Skip to main content

match()

DateParser.match() is the lower-level API. It returns the pyparsing result and the pattern metadata instead of building a datetime.

parser.match(text) -> dict | None

Signature​

match(
    text,
    noprefix=False,
    noyear=True,
    nocharsetfilter=False,
    noseparatorfilter=False,
    noyearformatfilter=False,
    nolanguagefilter=False,
)

Arguments match parse(), plus:

ArgumentDefaultMeaning
noyearTrueInclude patterns that omit the year (current year is filled later by parse())

Set noyear=False to skip day-month-only patterns and reduce false positives on short numeric fragments.

Return value​

On success, a dict with two keys:

  • values — pyparsing ParseResults; items are int()-able (day, month, year, and optionally time fields)
  • pattern — the matched pattern metadata dict, including key, name, format, length, and related flags

On failure, None.

result = parser.match("01.12.2009")
result["pattern"]["key"]   # e.g. 'dt:date:date_2'
dict(result["values"])     # {'day': '01', 'month': '12', 'year': '2009'}

Month values outside 1–12 and day values outside 1–31 are skipped; the next candidate pattern is tried.

When to use it​

  • Logging which pattern fired in a crawl
  • Building analytics over pattern["key"]
  • Diagnosing unexpected parses

Application code that only needs a timestamp should call parse().