genome annotation improvements

[jameshadfield]

For a detailed write-up of the bug which motivated this PR, see
#881.

The main work is in the 2nd commit:

By storing the (nucleotide) genome annotation in the node-data produced
from augur-ancestral we make this information available for export.
Previously this information was only exported by augur translate which
was problematic for workflows which didn't perform translation.

No changes are needed to augur export v2 (which may now process
multiple "annotations" blocks) due to the behavior of
NodeData.deep_add_or_update which will recurse into dicts in
annotation blocks and when confronted with non-dict values which already
exist overwrite them. This poses a potential problem where two node-data
JSONs which (e.g.) define different annotations['nuc'] coordinates
will not raise any error and the resulting coodinates are dependent on
the order the node-data JSONs were provided to augur export v2.

Closes #881.

[jameshadfield]

The following test is failing (locally & on CI), but it's also failing for me on current master (where CI passes...). @victorlin any ideas?

$ cram --shell=/bin/bash tests/functional/frequencies.t

@@ -51,7 +51,7 @@
   >  --exclude-paths "root['generated_by']['version']" "root['pivots']" -- \
   >  "frequencies/zika_tip-frequencies_with_relative_dates.json" \
   >  "$TMP/tip-frequencies.json"
-  {}
+  {'iterable_item_added': {"root['BRA/2016/FC_6706']['frequencies'][2]": 0.0, "root['COL/FLR_00008/2015']['frequencies'][2]": 0.0, "root['Colombia/2016/ZC204Se']['frequencies'][2]": 0.0, "root['DOM/2016/BB_0183']['frequencies'][2]": 0.0, "root['EcEs062_16']['frequencies'][2]": 0.0, "root['HND/2016/HU_ME59']['frequencies'][2]": 0.0, "root['PAN/CDC_259359_V1_V3/2015']['frequencies'][2]": 0.0, "root['PRVABC59']['frequencies'][2]": 0.0, "root['VEN/UF_1/2016']['frequencies'][2]": 0.0, "root['ZKC2/2016']['frequencies'][2]": 0.0}}

[victorlin]

CI is temp patched with #962, just need to rebase onto latest master here and other PRs. Or run again when it's not the first of the month 🙃

[jameshadfield]

(No changes, but force-pushed with a new SHA to trigger CI)

[codecov]

Codecov Report

Merging #961 (8ef9753) into master (446b39d) will increase coverage by 2.17%.
The diff coverage is 25.00%.

@@            Coverage Diff             @@
##           master     #961      +/-   ##
==========================================
+ Coverage   59.36%   61.53%   +2.17%     
==========================================
  Files          43       52       +9     
  Lines        6014     6794     +780     
  Branches     1539     1768     +229     
==========================================
+ Hits         3570     4181     +611     
- Misses       2185     2348     +163     
- Partials      259      265       +6     

Impacted Files	Coverage Δ
augur/ancestral.py	66.37% <25.00%> (-5.18%)	⬇️
augur/io/__init__.py	100.00% <0.00%> (ø)
augur/io/metadata.py	100.00% <0.00%> (ø)
augur/measurements.py	31.25% <0.00%> (ø)
augur/io/sequences.py	100.00% <0.00%> (ø)
augur/util_support/metadata_file.py	100.00% <0.00%> (ø)
augur/io/shell_command_runner.py	91.89% <0.00%> (ø)
augur/io/print.py	100.00% <0.00%> (ø)
augur/io/vcf.py	42.50% <0.00%> (ø)
... and 5 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 446b39d...8ef9753. Read the comment docs.