[TAAS-84] omit hub and minor updates

Dockerfile

@@ -1,4 +1,4 @@
-FROM unocha/alpine-base-s6-python2:latest
+FROM unocha/alpine-base-s6-python2:3.8
 
 MAINTAINER Paul Fenwick "paul@humanitarianresponse.info"
 
@@ -16,11 +16,9 @@ LABEL org.label-schema.schema-version="1.0" \
       org.label-schema.vcs-url=$VCS_URL \
       org.label-schema.vcs-ref=$VCS_REF
 
-# Git is everyone's favourite source control manager
-# Hub actually lets us make PRs to github
 # openssh is the transport needed to make PRs to github
 RUN apk add --update --no-cache \
-      git hub openssh \
+      git openssh \
       make \
       py-virtualenv \
       # These are needed so python can build its things.

Makefile

@@ -50,7 +50,7 @@ tag:
 	docker tag $(IMAGE_HASH) $(ORGANISATION)/$(IMAGE):$(VERSION)
 
 clean:
-	rm -rf venv *.pyc taas/*.pyc .cache tests/__pycache__ .coverage taas.egg-info htmlcov buidlog.txt
+	rm -rf venv *.pyc taas/*.pyc .cache tests/__pycache__ .coverage taas.egg-info htmlcov buildlog.txt
 
 # Magical line that allows us to force execution of a target
 FORCE:

docs/Config.md

@@ -30,6 +30,8 @@ sources:
 
 The `url` field *must* contain a URL complete with the `gid=` parameter at the end. For spreadsheets that consist of multiple sheets, the gid parameter specifies the individual sheet to be read. The `/edit` part of the URL is optional. In almost all cases a simple copy-and-paste will work correctly.
 
+Permissions on the spreadsheet must allow 'Anyone with the link' to view it.
+
 A source *may* contain a `fragment_key`. If set, JSON fragments (individual files containing only a single record each) will be emitted in addition to the main file, using the contents of the mapping field specified. For example, if we're using numeric IDs, the example above would create files named `beta-v1/functional_roles/1`, `beta-v1/function_roles/2`, etc.
 
 To match how existing APIs present their records, fragments do not have the `.json` or any other extension.

taas/__init__.py

@@ -115,7 +115,7 @@ def read_config(path=None):
 
     if os.path.isfile(path):
         with open(path) as stream:
-            return yaml.load(stream)
+            return yaml.safe_load(stream)
 
     # If we're here, we have a directory. (Or we're about
     # to throw an exception, because we don't.)
@@ -132,7 +132,7 @@ def get_directory_config(path):
     config = {}
     for file in glob.glob(os.path.join(path, "*.yml")):
         with open(file) as stream:
-            sub_config = yaml.load(stream)
+            sub_config = yaml.safe_load(stream)
             config = _merge_config(sub_config, config)
 
     return config

taas/cmdline.py

@@ -69,9 +69,6 @@ def push_to_github(label):
 
     run(["git", "push", "--set-upstream", "origin", branch_name])
 
-    # NB: `hub` is available from github
-    run(["hub", "pull-request", "-m", "Automated update: "+label])
-
 
 def gss2json():
     """Console script for processing google spreadsheet(s) to JSON."""