Merge pull request #294 from dartmouth-pbs/bf-pretty-json-dates

Fixing up the abomination -- the regex to remove multiple spaces before numbers

Author: mgxd

heudiconv/tests/test_utils.py

@@ -52,3 +52,10 @@ def test_json_dumps_pretty():
         == '{\n  "a": -1,\n  "b": "123",\n  "c": [1, 2, 3],\n  "d": ["1.0", "2.0"]\n}'
     assert pretty({'a': ["0.3", "-1.9128906358217845e-12", "0.2"]}) \
         == '{\n  "a": ["0.3", "-1.9128906358217845e-12", "0.2"]\n}'
+    # original, longer string
+    tstr = 'f9a7d4be-a7d7-47d2-9de0-b21e9cd10755||' \
+          'Sequence: ve11b/master r/50434d5; ' \
+          'Mar  3 2017 10:46:13 by eja'
+    # just the date which reveals the issue
+    # tstr = 'Mar  3 2017 10:46:13 by eja'
+    assert pretty({'WipMemBlock': tstr}) == '{\n  "WipMemBlock": "%s"\n}' % tstr

heudiconv/utils.py

@@ -217,8 +217,16 @@ def json_dumps_pretty(j, indent=2, sort_keys=True):
     # uniform no spaces before ]
     js_ = re.sub(" *\]", "]", js_)
     # uniform spacing before numbers
-    js_ = re.sub('  *("?[-+.0-9e]+"?)(?P<space> ?)[ \n]*',
-                 r' \1\g<space>', js_)
+    # But that thing could screw up dates within strings which would have 2 spaces
+    # in a date like Mar  3 2017, so we do negative lookahead to avoid changing
+    # in those cases
+    #import pdb; pdb.set_trace()
+    js_ = re.sub(
+        '(?<!\w{3})'    # negative lookbehind for the month
+        '  *("?[-+.0-9e]+"?)'
+        '(?! [123]\d{3})'  # negative lookahead for a year
+        '(?P<space> ?)[ \n]*',
+        r' \1\g<space>', js_)
     # no spaces after [
     js_ = re.sub('\[ ', '[', js_)
     # the load from the original dump and reload from tuned up