fixups

Author: Marco Gorelli

pandas/_libs/tslib.pyx

@@ -624,7 +624,7 @@ cpdef array_to_datetime(
                                 continue
                             elif is_raise:
                                 raise ValueError(
-                                    f"time data \"{val}\" at position {i} doesn't match \"{format}\""
+                                    f"time data \"{val}\" at position {i} doesn't match format \"{format}\""
                                 )
                             return values, tz_out
 

pandas/_libs/tslibs/np_datetime.pyx

@@ -53,7 +53,7 @@ cdef extern from "src/datetime/np_datetime_strings.h":
                                 npy_datetimestruct *out,
                                 NPY_DATETIMEUNIT *out_bestunit,
                                 int *out_local, int *out_tzoffset,
-                                const char *format, int exact)
+                                const char *format, int format_len, int exact)
 
 
 # ----------------------------------------------------------------------
@@ -287,7 +287,7 @@ cdef inline int string_to_dts(
     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, exact)
+                                   format_buf, format_length, exact)
 
 
 cpdef ndarray astype_overflowsafe(

pandas/_libs/tslibs/src/datetime/np_datetime_strings.c

@@ -66,28 +66,40 @@ This file implements string parsing and creation for NumPy datetime.
  *      they aren't using the computer's local timezone offset.
  *
  * Returns 0 on success, -1 on failure.
+
+so, basically:
+    - if exact, need to check that it matches and that we're not at the end of the string
+    - if not exact, then either it matches, or we're at the end of the string
+
  */
 
 #define FORMAT_STARTSWITH(ch) \
-  if (format_len > 0){   \
-  if (*format != ch) {   \
-    goto parse_error;    \
-  }                      \
-  ++format;              \
+  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,
-                            const char* format, int exact) {
-    printf("entering\n");
+                            const char* format, int format_len, int exact) {
     int year_leap = 0;
     int i, numdigits;
     const char *substr;
     int sublen;
     NPY_DATETIMEUNIT bestunit = NPY_FR_GENERIC;
-    int format_len = strlen(format);
     printf("format len: %d", format_len);
 
     /* If year-month-day are separated by a valid separator,
@@ -118,26 +130,14 @@ int parse_iso_8601_datetime(const char *str, int len, int want_exc,
     while (sublen > 0 && isspace(*substr)) {
         ++substr;
         --sublen;
-        if (!isspace(*format)) {
-          goto parse_error;
-        }
-        ++format;
-        if (!*format) {
-          goto parse_error;
-        }
+        FORMAT_STARTSWITH(' ');
     }
 
     /* Leading '-' sign for negative year */
     if (*substr == '-') {
         ++substr;
         --sublen;
-        if (*format != '-') {
-          goto parse_error;
-        }
-        ++format;
-        if (!*format) {
-          goto parse_error;
-        }
+        FORMAT_STARTSWITH('-');
     }
 
     FORMAT_STARTSWITH('%');

pandas/_libs/tslibs/src/datetime/np_datetime_strings.h

@@ -60,6 +60,7 @@ parse_iso_8601_datetime(const char *str, int len, int want_exc,
                         int *out_local,
                         int *out_tzoffset,
                         const char* format,
+                        int format_len,
                         int exact);
 
 /*

pandas/core/arrays/datetimes.py

@@ -2211,7 +2211,7 @@ def objects_to_datetime64ns(
             yearfirst=yearfirst,
             require_iso8601=require_iso8601,
             allow_mixed=allow_mixed,
-            format=format,
+            format=format or "",
             exact=exact
         )
         result = result.reshape(data.shape, order=order)

pandas/core/tools/datetimes.py

@@ -443,7 +443,8 @@ def _convert_listlike_datetimes(
         errors=errors,
         require_iso8601=require_iso8601,
         allow_object=True,
-        format=format, exact=False
+        format=format,
+        exact=exact,
     )
 
     if tz_parsed is not None:

pandas/tests/tools/test_to_datetime.py

@@ -1767,7 +1767,7 @@ def test_to_datetime_iso8601_fails(self, input, format, exact):
         # https://github.com/pandas-dev/pandas/issues/12649
         with pytest.raises(
             ValueError,
-            match=rf"time data \"{input}\" at position 0 doesn't match format {format}",
+            match=rf"time data \"{input}\" at position 0 doesn't match format \"{format}\"",
         ):
             to_datetime(input, format=format, exact=exact)
 
@@ -1785,7 +1785,7 @@ def test_to_datetime_iso8601_exact_fails(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 {format}",
+            match=rf"time data \"{input}\" at position 0 doesn't match format \"{format}\"",
         ):
             to_datetime(input, format=format)
 
@@ -1822,7 +1822,7 @@ def test_to_datetime_iso8601_separator(self, input, format):
         with pytest.raises(
             ValueError,
             match=(
-                rf"time data \"{input}\" at position 0 doesn\'t match format {format}"
+                rf"time data \"{input}\" at position 0 doesn\'t match format \"{format}\""
             ),
         ):
             to_datetime(input, format=format)