From 63cd539d04f703a269dd4108cd7a7d7cacd496c9 Mon Sep 17 00:00:00 2001
From: ines <ines@ines.io>
Date: Sun, 4 Jun 2017 20:52:10 +0200
Subject: [PATCH] Add more details on model packages and requirements.txt (see
 #1099)

---
 website/docs/usage/models.jade         | 10 ++++
 website/docs/usage/production-use.jade | 69 ++++++++++++++++++++++++++
 2 files changed, 79 insertions(+)

diff --git a/website/docs/usage/models.jade b/website/docs/usage/models.jade
index c091c9489..51eea37d5 100644
--- a/website/docs/usage/models.jade
+++ b/website/docs/usage/models.jade
@@ -104,6 +104,16 @@ p
     |  recommend using pip with a direct link, instead of relying on spaCy's
     |  #[+api("cli#download") #[code download]] command.
 
+p
+    |  You can also add the direct download link to your application's
+    |  #[code requirements.txt]. For more information on this, see the
+    |  #[+a("https://pip.pypa.io/en/latest/reference/pip_install/#requirements-file-format") pip documentation].
+    |  This will only install the package and not trigger any of spaCy's internal
+    |  commands like #[code download] or #[code link]. So you'll have to make
+    |  sure to create a link for your model manually, or
+    |  #[+a("#usage-import") import it as a module] instead.
+
+
 +h(3, "download-manual") Manual download and installation
 
 p
diff --git a/website/docs/usage/production-use.jade b/website/docs/usage/production-use.jade
index e9fd4a30f..70227e648 100644
--- a/website/docs/usage/production-use.jade
+++ b/website/docs/usage/production-use.jade
@@ -76,3 +76,72 @@ p
     |  attributes to set the part-of-speech tags, syntactic dependencies, named
     |  entities and other attributes. For details, see the respective usage
     |  pages.
+
++h(2, "models") Working with models
+
+p
+    |  If your application depends on one or more #[+a("/docs/usage/models") models],
+    |  you'll usually want to integrate them into your continuous integration
+    |  workflow and build process. While spaCy provides a range of useful helpers
+    |  for downloading, linking and loading models, the underlying functionality
+    |  is entirely based on native Python packages. This allows your application
+    |  to handle a model like any other package dependency.
+
++h(3, "models-download") Downloading and requiring model dependencies
+
+p
+    |  spaCy's built-in #[+api("cli#download") #[code download]] command
+    |  is mostly intended as a convenient, interactive wrapper. It performs
+    |  compatibility checks and prints detailed error messages and warnings.
+    |  However, if you're downloading models as part of an automated build
+    |  process, this only adds an unecessary layer of complexity. If you know
+    |  which models your application needs, you should be specifying them directly.
+
+p
+    |  Because all models are valid Python packages, you can add them to your
+    |  application's #[code requirements.txt]. If you're running your own
+    |  internal PyPi installation, you can simply upload the models there. pip's
+    |  #[+a("https://pip.pypa.io/en/latest/reference/pip_install/#requirements-file-format") requirements file format]
+    |  supports both package names to download via a PyPi server, as well as direct
+    |  URLs.
+
++code("requirements.txt", "text").
+    spacy&gt;=2.0.0,&lt;3.0.0
+    -e #{gh("spacy-models")}/releases/download/en_core_web_sm-2.0.0/en_core_web_sm-2.0.0.tar.gz
+
+p
+    |  All models are versioned and specify their spaCy dependency. This ensures
+    |  cross-compatibility and lets you specify exact version requirements for
+    |  each model. If you've trained your own model, you can use the
+    |  #[+api("cli#package") #[code package]] command to generate the required
+    |  meta data and turn it into a loadable package.
+
++h(3, "models-loading") Loading and testing models
+
+p
+    |  Downloading models directly via pip won't call spaCy's link
+    |  #[+api("cli#link") #[code link]] command, which creates
+    |  symlinks for model shortcuts. This means that you'll have to run this
+    |  command separately, or use the native #[code import] syntax to load the
+    |  models:
+
++code.
+    import en_core_web_sm
+    nlp = en_core_web_sm.load()
+
+p
+    |  In general, this approach is recommended for larger code bases, as it's
+    |  more "native", and doesn't depend on symlinks or rely on spaCy's loader
+    |  to resolve string names to model packages. If a model can't be
+    |  imported, Python will raise an #[code ImportError] immediately. And if a
+    |  model is imported but not used, any linter will catch that.
+
+p
+    |  Similarly, it'll give you more flexibility when writing tests that
+    |  require loading models. For example, instead of writing your own
+    |  #[code try] and #[code except] logic around spaCy's loader, you can use
+    |  #[+a("http://pytest.readthedocs.io/en/latest/") pytest]'s
+    |  #[code importorskip()] method to only run a test if a specific model or
+    |  model version is installed. Each model package exposes a #[code __version__]
+    |  attribute which you can also use to perform your own version compatibility
+    |  checks before loading a model.