cdiscbuilder 1.1.5__py3-none-any.whl → 1.1.7__py3-none-any.whl

cdiscbuilder/adam/adam_derivation/derivations/sql_derivation.py

@@ -149,9 +149,13 @@ class SQLDerivation(BaseDerivation):
                 # The original code raised ValueError here.
                 raise ValueError(f"Column {source} not found in target dataset")
 
-            # Apply mapping if present (legacy / optimization for local cols)
-            if "mapping" in derivation:
-                series = self._apply_mapping(series, derivation["mapping"])
+        # Auto-Strip Whitespace
+        if series.dtype == pl.Utf8:
+             series = series.str.strip_chars()
+
+        # Apply mapping if present (legacy / optimization for local cols)
+        if "mapping" in derivation:
+            series = self._apply_mapping(series, derivation["mapping"])
 
         return series
 
@@ -425,16 +429,22 @@ class SQLDerivation(BaseDerivation):
                 # Calculate distance to target date
                 if target_col in subject_data.columns and date_col in subject_data.columns:
                     # Get target date (should be same for all rows of this subject)
-                    target_date = subject_data[target_col][0]
-
                     # Calculate date differences and find closest
-                    # Handle partial dates by using strptime with appropriate format
+                    # Handle multiple formats: "YYYY-MM-DD HH:MM:SS", "YYYY-MM-DDTHH:MM:SS" (ISO), or "YYYY-MM-DD"
+                    # We convert everything to Datetime for precision comparison (ms level)
+                    def parse_to_datetime(expr: pl.Expr) -> pl.Expr:
+                        return (
+                            expr.str.strptime(pl.Datetime, "%Y-%m-%d %H:%M:%S", strict=False)
+                            .fill_null(expr.str.strptime(pl.Datetime, "%Y-%m-%dT%H:%M:%S", strict=False))
+                            .fill_null(expr.str.strptime(pl.Date, "%Y-%m-%d", strict=False).cast(pl.Datetime))
+                        )
+
                     with_diff = subject_data.with_columns(
                         (
-                            pl.col(date_col).str.strptime(pl.Date, "%Y-%m-%d", strict=False)
-                            - pl.lit(target_date).str.strptime(pl.Date, "%Y-%m-%d", strict=False)
+                            parse_to_datetime(pl.col(date_col))
+                            - parse_to_datetime(pl.lit(target_date))
                         )
-                        .dt.total_days()
+                        .dt.total_milliseconds()
                         .abs()
                         .alias("date_diff")
                     )

cdiscbuilder/sdtm/engine/classes/finding.py

@@ -111,6 +111,10 @@ class FindingProcessor:
                     elif source_expr in source_df.columns:
                          series = source_df[source_expr]
                     
+                    # Auto-Strip
+                    if series is not None and pd.api.types.is_object_dtype(series):
+                         series = series.astype(str).str.strip().replace('nan', None)
+                    
                     if regex_extract and series is not None:
                          # Extract group 1
                          series = series.astype(str).str.extract(regex_extract)[0]

cdiscbuilder/sdtm/engine/classes/general.py

@@ -185,6 +185,13 @@ class GeneralProcessor:
                         # Source defined but not found.
                         print(f"Warning: Source column '{source_expr}' not found for '{domain_name}.{target_col}'. Filling with NaN.")
                         series = pd.Series([None] * len(pivoted))
+                    
+                    # Auto-Strip Whitespace for strings
+                    if series is not None and pd.api.types.is_object_dtype(series):
+                        try:
+                            series = series.astype(str).str.strip().replace('nan', None)
+                        except:
+                            pass
                 else:
                     print(f"Warning: No source or literal defined for '{domain_name}.{target_col}'. Filling with NaN.")
                     series = pd.Series([None] * len(pivoted))

{cdiscbuilder-1.1.5.dist-info → cdiscbuilder-1.1.7.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: cdiscbuilder
-Version: 1.1.5
+Version: 1.1.7
 Summary: A package to convert ODM XML to SDTM/ADaM Datasets
 Author-email: Ming-Chun Chen <hellomingchun@gmail.com>
 Requires-Python: >=3.8

{cdiscbuilder-1.1.5.dist-info → cdiscbuilder-1.1.7.dist-info}/RECORD

@@ -7,7 +7,7 @@ cdiscbuilder/adam/adam_derivation/engine.py,sha256=5Wk_tc3c6nRKBiiL5gE7YXxvdPgXV
 cdiscbuilder/adam/adam_derivation/derivations/__init__.py,sha256=iu7xzlQF_uK_5nwYw5ovEsBDNdaX0S8oVFYYxcxheTc,213
 cdiscbuilder/adam/adam_derivation/derivations/base.py,sha256=thbi0UD17Z8hUA5G6np_wWDcFth-UmoY7moS2KzvIps,1242
 cdiscbuilder/adam/adam_derivation/derivations/function_derivation.py,sha256=mPBet5JR10RS9euaS9ZLE1GCMRS3Grp9Gc3hdiQOVDI,7167
-cdiscbuilder/adam/adam_derivation/derivations/sql_derivation.py,sha256=rXk7vvvMozgAQqFZs2IM5VPGreRFoDqxRkmJ8T62x9s,22603
+cdiscbuilder/adam/adam_derivation/derivations/sql_derivation.py,sha256=vy8u8Et9-DF-rObHjxTWsoaMp95z5cZdqTtP69HGbyM,23166
 cdiscbuilder/adam/adam_derivation/functions/__init__.py,sha256=Cv4pLpgHTr9i2zf7owbpZwMyxm8JLFmX4J31dvX0jYg,1773
 cdiscbuilder/adam/adam_derivation/functions/get_bmi.py,sha256=RR1lQpRSDO4cEkhxjnPutAy_xfVvfEwmhYR90XK2eMM,229
 cdiscbuilder/adam/adam_derivation/loaders/__init__.py,sha256=eUnUKzyahIaLIqjn39sYzXvzPpfabZoipkPdQ5mA9IE,116
@@ -42,16 +42,16 @@ cdiscbuilder/sdtm/engine/functions.py,sha256=SdyhFftFoxIzL-pzIcjVsf9CbZmxKKo7J1J
 cdiscbuilder/sdtm/engine/processor.py,sha256=7ohnIY8KEGxUA0ENZm4_3yKkxaJzkboRTeL3d5AX-Ac,5040
 cdiscbuilder/sdtm/engine/validate.py,sha256=vGUirL3yPk6a9D6MlRpmHf54tB-PMNH7FMKCxsXROgQ,3799
 cdiscbuilder/sdtm/engine/classes/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-cdiscbuilder/sdtm/engine/classes/finding.py,sha256=DtOdQWZU_68jVOyZNec2HXnMl8UfNBK1xiI781AYuoo,8302
-cdiscbuilder/sdtm/engine/classes/general.py,sha256=0-KP2mNfwRjeunDt-ZY0q2t-_k7bSwCglXrotsPN3kI,17082
+cdiscbuilder/sdtm/engine/classes/finding.py,sha256=g_16w_PBHIZLe8jH1bpxY-QKzaSAXrtAhFYOIetDCRU,8530
+cdiscbuilder/sdtm/engine/classes/general.py,sha256=htUgE12XL_hGRW5VEQrfXlX_-j5KQwvEHgoXIa4TbCE,17433
 cdiscbuilder/sdtm/loader/__init__.py,sha256=ZL0EV0TnNNvOslawOpRK0wkzIjVeiPzY2IpfyiN0lgw,104
 cdiscbuilder/sdtm/loader/load.py,sha256=30x9fAEjnuRq313SRq9I7mBwrWl4sEhiKkmrr9x1cjo,6171
 cdiscbuilder/sdtm/loader/tests/__init__.py,sha256=ujCd31KMb52gvkWkVBBogLBda3aiRmbqS0nRS-yTnk0,69
 cdiscbuilder/sdtm/loader/tests/test_load.py,sha256=yrXFQlzH_EkN6HxJc7ItFie8xqwrs2XxwNu2wGaxjok,11180
 cdiscbuilder/tlf/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-cdiscbuilder-1.1.5.dist-info/licenses/LICENSE,sha256=Da4Pm8tEdB1ycKkAkF2EwfHd-8l0wrJ_uwejzfBRHFE,1092
-cdiscbuilder-1.1.5.dist-info/METADATA,sha256=EvG3y-fCShEnUX2cPuOVOSj-PQBH7FpMXEu6-J4SF-Y,356
-cdiscbuilder-1.1.5.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
-cdiscbuilder-1.1.5.dist-info/entry_points.txt,sha256=KsuiJD4hsfjRqDkdvWgLFPeQCz7QBl73RA665BmF0Pw,53
-cdiscbuilder-1.1.5.dist-info/top_level.txt,sha256=3x5ENgnXBENgXKp5Zsht_4_OmA5wLVvDuk0UEkYOGNw,13
-cdiscbuilder-1.1.5.dist-info/RECORD,,
+cdiscbuilder-1.1.7.dist-info/licenses/LICENSE,sha256=Da4Pm8tEdB1ycKkAkF2EwfHd-8l0wrJ_uwejzfBRHFE,1092
+cdiscbuilder-1.1.7.dist-info/METADATA,sha256=P20HXeNmjdmImD2_DfSLw3FlBddDH0MNO4CHJu4XzAs,356
+cdiscbuilder-1.1.7.dist-info/WHEEL,sha256=wUyA8OaulRlbfwMtmQsvNngGrxQHAvkKcvRmdizlJi0,92
+cdiscbuilder-1.1.7.dist-info/entry_points.txt,sha256=KsuiJD4hsfjRqDkdvWgLFPeQCz7QBl73RA665BmF0Pw,53
+cdiscbuilder-1.1.7.dist-info/top_level.txt,sha256=3x5ENgnXBENgXKp5Zsht_4_OmA5wLVvDuk0UEkYOGNw,13
+cdiscbuilder-1.1.7.dist-info/RECORD,,