Influxdb error on writing CSV

[james-tanner]

In running formant tracks for Hastings, I get the following output whilst writing to CSV:

Traceback (most recent call last):
  File "formant_track.py", line 130, in <module>
    corpus_conf['speakers'], vowel_prototypes_path = vowel_prototypes_path)
  File "formant_track.py", line 85, in formant_track_export
    q.to_csv(csv_path)
  File "/data/iscan-spade-server/src/polyglotdb/polyglotdb/query/annotations/query.py", line 404, in to_csv
    r.to_csv(path, mode=mode)
  File "/data/iscan-spade-server/src/polyglotdb/polyglotdb/query/annotations/results.py", line 227, in to_csv
    super(QueryResults, self).to_csv(path, mode=mode)
  File "/data/iscan-spade-server/src/polyglotdb/polyglotdb/query/base/results.py", line 122, in to_csv
    save_results(self.rows_for_csv(), path, header=self.columns, mode=mode)
  File "/data/iscan-spade-server/src/polyglotdb/polyglotdb/io/exporters/csv.py", line 50, in save_results
    for line in results:
  File "/data/iscan-spade-server/src/polyglotdb/polyglotdb/query/annotations/results.py", line 212, in rows_for_csv
    for line in self:
  File "/data/iscan-spade-server/src/polyglotdb/polyglotdb/query/base/results.py", line 111, in __iter__
    r = self._sanitize_record(r)
  File "/data/iscan-spade-server/src/polyglotdb/polyglotdb/query/annotations/results.py", line 196, in _sanitize_record
    data = self.corpus.get_utterance_acoustics(a.attribute.label, utterance_id, discourse, speaker)
  File "/data/iscan-spade-server/src/polyglotdb/polyglotdb/corpus/audio.py", line 692, in get_utterance_acoustics
    result = client.query(query)
  File "/home/linguistics/jtanner/miniconda3/lib/python3.6/site-packages/influxdb/client.py", line 416, in query
    expected_response_code=expected_response_code
  File "/home/linguistics/jtanner/miniconda3/lib/python3.6/site-packages/influxdb/client.py", line 286, in request
    raise InfluxDBClientError(response.content, response.status_code)
influxdb.exceptions.InfluxDBClientError: 400: {"error":"error parsing query: found s, expected ; at line 4, char 48"}

I tried to do some digging online as to what this would mean, but it's still pretty opaque to me.

[msonderegger]

@james-tanner I think for debugging the exact query being error'ed on will be needed. If you're running on oka, I think (?) you can see the exact queries sent in the celery (?) log -- could you post here?

(On roquefort I'm less sure what you'd do. You can consult MG on Slack for faster response on how to see the actual queries being run.)

[james-tanner]

I'm not sure where to find this. the influxdb.log looks fine, and I can't see anywhere where celery logs would be stored (the screen running celery hasn't printed anything new). I've checked the polyglot_data directory to make sure the corpus itself has been parsed correctly, and it looks fine (each file has multiple utterances, phones, words, etc).

Any advice about where to look for debugging this?

[msonderegger]

Nope, no idea. I'd ping @mmcauliffe about this on Slack iscan-dev.

[mmcauliffe]

Was an issue with not escaping apostrophes in speaker names, fixed now.