csv-detective 0.7.5.dev1298__py3-none-any.whl → 0.7.5.dev1320__py3-none-any.whl

csv_detective/detect_fields/FR/other/code_import/__init__.py

@@ -0,0 +1,9 @@
+import re
+
+PROPORTION = 0.9
+regex = r"^(\d{3}[SP]\d{4,10}(.\w{1,3}\d{0,5})?|\d[A-Z0-9]\d[SP]\w(\w-?\w{0,2}\d{0,6})?)$"
+
+
+def _is(val):
+    """Repere le code Import (ancien RNA)"""
+    return isinstance(val, str) and bool(re.match(regex, val))

csv_detective/detect_fields/FR/other/code_waldec/__init__.py

@@ -1,12 +1,9 @@
 import re
 
 PROPORTION = 0.9
-regex = (
-    r'^\d{3}\D\d{1,10}$|^\d\D\d\D\d{10}$|^\d{3}\D{3}\d{1,10}$|^\d{3}\D\d{4}\D\d{1,10}'
-    r'$|^\d{3}\D\d{2}[-]\d{3}$|^\d\D\d\D\d{2}\D\d{1,8}$'
-)
+regex = r"^W\d[\dA-Z]\d{7}$"
 
 
 def _is(val):
-    '''Repere le code Waldec'''
+    """Repere le code Waldec"""
     return isinstance(val, str) and bool(re.match(regex, val))

csv_detective/detect_fields/__init__.py

@@ -9,6 +9,7 @@ from .FR.other import (
     siret,
     insee_ape700,
     date_fr,
+    code_import,
     code_waldec,
     code_rna,
 )

csv_detective/validate.py

@@ -42,7 +42,8 @@ def validate(
         any(col_name not in list(table.columns) for col_name in previous_analysis["columns"])
         or any(col_name not in list(previous_analysis["columns"].keys()) for col_name in table.columns)
     ):
-        logging.warning("> Columns do not match, proceeding with full analysis")
+        if verbose:
+            logging.warning("> Columns do not match, proceeding with full analysis")
         return False, table, analysis
     for col_name, args in previous_analysis["columns"].items():
         if verbose:
@@ -55,7 +56,8 @@ def validate(
         if skipna:
             col_data = col_data.loc[~col_data.isna()]
         if not col_data.apply(test_func).all():
-            logging.warning("> Test failed, proceeding with full analysis")
+            if verbose:
+                logging.warning("> Test failed, proceeding with full analysis")
             return False, table, analysis
     if verbose:
         logging.info("> All checks successful")

{csv_detective-0.7.5.dev1298.data → csv_detective-0.7.5.dev1320.data}/data/share/csv_detective/CHANGELOG.md

@@ -16,6 +16,7 @@
 - Add validation function and associated flow [#112](https://github.com/datagouv/csv-detective/pull/112)
 - Better float detection [#113](https://github.com/datagouv/csv-detective/pull/113)
 - Refactor fields tests [#114](https://github.com/datagouv/csv-detective/pull/114)
+- Better code waldec and add code import [#116](https://github.com/datagouv/csv-detective/pull/116)
 
 ## 0.7.4 (2024-11-15)
 

{csv_detective-0.7.5.dev1298.dist-info → csv_detective-0.7.5.dev1320.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: csv_detective
-Version: 0.7.5.dev1298
+Version: 0.7.5.dev1320
 Summary: Detect CSV column content
 Home-page: https://github.com/etalab/csv_detective
 Author: Etalab

{csv_detective-0.7.5.dev1298.dist-info → csv_detective-0.7.5.dev1320.dist-info}/RECORD

@@ -4,8 +4,8 @@ csv_detective/explore_csv.py,sha256=ocWlUEtuwZ-6bjDc6gfhC2-6DljMVhvXhHrfICCXGfQ,
 csv_detective/load_tests.py,sha256=GILvfkd4OVI-72mA4nzbPlZqgcXZ4wznOhGfZ1ucWkM,2385
 csv_detective/s3_utils.py,sha256=1cIVdQUYY2ovErbMwp72Gqtqx2bkB8nfVhn-QaOFTT0,1451
 csv_detective/utils.py,sha256=Bx_1k4Sdpd5PCjuAy4AeayCmmw7TMR_zgtKIHNLi5g0,1157
-csv_detective/validate.py,sha256=o4Qulf8E-x1zsWT9OD4Fpw83Gku1WA3JlX83j7bu0DA,2314
-csv_detective/detect_fields/__init__.py,sha256=qkwT_o_S7qvLEsRssICpoGmCc3h5y2MVy1XI56LFcV0,959
+csv_detective/validate.py,sha256=0wSi5GgKPRW3m66413a-9Uti1vBRam5pQxVA9Dc5jQ8,2368
+csv_detective/detect_fields/__init__.py,sha256=7Tz0Niaz0BboA3YVsp_6WPA6ywciwDN4-lOy_Ie_0Y8,976
 csv_detective/detect_fields/FR/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 csv_detective/detect_fields/FR/geo/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 csv_detective/detect_fields/FR/geo/adresse/__init__.py,sha256=NqV8ULf9gY9iFnA1deKR-1Yobr96WwCsn5JfbP_MjiY,1675
@@ -26,8 +26,9 @@ csv_detective/detect_fields/FR/geo/region/__init__.py,sha256=JbFKDd4jAnd9yb7YqP3
 csv_detective/detect_fields/FR/other/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 csv_detective/detect_fields/FR/other/code_csp_insee/__init__.py,sha256=SRWJvg3Ikyjmop9iL14igTjxNGpO-QB3fpADI_bLYEY,566
 csv_detective/detect_fields/FR/other/code_csp_insee/code_csp_insee.txt,sha256=rbcjtMP6qTZ7BTU6ZegkiXKCruqY_m9Ep6ZgRabFS_E,2486
+csv_detective/detect_fields/FR/other/code_import/__init__.py,sha256=zJ9YfPa5p--uHNQFeO1gTjxDy2Um_r-MxQd29VBNjFw,243
 csv_detective/detect_fields/FR/other/code_rna/__init__.py,sha256=Z0RjMBt1--ZL7Jd1RsHAQCCbTAQk_BnlnTq8VF1o_VA,146
-csv_detective/detect_fields/FR/other/code_waldec/__init__.py,sha256=g9n5sOjRlk4I9YFZjdaTYrXf8ftXRDunGZOUpYhN4fA,295
+csv_detective/detect_fields/FR/other/code_waldec/__init__.py,sha256=41SYNzCzUFh4trQlwG-9UC0-1Wi4fTcv8Byi_dd9Lq4,168
 csv_detective/detect_fields/FR/other/csp_insee/__init__.py,sha256=lvcaVKgOPrCaZb-Y1-wYCbLYB_CQjCJFNAzfWDwtTVE,496
 csv_detective/detect_fields/FR/other/csp_insee/csp_insee.txt,sha256=kgKaKc-5PHu5U4--ugLjpFyMNtTU9CGdZ9ANU3YAsM4,32879
 csv_detective/detect_fields/FR/other/date_fr/__init__.py,sha256=kMV52djlG0y4o0ELEZuvTv_FvooYOgTnV1aWhycFJDc,284
@@ -147,19 +148,19 @@ csv_detective/parsing/csv.py,sha256=11mibDnJhIjykXLGZvA5ZEU5U7KgxIrbyO6BNv6jlro,
 csv_detective/parsing/excel.py,sha256=AslE2S1e67o8yTIAIhp-lAnJ6-XqeBBRz1-VMFqhZBM,7055
 csv_detective/parsing/load.py,sha256=SpP0pfxswOAPPpwbZfoP1blh0EKV5VMs0TpTgQJKzjs,3621
 csv_detective/parsing/text.py,sha256=rsfk66BCmdpsCOd0kDJ8tmqMsEWd-OeBkEisWc4Ej9k,1246
-csv_detective-0.7.5.dev1298.data/data/share/csv_detective/CHANGELOG.md,sha256=Y8aL18x5EGGvA9AqukEi4tn78se_Lzisa2J32kOSer8,7984
-csv_detective-0.7.5.dev1298.data/data/share/csv_detective/LICENSE.AGPL.txt,sha256=2N5ReRelkdqkR9a-KP-y-shmcD5P62XoYiG-miLTAzo,34519
-csv_detective-0.7.5.dev1298.data/data/share/csv_detective/README.md,sha256=Qr8xRXc-dxQ-tdXCpCTCKp1Uliqq84r0UOlPRNuGCpI,9506
-csv_detective-0.7.5.dev1298.dist-info/licenses/LICENSE.AGPL.txt,sha256=2N5ReRelkdqkR9a-KP-y-shmcD5P62XoYiG-miLTAzo,34519
+csv_detective-0.7.5.dev1320.data/data/share/csv_detective/CHANGELOG.md,sha256=aFDguybPGcPheztzpQNq-YVZZW1n8prG1txK4b32DhM,8084
+csv_detective-0.7.5.dev1320.data/data/share/csv_detective/LICENSE.AGPL.txt,sha256=2N5ReRelkdqkR9a-KP-y-shmcD5P62XoYiG-miLTAzo,34519
+csv_detective-0.7.5.dev1320.data/data/share/csv_detective/README.md,sha256=Qr8xRXc-dxQ-tdXCpCTCKp1Uliqq84r0UOlPRNuGCpI,9506
+csv_detective-0.7.5.dev1320.dist-info/licenses/LICENSE.AGPL.txt,sha256=2N5ReRelkdqkR9a-KP-y-shmcD5P62XoYiG-miLTAzo,34519
 tests/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 tests/test_example.py,sha256=JeHxSK0IVDcSrOhSZlNGSQv4JAc_r6mzvJM8PfmLTMw,2018
-tests/test_fields.py,sha256=0hce2XtDHY9dTLCYhrm2s4I41OeKsQbbaKmDZ4XctUw,9824
+tests/test_fields.py,sha256=E6kEsp6_W56WW6FXWUl7hggsJv-vsKuOaJ9JLoFmrUw,9964
 tests/test_file.py,sha256=9APE1d43lQ8Dk8lwJFNUK_YekYYsQ0ae2_fgpcPE9mk,8116
 tests/test_labels.py,sha256=6MOKrGznkwU5fjZ_3oiB6Scmb480Eu-9geBJs0UDLds,159
 tests/test_structure.py,sha256=bv-tjgXohvQAxwmxzH0BynFpK2TyPjcxvtIAmIRlZmA,1393
 tests/test_validation.py,sha256=VwtBcnGAQ_eSFrBibWnMSTDjuy6y2JLlqvc3Zb667NY,479
-csv_detective-0.7.5.dev1298.dist-info/METADATA,sha256=cy8kKhsbQVd8DQ2UMJe7z1nyxoGEvmFnQfsTdCTwXXc,1386
-csv_detective-0.7.5.dev1298.dist-info/WHEEL,sha256=SmOxYU7pzNKBqASvQJ7DjX3XGUF92lrGhMb3R6_iiqI,91
-csv_detective-0.7.5.dev1298.dist-info/entry_points.txt,sha256=JjweTReFqKJmuvkegzlew2j3D5pZzfxvbEGOtGVGmaY,56
-csv_detective-0.7.5.dev1298.dist-info/top_level.txt,sha256=M0Nv646VHo-49zWjPkwo2C48UmtfddV8_9mEZeIxy8Q,20
-csv_detective-0.7.5.dev1298.dist-info/RECORD,,
+csv_detective-0.7.5.dev1320.dist-info/METADATA,sha256=lxx-TBya3ciYiOlxVY6YGAd7MVv7D6ChKGtl6gJDkRE,1386
+csv_detective-0.7.5.dev1320.dist-info/WHEEL,sha256=DnLRTWE75wApRYVsjgc6wsVswC54sMSJhAEd4xhDpBk,91
+csv_detective-0.7.5.dev1320.dist-info/entry_points.txt,sha256=JjweTReFqKJmuvkegzlew2j3D5pZzfxvbEGOtGVGmaY,56
+csv_detective-0.7.5.dev1320.dist-info/top_level.txt,sha256=M0Nv646VHo-49zWjPkwo2C48UmtfddV8_9mEZeIxy8Q,20
+csv_detective-0.7.5.dev1320.dist-info/RECORD,,

{csv_detective-0.7.5.dev1298.dist-info → csv_detective-0.7.5.dev1320.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (79.0.1)
+Generator: setuptools (80.4.0)
 Root-Is-Purelib: true
 Tag: py3-none-any
 

tests/test_fields.py

@@ -24,6 +24,7 @@ from csv_detective.detect_fields.FR.geo import (
 from csv_detective.detect_fields.FR.other import (
     code_csp_insee,
     code_rna,
+    code_import,
     code_waldec,
     csp_insee,
     date_fr,
@@ -184,8 +185,12 @@ fields = {
             "Wa1#89sf&h",
         ],
     },
+    code_import: {
+        True: ["123S1871092288"],
+        False: ["AA751PEE00188854", "W123456789"],
+    },
     code_waldec: {
-        True: ["751P00188854"],
+        True: ["W123456789", "W2D1234567"],
         False: ["AA751PEE00188854"],
     },
     csp_insee: {