We should have a unit test that makes sure term labels are alphanumeric

[danbri]

i.e. complains if rdfs:label has chars that aren't in [^a-zA-Z0-9_ ].

• Probably we really want to say [a-zA-Z]?[a-zA-Z0-9_]*
• Tolerate whitespace at ends? in middle? for now. Do we want the ID in string form not a humanized version?
• Make sure we run this against the data/ext/* extension definitions too.

[danbri]

@RichardWallis - I don't think this is running against the data/ext/health-lifesci/*rdfa data - can we do that easily?

[danbri]

#!/usr/bin/env python

import os
import fnmatch

path = 'data/ext/'

files = [os.path.join(dirpath, f)
    for dirpath, dirnames, files in os.walk(path)
    for f in fnmatch.filter(files, '*.rdfa')]

for f in files:
   print f

Gives

data/ext/auto/auto.rdfa
data/ext/bib/bsdo-1.0.rdfa
data/ext/bib/comics.rdfa
data/ext/health-lifesci/med-health-core.rdfa
data/ext/health-lifesci/physical-activity-and-exercise.rdfa
data/ext/pending/issue-1004.rdfa
data/ext/pending/issue-1045.rdfa
data/ext/pending/issue-1050.rdfa
data/ext/pending/issue-1061.rdfa
data/ext/pending/issue-1062.rdfa
data/ext/pending/issue-1100.rdfa
data/ext/pending/issue-1125.rdfa
data/ext/pending/issue-243.rdfa
data/ext/pending/issue-271.rdfa
data/ext/pending/issue-447.rdfa
data/ext/pending/issue-743.rdfa
data/ext/pending/issue-template.rdfa
data/ext/test001/test001demo.rdfa

This could plug in around https://github.com/schemaorg/schemaorg/blob/sdo-deimos/tests/test_graphs.py#L51 ...?

[danbri]

This now runs across extension data too thanks to @RichardWallis 's #1168

[RichardWallis]

Fixed in #1168