Move to storing test narratives in the auspice repo

This commit signifies a shift to storing a set of narrative files in this repo, as opposed to obtaining the nextstrain.org narratives via a script.

The Nextstrain.org narratives are still obtainable (they are all stored in their own github repo) and the docs have been updated to show this.

By storing and maintaining "test narratives" in the auspice repo we can both provide examples of the capability of narratives and use them to test functionality and fix bugs. As the are stored in the `narratives` directory, they are available via Nextstrain Community URLs at nextstrain.org/community/narratives/nextstrain/auspice/...

.github/workflows/ci.yaml

@@ -30,7 +30,6 @@ jobs:
           node-version: ${{ matrix.node }}
       - run: npm ci
       - run: npm run get-data
-      - run: npm run get-narratives
       - run: npm run smoke-test:ci
   lint:
     runs-on: ubuntu-latest

.gitignore

@@ -6,7 +6,6 @@ big_flu
 /static/
 s3/
 /local_narratives/
-/narratives/
 /datasets/
 **/__diff_output__
 

DEV_DOCS.md

@@ -48,7 +48,7 @@ Run `npm run lint`. If there are issues run `npm run lint:fix`.
 
 > For integration tests to work, you'll need to have `git-lfs` installed (see below) as it stores the images that the snapshot tests will use.
 
-1. Fetch the datasets with `npm run get-data` and `npm run get-narratives`.
+1. Fetch the datasets with `npm run get-data`.
 2. Ensure you are **not** currently running the site locally, then run `npm run integration-test:ci`.
 
 
@@ -60,7 +60,7 @@ Run `npm run lint`. If there are issues run `npm run lint:fix`.
 
 #### For smoke tests
 
-1. Fetch the datasets with `npm run get-data` and `npm run get-narratives`.
+1. Fetch the datasets with `npm run get-data`.
 2. Ensure you are **not** currently running the site locally, then run `npm run smoke-test:ci`.
 
 

README.md

@@ -78,9 +78,16 @@ If you've installed auspice from source, we have helper scripts to make all the
 ```bash
 # from the auspice src directory
 npm run get-data
-npm run get-narratives
 ```
 
+### Obtain narratives to view locally
+
+This repository contains a number of "test narratives" which serve both to provide examples of the capability of narratives, as well as being used to test functionality and fix bugs. 
+These should work out of the box, assuming you have obtained the necessary datasets via the above script.
+
+If you wish to view the [nextstrain-maintained narratives](https://nextstrain.org/docs/contributing/sharing-data), then this can be done by checking out the [nextstrain/narratives github repo](github.com/nextstrain/narratives) and telling `auspice` to look for narratives there via the `--narrativeDir` argument.
+
+
 ### Run auspice
 
 ```bash

docs-src/docs/introduction/how-to-run.md

@@ -120,9 +120,9 @@ For narratives, please see [Writing a Narrative](narratives/how-to-write.md) for
 
 ## Obtaining a Set of Input Files
 
-If you'd like to download the datasets and narratives on [nextstrain.org](https://nextstrain.org) then there are two scripts which allow you to do this:
+If you'd like to download the dataset JSONs which are behind the core-datasets shown on [nextstrain.org](https://nextstrain.org), then you can run [this script](https://github.com/nextstrain/auspice/blob/master/scripts/get-data.sh) which will create a `./data` directory for you.
 
-* You can download the dataset JSONs by running [this script](https://github.com/nextstrain/auspice/blob/master/scripts/get-data.sh) which will create a `./data` directory for you.
-* You can download the narrative Markdown files by running [this script](https://github.com/nextstrain/auspice/blob/master/scripts/get-narratives.sh) which will create a `./narratives` directory for you.
+The nextstrain-maintained narratives are stored in the [nextstrain/narratives github repo](github.com/nextstrain/narratives).
+You can obtain these by cloning that repo.
 
-You can then run `auspice view --datasetDir data --narrativeDir narratives` to visualise all of the [nextstrain.org](https://nextstrain.org) datasets locally.
+You can then run `auspice view --datasetDir data --narrativeDir <path-to-narratives>` to visualise all of the [nextstrain.org](https://nextstrain.org) datasets locally.

docs-src/docs/introduction/install.md

@@ -62,9 +62,8 @@ auspice build
 auspice --version
 auspice --help
 
-# Obtain datasets & narratives to view locally (optional)
+# Obtain nextstrain.org datasets to view locally (optional)
 npm run get-data
-npm run get-narratives
 ```
 
 Updating Auspice should only require pulling the new version from GitHub -- it shouldn't require any `npm` commands.

narratives/test_parsing-error.md

@@ -0,0 +1,13 @@
+---
+title: "Example of a narrative which can't be parsed"
+date: "June 2020"
+dataset: "https://nextstrain.org/zika"
+abstract: "This narrative in incorrectly parsed by auspice v2.16.0 (and earlier) such that it doesn't even render."
+---
+
+# [Summary](https://nextstrain.org/community/andersen-lab/HCoV-19-Genomics-Nextstrain/hCoV-19/usa/sandiego?c=region&d=tree&legend=open&onlyPanels&p=full&sidebar=closed)
+
+The problem is the following lines -- adding a character(s) to each line following the link syntax fixes the problem.
+
+Table of Contents:  
+1. [Some link](https://nextstrain.org)

narratives/test_simultaneous-tree-updates.md

@@ -0,0 +1,34 @@
+---
+title: Simultaneous Updates to Tree State
+authors:
+  - James Hadfield
+authorLinks:
+  - https://twitter.com/hamesjadfield
+affiliations: "Fred Hutch"
+dataset: "http://localhost:4000/flu/seasonal/h3n2/ha/3y?d=tree"
+date: June 2020
+abstract: "A narrative to explore simultaneous changes to tree state. The aim is both to reveal what's possible, and to prevent future regressions via automated testing."
+---
+
+
+# [P1: Change color-by to num_date]("http://localhost:4000/flu/seasonal/h3n2/ha/3y?d=tree&c=num_date)
+Check that both the branches and tips update.
+
+# [P2: Zoom into clade A1]("http://localhost:4000/flu/seasonal/h3n2/ha/3y?d=tree&c=num_date&label=clade:A1)
+
+# [P3: Zoom into clade A1b _and_ change color]("http://localhost:4000/flu/seasonal/h3n2/ha/3y?d=tree&label=clade:A1b)
+Check that the coloring of the branches and tips update as we zoom in.
+
+# [P4: Lots of simultaneous changes](http://localhost:4000/flu/seasonal/h3n2/ha/3y?c=lbi&d=tree&dmin=2017-01-01&f_region=North%20America&label=clade:3c2.A&m=div)
+* Zoomed out to near the root (clade 3c2.A)
+* Changed the horizontal scale to divergence
+* Changed the color-by to LBI (and legend should be open)
+* Filtered to North American Samples (i.e. the majority of tips are not visible).
+* Time slice excludes pre-2017 samples (note that we don't have the grey overlay here since the axis is divergence)
+
+# [P5: Layout change (and others)](http://localhost:4000/flu/seasonal/h3n2/ha/3y?branchLabel=none&d=tree&l=radial&m=div)
+Underneath, this uses a different set of d3 calls where instead of transitioning all elements (branches, tips etc), which is too slow, we do a "in-parts" update: we hide the branches, transition the tips, then redraw the branches in their new position.
+This slide should
+* switch to radial view
+* zoom out to show the entire tree, and remove filters so all tips should be shown
+* change the coloring to clade

package.json

@@ -26,8 +26,7 @@
     "lint": "eslint src",
     "lint:fix": "eslint --fix src",
     "get-data": "env bash ./scripts/get-data.sh",
-    "get-narratives": "env bash ./scripts/get-narratives.sh",
-    "heroku-postbuild": "npm run build && npm run get-data && npm run get-narratives",
+    "heroku-postbuild": "npm run build && npm run get-data",
     "rebuild-docker-image": "env bash ./scripts/rebuild-docker-image.sh",
     "gzip-and-upload": "env bash ./scripts/gzip-and-upload.sh",
     "build-docs": "echo 'see ./docs-src/README.md'",