From 45f82c3be1fd147aebf23c8fc2de16361d93b7cb Mon Sep 17 00:00:00 2001
From: Nikita Vedeneev <nik@quansight.com>
Date: Wed, 26 Oct 2022 13:50:29 +0000
Subject: [PATCH 01/18] initial format support

Co-Authored-By: MarcoGorelli <>
Co-Authored-By: FDRocha <>
---
 doc/source/whatsnew/v2.0.0.rst                |   1 +
 pandas/_libs/tslib.pyx                        |  17 ++-
 pandas/_libs/tslibs/conversion.pyx            |   2 +-
 pandas/_libs/tslibs/np_datetime.pxd           |   2 +
 pandas/_libs/tslibs/np_datetime.pyx           |  11 +-
 pandas/_libs/tslibs/parsing.pyx               |   2 +-
 .../tslibs/src/datetime/np_datetime_strings.c |  74 +++++++++-
 .../tslibs/src/datetime/np_datetime_strings.h |   5 +-
 pandas/core/arrays/datetimes.py               |   4 +
 pandas/core/tools/datetimes.py                |   9 +-
 pandas/tests/tools/test_to_datetime.py        | 126 +++++++++++++++++-
 11 files changed, 239 insertions(+), 14 deletions(-)

diff --git a/doc/source/whatsnew/v2.0.0.rst b/doc/source/whatsnew/v2.0.0.rst
index b5c5a721324c6..e3cc6e3c9bc1e 100644
--- a/doc/source/whatsnew/v2.0.0.rst
+++ b/doc/source/whatsnew/v2.0.0.rst
@@ -333,6 +333,7 @@ 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
 ^^^^^^^
diff --git a/pandas/_libs/tslib.pyx b/pandas/_libs/tslib.pyx
index d7c0c91332e02..4fd015f15ed76 100644
--- a/pandas/_libs/tslib.pyx
+++ b/pandas/_libs/tslib.pyx
@@ -89,7 +89,7 @@ def _test_parse_iso8601(ts: str):
     elif ts == 'today':
         return Timestamp.now().normalize()
 
-    string_to_dts(ts, &obj.dts, &out_bestunit, &out_local, &out_tzoffset, True)
+    string_to_dts(ts, &obj.dts, &out_bestunit, &out_local, &out_tzoffset, True, "", False)
     obj.value = npy_datetimestruct_to_datetime(NPY_FR_ns, &obj.dts)
     check_dts_bounds(&obj.dts)
     if out_local == 1:
@@ -445,6 +445,8 @@ cpdef array_to_datetime(
     bint utc=False,
     bint require_iso8601=False,
     bint allow_mixed=False,
+    str format="",
+    bint exact=False,
 ):
     """
     Converts a 1D array of date-like values to a numpy array of either:
@@ -564,6 +566,15 @@ 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 match format \"{format}\""
+                            )
+                        return values, tz_out
                     # these must be ns unit by-definition
                     seen_integer = True
 
@@ -594,7 +605,7 @@ 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
@@ -609,7 +620,7 @@ cpdef array_to_datetime(
                                 continue
                             elif is_raise:
                                 raise ValueError(
-                                    f"time data \"{val}\" at position {i} doesn't match format specified"
+                                    f"time data \"{val}\" at position {i} doesn't match format \"{format}\""
                                 )
                             return values, tz_out
 
diff --git a/pandas/_libs/tslibs/conversion.pyx b/pandas/_libs/tslibs/conversion.pyx
index 923dfa3c54d26..481f66bc256bf 100644
--- a/pandas/_libs/tslibs/conversion.pyx
+++ b/pandas/_libs/tslibs/conversion.pyx
@@ -488,7 +488,7 @@ cdef _TSObject _convert_str_to_tsobject(object ts, tzinfo tz, str unit,
     else:
         string_to_dts_failed = string_to_dts(
             ts, &dts, &out_bestunit, &out_local,
-            &out_tzoffset, False
+            &out_tzoffset, False, "", False
         )
         if not string_to_dts_failed:
             try:
diff --git a/pandas/_libs/tslibs/np_datetime.pxd b/pandas/_libs/tslibs/np_datetime.pxd
index e51bbd4e074e1..770b1f7e46c12 100644
--- a/pandas/_libs/tslibs/np_datetime.pxd
+++ b/pandas/_libs/tslibs/np_datetime.pxd
@@ -95,6 +95,8 @@ cdef int string_to_dts(
     int* out_local,
     int* out_tzoffset,
     bint want_exc,
+    str format,
+    bint exact
 ) except? -1
 
 cdef NPY_DATETIMEUNIT get_unit_from_dtype(cnp.dtype dtype)
diff --git a/pandas/_libs/tslibs/np_datetime.pyx b/pandas/_libs/tslibs/np_datetime.pyx
index 07872050dc822..8acd51d8a4431 100644
--- a/pandas/_libs/tslibs/np_datetime.pyx
+++ b/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)
 
 
 # ----------------------------------------------------------------------
@@ -273,14 +274,20 @@ cdef inline int string_to_dts(
     int* out_local,
     int* out_tzoffset,
     bint want_exc,
+    str format,
+    bint exact,
 ) 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)
+    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(
diff --git a/pandas/_libs/tslibs/parsing.pyx b/pandas/_libs/tslibs/parsing.pyx
index 469e0721f1207..eeb738bdc3c1a 100644
--- a/pandas/_libs/tslibs/parsing.pyx
+++ b/pandas/_libs/tslibs/parsing.pyx
@@ -409,7 +409,7 @@ cdef parse_datetime_string_with_reso(
     # TODO: does this render some/all of parse_delimited_date redundant?
     string_to_dts_failed = string_to_dts(
         date_string, &dts, &out_bestunit, &out_local,
-        &out_tzoffset, False
+        &out_tzoffset, False, "", False
     )
     if not string_to_dts_failed:
         if dts.ps != 0 or out_local:
diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index cfbaed01b57c9..9a95bf44ddd6d 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -66,10 +66,29 @@ This file implements string parsing and creation for NumPy datetime.
  *
  * Returns 0 on success, -1 on failure.
  */
+
+#define FORMAT_STARTSWITH(ch) \
+  if (exact) { \
+    if (!format_len || *format != ch) { \
+        goto parse_error; \
+    } \
+    ++format; \
+    --format_len; \
+  } else { \
+    if (format_len > 0) { \
+        if (*format != ch) { \
+            goto parse_error;  \
+        } \
+        ++format; \
+        --format_len; \
+    } \
+  } \
+
 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) {
     int year_leap = 0;
     int i, numdigits;
     const char *substr;
@@ -104,14 +123,19 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
+        FORMAT_STARTSWITH(' ');
     }
 
     /* Leading '-' sign for negative year */
     if (*substr == '-') {
         ++substr;
         --sublen;
+        FORMAT_STARTSWITH('-');
     }
 
+    FORMAT_STARTSWITH('%');
+    FORMAT_STARTSWITH('Y');
+
     if (sublen == 0) {
         goto parse_error;
     }
@@ -139,6 +163,9 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (out_local != NULL) {
             *out_local = 0;
         }
+        if (format_len) {
+          goto parse_error;
+        }
         bestunit = NPY_FR_Y;
         goto finish;
     }
@@ -156,6 +183,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ymd_sep = valid_ymd_sep[i];
         ++substr;
         --sublen;
+        FORMAT_STARTSWITH(ymd_sep);
         /* Cannot have trailing separator */
         if (sublen == 0 || !isdigit(*substr)) {
             goto parse_error;
@@ -167,6 +195,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     out->month = (*substr - '0');
     ++substr;
     --sublen;
+    FORMAT_STARTSWITH('%');
+    FORMAT_STARTSWITH('m');
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->month = 10 * out->month + (*substr - '0');
@@ -190,6 +220,9 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (!has_ymd_sep) {
             goto parse_error;
         }
+        if (format_len) {
+          goto parse_error;
+        }
         if (out_local != NULL) {
             *out_local = 0;
         }
@@ -203,6 +236,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         }
         ++substr;
         --sublen;
+        FORMAT_STARTSWITH(ymd_sep);
     }
 
     /* PARSE THE DAY */
@@ -213,6 +247,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     out->day = (*substr - '0');
     ++substr;
     --sublen;
+    FORMAT_STARTSWITH('%');
+    FORMAT_STARTSWITH('d');
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->day = 10 * out->day + (*substr - '0');
@@ -235,6 +271,9 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (out_local != NULL) {
             *out_local = 0;
         }
+        if (format_len) {
+          goto parse_error;
+        }
         bestunit = NPY_FR_D;
         goto finish;
     }
@@ -242,6 +281,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if ((*substr != 'T' && *substr != ' ') || sublen == 1) {
         goto parse_error;
     }
+    FORMAT_STARTSWITH(*substr);
     ++substr;
     --sublen;
 
@@ -250,6 +290,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if (!isdigit(*substr)) {
         goto parse_error;
     }
+    FORMAT_STARTSWITH('%');
+    FORMAT_STARTSWITH('H');
     out->hour = (*substr - '0');
     ++substr;
     --sublen;
@@ -274,6 +316,9 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (!hour_was_2_digits) {
             goto parse_error;
         }
+        if (format_len) {
+          goto parse_error;
+        }
         bestunit = NPY_FR_h;
         goto finish;
     }
@@ -286,6 +331,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (sublen == 0 || !isdigit(*substr)) {
             goto parse_error;
         }
+        FORMAT_STARTSWITH(':');
     } else if (!isdigit(*substr)) {
         if (!hour_was_2_digits) {
             goto parse_error;
@@ -298,6 +344,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     out->min = (*substr - '0');
     ++substr;
     --sublen;
+    FORMAT_STARTSWITH('%');
+    FORMAT_STARTSWITH('M');
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->min = 10 * out->min + (*substr - '0');
@@ -317,12 +365,16 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
 
     if (sublen == 0) {
         bestunit = NPY_FR_m;
+        if (format_len) {
+          goto parse_error;
+        }
         goto finish;
     }
 
     /* If we make it through this condition block, then the next
      * character is a digit. */
     if (has_hms_sep && *substr == ':') {
+        FORMAT_STARTSWITH(':');
         ++substr;
         --sublen;
         /* Cannot have a trailing ':' */
@@ -339,6 +391,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     out->sec = (*substr - '0');
     ++substr;
     --sublen;
+    FORMAT_STARTSWITH('%');
+    FORMAT_STARTSWITH('S');
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->sec = 10 * out->sec + (*substr - '0');
@@ -360,12 +414,15 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if (sublen > 0 && *substr == '.') {
         ++substr;
         --sublen;
+        FORMAT_STARTSWITH('.');
     } else {
         bestunit = NPY_FR_s;
         goto parse_timezone;
     }
 
     /* PARSE THE MICROSECONDS (0 to 6 digits) */
+    FORMAT_STARTSWITH('%');
+    FORMAT_STARTSWITH('f');
     numdigits = 0;
     for (i = 0; i < 6; ++i) {
         out->us *= 10;
@@ -430,15 +487,22 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
+        FORMAT_STARTSWITH(' ');
     }
 
     if (sublen == 0) {
         // Unlike NumPy, treating no time zone as naive
+        if (format_len > 0) {
+            goto parse_error;
+        }
         goto finish;
     }
 
     /* UTC specifier */
     if (*substr == 'Z') {
+        FORMAT_STARTSWITH('%');
+        FORMAT_STARTSWITH('Z');
+
         /* "Z" should be equivalent to tz offset "+00:00" */
         if (out_local != NULL) {
             *out_local = 1;
@@ -449,12 +513,17 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         }
 
         if (sublen == 1) {
+            if (format_len > 0) {
+                goto parse_error;
+            }
             goto finish;
         } else {
             ++substr;
             --sublen;
         }
     } else if (*substr == '-' || *substr == '+') {
+        FORMAT_STARTSWITH('%');
+        FORMAT_STARTSWITH('z');
         /* Time zone offset */
         int offset_neg = 0, offset_hour = 0, offset_minute = 0;
 
@@ -538,9 +607,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
+        FORMAT_STARTSWITH(' ');
     }
 
-    if (sublen != 0) {
+    if ((sublen != 0) || (format_len != 0)) {
         goto parse_error;
     }
 
diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.h b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.h
index 511d9a401fed2..734f7daceba05 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.h
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.h
@@ -58,7 +58,10 @@ 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_tzoffset,
+                        const char* format,
+                        int format_len,
+                        int exact);
 
 /*
  * Provides a string length to use for converting datetime
diff --git a/pandas/core/arrays/datetimes.py b/pandas/core/arrays/datetimes.py
index 1768bb7507dd9..fc4234ccd8204 100644
--- a/pandas/core/arrays/datetimes.py
+++ b/pandas/core/arrays/datetimes.py
@@ -2161,6 +2161,8 @@ def objects_to_datetime64ns(
     require_iso8601: bool = False,
     allow_object: bool = False,
     allow_mixed: bool = False,
+    format: str = "",
+    exact: bool = True,
 ):
     """
     Convert data to array of timestamps.
@@ -2208,6 +2210,8 @@ def objects_to_datetime64ns(
             yearfirst=yearfirst,
             require_iso8601=require_iso8601,
             allow_mixed=allow_mixed,
+            format=format,
+            exact=exact,
         )
         result = result.reshape(data.shape, order=order)
     except OverflowError as err:
diff --git a/pandas/core/tools/datetimes.py b/pandas/core/tools/datetimes.py
index 7791ea804a52a..52c5d557dff9c 100644
--- a/pandas/core/tools/datetimes.py
+++ b/pandas/core/tools/datetimes.py
@@ -426,16 +426,17 @@ def _convert_listlike_datetimes(
         format_is_iso8601 = format_is_iso(format)
         if format_is_iso8601:
             require_iso8601 = not infer_datetime_format
-            format = None
 
-    if format is not None:
+    if format is not None and not require_iso8601:
         res = _to_datetime_with_format(
             arg, orig_arg, name, tz, format, exact, errors, infer_datetime_format
         )
         if res is not None:
             return res
+    elif format is None:
+        format = ""
+        exact = False
 
-    assert format is None or infer_datetime_format
     utc = tz == "utc"
     result, tz_parsed = objects_to_datetime64ns(
         arg,
@@ -445,6 +446,8 @@ def _convert_listlike_datetimes(
         errors=errors,
         require_iso8601=require_iso8601,
         allow_object=True,
+        format=format,
+        exact=exact,
     )
 
     if tz_parsed is not None:
diff --git a/pandas/tests/tools/test_to_datetime.py b/pandas/tests/tools/test_to_datetime.py
index c3b4159c2cbfc..1b08faba2eb67 100644
--- a/pandas/tests/tools/test_to_datetime.py
+++ b/pandas/tests/tools/test_to_datetime.py
@@ -1741,6 +1741,130 @@ def test_to_datetime_iso8601(self, cache, arg, exp_str):
         exp = Timestamp(exp_str)
         assert result[0] == exp
 
+    @pytest.mark.parametrize(
+        "input, format",
+        [
+            ("2012", "%Y-%m"),
+            ("2012-01", "%Y-%m-%d"),
+            ("2012-01-01", "%Y-%m-%d %H"),
+            ("2012-01-01 10", "%Y-%m-%d %H:%M"),
+            ("2012-01-01 10:00", "%Y-%m-%d %H:%M:%S"),
+            ("2012-01-01 10:00:00", "%Y-%m-%d %H:%M:%S.%f"),
+            ("2012-01-01 10:00:00.123", "%Y-%m-%d %H:%M:%S.%f%Z"),
+            ("2012-01-01 10:00:00.123", "%Y-%m-%d %H:%M:%S.%f%z"),
+            (0, "%Y-%m-%d"),
+        ],
+    )
+    @pytest.mark.parametrize("exact", [True, False])
+    def test_to_datetime_iso8601_fails(self, input, format, exact):
+        # https://github.com/pandas-dev/pandas/issues/12649
+        # `format` is longer than the string, so this fails regardless of `exact`
+        with pytest.raises(
+            ValueError,
+            match=(
+                rf"time data \"{input}\" at position 0 doesn't match format "
+                rf"\"{format}\""
+            ),
+        ):
+            to_datetime(input, format=format, exact=exact)
+
+    @pytest.mark.parametrize(
+        "input, format",
+        [
+            ("2012-01-01", "%Y-%m"),
+            ("2012-01-01 10", "%Y-%m-%d"),
+            ("2012-01-01 10:00", "%Y-%m-%d %H"),
+            ("2012-01-01 10:00:00", "%Y-%m-%d %H:%M"),
+            (0, "%Y-%m-%d"),
+        ],
+    )
+    def test_to_datetime_iso8601_exact_fails(self, input, format):
+        # https://github.com/pandas-dev/pandas/issues/12649
+        # `format` is shorter than the date string, so only fails with `exact=True`
+        with pytest.raises(
+            ValueError,
+            match=(
+                rf"time data \"{input}\" at position 0 doesn't match format "
+                rf"\"{format}\""
+            ),
+        ):
+            to_datetime(input, format=format)
+
+    @pytest.mark.parametrize(
+        "input, format",
+        [
+            ("2012-01-01", "%Y-%m"),
+            ("2012-01-01 00", "%Y-%m-%d"),
+            ("2012-01-01 00:00", "%Y-%m-%d %H"),
+            ("2012-01-01 00:00:00", "%Y-%m-%d %H:%M"),
+        ],
+    )
+    def test_to_datetime_iso8601_non_exact(self, input, format):
+        # https://github.com/pandas-dev/pandas/issues/12649
+        expected = Timestamp(2012, 1, 1)
+        result = to_datetime(input, format=format, exact=False)
+        assert result == expected
+
+    @pytest.mark.parametrize(
+        "input, format",
+        [
+            ("2020-01", "%Y/%m"),
+            ("2020-01-01", "%Y/%m/%d"),
+            ("2020-01-01 00", "%Y/%m/%dT%H"),
+            ("2020-01-01T00", "%Y/%m/%d %H"),
+            ("2020-01-01 00:00", "%Y/%m/%dT%H:%M"),
+            ("2020-01-01T00:00", "%Y/%m/%d %H:%M"),
+            ("2020-01-01 00:00:00", "%Y/%m/%dT%H:%M:%S"),
+            ("2020-01-01T00:00:00", "%Y/%m/%d %H:%M:%S"),
+        ],
+    )
+    def test_to_datetime_iso8601_separator(self, input, format):
+        # https://github.com/pandas-dev/pandas/issues/12649
+        with pytest.raises(
+            ValueError,
+            match=(
+                rf"time data \"{input}\" at position 0 doesn\'t match format "
+                rf"\"{format}\""
+            ),
+        ):
+            to_datetime(input, format=format)
+
+    @pytest.mark.parametrize(
+        "input, format",
+        [
+            ("2020-01", "%Y-%m"),
+            ("2020-01-01", "%Y-%m-%d"),
+            ("2020-01-01 00", "%Y-%m-%d %H"),
+            ("2020-01-01T00", "%Y-%m-%dT%H"),
+            ("2020-01-01 00:00", "%Y-%m-%d %H:%M"),
+            ("2020-01-01T00:00", "%Y-%m-%dT%H:%M"),
+            ("2020-01-01 00:00:00", "%Y-%m-%d %H:%M:%S"),
+            ("2020-01-01T00:00:00", "%Y-%m-%dT%H:%M:%S"),
+            ("2020-01-01T00:00:00.000", "%Y-%m-%dT%H:%M:%S.%f"),
+            ("2020-01-01T00:00:00.000000", "%Y-%m-%dT%H:%M:%S.%f"),
+            ("2020-01-01T00:00:00.000000000", "%Y-%m-%dT%H:%M:%S.%f"),
+        ],
+    )
+    def test_to_datetime_iso8601_valid(self, input, format):
+        # https://github.com/pandas-dev/pandas/issues/12649
+        expected = Timestamp(2020, 1, 1)
+        result = to_datetime(input, format=format)
+        assert result == expected
+
+    @pytest.mark.parametrize(
+        "input, format",
+        [
+            ("2020-01-01T00:00:00.000000000+00:00", "%Y-%m-%dT%H:%M:%S.%f%z"),
+            ("2020-01-01T00:00:00+00:00", "%Y-%m-%dT%H:%M:%S%z"),
+            ("2020-01-01T00:00:00Z", "%Y-%m-%dT%H:%M:%S%Z"),
+        ],
+    )
+    def test_to_datetime_iso8601_with_timezone_valid(self, input, format):
+        # https://github.com/pandas-dev/pandas/issues/12649
+        expected = Timestamp(2020, 1, 1, tzinfo=pytz.UTC)
+        result = to_datetime(input, format=format)
+        assert result == expected
+
     def test_to_datetime_default(self, cache):
         rs = to_datetime("2001", cache=cache)
         xp = datetime(2001, 1, 1)
@@ -2254,7 +2378,7 @@ def test_day_not_in_month_raise(self, cache):
 
     @pytest.mark.parametrize("arg", ["2015-02-29", "2015-02-32", "2015-04-31"])
     def test_day_not_in_month_raise_value(self, cache, arg):
-        msg = f'time data "{arg}" at position 0 doesn\'t match format specified'
+        msg = f'time data "{arg}" at position 0 doesn\'t match format "%Y-%m-%d"'
         with pytest.raises(ValueError, match=msg):
             to_datetime(arg, errors="raise", format="%Y-%m-%d", cache=cache)
 

From e473adba8b5e9adf4e8667fab0ded57f3e9f3c1b Mon Sep 17 00:00:00 2001
From: Marco Gorelli <>
Date: Fri, 28 Oct 2022 11:16:00 +0200
Subject: [PATCH 02/18] set exact=False default in objects_to_datetime

---
 pandas/core/arrays/datetimes.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pandas/core/arrays/datetimes.py b/pandas/core/arrays/datetimes.py
index fc4234ccd8204..3cb1267029287 100644
--- a/pandas/core/arrays/datetimes.py
+++ b/pandas/core/arrays/datetimes.py
@@ -2162,7 +2162,7 @@ def objects_to_datetime64ns(
     allow_object: bool = False,
     allow_mixed: bool = False,
     format: str = "",
-    exact: bool = True,
+    exact: bool = False,
 ):
     """
     Convert data to array of timestamps.

From a6ea6d0a3ba6a849f293cb4f367ba7f3715c2afc Mon Sep 17 00:00:00 2001
From: Marco Gorelli <>
Date: Fri, 28 Oct 2022 12:22:58 +0200
Subject: [PATCH 03/18] :label: typing

---
 pandas/_libs/tslib.pyi | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/pandas/_libs/tslib.pyi b/pandas/_libs/tslib.pyi
index 8fec9ecf27f30..4228cb3eb9354 100644
--- a/pandas/_libs/tslib.pyi
+++ b/pandas/_libs/tslib.pyi
@@ -25,6 +25,8 @@ def array_to_datetime(
     utc: bool = ...,
     require_iso8601: bool = ...,
     allow_mixed: bool = ...,
+    format: str = ...,
+    exact: bool = ...,
 ) -> tuple[np.ndarray, tzinfo | None]: ...
 
 # returned ndarray may be object dtype or datetime64[ns]

From 8fede1fa9f2a3b7324e376ec6f751f9fcdf96365 Mon Sep 17 00:00:00 2001
From: Marco Gorelli <>
Date: Fri, 28 Oct 2022 15:51:32 +0200
Subject: [PATCH 04/18] simplify

---
 .../tslibs/src/datetime/np_datetime_strings.c | 19 +++++++------------
 1 file changed, 7 insertions(+), 12 deletions(-)

diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index 9a95bf44ddd6d..7f1a82477e144 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -68,21 +68,16 @@ This file implements string parsing and creation for NumPy datetime.
  */
 
 #define FORMAT_STARTSWITH(ch) \
-  if (exact) { \
-    if (!format_len || *format != ch) { \
+    /* Always error on character mismatch conditioned on non-exhausted format, \
+       or when format is exhausted in the exact case. */ \
+    if ((format_len && *format != ch) || (exact && !format_len)){ \
         goto parse_error; \
     } \
-    ++format; \
-    --format_len; \
-  } else { \
-    if (format_len > 0) { \
-        if (*format != ch) { \
-            goto parse_error;  \
-        } \
-        ++format; \
-        --format_len; \
+    /* Advance if format is not exhausted */ \
+    if (format_len) { \
+      ++format; \
+      --format_len; \
     } \
-  } \
 
 int parse_iso_8601_datetime(const char *str, int len, int want_exc,
                             npy_datetimestruct *out,

From 12721b07faa2dfe8bb45ee3d5d249f819fd6ee6e Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Sun, 30 Oct 2022 08:52:59 +0000
Subject: [PATCH 05/18] replace macro with function

---
 .pre-commit-config.yaml                       |   9 +-
 .../tslibs/src/datetime/np_datetime_strings.c | 102 +++++++++++-------
 2 files changed, 71 insertions(+), 40 deletions(-)

diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index 8ff7526b87521..97fd0b11c8eb4 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -48,7 +48,14 @@ 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',
+            '--linelength=88'
+        ]
 -   repo: https://github.com/PyCQA/flake8
     rev: 5.0.4
     hooks:
diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index 7f1a82477e144..79dc730c25906 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -67,19 +67,15 @@ This file implements string parsing and creation for NumPy datetime.
  * Returns 0 on success, -1 on failure.
  */
 
-#define FORMAT_STARTSWITH(ch) \
-    /* Always error on character mismatch conditioned on non-exhausted format, \
-       or when format is exhausted in the exact case. */ \
-    if ((format_len && *format != ch) || (exact && !format_len)){ \
-        goto parse_error; \
-    } \
-    /* Advance if format is not exhausted */ \
-    if (format_len) { \
-      ++format; \
-      --format_len; \
-    } \
+int format_startswith(char ch, int format_len, char format, int exact) {
+    if ((format_len && format != ch) || (exact && !format_len)) {
+        return 0;
+    }
+    return 1;
+}
 
 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,
@@ -118,18 +114,22 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        FORMAT_STARTSWITH(' ');
+        if (!format_startswith(' ', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
     }
 
     /* Leading '-' sign for negative year */
     if (*substr == '-') {
         ++substr;
         --sublen;
-        FORMAT_STARTSWITH('-');
+        if (!format_startswith('-', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
     }
 
-    FORMAT_STARTSWITH('%');
-    FORMAT_STARTSWITH('Y');
+    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
+    if (!format_startswith('Y', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
 
     if (sublen == 0) {
         goto parse_error;
@@ -178,7 +178,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ymd_sep = valid_ymd_sep[i];
         ++substr;
         --sublen;
-        FORMAT_STARTSWITH(ymd_sep);
+        if (!format_startswith(ymd_sep, format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
         /* Cannot have trailing separator */
         if (sublen == 0 || !isdigit(*substr)) {
             goto parse_error;
@@ -190,8 +191,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     out->month = (*substr - '0');
     ++substr;
     --sublen;
-    FORMAT_STARTSWITH('%');
-    FORMAT_STARTSWITH('m');
+    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
+    if (!format_startswith('m', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->month = 10 * out->month + (*substr - '0');
@@ -231,7 +234,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         }
         ++substr;
         --sublen;
-        FORMAT_STARTSWITH(ymd_sep);
+        if (!format_startswith(ymd_sep, format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
     }
 
     /* PARSE THE DAY */
@@ -242,8 +246,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     out->day = (*substr - '0');
     ++substr;
     --sublen;
-    FORMAT_STARTSWITH('%');
-    FORMAT_STARTSWITH('d');
+    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
+    if (!format_startswith('d', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->day = 10 * out->day + (*substr - '0');
@@ -276,7 +282,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if ((*substr != 'T' && *substr != ' ') || sublen == 1) {
         goto parse_error;
     }
-    FORMAT_STARTSWITH(*substr);
+    if (!format_startswith(*substr, format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
     ++substr;
     --sublen;
 
@@ -285,8 +292,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if (!isdigit(*substr)) {
         goto parse_error;
     }
-    FORMAT_STARTSWITH('%');
-    FORMAT_STARTSWITH('H');
+    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
+    if (!format_startswith('H', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
     out->hour = (*substr - '0');
     ++substr;
     --sublen;
@@ -326,7 +335,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (sublen == 0 || !isdigit(*substr)) {
             goto parse_error;
         }
-        FORMAT_STARTSWITH(':');
+        if (!format_startswith(':', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
     } else if (!isdigit(*substr)) {
         if (!hour_was_2_digits) {
             goto parse_error;
@@ -339,8 +349,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     out->min = (*substr - '0');
     ++substr;
     --sublen;
-    FORMAT_STARTSWITH('%');
-    FORMAT_STARTSWITH('M');
+    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
+    if (!format_startswith('M', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->min = 10 * out->min + (*substr - '0');
@@ -369,7 +381,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     /* If we make it through this condition block, then the next
      * character is a digit. */
     if (has_hms_sep && *substr == ':') {
-        FORMAT_STARTSWITH(':');
+        if (!format_startswith(':', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
         ++substr;
         --sublen;
         /* Cannot have a trailing ':' */
@@ -386,8 +399,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     out->sec = (*substr - '0');
     ++substr;
     --sublen;
-    FORMAT_STARTSWITH('%');
-    FORMAT_STARTSWITH('S');
+    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
+    if (!format_startswith('S', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->sec = 10 * out->sec + (*substr - '0');
@@ -409,15 +424,18 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if (sublen > 0 && *substr == '.') {
         ++substr;
         --sublen;
-        FORMAT_STARTSWITH('.');
+        if (!format_startswith('.', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
     } else {
         bestunit = NPY_FR_s;
         goto parse_timezone;
     }
 
     /* PARSE THE MICROSECONDS (0 to 6 digits) */
-    FORMAT_STARTSWITH('%');
-    FORMAT_STARTSWITH('f');
+    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
+    if (!format_startswith('f', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
     numdigits = 0;
     for (i = 0; i < 6; ++i) {
         out->us *= 10;
@@ -482,7 +500,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        FORMAT_STARTSWITH(' ');
+        if (!format_startswith(' ', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
     }
 
     if (sublen == 0) {
@@ -495,8 +514,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
 
     /* UTC specifier */
     if (*substr == 'Z') {
-        FORMAT_STARTSWITH('%');
-        FORMAT_STARTSWITH('Z');
+        if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
+        if (!format_startswith('Z', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
 
         /* "Z" should be equivalent to tz offset "+00:00" */
         if (out_local != NULL) {
@@ -517,8 +538,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
             --sublen;
         }
     } else if (*substr == '-' || *substr == '+') {
-        FORMAT_STARTSWITH('%');
-        FORMAT_STARTSWITH('z');
+        if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
+        if (!format_startswith('z', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
         /* Time zone offset */
         int offset_neg = 0, offset_hour = 0, offset_minute = 0;
 
@@ -602,7 +625,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        FORMAT_STARTSWITH(' ');
+        if (!format_startswith(' ', format_len, *format, exact)) goto parse_error;
+        if (format_len) {++format; --format_len;}
     }
 
     if ((sublen != 0) || (format_len != 0)) {

From 05319671ee4eba17f23c34637ac9285f20d9b8c1 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Sun, 30 Oct 2022 09:10:48 +0000
Subject: [PATCH 06/18] clean up

---
 .../tslibs/src/datetime/np_datetime_strings.c | 52 +++++++++----------
 1 file changed, 25 insertions(+), 27 deletions(-)

diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index 79dc730c25906..d3c9b330e0e63 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -75,7 +75,6 @@ int format_startswith(char ch, int format_len, char format, int exact) {
 }
 
 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,
@@ -126,16 +125,16 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (format_len) {++format; --format_len;}
     }
 
-    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
-    if (!format_startswith('Y', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
-
     if (sublen == 0) {
         goto parse_error;
     }
 
     /* PARSE THE YEAR (4 digits) */
+    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
+    if (!format_startswith('Y', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
+
     out->year = 0;
     if (sublen >= 4 && isdigit(substr[0]) && isdigit(substr[1]) &&
         isdigit(substr[2]) && isdigit(substr[3])) {
@@ -187,14 +186,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE MONTH */
-    /* First digit required */
-    out->month = (*substr - '0');
-    ++substr;
-    --sublen;
     if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
     if (format_len) {++format; --format_len;}
     if (!format_startswith('m', format_len, *format, exact)) goto parse_error;
     if (format_len) {++format; --format_len;}
+    /* First digit required */
+    out->month = (*substr - '0');
+    ++substr;
+    --sublen;
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->month = 10 * out->month + (*substr - '0');
@@ -239,6 +238,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE DAY */
+    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
+    if (!format_startswith('d', format_len, *format, exact)) goto parse_error;
+    if (format_len) {++format; --format_len;}
     /* First digit required */
     if (!isdigit(*substr)) {
         goto parse_error;
@@ -246,10 +249,6 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     out->day = (*substr - '0');
     ++substr;
     --sublen;
-    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
-    if (!format_startswith('d', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->day = 10 * out->day + (*substr - '0');
@@ -288,14 +287,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     --sublen;
 
     /* PARSE THE HOURS */
-    /* First digit required */
-    if (!isdigit(*substr)) {
-        goto parse_error;
-    }
     if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
     if (format_len) {++format; --format_len;}
     if (!format_startswith('H', format_len, *format, exact)) goto parse_error;
     if (format_len) {++format; --format_len;}
+    /* First digit required */
+    if (!isdigit(*substr)) {
+        goto parse_error;
+    }
     out->hour = (*substr - '0');
     ++substr;
     --sublen;
@@ -345,14 +344,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE MINUTES */
-    /* First digit required */
-    out->min = (*substr - '0');
-    ++substr;
-    --sublen;
     if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
     if (format_len) {++format; --format_len;}
     if (!format_startswith('M', format_len, *format, exact)) goto parse_error;
     if (format_len) {++format; --format_len;}
+    /* First digit required */
+    out->min = (*substr - '0');
+    ++substr;
+    --sublen;
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->min = 10 * out->min + (*substr - '0');
@@ -395,14 +394,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE SECONDS */
-    /* First digit required */
-    out->sec = (*substr - '0');
-    ++substr;
-    --sublen;
     if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
     if (format_len) {++format; --format_len;}
     if (!format_startswith('S', format_len, *format, exact)) goto parse_error;
     if (format_len) {++format; --format_len;}
+    /* First digit required */
+    out->sec = (*substr - '0');
+    ++substr;
+    --sublen;
     /* Second digit optional if there was a separator */
     if (isdigit(*substr)) {
         out->sec = 10 * out->sec + (*substr - '0');
@@ -518,7 +517,6 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (format_len) {++format; --format_len;}
         if (!format_startswith('Z', format_len, *format, exact)) goto parse_error;
         if (format_len) {++format; --format_len;}
-
         /* "Z" should be equivalent to tz offset "+00:00" */
         if (out_local != NULL) {
             *out_local = 1;

From a57175323dd08dfd11ba1cc4b003b7520d5823c6 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Sun, 30 Oct 2022 09:33:53 +0000
Subject: [PATCH 07/18] :memo: restore docstring

---
 pandas/_libs/tslibs/src/datetime/np_datetime_strings.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index d3c9b330e0e63..1d1beea608867 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -68,6 +68,12 @@ This file implements string parsing and creation for NumPy datetime.
  */
 
 int format_startswith(char ch, int format_len, char format, int exact) {
+    /* Check if the current character in `format` is `ch`.
+
+    Always error on character mismatch conditioned on non-exhausted format,
+    or when format is exhausted in the exact case.
+    Note that if `format` hasn't been exhausted, it should be advanced
+    outside of this function. */
     if ((format_len && format != ch) || (exact && !format_len)) {
         return 0;
     }

From e814a2ed4e400e2b7452c932268198100c26d678 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Sun, 30 Oct 2022 12:24:44 +0000
Subject: [PATCH 08/18] inline

---
 pandas/_libs/tslibs/src/datetime/np_datetime_strings.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index 1d1beea608867..cec770027b868 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -67,7 +67,7 @@ This file implements string parsing and creation for NumPy datetime.
  * Returns 0 on success, -1 on failure.
  */
 
-int format_startswith(char ch, int format_len, char format, int exact) {
+inline int format_startswith(char ch, int format_len, char format, int exact) {
     /* Check if the current character in `format` is `ch`.
 
     Always error on character mismatch conditioned on non-exhausted format,

From 19c34f8eab9bf03631f672b909b75905b55f8718 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Mon, 31 Oct 2022 14:40:56 +0000
Subject: [PATCH 09/18] set format default to None

---
 pandas/_libs/tslib.pyx              |  6 +++---
 pandas/_libs/tslibs/conversion.pyx  |  2 +-
 pandas/_libs/tslibs/np_datetime.pxd |  4 ++--
 pandas/_libs/tslibs/np_datetime.pyx | 11 ++++++++---
 pandas/_libs/tslibs/parsing.pyx     |  2 +-
 pandas/core/tools/datetimes.py      |  3 ---
 6 files changed, 15 insertions(+), 13 deletions(-)

diff --git a/pandas/_libs/tslib.pyx b/pandas/_libs/tslib.pyx
index 4fd015f15ed76..b86d0453bd2e1 100644
--- a/pandas/_libs/tslib.pyx
+++ b/pandas/_libs/tslib.pyx
@@ -89,7 +89,7 @@ def _test_parse_iso8601(ts: str):
     elif ts == 'today':
         return Timestamp.now().normalize()
 
-    string_to_dts(ts, &obj.dts, &out_bestunit, &out_local, &out_tzoffset, True, "", False)
+    string_to_dts(ts, &obj.dts, &out_bestunit, &out_local, &out_tzoffset, True)
     obj.value = npy_datetimestruct_to_datetime(NPY_FR_ns, &obj.dts)
     check_dts_bounds(&obj.dts)
     if out_local == 1:
@@ -445,8 +445,8 @@ cpdef array_to_datetime(
     bint utc=False,
     bint require_iso8601=False,
     bint allow_mixed=False,
-    str format="",
-    bint exact=False,
+    format: str | None=None,
+    bint exact=True,
 ):
     """
     Converts a 1D array of date-like values to a numpy array of either:
diff --git a/pandas/_libs/tslibs/conversion.pyx b/pandas/_libs/tslibs/conversion.pyx
index 481f66bc256bf..0db605d99a0a7 100644
--- a/pandas/_libs/tslibs/conversion.pyx
+++ b/pandas/_libs/tslibs/conversion.pyx
@@ -488,7 +488,7 @@ cdef _TSObject _convert_str_to_tsobject(object ts, tzinfo tz, str unit,
     else:
         string_to_dts_failed = string_to_dts(
             ts, &dts, &out_bestunit, &out_local,
-            &out_tzoffset, False, "", False
+            &out_tzoffset, False,
         )
         if not string_to_dts_failed:
             try:
diff --git a/pandas/_libs/tslibs/np_datetime.pxd b/pandas/_libs/tslibs/np_datetime.pxd
index 770b1f7e46c12..de81c611c9ee9 100644
--- a/pandas/_libs/tslibs/np_datetime.pxd
+++ b/pandas/_libs/tslibs/np_datetime.pxd
@@ -95,8 +95,8 @@ cdef int string_to_dts(
     int* out_local,
     int* out_tzoffset,
     bint want_exc,
-    str format,
-    bint exact
+    format: str | None = *,
+    bint exact = *
 ) except? -1
 
 cdef NPY_DATETIMEUNIT get_unit_from_dtype(cnp.dtype dtype)
diff --git a/pandas/_libs/tslibs/np_datetime.pyx b/pandas/_libs/tslibs/np_datetime.pyx
index 8acd51d8a4431..ec45fac6a5c69 100644
--- a/pandas/_libs/tslibs/np_datetime.pyx
+++ b/pandas/_libs/tslibs/np_datetime.pyx
@@ -274,8 +274,8 @@ cdef inline int string_to_dts(
     int* out_local,
     int* out_tzoffset,
     bint want_exc,
-    str format,
-    bint exact,
+    format: str | None = None,
+    bint exact = True,
 ) except? -1:
     cdef:
         Py_ssize_t length
@@ -284,7 +284,12 @@ cdef inline int string_to_dts(
         const char* format_buf
 
     buf = get_c_string_buf_and_size(val, &length)
-    format_buf = get_c_string_buf_and_size(format, &format_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,
                                    format_buf, format_length, exact)
diff --git a/pandas/_libs/tslibs/parsing.pyx b/pandas/_libs/tslibs/parsing.pyx
index 965538044fd72..46042cabe97a1 100644
--- a/pandas/_libs/tslibs/parsing.pyx
+++ b/pandas/_libs/tslibs/parsing.pyx
@@ -409,7 +409,7 @@ cdef parse_datetime_string_with_reso(
     # TODO: does this render some/all of parse_delimited_date redundant?
     string_to_dts_failed = string_to_dts(
         date_string, &dts, &out_bestunit, &out_local,
-        &out_tzoffset, False, "", False
+        &out_tzoffset, False,
     )
     if not string_to_dts_failed:
         if dts.ps != 0 or out_local:
diff --git a/pandas/core/tools/datetimes.py b/pandas/core/tools/datetimes.py
index b7a4700b74e0c..1800c91ee16f0 100644
--- a/pandas/core/tools/datetimes.py
+++ b/pandas/core/tools/datetimes.py
@@ -433,9 +433,6 @@ def _convert_listlike_datetimes(
         )
         if res is not None:
             return res
-    elif format is None:
-        format = ""
-        exact = False
 
     utc = tz == "utc"
     result, tz_parsed = objects_to_datetime64ns(

From eb50dfbb50126f2a4b4ebcdc0543f2d8f9b41463 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Mon, 31 Oct 2022 14:47:28 +0000
Subject: [PATCH 10/18] clean up

---
 pandas/_libs/tslibs/conversion.pyx | 2 +-
 pandas/_libs/tslibs/parsing.pyx    | 2 +-
 pandas/core/arrays/datetimes.py    | 4 ++--
 3 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/pandas/_libs/tslibs/conversion.pyx b/pandas/_libs/tslibs/conversion.pyx
index 0db605d99a0a7..923dfa3c54d26 100644
--- a/pandas/_libs/tslibs/conversion.pyx
+++ b/pandas/_libs/tslibs/conversion.pyx
@@ -488,7 +488,7 @@ cdef _TSObject _convert_str_to_tsobject(object ts, tzinfo tz, str unit,
     else:
         string_to_dts_failed = string_to_dts(
             ts, &dts, &out_bestunit, &out_local,
-            &out_tzoffset, False,
+            &out_tzoffset, False
         )
         if not string_to_dts_failed:
             try:
diff --git a/pandas/_libs/tslibs/parsing.pyx b/pandas/_libs/tslibs/parsing.pyx
index 46042cabe97a1..1312124cfb77b 100644
--- a/pandas/_libs/tslibs/parsing.pyx
+++ b/pandas/_libs/tslibs/parsing.pyx
@@ -409,7 +409,7 @@ cdef parse_datetime_string_with_reso(
     # TODO: does this render some/all of parse_delimited_date redundant?
     string_to_dts_failed = string_to_dts(
         date_string, &dts, &out_bestunit, &out_local,
-        &out_tzoffset, False,
+        &out_tzoffset, False
     )
     if not string_to_dts_failed:
         if dts.ps != 0 or out_local:
diff --git a/pandas/core/arrays/datetimes.py b/pandas/core/arrays/datetimes.py
index e1c7647fc7295..b52b31d4ecf2f 100644
--- a/pandas/core/arrays/datetimes.py
+++ b/pandas/core/arrays/datetimes.py
@@ -2133,8 +2133,8 @@ def objects_to_datetime64ns(
     require_iso8601: bool = False,
     allow_object: bool = False,
     allow_mixed: bool = False,
-    format: str = "",
-    exact: bool = False,
+    format: str | None = None,
+    exact: bool = True,
 ):
     """
     Convert data to array of timestamps.

From 0dd74072ce014df7402b96b042f3491bb829efd7 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Sun, 6 Nov 2022 13:02:01 +0000
Subject: [PATCH 11/18] remove function, perform check inline

---
 .../tslibs/src/datetime/np_datetime_strings.c | 135 +++++++++---------
 1 file changed, 66 insertions(+), 69 deletions(-)

diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index cec770027b868..e5a14b64da266 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -67,19 +67,6 @@ This file implements string parsing and creation for NumPy datetime.
  * Returns 0 on success, -1 on failure.
  */
 
-inline int format_startswith(char ch, int format_len, char format, int exact) {
-    /* Check if the current character in `format` is `ch`.
-
-    Always error on character mismatch conditioned on non-exhausted format,
-    or when format is exhausted in the exact case.
-    Note that if `format` hasn't been exhausted, it should be advanced
-    outside of this function. */
-    if ((format_len && format != ch) || (exact && !format_len)) {
-        return 0;
-    }
-    return 1;
-}
-
 int parse_iso_8601_datetime(const char *str, int len, int want_exc,
                             npy_datetimestruct *out,
                             NPY_DATETIMEUNIT *out_bestunit,
@@ -119,16 +106,18 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        if (!format_startswith(' ', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 1 && exact) goto parse_error;
+        if (*format++ != ' ') goto parse_error;
+        if (format_len) --format_len;
     }
 
     /* Leading '-' sign for negative year */
     if (*substr == '-') {
         ++substr;
         --sublen;
-        if (!format_startswith('-', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 1 && exact) goto parse_error;
+        if (*format++ != '-') goto parse_error;
+        if (format_len) --format_len;
     }
 
     if (sublen == 0) {
@@ -136,10 +125,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE YEAR (4 digits) */
-    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
-    if (!format_startswith('Y', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
+    if (format_len < 2 && exact) goto parse_error;
+    if (*format++ != '%') goto parse_error;
+    if (*format++ != 'Y') goto parse_error;
+    if (format_len) format_len -= 2;
 
     out->year = 0;
     if (sublen >= 4 && isdigit(substr[0]) && isdigit(substr[1]) &&
@@ -183,8 +172,9 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ymd_sep = valid_ymd_sep[i];
         ++substr;
         --sublen;
-        if (!format_startswith(ymd_sep, format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 1 && exact) goto parse_error;
+        if (*format++ != ymd_sep) goto parse_error;
+        if (format_len) --format_len;
         /* Cannot have trailing separator */
         if (sublen == 0 || !isdigit(*substr)) {
             goto parse_error;
@@ -192,10 +182,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE MONTH */
-    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
-    if (!format_startswith('m', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
+    if (format_len < 2 && exact) goto parse_error;
+    if (*format++ != '%') goto parse_error;
+    if (*format++ != 'm') goto parse_error;
+    if (format_len) format_len -= 2;
     /* First digit required */
     out->month = (*substr - '0');
     ++substr;
@@ -239,15 +229,16 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         }
         ++substr;
         --sublen;
-        if (!format_startswith(ymd_sep, format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 1 && exact) goto parse_error;
+        if (*format++ != ymd_sep) goto parse_error;
+        if (format_len) --format_len;
     }
 
     /* PARSE THE DAY */
-    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
-    if (!format_startswith('d', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
+    if (format_len < 2 && exact) goto parse_error;
+    if (*format++ != '%') goto parse_error;
+    if (*format++ != 'd') goto parse_error;
+    if (format_len) format_len -= 2;
     /* First digit required */
     if (!isdigit(*substr)) {
         goto parse_error;
@@ -284,19 +275,20 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         goto finish;
     }
 
+    if (format_len < 1 && exact) goto parse_error;
+    if (*format++ != *substr) goto parse_error;
+    if (format_len) --format_len;
     if ((*substr != 'T' && *substr != ' ') || sublen == 1) {
         goto parse_error;
     }
-    if (!format_startswith(*substr, format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
     ++substr;
     --sublen;
 
     /* PARSE THE HOURS */
-    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
-    if (!format_startswith('H', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
+    if (format_len < 2 && exact) goto parse_error;
+    if (*format++ != '%') goto parse_error;
+    if (*format++ != 'H') goto parse_error;
+    if (format_len) format_len -= 2;
     /* First digit required */
     if (!isdigit(*substr)) {
         goto parse_error;
@@ -340,8 +332,9 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (sublen == 0 || !isdigit(*substr)) {
             goto parse_error;
         }
-        if (!format_startswith(':', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 1 && exact) goto parse_error;
+        if (*format++ != ':') goto parse_error;
+        if (format_len) --format_len;
     } else if (!isdigit(*substr)) {
         if (!hour_was_2_digits) {
             goto parse_error;
@@ -350,10 +343,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE MINUTES */
-    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
-    if (!format_startswith('M', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
+    if (format_len < 2 && exact) goto parse_error;
+    if (*format++ != '%') goto parse_error;
+    if (*format++ != 'M') goto parse_error;
+    if (format_len) format_len -= 2;
     /* First digit required */
     out->min = (*substr - '0');
     ++substr;
@@ -386,8 +379,9 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     /* If we make it through this condition block, then the next
      * character is a digit. */
     if (has_hms_sep && *substr == ':') {
-        if (!format_startswith(':', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 1 && exact) goto parse_error;
+        if (*format++ != ':') goto parse_error;
+        if (format_len) --format_len;
         ++substr;
         --sublen;
         /* Cannot have a trailing ':' */
@@ -400,10 +394,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE SECONDS */
-    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
-    if (!format_startswith('S', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
+    if (format_len < 2 && exact) goto parse_error;
+    if (*format++ != '%') goto parse_error;
+    if (*format++ != 'S') goto parse_error;
+    if (format_len) format_len -= 2;
     /* First digit required */
     out->sec = (*substr - '0');
     ++substr;
@@ -429,18 +423,19 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if (sublen > 0 && *substr == '.') {
         ++substr;
         --sublen;
-        if (!format_startswith('.', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 1 && exact) goto parse_error;
+        if (*format++ != '.') goto parse_error;
+        if (format_len) --format_len;
     } else {
         bestunit = NPY_FR_s;
         goto parse_timezone;
     }
 
     /* PARSE THE MICROSECONDS (0 to 6 digits) */
-    if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
-    if (!format_startswith('f', format_len, *format, exact)) goto parse_error;
-    if (format_len) {++format; --format_len;}
+    if (format_len < 2 && exact) goto parse_error;
+    if (*format++ != '%') goto parse_error;
+    if (*format++ != 'S') goto parse_error;
+    if (format_len) format_len -= 2;
     numdigits = 0;
     for (i = 0; i < 6; ++i) {
         out->us *= 10;
@@ -505,8 +500,9 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        if (!format_startswith(' ', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 1 && exact) goto parse_error;
+        if (*format++ != ' ') goto parse_error;
+        if (format_len) --format_len;
     }
 
     if (sublen == 0) {
@@ -519,10 +515,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
 
     /* UTC specifier */
     if (*substr == 'Z') {
-        if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
-        if (!format_startswith('Z', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 2 && exact) goto parse_error;
+        if (*format++ != '%') goto parse_error;
+        if (*format++ != 'Z') goto parse_error;
+        if (format_len) format_len -= 2;
         /* "Z" should be equivalent to tz offset "+00:00" */
         if (out_local != NULL) {
             *out_local = 1;
@@ -542,10 +538,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
             --sublen;
         }
     } else if (*substr == '-' || *substr == '+') {
-        if (!format_startswith('%', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
-        if (!format_startswith('z', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 2 && exact) goto parse_error;
+        if (*format++ != '%') goto parse_error;
+        if (*format++ != 'z') goto parse_error;
+        if (format_len) format_len -= 2;
         /* Time zone offset */
         int offset_neg = 0, offset_hour = 0, offset_minute = 0;
 
@@ -629,8 +625,9 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        if (!format_startswith(' ', format_len, *format, exact)) goto parse_error;
-        if (format_len) {++format; --format_len;}
+        if (format_len < 1 && exact) goto parse_error;
+        if (*format++ != ' ') goto parse_error;
+        if (format_len) --format_len;
     }
 
     if ((sublen != 0) || (format_len != 0)) {

From 3acfdf6190a0b30183846b15986c61a4f8eafd26 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Mon, 7 Nov 2022 15:47:06 +0000
Subject: [PATCH 12/18] only compare *format++ if format_len

---
 .../tslibs/src/datetime/np_datetime_strings.c | 60 +++++++++----------
 1 file changed, 30 insertions(+), 30 deletions(-)

diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index e5a14b64da266..f0264db102045 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -107,7 +107,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ++substr;
         --sublen;
         if (format_len < 1 && exact) goto parse_error;
-        if (*format++ != ' ') goto parse_error;
+        if (format_len && *format++ != ' ') goto parse_error;
         if (format_len) --format_len;
     }
 
@@ -116,7 +116,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ++substr;
         --sublen;
         if (format_len < 1 && exact) goto parse_error;
-        if (*format++ != '-') goto parse_error;
+        if (format_len && *format++ != '-') goto parse_error;
         if (format_len) --format_len;
     }
 
@@ -126,8 +126,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
 
     /* PARSE THE YEAR (4 digits) */
     if (format_len < 2 && exact) goto parse_error;
-    if (*format++ != '%') goto parse_error;
-    if (*format++ != 'Y') goto parse_error;
+    if (format_len && *format++ != '%') goto parse_error;
+    if (format_len && *format++ != 'Y') goto parse_error;
     if (format_len) format_len -= 2;
 
     out->year = 0;
@@ -173,7 +173,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ++substr;
         --sublen;
         if (format_len < 1 && exact) goto parse_error;
-        if (*format++ != ymd_sep) goto parse_error;
+        if (format_len && *format++ != ymd_sep) goto parse_error;
         if (format_len) --format_len;
         /* Cannot have trailing separator */
         if (sublen == 0 || !isdigit(*substr)) {
@@ -183,8 +183,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
 
     /* PARSE THE MONTH */
     if (format_len < 2 && exact) goto parse_error;
-    if (*format++ != '%') goto parse_error;
-    if (*format++ != 'm') goto parse_error;
+    if (format_len && *format++ != '%') goto parse_error;
+    if (format_len && *format++ != 'm') goto parse_error;
     if (format_len) format_len -= 2;
     /* First digit required */
     out->month = (*substr - '0');
@@ -230,14 +230,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ++substr;
         --sublen;
         if (format_len < 1 && exact) goto parse_error;
-        if (*format++ != ymd_sep) goto parse_error;
+        if (format_len && *format++ != ymd_sep) goto parse_error;
         if (format_len) --format_len;
     }
 
     /* PARSE THE DAY */
     if (format_len < 2 && exact) goto parse_error;
-    if (*format++ != '%') goto parse_error;
-    if (*format++ != 'd') goto parse_error;
+    if (format_len && *format++ != '%') goto parse_error;
+    if (format_len && *format++ != 'd') goto parse_error;
     if (format_len) format_len -= 2;
     /* First digit required */
     if (!isdigit(*substr)) {
@@ -275,19 +275,19 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         goto finish;
     }
 
-    if (format_len < 1 && exact) goto parse_error;
-    if (*format++ != *substr) goto parse_error;
-    if (format_len) --format_len;
     if ((*substr != 'T' && *substr != ' ') || sublen == 1) {
         goto parse_error;
     }
+    if (format_len < 1 && exact) goto parse_error;
+    if (format_len && *format++ != *substr) goto parse_error;
+    if (format_len) --format_len;
     ++substr;
     --sublen;
 
     /* PARSE THE HOURS */
     if (format_len < 2 && exact) goto parse_error;
-    if (*format++ != '%') goto parse_error;
-    if (*format++ != 'H') goto parse_error;
+    if (format_len && *format++ != '%') goto parse_error;
+    if (format_len && *format++ != 'H') goto parse_error;
     if (format_len) format_len -= 2;
     /* First digit required */
     if (!isdigit(*substr)) {
@@ -333,7 +333,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
             goto parse_error;
         }
         if (format_len < 1 && exact) goto parse_error;
-        if (*format++ != ':') goto parse_error;
+        if (format_len && *format++ != ':') goto parse_error;
         if (format_len) --format_len;
     } else if (!isdigit(*substr)) {
         if (!hour_was_2_digits) {
@@ -344,8 +344,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
 
     /* PARSE THE MINUTES */
     if (format_len < 2 && exact) goto parse_error;
-    if (*format++ != '%') goto parse_error;
-    if (*format++ != 'M') goto parse_error;
+    if (format_len && *format++ != '%') goto parse_error;
+    if (format_len && *format++ != 'M') goto parse_error;
     if (format_len) format_len -= 2;
     /* First digit required */
     out->min = (*substr - '0');
@@ -380,7 +380,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
      * character is a digit. */
     if (has_hms_sep && *substr == ':') {
         if (format_len < 1 && exact) goto parse_error;
-        if (*format++ != ':') goto parse_error;
+        if (format_len && *format++ != ':') goto parse_error;
         if (format_len) --format_len;
         ++substr;
         --sublen;
@@ -395,8 +395,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
 
     /* PARSE THE SECONDS */
     if (format_len < 2 && exact) goto parse_error;
-    if (*format++ != '%') goto parse_error;
-    if (*format++ != 'S') goto parse_error;
+    if (format_len && *format++ != '%') goto parse_error;
+    if (format_len && *format++ != 'S') goto parse_error;
     if (format_len) format_len -= 2;
     /* First digit required */
     out->sec = (*substr - '0');
@@ -424,7 +424,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ++substr;
         --sublen;
         if (format_len < 1 && exact) goto parse_error;
-        if (*format++ != '.') goto parse_error;
+        if (format_len && *format++ != '.') goto parse_error;
         if (format_len) --format_len;
     } else {
         bestunit = NPY_FR_s;
@@ -433,8 +433,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
 
     /* PARSE THE MICROSECONDS (0 to 6 digits) */
     if (format_len < 2 && exact) goto parse_error;
-    if (*format++ != '%') goto parse_error;
-    if (*format++ != 'S') goto parse_error;
+    if (format_len && *format++ != '%') goto parse_error;
+    if (format_len && *format++ != 'f') goto parse_error;
     if (format_len) format_len -= 2;
     numdigits = 0;
     for (i = 0; i < 6; ++i) {
@@ -501,7 +501,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ++substr;
         --sublen;
         if (format_len < 1 && exact) goto parse_error;
-        if (*format++ != ' ') goto parse_error;
+        if (format_len && *format++ != ' ') goto parse_error;
         if (format_len) --format_len;
     }
 
@@ -516,8 +516,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     /* UTC specifier */
     if (*substr == 'Z') {
         if (format_len < 2 && exact) goto parse_error;
-        if (*format++ != '%') goto parse_error;
-        if (*format++ != 'Z') goto parse_error;
+        if (format_len && *format++ != '%') goto parse_error;
+        if (format_len && *format++ != 'Z') goto parse_error;
         if (format_len) format_len -= 2;
         /* "Z" should be equivalent to tz offset "+00:00" */
         if (out_local != NULL) {
@@ -539,8 +539,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         }
     } else if (*substr == '-' || *substr == '+') {
         if (format_len < 2 && exact) goto parse_error;
-        if (*format++ != '%') goto parse_error;
-        if (*format++ != 'z') goto parse_error;
+        if (format_len && *format++ != '%') goto parse_error;
+        if (format_len && *format++ != 'z') goto parse_error;
         if (format_len) format_len -= 2;
         /* Time zone offset */
         int offset_neg = 0, offset_hour = 0, offset_minute = 0;
@@ -626,7 +626,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ++substr;
         --sublen;
         if (format_len < 1 && exact) goto parse_error;
-        if (*format++ != ' ') goto parse_error;
+        if (format_len && *format++ != ' ') goto parse_error;
         if (format_len) --format_len;
     }
 

From f3060c9afab7f8cd8e4362324cc3978d71314a73 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Mon, 7 Nov 2022 15:55:41 +0000
Subject: [PATCH 13/18] clean up

---
 .pre-commit-config.yaml                               | 11 ++---------
 pandas/_libs/tslibs/np_datetime.pyx                   |  4 ++--
 .../_libs/tslibs/src/datetime/np_datetime_strings.c   |  1 -
 3 files changed, 4 insertions(+), 12 deletions(-)

diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index 6ccaaa3c8eb25..6aa1f5659365f 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -27,7 +27,7 @@ repos:
     -   id: codespell
         types_or: [python, rst, markdown]
 -   repo: https://github.com/MarcoGorelli/cython-lint
-    rev: v0.4.0
+    rev: v0.2.1
     hooks:
     -   id: cython-lint
 -   repo: https://github.com/pre-commit/pre-commit-hooks
@@ -48,14 +48,7 @@ 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',
-            '--linelength=88'
-        ]
+        args: [--quiet, '--extensions=c,h', '--headers=h', --recursive, '--filter=-readability/casting,-runtime/int,-build/include_subdir']
 -   repo: https://github.com/PyCQA/flake8
     rev: 5.0.4
     hooks:
diff --git a/pandas/_libs/tslibs/np_datetime.pyx b/pandas/_libs/tslibs/np_datetime.pyx
index 08d782ec0de06..d49c41e54764f 100644
--- a/pandas/_libs/tslibs/np_datetime.pyx
+++ b/pandas/_libs/tslibs/np_datetime.pyx
@@ -278,8 +278,8 @@ cdef inline int string_to_dts(
     int* out_local,
     int* out_tzoffset,
     bint want_exc,
-    format: str | None = None,
-    bint exact = True,
+    format: str | None=None,
+    bint exact=True,
 ) except? -1:
     cdef:
         Py_ssize_t length
diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index f0264db102045..e1893d1819cb2 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -66,7 +66,6 @@ This file implements string parsing and creation for NumPy datetime.
  *
  * Returns 0 on success, -1 on failure.
  */
-
 int parse_iso_8601_datetime(const char *str, int len, int want_exc,
                             npy_datetimestruct *out,
                             NPY_DATETIMEUNIT *out_bestunit,

From 7310e139e0df5207c958e5f213614a3a5962a78e Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Mon, 7 Nov 2022 17:17:42 +0000
Subject: [PATCH 14/18] typing

---
 pandas/_libs/tslib.pyi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/pandas/_libs/tslib.pyi b/pandas/_libs/tslib.pyi
index b8dc696b4d66c..bb70b07dd4908 100644
--- a/pandas/_libs/tslib.pyi
+++ b/pandas/_libs/tslib.pyi
@@ -25,7 +25,7 @@ def array_to_datetime(
     utc: bool = ...,
     require_iso8601: bool = ...,
     allow_mixed: bool = ...,
-    format: str = ...,
+    format: str | None = ...,
     exact: bool = ...,
 ) -> tuple[np.ndarray, tzinfo | None]: ...
 

From 3ceb1ee1c5291536c205853a996d725ffab107d0 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Wed, 9 Nov 2022 10:38:02 +0000
Subject: [PATCH 15/18] split out branches

---
 .pre-commit-config.yaml                       |   8 +-
 .../tslibs/src/datetime/np_datetime_strings.c | 220 ++++++++++++------
 2 files changed, 161 insertions(+), 67 deletions(-)

diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index 6aa1f5659365f..d7f1549bd3928 100644
--- a/.pre-commit-config.yaml
+++ b/.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:
diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index e1893d1819cb2..220baf499e1c7 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -71,6 +71,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
                             NPY_DATETIMEUNIT *out_bestunit,
                             int *out_local, int *out_tzoffset,
                             const char* format, int format_len, int exact) {
+    if (len < 0 || format_len < 0)
+        goto parse_error;
     int year_leap = 0;
     int i, numdigits;
     const char *substr;
@@ -105,18 +107,28 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        if (format_len < 1 && exact) goto parse_error;
-        if (format_len && *format++ != ' ') goto parse_error;
-        if (format_len) --format_len;
+        if (format_len < 1) {
+            if (exact)
+                goto parse_error;
+        } else {
+            if (*format++ != ' ')
+                goto parse_error;
+            format_len -= 1;
+        }
     }
 
     /* Leading '-' sign for negative year */
     if (*substr == '-') {
         ++substr;
         --sublen;
-        if (format_len < 1 && exact) goto parse_error;
-        if (format_len && *format++ != '-') goto parse_error;
-        if (format_len) --format_len;
+        if (format_len < 1) {
+            if (exact)
+                goto parse_error;
+        } else {
+            if (*format++ != '-')
+                goto parse_error;
+            format_len -= 1;
+        }
     }
 
     if (sublen == 0) {
@@ -124,10 +136,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE YEAR (4 digits) */
-    if (format_len < 2 && exact) goto parse_error;
-    if (format_len && *format++ != '%') goto parse_error;
-    if (format_len && *format++ != 'Y') goto parse_error;
-    if (format_len) format_len -= 2;
+    if (format_len < 2) {
+        if (exact || format_len)
+            goto parse_error;
+    } else {
+        if (*format++ != '%' || *format++ != 'Y')
+            goto parse_error;
+        format_len -= 2;
+    }
 
     out->year = 0;
     if (sublen >= 4 && isdigit(substr[0]) && isdigit(substr[1]) &&
@@ -171,9 +187,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ymd_sep = valid_ymd_sep[i];
         ++substr;
         --sublen;
-        if (format_len < 1 && exact) goto parse_error;
-        if (format_len && *format++ != ymd_sep) goto parse_error;
-        if (format_len) --format_len;
+        if (format_len < 1) {
+            if (exact)
+                goto parse_error;
+        } else {
+            if (*format++ != ymd_sep)
+                goto parse_error;
+            format_len -= 1;
+        }
         /* Cannot have trailing separator */
         if (sublen == 0 || !isdigit(*substr)) {
             goto parse_error;
@@ -181,10 +202,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE MONTH */
-    if (format_len < 2 && exact) goto parse_error;
-    if (format_len && *format++ != '%') goto parse_error;
-    if (format_len && *format++ != 'm') goto parse_error;
-    if (format_len) format_len -= 2;
+    if (format_len < 2) {
+        if (exact || format_len)
+            goto parse_error;
+    } else {
+        if (*format++ != '%' || *format++ != 'm')
+            goto parse_error;
+        format_len -= 2;
+    }
     /* First digit required */
     out->month = (*substr - '0');
     ++substr;
@@ -228,16 +253,25 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         }
         ++substr;
         --sublen;
-        if (format_len < 1 && exact) goto parse_error;
-        if (format_len && *format++ != ymd_sep) goto parse_error;
-        if (format_len) --format_len;
+        if (format_len < 1) {
+            if (exact)
+                goto parse_error;
+        } else {
+            if (*format++ != ymd_sep)
+                goto parse_error;
+            format_len -= 1;
+        }
     }
 
     /* PARSE THE DAY */
-    if (format_len < 2 && exact) goto parse_error;
-    if (format_len && *format++ != '%') goto parse_error;
-    if (format_len && *format++ != 'd') goto parse_error;
-    if (format_len) format_len -= 2;
+    if (format_len < 2) {
+        if (exact || format_len)
+            goto parse_error;
+    } else {
+        if (*format++ != '%' || *format++ != 'd')
+            goto parse_error;
+        format_len -= 2;
+    }
     /* First digit required */
     if (!isdigit(*substr)) {
         goto parse_error;
@@ -277,17 +311,26 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if ((*substr != 'T' && *substr != ' ') || sublen == 1) {
         goto parse_error;
     }
-    if (format_len < 1 && exact) goto parse_error;
-    if (format_len && *format++ != *substr) goto parse_error;
-    if (format_len) --format_len;
+        if (format_len < 1) {
+            if (exact)
+                goto parse_error;
+        } else {
+            if (*format++ != *substr)
+                goto parse_error;
+            format_len -= 1;
+        }
     ++substr;
     --sublen;
 
     /* PARSE THE HOURS */
-    if (format_len < 2 && exact) goto parse_error;
-    if (format_len && *format++ != '%') goto parse_error;
-    if (format_len && *format++ != 'H') goto parse_error;
-    if (format_len) format_len -= 2;
+    if (format_len < 2) {
+        if (exact || format_len)
+            goto parse_error;
+    } else {
+        if (*format++ != '%' || *format++ != 'H')
+            goto parse_error;
+        format_len -= 2;
+    }
     /* First digit required */
     if (!isdigit(*substr)) {
         goto parse_error;
@@ -331,9 +374,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (sublen == 0 || !isdigit(*substr)) {
             goto parse_error;
         }
-        if (format_len < 1 && exact) goto parse_error;
-        if (format_len && *format++ != ':') goto parse_error;
-        if (format_len) --format_len;
+        if (format_len < 1) {
+            if (exact)
+                goto parse_error;
+        } else {
+            if (*format++ != ':')
+                goto parse_error;
+            format_len -= 1;
+        }
     } else if (!isdigit(*substr)) {
         if (!hour_was_2_digits) {
             goto parse_error;
@@ -342,10 +390,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE MINUTES */
-    if (format_len < 2 && exact) goto parse_error;
-    if (format_len && *format++ != '%') goto parse_error;
-    if (format_len && *format++ != 'M') goto parse_error;
-    if (format_len) format_len -= 2;
+    if (format_len < 2) {
+        if (exact || format_len)
+            goto parse_error;
+    } else {
+        if (*format++ != '%' || *format++ != 'M')
+            goto parse_error;
+        format_len -= 2;
+    }
     /* First digit required */
     out->min = (*substr - '0');
     ++substr;
@@ -378,9 +430,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     /* If we make it through this condition block, then the next
      * character is a digit. */
     if (has_hms_sep && *substr == ':') {
-        if (format_len < 1 && exact) goto parse_error;
-        if (format_len && *format++ != ':') goto parse_error;
-        if (format_len) --format_len;
+        if (format_len < 1) {
+            if (exact)
+                goto parse_error;
+        } else {
+            if (*format++ != ':')
+                goto parse_error;
+            format_len -= 1;
+        }
         ++substr;
         --sublen;
         /* Cannot have a trailing ':' */
@@ -393,10 +450,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE SECONDS */
-    if (format_len < 2 && exact) goto parse_error;
-    if (format_len && *format++ != '%') goto parse_error;
-    if (format_len && *format++ != 'S') goto parse_error;
-    if (format_len) format_len -= 2;
+    if (format_len < 2) {
+        if (exact || format_len)
+            goto parse_error;
+    } else {
+        if (*format++ != '%' || *format++ != 'S')
+            goto parse_error;
+        format_len -= 2;
+    }
     /* First digit required */
     out->sec = (*substr - '0');
     ++substr;
@@ -422,19 +483,28 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if (sublen > 0 && *substr == '.') {
         ++substr;
         --sublen;
-        if (format_len < 1 && exact) goto parse_error;
-        if (format_len && *format++ != '.') goto parse_error;
-        if (format_len) --format_len;
+        if (format_len < 1) {
+            if (exact)
+                goto parse_error;
+        } else {
+            if (*format++ != '.')
+                goto parse_error;
+            format_len -= 1;
+        }
     } else {
         bestunit = NPY_FR_s;
         goto parse_timezone;
     }
 
     /* PARSE THE MICROSECONDS (0 to 6 digits) */
-    if (format_len < 2 && exact) goto parse_error;
-    if (format_len && *format++ != '%') goto parse_error;
-    if (format_len && *format++ != 'f') goto parse_error;
-    if (format_len) format_len -= 2;
+    if (format_len < 2) {
+        if (exact || format_len)
+            goto parse_error;
+    } else {
+        if (*format++ != '%' || *format++ != 'f')
+            goto parse_error;
+        format_len -= 2;
+    }
     numdigits = 0;
     for (i = 0; i < 6; ++i) {
         out->us *= 10;
@@ -499,9 +569,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        if (format_len < 1 && exact) goto parse_error;
-        if (format_len && *format++ != ' ') goto parse_error;
-        if (format_len) --format_len;
+        if (format_len < 1) {
+            if (exact)
+                goto parse_error;
+        } else {
+            if (*format++ != ' ')
+                goto parse_error;
+            format_len -= 1;
+        }
     }
 
     if (sublen == 0) {
@@ -514,10 +589,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
 
     /* UTC specifier */
     if (*substr == 'Z') {
-        if (format_len < 2 && exact) goto parse_error;
-        if (format_len && *format++ != '%') goto parse_error;
-        if (format_len && *format++ != 'Z') goto parse_error;
-        if (format_len) format_len -= 2;
+        if (format_len < 2) {
+            if (exact || format_len)
+                goto parse_error;
+        } else {
+            if (*format++ != '%' || *format++ != 'Z')
+                goto parse_error;
+            format_len -= 2;
+        }
         /* "Z" should be equivalent to tz offset "+00:00" */
         if (out_local != NULL) {
             *out_local = 1;
@@ -537,10 +616,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
             --sublen;
         }
     } else if (*substr == '-' || *substr == '+') {
-        if (format_len < 2 && exact) goto parse_error;
-        if (format_len && *format++ != '%') goto parse_error;
-        if (format_len && *format++ != 'z') goto parse_error;
-        if (format_len) format_len -= 2;
+        if (format_len < 2) {
+            if (exact || format_len)
+                goto parse_error;
+        } else {
+            if (*format++ != '%' || *format++ != 'z')
+                goto parse_error;
+            format_len -= 2;
+        }
         /* Time zone offset */
         int offset_neg = 0, offset_hour = 0, offset_minute = 0;
 
@@ -624,9 +707,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        if (format_len < 1 && exact) goto parse_error;
-        if (format_len && *format++ != ' ') goto parse_error;
-        if (format_len) --format_len;
+        if (format_len < 1) {
+            if (exact)
+                goto parse_error;
+        } else {
+            if (*format++ != ' ')
+                goto parse_error;
+            format_len -= 1;
+        }
     }
 
     if ((sublen != 0) || (format_len != 0)) {

From 080f0183b0f96879dab0540a9661adf7b672f7c1 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Sun, 13 Nov 2022 12:35:20 +0000
Subject: [PATCH 16/18] use compare_format function

---
 .pre-commit-config.yaml                       |   8 +-
 .../tslibs/src/datetime/np_datetime_strings.c | 206 ++++++------------
 2 files changed, 73 insertions(+), 141 deletions(-)

diff --git a/.pre-commit-config.yaml b/.pre-commit-config.yaml
index d7f1549bd3928..6aa1f5659365f 100644
--- a/.pre-commit-config.yaml
+++ b/.pre-commit-config.yaml
@@ -48,13 +48,7 @@ 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,-readability/fn_size'
-        ]
+        args: [--quiet, '--extensions=c,h', '--headers=h', --recursive, '--filter=-readability/casting,-runtime/int,-build/include_subdir']
 -   repo: https://github.com/PyCQA/flake8
     rev: 5.0.4
     hooks:
diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index 220baf499e1c7..76e3e1147df00 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -66,6 +66,38 @@ This file implements string parsing and creation for NumPy datetime.
  *
  * Returns 0 on success, -1 on failure.
  */
+
+// This function will advance the pointer on format
+// and decrement characters_remaining by n on success
+// On failure will return -1 without incrementing
+int compare_format(const char **format, int *characters_remaining,
+                 const char *compare_to, int n, const int exact) {
+  if (*characters_remaining < n) {
+    if (exact) {
+      // TODO(pandas-dev): in the future we should set a PyErr here
+      // to be very clear about what went wrong
+      return -1;
+    } else if (*characters_remaining) {
+      // TODO(pandas-dev): same return value in this function as
+      // above branch, but stub out a future where
+      // we have a better error message
+      return -1;
+    } else {
+        return 0;
+    }
+  } else {
+    if (strncmp(*format, compare_to, n)) {
+      // TODO(pandas-dev): PyErr to differentiate what went wrong
+      return -1;
+    } else {
+      *format += n;
+      *characters_remaining -= n;
+      return 0;
+    }
+  }
+  return 0;
+}
+
 int parse_iso_8601_datetime(const char *str, int len, int want_exc,
                             npy_datetimestruct *out,
                             NPY_DATETIMEUNIT *out_bestunit,
@@ -107,13 +139,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        if (format_len < 1) {
-            if (exact)
-                goto parse_error;
-        } else {
-            if (*format++ != ' ')
-                goto parse_error;
-            format_len -= 1;
+        if (compare_format(&format, &format_len, " ", 1, exact)) {
+            goto parse_error;
         }
     }
 
@@ -121,14 +148,6 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if (*substr == '-') {
         ++substr;
         --sublen;
-        if (format_len < 1) {
-            if (exact)
-                goto parse_error;
-        } else {
-            if (*format++ != '-')
-                goto parse_error;
-            format_len -= 1;
-        }
     }
 
     if (sublen == 0) {
@@ -136,13 +155,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE YEAR (4 digits) */
-    if (format_len < 2) {
-        if (exact || format_len)
-            goto parse_error;
-    } else {
-        if (*format++ != '%' || *format++ != 'Y')
-            goto parse_error;
-        format_len -= 2;
+    if (compare_format(&format, &format_len, "%Y", 2, exact)) {
+        goto parse_error;
     }
 
     out->year = 0;
@@ -187,13 +201,10 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ymd_sep = valid_ymd_sep[i];
         ++substr;
         --sublen;
-        if (format_len < 1) {
-            if (exact)
-                goto parse_error;
-        } else {
-            if (*format++ != ymd_sep)
-                goto parse_error;
-            format_len -= 1;
+
+        const char tmp[1] = {ymd_sep};
+        if (compare_format(&format, &format_len, tmp, 1, exact)) {
+            goto parse_error;
         }
         /* Cannot have trailing separator */
         if (sublen == 0 || !isdigit(*substr)) {
@@ -202,13 +213,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE MONTH */
-    if (format_len < 2) {
-        if (exact || format_len)
-            goto parse_error;
-    } else {
-        if (*format++ != '%' || *format++ != 'm')
-            goto parse_error;
-        format_len -= 2;
+    if (compare_format(&format, &format_len, "%m", 2, exact)) {
+        goto parse_error;
     }
     /* First digit required */
     out->month = (*substr - '0');
@@ -253,24 +259,15 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         }
         ++substr;
         --sublen;
-        if (format_len < 1) {
-            if (exact)
-                goto parse_error;
-        } else {
-            if (*format++ != ymd_sep)
-                goto parse_error;
-            format_len -= 1;
+        const char tmp[1] = {ymd_sep};
+        if (compare_format(&format, &format_len, tmp, 1, exact)) {
+            goto parse_error;
         }
     }
 
     /* PARSE THE DAY */
-    if (format_len < 2) {
-        if (exact || format_len)
-            goto parse_error;
-    } else {
-        if (*format++ != '%' || *format++ != 'd')
-            goto parse_error;
-        format_len -= 2;
+    if (compare_format(&format, &format_len, "%d", 2, exact)) {
+        goto parse_error;
     }
     /* First digit required */
     if (!isdigit(*substr)) {
@@ -311,25 +308,16 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if ((*substr != 'T' && *substr != ' ') || sublen == 1) {
         goto parse_error;
     }
-        if (format_len < 1) {
-            if (exact)
-                goto parse_error;
-        } else {
-            if (*format++ != *substr)
-                goto parse_error;
-            format_len -= 1;
+        const char tmp[1] = {*substr};
+        if (compare_format(&format, &format_len, tmp, 1, exact)) {
+            goto parse_error;
         }
     ++substr;
     --sublen;
 
     /* PARSE THE HOURS */
-    if (format_len < 2) {
-        if (exact || format_len)
-            goto parse_error;
-    } else {
-        if (*format++ != '%' || *format++ != 'H')
-            goto parse_error;
-        format_len -= 2;
+    if (compare_format(&format, &format_len, "%H", 2, exact)) {
+        goto parse_error;
     }
     /* First digit required */
     if (!isdigit(*substr)) {
@@ -374,13 +362,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         if (sublen == 0 || !isdigit(*substr)) {
             goto parse_error;
         }
-        if (format_len < 1) {
-            if (exact)
-                goto parse_error;
-        } else {
-            if (*format++ != ':')
-                goto parse_error;
-            format_len -= 1;
+        if (compare_format(&format, &format_len, ":", 1, exact)) {
+            goto parse_error;
         }
     } else if (!isdigit(*substr)) {
         if (!hour_was_2_digits) {
@@ -390,13 +373,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE MINUTES */
-    if (format_len < 2) {
-        if (exact || format_len)
-            goto parse_error;
-    } else {
-        if (*format++ != '%' || *format++ != 'M')
-            goto parse_error;
-        format_len -= 2;
+    if (compare_format(&format, &format_len, "%M", 2, exact)) {
+        goto parse_error;
     }
     /* First digit required */
     out->min = (*substr - '0');
@@ -430,13 +408,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     /* If we make it through this condition block, then the next
      * character is a digit. */
     if (has_hms_sep && *substr == ':') {
-        if (format_len < 1) {
-            if (exact)
-                goto parse_error;
-        } else {
-            if (*format++ != ':')
-                goto parse_error;
-            format_len -= 1;
+        if (compare_format(&format, &format_len, ":", 1, exact)) {
+            goto parse_error;
         }
         ++substr;
         --sublen;
@@ -450,13 +423,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE SECONDS */
-    if (format_len < 2) {
-        if (exact || format_len)
-            goto parse_error;
-    } else {
-        if (*format++ != '%' || *format++ != 'S')
-            goto parse_error;
-        format_len -= 2;
+    if (compare_format(&format, &format_len, "%S", 2, exact)) {
+        goto parse_error;
     }
     /* First digit required */
     out->sec = (*substr - '0');
@@ -483,13 +451,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if (sublen > 0 && *substr == '.') {
         ++substr;
         --sublen;
-        if (format_len < 1) {
-            if (exact)
-                goto parse_error;
-        } else {
-            if (*format++ != '.')
-                goto parse_error;
-            format_len -= 1;
+        if (compare_format(&format, &format_len, ".", 1, exact)) {
+            goto parse_error;
         }
     } else {
         bestunit = NPY_FR_s;
@@ -497,13 +460,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     }
 
     /* PARSE THE MICROSECONDS (0 to 6 digits) */
-    if (format_len < 2) {
-        if (exact || format_len)
-            goto parse_error;
-    } else {
-        if (*format++ != '%' || *format++ != 'f')
-            goto parse_error;
-        format_len -= 2;
+    if (compare_format(&format, &format_len, "%f", 2, exact)) {
+        goto parse_error;
     }
     numdigits = 0;
     for (i = 0; i < 6; ++i) {
@@ -569,13 +527,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        if (format_len < 1) {
-            if (exact)
-                goto parse_error;
-        } else {
-            if (*format++ != ' ')
-                goto parse_error;
-            format_len -= 1;
+        if (compare_format(&format, &format_len, " ", 1, exact)) {
+            goto parse_error;
         }
     }
 
@@ -589,13 +542,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
 
     /* UTC specifier */
     if (*substr == 'Z') {
-        if (format_len < 2) {
-            if (exact || format_len)
-                goto parse_error;
-        } else {
-            if (*format++ != '%' || *format++ != 'Z')
-                goto parse_error;
-            format_len -= 2;
+        if (compare_format(&format, &format_len, "%Z", 2, exact)) {
+            goto parse_error;
         }
         /* "Z" should be equivalent to tz offset "+00:00" */
         if (out_local != NULL) {
@@ -616,13 +564,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
             --sublen;
         }
     } else if (*substr == '-' || *substr == '+') {
-        if (format_len < 2) {
-            if (exact || format_len)
-                goto parse_error;
-        } else {
-            if (*format++ != '%' || *format++ != 'z')
-                goto parse_error;
-            format_len -= 2;
+        if (compare_format(&format, &format_len, "%z", 2, exact)) {
+            goto parse_error;
         }
         /* Time zone offset */
         int offset_neg = 0, offset_hour = 0, offset_minute = 0;
@@ -707,13 +650,8 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        if (format_len < 1) {
-            if (exact)
-                goto parse_error;
-        } else {
-            if (*format++ != ' ')
-                goto parse_error;
-            format_len -= 1;
+        if (compare_format(&format, &format_len, " ", 1, exact)) {
+            goto parse_error;
         }
     }
 

From 031e0e3660b825cdcde2999f90b8429f640afff0 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Sun, 13 Nov 2022 14:39:44 +0000
Subject: [PATCH 17/18] remove tmp variable

---
 .../_libs/tslibs/src/datetime/np_datetime_strings.c   | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)

diff --git a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
index 76e3e1147df00..597a2aae7a2a3 100644
--- a/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
+++ b/pandas/_libs/tslibs/src/datetime/np_datetime_strings.c
@@ -70,7 +70,7 @@ This file implements string parsing and creation for NumPy datetime.
 // This function will advance the pointer on format
 // and decrement characters_remaining by n on success
 // On failure will return -1 without incrementing
-int compare_format(const char **format, int *characters_remaining,
+static int compare_format(const char **format, int *characters_remaining,
                  const char *compare_to, int n, const int exact) {
   if (*characters_remaining < n) {
     if (exact) {
@@ -202,8 +202,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         ++substr;
         --sublen;
 
-        const char tmp[1] = {ymd_sep};
-        if (compare_format(&format, &format_len, tmp, 1, exact)) {
+        if (compare_format(&format, &format_len, &ymd_sep, 1, exact)) {
             goto parse_error;
         }
         /* Cannot have trailing separator */
@@ -259,8 +258,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
         }
         ++substr;
         --sublen;
-        const char tmp[1] = {ymd_sep};
-        if (compare_format(&format, &format_len, tmp, 1, exact)) {
+        if (compare_format(&format, &format_len, &ymd_sep, 1, exact)) {
             goto parse_error;
         }
     }
@@ -308,8 +306,7 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     if ((*substr != 'T' && *substr != ' ') || sublen == 1) {
         goto parse_error;
     }
-        const char tmp[1] = {*substr};
-        if (compare_format(&format, &format_len, tmp, 1, exact)) {
+        if (compare_format(&format, &format_len, substr, 1, exact)) {
             goto parse_error;
         }
     ++substr;

From c1e6bc2573ebcd6ec9833cdf597d3c890d0f8503 Mon Sep 17 00:00:00 2001
From: MarcoGorelli <>
Date: Thu, 17 Nov 2022 10:08:21 +0000
Subject: [PATCH 18/18] Add co-authors > > Co-authored-by: fdrocha <>
 Co-authored-by: nikitaved <>