *pip_args):
     """
     Download compatible model from default download path using pip. Model
     can be shortcut, model name or, if --direct flag is set, full model name
     with version.
-    The --insecure optional flag can be used to disable ssl verification
-    The --ca-file option can be used to provide a local CA file
-    used for certificate verification.
     """
-
-    # ssl_verify is the argument handled to the 'verify' parameter
-    # of requests package. It must be either None, a boolean,
-    # or a string containing the path to CA file
-    ssl_verify = None
-    if insecure:
-        ca_file = None
-        ssl_verify = False
-    else:
-        if ca_file is not None:
-            ssl_verify = ca_file
-
-    # Download the model
     if direct:
-        dl = download_model('{m}/{m}.tar.gz'.format(m=model))
+        dl = download_model('{m}/{m}.tar.gz#egg={m}'.format(m=model), pip_args)
     else:
-        shortcuts = get_json(about.__shortcuts__, "available shortcuts", ssl_verify)
+        shortcuts = get_json(about.__shortcuts__, "available shortcuts")
         model_name = shortcuts.get(model, model)
-        compatibility = get_compatibility(ssl_verify)
+        compatibility = get_compatibility()
         version = get_version(model_name, compatibility)
-        dl = download_model('{m}-{v}/{m}-{v}.tar.gz'.format(m=model_name,
-                                                            v=version))
+        dl = download_model('{m}-{v}/{m}-{v}.tar.gz#egg={m}=={v}'
+                            .format(m=model_name, v=version), pip_args)
         if dl != 0:  # if download subprocess doesn't return 0, exit
             sys.exit(dl)
         try:
@@ -69,19 +50,18 @@ def download(model, direct=False, insecure=False, ca_file=None):
             prints(Messages.M001.format(name=model_name), title=Messages.M002)
 
 
-def get_json(url, desc, ssl_verify):
-    try:
-        data = url_read(url, verify=ssl_verify)
-    except HTTPError as e:
-        prints(Messages.M004.format(desc, about.__version__),
-               title=Messages.M003.format(e.code, e.reason), exits=1)
-    return ujson.loads(data)
+def get_json(url, desc):
+    r = requests.get(url)
+    if r.status_code != 200:
+        prints(Messages.M004.format(desc=desc, version=about.__version__),
+               title=Messages.M003.format(code=r.status_code), exits=1)
+    return r.json()
 
 
-def get_compatibility(ssl_verify):
+def get_compatibility():
     version = about.__version__
     version = version.rsplit('.dev', 1)[0]
-    comp_table = get_json(about.__compatibility__, "compatibility table", ssl_verify)
+    comp_table = get_json(about.__compatibility__, "compatibility table")
     comp = comp_table['spacy']
     if version not in comp:
         prints(Messages.M006.format(version=version), title=Messages.M005,
@@ -97,8 +77,10 @@ def get_version(model, comp):
     return comp[model][0]
 
 
-def download_model(filename):
+def download_model(filename, user_pip_args=None):
     download_url = about.__download_url__ + '/' + filename
-    return subprocess.call(
-        [sys.executable, '-m', 'pip', 'install', '--no-cache-dir', '--no-deps',
-         download_url], env=os.environ.copy())
+    pip_args = ['--no-cache-dir', '--no-deps']
+    if user_pip_args:
+        pip_args.extend(user_pip_args)
+    cmd = [sys.executable, '-m', 'pip', 'install'] + pip_args + [download_url]
+    return subprocess.call(cmd, env=os.environ.copy())

spacy/cli/validate.py

@@ -5,9 +5,10 @@ import pkg_resources
 from pathlib import Path
 import sys
 import ujson
+import requests
 
 from ._messages import Messages
-from ..compat import path2str, locale_escape, url_read, HTTPError
+from ..compat import path2str, locale_escape
 from ..util import prints, get_data_path, read_json
 from .. import about
 
@@ -16,12 +17,11 @@ def validate():
     """Validate that the currently installed version of spaCy is compatible
     with the installed models. Should be run after `pip install -U spacy`.
     """
-    try:
-        data = url_read(about.__compatibility__)
-    except HTTPError as e:
-        title = Messages.M003.format(code=e.code, desc=e.reason)
-        prints(Messages.M021, title=title, exits=1)
-    compat = ujson.loads(data)['spacy']
+    r = requests.get(about.__compatibility__)
+    if r.status_code != 200:
+        prints(Messages.M021, title=Messages.M003.format(code=r.status_code),
+               exits=1)
+    compat = r.json()['spacy']
     current_compat = compat.get(about.__version__)
     if not current_compat:
         prints(about.__compatibility__, exits=1,

spacy/compat.py

@@ -33,16 +33,6 @@ try:
 except ImportError:
     from thinc.neural.optimizers import Adam as Optimizer
 
-try:
-    import urllib.request
-except ImportError:
-    import urllib2 as urllib
-
-try:
-    from urllib.error import HTTPError
-except ImportError:
-    from urllib2 import HTTPError
-
 pickle = pickle
 copy_reg = copy_reg
 CudaStream = CudaStream
@@ -66,7 +56,6 @@ if is_python2:
     input_ = raw_input  # noqa: F821
     json_dumps = lambda data: ujson.dumps(data, indent=2, escape_forward_slashes=False).decode('utf8')
     path2str = lambda path: str(path).decode('utf8')
-    url_open = urllib.urlopen
 
 elif is_python3:
     bytes_ = bytes
@@ -75,16 +64,6 @@ elif is_python3:
     input_ = input
     json_dumps = lambda data: ujson.dumps(data, indent=2, escape_forward_slashes=False)
     path2str = lambda path: str(path)
-    url_open = urllib.request.urlopen
-
-
-def url_read(url):
-    file_ = url_open(url)
-    code = file_.getcode()
-    if code != 200:
-        raise HTTPError(url, code, "Cannot GET url", [], file_)
-    data = file_.read()
-    return data
 
 
 def b_to_str(b_str):

website/api/cli.jade

@@ -15,7 +15,8 @@ p
     |  package and automatically creates a
     |  #[+a("/usage/models#usage") shortcut link] to load the model by name.
     |  Direct downloads don't perform any compatibility checks and require the
-    |  model name to be specified with its version (e.g., #[code en_core_web_sm-1.2.0]).
+    |  model name to be specified with its version (e.g.
+    |  #[code en_core_web_sm-2.0.0]).
 
 +aside("Downloading best practices")
     |  The #[code download] command is mostly intended as a convenient,
@@ -35,13 +36,23 @@ p
     +row
         +cell #[code model]
         +cell positional
-        +cell Model name or shortcut (#[code en], #[code de], #[code vectors]).
+        +cell
+            |  Model name or shortcut (#[code en], #[code de],
+            |  #[code en_core_web_sm]).
 
     +row
         +cell #[code --direct], #[code -d]
         +cell flag
         +cell Force direct download of exact model version.
 
+    +row
+        +cell other
+        +cell -
+        +cell
+            |  Additional installation options to be passed to
+            |  #[code pip install] when installing the model package. For
+            |  example, #[code --user] to install to the user home directory.
+
     +row
         +cell #[code --help], #[code -h]
         +cell flag