csv-detective 0.9.3.dev2486__py3-none-any.whl → 0.9.3.dev2500__py3-none-any.whl

csv_detective/formats/float.py

@@ -1,22 +1,29 @@
+import re
+
 proportion = 1
 tags = ["type"]
 labels = ["part", "ratio", "taux"]
 
+scientific_notation_pattern = r"\d+\.\d+[e|E][+|-]?\d+"
+
 
 def float_casting(val: str) -> float:
     return float(val.replace(",", "."))
 
 
 def _is(val):
-    """Detects floats, assuming that tables will not have scientific
-    notations (3e6) or "+" in the string. "-" is still accepted."""
+    """Detects floats (including scientific notation), unless there is an underscore or a plus sign (bad practice)."""
     try:
         if (
             not isinstance(val, str)
-            or any([k in val for k in ["_", "+", "e", "E"]])
+            or "_" in val
             or (val.startswith("0") and len(val) > 1 and val[1] not in [".", ","])
         ):
             return False
+        elif any([k in val for k in ["+", "e", "E"]]) and not re.match(
+            scientific_notation_pattern, val
+        ):
+            return False
         float_casting(val)
         return True
     except ValueError:
@@ -24,6 +31,6 @@ def _is(val):
 
 
 _test_values = {
-    True: ["1", "0", "1764", "-24", "1.2", "1863.23", "-12.7", "0.1"],
+    True: ["1", "0", "1764", "-24", "1.2", "1863.23", "-12.7", "0.1", "1.9764E-1", "19.01e-29"],
     False: ["01053", "01053.89", "1e3", "123_456", "123_456.78", "+35", "+35.9"],
 }

csv_detective/parsing/columns.py

@@ -48,11 +48,10 @@ def test_col_val(
             for _range in [
                 min(1, ser_len),
                 min(5, ser_len),
-                ser_len,
             ]:
                 if not all(apply_test_func(serie, format.func, _range)):
                     return 0.0
-            return 1.0
+            return float(serie.apply(format.func).sum() == ser_len)
     finally:
         if verbose and time() - start > 3:
             display_logs_depending_process_time(

{csv_detective-0.9.3.dev2486.dist-info → csv_detective-0.9.3.dev2500.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: csv-detective
-Version: 0.9.3.dev2486
+Version: 0.9.3.dev2500
 Summary: Detect tabular files column content
 Keywords: CSV,data processing,encoding,guess,parser,tabular
 Author: data.gouv.fr

{csv_detective-0.9.3.dev2486.dist-info → csv_detective-0.9.3.dev2500.dist-info}/RECORD

@@ -38,7 +38,7 @@ csv_detective/formats/datetime_naive.py,sha256=nvA8qT1fb2RmpXN5_Cw9YZA6pC4BryX_B
 csv_detective/formats/datetime_rfc822.py,sha256=l-SLb34hSuHxC2JQ-9SD-nG38JqzoozwUZiGtoybb0A,601
 csv_detective/formats/departement.py,sha256=UP9UF23BFq_-mIS8N10K5XkoCXwPmDeSoa_7lCAkI4w,768
 csv_detective/formats/email.py,sha256=Qen2EBDYY5TtWXwxrrTGWRrbIybz0ySlVpl4ZRk8pzA,517
-csv_detective/formats/float.py,sha256=tWs_tW64OuacNQENu3uk5GOEVQMQls2iiteFOacQRAQ,832
+csv_detective/formats/float.py,sha256=DF8CwBC4Vk-PFRlIawDr6OUPTtZjAiKYguvilDGUcmY,1033
 csv_detective/formats/geojson.py,sha256=udbBxCBRmb0o6TD8z5ryemfqdinBz6njNJU0XcbfMig,757
 csv_detective/formats/insee_ape700.py,sha256=cLs3Eersqm4wX6oqsqp0Vb3WGPJb2xY5Za_vh0uLgKc,780
 csv_detective/formats/insee_canton.py,sha256=Q5jczsOmh1wPP2KtDkcmqZ7Hlv50Zz9YvPIbxy46qs0,531
@@ -78,7 +78,7 @@ csv_detective/output/profile.py,sha256=VUQp0VJ22dfY4R5TybTpuQW_TOX_rLEp98cOzu-Jf
 csv_detective/output/schema.py,sha256=XoKljXPXP00DfqPCiz1ydwTHYGAFsvNxnaPCNBuuBIo,10443
 csv_detective/output/utils.py,sha256=tbji3dEH7bDc6gLCeVSVquqU3xaHA1CQOMuaJT4Hub8,3297
 csv_detective/parsing/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-csv_detective/parsing/columns.py,sha256=WwivsR4r-SAkugzVSmYeUkgbNXz3CWXnEl2ZmoX_tcs,9238
+csv_detective/parsing/columns.py,sha256=rb5JywbKnYCT3Jb0ZaG1BnyPVtB3gy5mSD-K7qcOl8I,9257
 csv_detective/parsing/compression.py,sha256=Fnw5tj-PpBNI8NYsWj5gD-DUoWcVLnsVpiKm9MpxmIA,350
 csv_detective/parsing/csv.py,sha256=5rw6gXZFQC1T4NT9CnW0AumidrYOkF8kjrfWGmk949I,1716
 csv_detective/parsing/excel.py,sha256=tb65I78tdYlZci_tzvvQt8U6bZSYKjeVdn2CEvsET1o,6972
@@ -86,7 +86,7 @@ csv_detective/parsing/load.py,sha256=f-8aKiNpy_47qg4Lq-UZUR4NNrbJ_-KEGvcUQZ8cmb0
 csv_detective/parsing/text.py,sha256=uz8wfmNTQnOd_4fjrIZ_5rxmFmgrg343hJh2szB73Hc,1770
 csv_detective/utils.py,sha256=RJ_zFOJ1DRY8HtDrKPiCdNk5gU6-KwOrOKOyfSkBZZY,1118
 csv_detective/validate.py,sha256=CjZXhhDP-n6wGgEqbwrGRqebU8L5bidwnvQp-TbnvFA,5424
-csv_detective-0.9.3.dev2486.dist-info/WHEEL,sha256=z-mOpxbJHqy3cq6SvUThBZdaLGFZzdZPtgWLcP2NKjQ,79
-csv_detective-0.9.3.dev2486.dist-info/entry_points.txt,sha256=1J86TQNCanjsLMboAufdEUla03qEQaC9QmVGYgt2FCQ,57
-csv_detective-0.9.3.dev2486.dist-info/METADATA,sha256=XOx2vmkCJGFTCAc0znJ3bcz7V9mAtq7qXqh0GF8ERys,11063
-csv_detective-0.9.3.dev2486.dist-info/RECORD,,
+csv_detective-0.9.3.dev2500.dist-info/WHEEL,sha256=z-mOpxbJHqy3cq6SvUThBZdaLGFZzdZPtgWLcP2NKjQ,79
+csv_detective-0.9.3.dev2500.dist-info/entry_points.txt,sha256=1J86TQNCanjsLMboAufdEUla03qEQaC9QmVGYgt2FCQ,57
+csv_detective-0.9.3.dev2500.dist-info/METADATA,sha256=Xei7oHRc7gmW58t75DcUOF8Jp43mZ6yWZ_WJCj2RHxo,11063
+csv_detective-0.9.3.dev2500.dist-info/RECORD,,