void pandas_datetime_to_datetimestruct(npy_datetime val, PANDAS_DATETIMEUNIT fr,
pandas_datetimestruct *result) {
pandas_datetime_metadata meta;
meta.base = fr;
meta.num = 1;
convert_datetime_to_datetimestruct(&meta, val, result);
}
void pandas_datetime_to_datetimestruct(npy_datetime val, PANDAS_DATETIMEUNIT fr,
pandas_datetimestruct *result) {
convert_datetime_to_datetimestruct(fr, val, result);
}
/*
* Parses (almost) standard ISO 8601 date strings. The differences are:
*
* + The date "20100312" is parsed as the year 20100312, not as
* equivalent to "2010-03-12". The '-' in the dates are not optional.
* + Only seconds may have a decimal point, with up to 18 digits after it
* (maximum attoseconds precision).
* + Either a 'T' as in ISO 8601 or a ' ' may be used to separate
* the date and the time. Both are treated equivalently.
* + Doesn't (yet) handle the "YYYY-DDD" or "YYYY-Www" formats.
* + Doesn't handle leap seconds (seconds value has 60 in these cases).
* + Doesn't handle 24:00:00 as synonym for midnight (00:00:00) tomorrow
* + Accepts special values "NaT" (not a time), "Today", (current
* day according to local time) and "Now" (current time in UTC).
*
* 'str' must be a NULL-terminated string, and 'len' must be its length.
* 'unit' should contain -1 if the unit is unknown, or the unit
* which will be used if it is.
* 'casting' controls how the detected unit from the string is allowed
* to be cast to the 'unit' parameter.
*
* 'out' gets filled with the parsed date-time.
* 'out_local' gets set to 1 if the parsed time was in local time,
* to 0 otherwise. The values 'now' and 'today' don't get counted
* as local, and neither do UTC +/-#### timezone offsets, because
* they aren't using the computer's local timezone offset.
* 'out_bestunit' gives a suggested unit based on the amount of
* resolution provided in the string, or -1 for NaT.
* 'out_special' gets set to 1 if the parsed time was 'today',
* 'now', or ''/'NaT'. For 'today', the unit recommended is
* 'D', for 'now', the unit recommended is 's', and for 'NaT'
* the unit recommended is 'Y'.
*
* Returns 0 on success, -1 on failure.
*/
int
parse_iso_8601_datetime(char *str, int len,
PANDAS_DATETIMEUNIT unit,
NPY_CASTING casting,
pandas_datetimestruct *out,
npy_bool *out_local,
PANDAS_DATETIMEUNIT *out_bestunit,
npy_bool *out_special)
{
int year_leap = 0;
int i, numdigits;
char *substr, sublen;
PANDAS_DATETIMEUNIT bestunit;
/* Initialize the output to all zeros */
memset(out, 0, sizeof(pandas_datetimestruct));
out->month = 1;
out->day = 1;
/*
* The string "today" means take today's date in local time, and
* convert it to a date representation. This date representation, if
* forced into a time unit, will be at midnight UTC.
* This is perhaps a little weird, but done so that the
* 'datetime64[D]' type produces the date you expect, rather than
* switching to an adjacent day depending on the current time and your
* timezone.
*/
if (len == 5 && tolower(str[0]) == 't' &&
tolower(str[1]) == 'o' &&
tolower(str[2]) == 'd' &&
tolower(str[3]) == 'a' &&
tolower(str[4]) == 'y') {
NPY_TIME_T rawtime = 0;
struct tm tm_;
time(&rawtime);
if (get_localtime(&rawtime, &tm_) < 0) {
return -1;
}
out->year = tm_.tm_year + 1900;
out->month = tm_.tm_mon + 1;
out->day = tm_.tm_mday;
bestunit = PANDAS_FR_D;
/*
* Indicate that this was a special value, and
* is a date (unit 'D').
*/
if (out_local != NULL) {
*out_local = 0;
}
if (out_bestunit != NULL) {
*out_bestunit = bestunit;
}
if (out_special != NULL) {
*out_special = 1;
}
/* Check the casting rule */
if (unit != -1 && !can_cast_datetime64_units(bestunit, unit,
casting)) {
PyErr_Format(PyExc_TypeError, "Cannot parse \"%s\" as unit "
"'%s' using casting rule %s",
str, _datetime_strings[unit],
npy_casting_to_string(casting));
return -1;
}
return 0;
}
/* The string "now" resolves to the current UTC time */
if (len == 3 && tolower(str[0]) == 'n' &&
tolower(str[1]) == 'o' &&
tolower(str[2]) == 'w') {
NPY_TIME_T rawtime = 0;
pandas_datetime_metadata meta;
time(&rawtime);
/* Set up a dummy metadata for the conversion */
meta.base = PANDAS_FR_s;
meta.num = 1;
bestunit = PANDAS_FR_s;
/*
* Indicate that this was a special value, and
* use 's' because the time() function has resolution
* seconds.
*/
if (out_local != NULL) {
*out_local = 0;
}
if (out_bestunit != NULL) {
*out_bestunit = bestunit;
}
if (out_special != NULL) {
*out_special = 1;
}
/* Check the casting rule */
if (unit != -1 && !can_cast_datetime64_units(bestunit, unit,
casting)) {
PyErr_Format(PyExc_TypeError, "Cannot parse \"%s\" as unit "
"'%s' using casting rule %s",
str, _datetime_strings[unit],
npy_casting_to_string(casting));
return -1;
}
return convert_datetime_to_datetimestruct(&meta, rawtime, out);
}
/* Anything else isn't a special value */
if (out_special != NULL) {
*out_special = 0;
}
substr = str;
sublen = len;
/* Skip leading whitespace */
while (sublen > 0 && isspace(*substr)) {
++substr;
--sublen;
}
/* Leading '-' sign for negative year */
if (*substr == '-') {
++substr;
--sublen;
}
if (sublen == 0) {
goto parse_error;
}
/* PARSE THE YEAR (digits until the '-' character) */
out->year = 0;
while (sublen > 0 && isdigit(*substr)) {
out->year = 10 * out->year + (*substr - '0');
++substr;
--sublen;
}
/* Negate the year if necessary */
if (str[0] == '-') {
out->year = -out->year;
}
/* Check whether it's a leap-year */
year_leap = is_leapyear(out->year);
/* Next character must be a '-' or the end of the string */
if (sublen == 0) {
if (out_local != NULL) {
*out_local = 0;
}
bestunit = PANDAS_FR_Y;
goto finish;
}
else if (*substr == '-') {
++substr;
--sublen;
}
else {
goto parse_error;
}
/* Can't have a trailing '-' */
if (sublen == 0) {
goto parse_error;
}
/* PARSE THE MONTH (2 digits) */
if (sublen >= 2 && isdigit(substr[0]) && isdigit(substr[1])) {
out->month = 10 * (substr[0] - '0') + (substr[1] - '0');
if (out->month < 1 || out->month > 12) {
PyErr_Format(PyExc_ValueError,
"Month out of range in datetime string \"%s\"", str);
goto error;
}
substr += 2;
sublen -= 2;
}
else {
goto parse_error;
}
/* Next character must be a '-' or the end of the string */
if (sublen == 0) {
if (out_local != NULL) {
*out_local = 0;
}
bestunit = PANDAS_FR_M;
goto finish;
}
else if (*substr == '-') {
++substr;
--sublen;
}
else {
goto parse_error;
}
/* Can't have a trailing '-' */
if (sublen == 0) {
goto parse_error;
}
/* PARSE THE DAY (2 digits) */
if (sublen >= 2 && isdigit(substr[0]) && isdigit(substr[1])) {
out->day = 10 * (substr[0] - '0') + (substr[1] - '0');
if (out->day < 1 ||
out->day > _days_per_month_table[year_leap][out->month-1]) {
PyErr_Format(PyExc_ValueError,
"Day out of range in datetime string \"%s\"", str);
goto error;
}
substr += 2;
sublen -= 2;
}
else {
goto parse_error;
}
/* Next character must be a 'T', ' ', or end of string */
if (sublen == 0) {
if (out_local != NULL) {
*out_local = 0;
}
bestunit = PANDAS_FR_D;
goto finish;
}
else if (*substr != 'T' && *substr != ' ') {
goto parse_error;
}
else {
++substr;
--sublen;
}
/* PARSE THE HOURS (2 digits) */
if (sublen >= 2 && isdigit(substr[0]) && isdigit(substr[1])) {
out->hour = 10 * (substr[0] - '0') + (substr[1] - '0');
if (out->hour < 0 || out->hour >= 24) {
PyErr_Format(PyExc_ValueError,
"Hours out of range in datetime string \"%s\"", str);
goto error;
}
substr += 2;
sublen -= 2;
}
else {
goto parse_error;
}
/* Next character must be a ':' or the end of the string */
if (sublen > 0 && *substr == ':') {
++substr;
--sublen;
}
else {
bestunit = PANDAS_FR_h;
goto parse_timezone;
}
/* Can't have a trailing ':' */
if (sublen == 0) {
goto parse_error;
}
/* PARSE THE MINUTES (2 digits) */
if (sublen >= 2 && isdigit(substr[0]) && isdigit(substr[1])) {
out->min = 10 * (substr[0] - '0') + (substr[1] - '0');
if (out->hour < 0 || out->min >= 60) {
PyErr_Format(PyExc_ValueError,
"Minutes out of range in datetime string \"%s\"", str);
goto error;
}
substr += 2;
sublen -= 2;
}
else {
goto parse_error;
}
/* Next character must be a ':' or the end of the string */
if (sublen > 0 && *substr == ':') {
++substr;
--sublen;
}
else {
bestunit = PANDAS_FR_m;
goto parse_timezone;
}
/* Can't have a trailing ':' */
if (sublen == 0) {
goto parse_error;
}
/* PARSE THE SECONDS (2 digits) */
if (sublen >= 2 && isdigit(substr[0]) && isdigit(substr[1])) {
out->sec = 10 * (substr[0] - '0') + (substr[1] - '0');
if (out->sec < 0 || out->sec >= 60) {
PyErr_Format(PyExc_ValueError,
"Seconds out of range in datetime string \"%s\"", str);
goto error;
}
substr += 2;
sublen -= 2;
}
else {
goto parse_error;
}
/* Next character may be a '.' indicating fractional seconds */
if (sublen > 0 && *substr == '.') {
++substr;
--sublen;
}
else {
bestunit = PANDAS_FR_s;
goto parse_timezone;
}
/* PARSE THE MICROSECONDS (0 to 6 digits) */
numdigits = 0;
for (i = 0; i < 6; ++i) {
out->us *= 10;
if (sublen > 0 && isdigit(*substr)) {
out->us += (*substr - '0');
++substr;
--sublen;
++numdigits;
}
}
if (sublen == 0 || !isdigit(*substr)) {
if (numdigits > 3) {
bestunit = PANDAS_FR_us;
}
else {
bestunit = PANDAS_FR_ms;
}
goto parse_timezone;
}
/* PARSE THE PICOSECONDS (0 to 6 digits) */
numdigits = 0;
for (i = 0; i < 6; ++i) {
out->ps *= 10;
if (sublen > 0 && isdigit(*substr)) {
out->ps += (*substr - '0');
++substr;
--sublen;
++numdigits;
}
}
if (sublen == 0 || !isdigit(*substr)) {
if (numdigits > 3) {
bestunit = PANDAS_FR_ps;
}
else {
bestunit = PANDAS_FR_ns;
}
goto parse_timezone;
}
/* PARSE THE ATTOSECONDS (0 to 6 digits) */
numdigits = 0;
for (i = 0; i < 6; ++i) {
out->as *= 10;
if (sublen > 0 && isdigit(*substr)) {
out->as += (*substr - '0');
++substr;
--sublen;
++numdigits;
}
}
if (numdigits > 3) {
bestunit = PANDAS_FR_as;
}
else {
bestunit = PANDAS_FR_fs;
}
parse_timezone:
if (sublen == 0) {
// Unlike NumPy, treating no time zone as naive
goto finish;
/*
if (convert_datetimestruct_local_to_utc(out, out) < 0) {
goto error;
}
// Since neither "Z" nor a time-zone was specified, it's local
if (out_local != NULL) {
*out_local = 1;
}
goto finish;
*/
}
/* UTC specifier */
if (*substr == 'Z') {
/* "Z" means not local */
if (out_local != NULL) {
*out_local = 0;
}
if (sublen == 1) {
goto finish;
}
else {
++substr;
--sublen;
}
}
/* Time zone offset */
else if (*substr == '-' || *substr == '+') {
int offset_neg = 0, offset_hour = 0, offset_minute = 0;
/*
* Since "local" means local with respect to the current
* machine, we say this is non-local.
*/
if (out_local != NULL) {
*out_local = 0;
}
if (*substr == '-') {
offset_neg = 1;
}
++substr;
--sublen;
/* The hours offset */
if (sublen >= 2 && isdigit(substr[0]) && isdigit(substr[1])) {
offset_hour = 10 * (substr[0] - '0') + (substr[1] - '0');
substr += 2;
sublen -= 2;
if (offset_hour >= 24) {
PyErr_Format(PyExc_ValueError,
"Timezone hours offset out of range "
"in datetime string \"%s\"", str);
goto error;
}
}
else {
goto parse_error;
}
/* The minutes offset is optional */
if (sublen > 0) {
/* Optional ':' */
if (*substr == ':') {
++substr;
--sublen;
}
/* The minutes offset (at the end of the string) */
if (sublen >= 2 && isdigit(substr[0]) && isdigit(substr[1])) {
offset_minute = 10 * (substr[0] - '0') + (substr[1] - '0');
substr += 2;
sublen -= 2;
if (offset_minute >= 60) {
PyErr_Format(PyExc_ValueError,
"Timezone minutes offset out of range "
"in datetime string \"%s\"", str);
goto error;
}
}
else {
goto parse_error;
}
}
/* Apply the time zone offset */
if (offset_neg) {
offset_hour = -offset_hour;
offset_minute = -offset_minute;
}
add_minutes_to_datetimestruct(out, -60 * offset_hour - offset_minute);
}
/* Skip trailing whitespace */
while (sublen > 0 && isspace(*substr)) {
++substr;
--sublen;
}
if (sublen != 0) {
goto parse_error;
}
finish:
if (out_bestunit != NULL) {
*out_bestunit = bestunit;
}
/* Check the casting rule */
if (unit != -1 && !can_cast_datetime64_units(bestunit, unit,
casting)) {
PyErr_Format(PyExc_TypeError, "Cannot parse \"%s\" as unit "
"'%s' using casting rule %s",
str, _datetime_strings[unit],
npy_casting_to_string(casting));
return -1;
}
return 0;
parse_error:
PyErr_Format(PyExc_ValueError,
"Error parsing datetime string \"%s\" at position %d",
str, (int)(substr-str));
return -1;
error:
return -1;
}