BUG: pandas.to_datetime() does not respect exact format string with ISO8601

.pre-commit-config.yaml

@@ -48,7 +48,13 @@ repos:
         # this particular codebase (e.g. src/headers, src/klib). However,
         # we can lint all header files since they aren't "generated" like C files are.
         exclude: ^pandas/_libs/src/(klib|headers)/
-        args: [--quiet, '--extensions=c,h', '--headers=h', --recursive, '--filter=-readability/casting,-runtime/int,-build/include_subdir']
+        args: [
+            --quiet,
+            '--extensions=c,h',
+            '--headers=h',
+            --recursive,
+            '--filter=-readability/casting,-runtime/int,-build/include_subdir,-readability/fn_size'
+        ]
 -   repo: https://github.com/PyCQA/flake8
     rev: 5.0.4
     hooks:

doc/source/whatsnew/v2.0.0.rst

@@ -576,6 +576,7 @@ Conversion
 - Bug in :meth:`Series.convert_dtypes` not converting dtype to nullable dtype when :class:`Series` contains ``NA`` and has dtype ``object`` (:issue:`48791`)
 - Bug where any :class:`ExtensionDtype` subclass with ``kind="M"`` would be interpreted as a timezone type (:issue:`34986`)
 - Bug in :class:`.arrays.ArrowExtensionArray` that would raise ``NotImplementedError`` when passed a sequence of strings or binary (:issue:`49172`)
+- Bug in :func:`to_datetime` was not respecting ``exact`` argument when ``format`` was an ISO8601 format (:issue:`12649`)
 
 Strings
 ^^^^^^^

pandas/_libs/tslib.pyi

@@ -25,6 +25,8 @@ def array_to_datetime(
     utc: bool = ...,
     require_iso8601: bool = ...,
     allow_mixed: bool = ...,
+    format: str | None = ...,
+    exact: bool = ...,
 ) -> tuple[np.ndarray, tzinfo | None]: ...
 
 # returned ndarray may be object dtype or datetime64[ns]

pandas/_libs/tslib.pyx

@@ -447,6 +447,8 @@ cpdef array_to_datetime(
     bint utc=False,
     bint require_iso8601=False,
     bint allow_mixed=False,
+    format: str | None=None,
+    bint exact=True,
 ):
     """
     Converts a 1D array of date-like values to a numpy array of either:
@@ -566,6 +568,16 @@ cpdef array_to_datetime(
                     iresult[i] = get_datetime64_nanos(val, NPY_FR_ns)
 
                 elif is_integer_object(val) or is_float_object(val):
+                    if require_iso8601:
+                        if is_coerce:
+                            iresult[i] = NPY_NAT
+                            continue
+                        elif is_raise:
+                            raise ValueError(
+                                f"time data \"{val}\" at position {i} doesn't "
+                                f"match format \"{format}\""
+                            )
+                        return values, tz_out
                     # these must be ns unit by-definition
                     seen_integer = True
 
@@ -596,7 +608,7 @@ cpdef array_to_datetime(
 
                     string_to_dts_failed = string_to_dts(
                         val, &dts, &out_bestunit, &out_local,
-                        &out_tzoffset, False
+                        &out_tzoffset, False, format, exact
                     )
                     if string_to_dts_failed:
                         # An error at this point is a _parsing_ error
@@ -612,7 +624,7 @@ cpdef array_to_datetime(
                             elif is_raise:
                                 raise ValueError(
                                     f"time data \"{val}\" at position {i} doesn't "
-                                    "match format specified"
+                                    f"match format \"{format}\""
                                 )
                             return values, tz_out
 

pandas/_libs/tslibs/np_datetime.pxd

@@ -95,6 +95,8 @@ cdef int string_to_dts(
     int* out_local,
     int* out_tzoffset,
     bint want_exc,
+    format: str | None = *,
+    bint exact = *
 ) except? -1
 
 cdef NPY_DATETIMEUNIT get_unit_from_dtype(cnp.dtype dtype)

pandas/_libs/tslibs/np_datetime.pyx

@@ -52,7 +52,8 @@ cdef extern from "src/datetime/np_datetime_strings.h":
     int parse_iso_8601_datetime(const char *str, int len, int want_exc,
                                 npy_datetimestruct *out,
                                 NPY_DATETIMEUNIT *out_bestunit,
-                                int *out_local, int *out_tzoffset)
+                                int *out_local, int *out_tzoffset,
+                                const char *format, int format_len, int exact)
 
 
 # ----------------------------------------------------------------------
@@ -277,14 +278,25 @@ cdef inline int string_to_dts(
     int* out_local,
     int* out_tzoffset,
     bint want_exc,
+    format: str | None=None,
+    bint exact=True,
 ) except? -1:
     cdef:
         Py_ssize_t length
         const char* buf
+        Py_ssize_t format_length
+        const char* format_buf
 
     buf = get_c_string_buf_and_size(val, &length)
+    if format is None:
+        format_buf = b''
+        format_length = 0
+        exact = False
+    else:
+        format_buf = get_c_string_buf_and_size(format, &format_length)
     return parse_iso_8601_datetime(buf, length, want_exc,
-                                   dts, out_bestunit, out_local, out_tzoffset)
+                                   dts, out_bestunit, out_local, out_tzoffset,
+                                   format_buf, format_length, exact)
 
 
 cpdef ndarray astype_overflowsafe(