Date Normalization¶

Overview¶

Date normalization converts written date expressions into their spoken form for TTS. It supports multiple common date formats, bilingual output for English and Malay, and automatic disambiguation between DD/MM and MM/DD formats.

Dates are extracted as entities early in the pipeline to prevent the fraction normalizer from misinterpreting date separators (e.g., 15/08 as a fraction).

Supported Formats¶

Format	Example	Notes
DD/MM/YYYY	`15/08/2025`	Day first when day > 12
MM/DD/YYYY	`08/15/2025`	Month first when month <= 12
YYYY-MM-DD	`2025-08-15`	ISO format
DD Month YYYY	`15 August 2025`	Full or abbreviated month names
Month DD, YYYY	`August 15, 2025`	Full or abbreviated month names

Month abbreviations are also recognized: Jan, Feb, Mar, Apr, May, Jun, Jul, Aug, Sep, Oct, Nov, Dec.

DD/MM vs MM/DD Ambiguity¶

When the slash format is used and the first number is greater than 12, it is unambiguously treated as DD/MM/YYYY (day/month/year). When both numbers are 12 or less, the format is treated as MM/DD/YYYY.

Input	Interpretation	Reason
`15/08/2025`	DD/MM/YYYY	15 cannot be a month
`08/15/2025`	MM/DD/YYYY	15 cannot be a month
`05/08/2025`	MM/DD/YYYY	Ambiguous -- defaults to MM/DD

English Output¶

from revo_norm import normalize_text

# Slash format (DD/MM, day > 12)
normalize_text("15/08/2025", language="en")
# "fifteen of August two thousand twenty five"

# Slash format (MM/DD, ambiguous)
normalize_text("08/15/2025", language="en")
# "August fifteen, two thousand twenty five"

# ISO format (YYYY-MM-DD)
normalize_text("2025-08-15", language="en")
# "August the fifteenth, two thousand twenty five"

Malay Output¶

Malay uses localized month names and Malay number-to-words conversion.

Malay Month Names¶

Number	English	Malay
01	January	Januari
02	February	Februari
03	March	Mac
04	April	April
05	May	Mei
06	June	Jun
07	July	Julai
08	August	Ogos
09	September	September
10	October	Oktober
11	November	November
12	December	Disember

from revo_norm import normalize_text

# Slash format (DD/MM)
normalize_text("15/08/2025", language="ms")
# "lima belas Ogos dua ribu dua puluh lima"

# Slash format (MM/DD, ambiguous)
normalize_text("08/15/2025", language="ms")
# "Ogos lima belas, dua ribu dua puluh lima"

# ISO format (YYYY-MM-DD)
normalize_text("2025-08-15", language="ms")
# "lima belas Ogos dua ribu dua puluh lima"

How to Disable¶

from revo_norm import normalize_text

# Disable date normalization
result = normalize_text("15/08/2025", language="en", disable=["dates"])
# The date is still extracted as an entity for protection,
# but restored as the original text instead of spoken form.

# Use minimal profile (dates not spoken)
result = normalize_text("15/08/2025", language="en", profile="minimal")

When dates are disabled, they are still extracted from the text to protect them from being mangled by other normalizers (e.g., the fraction normalizer). However, they are restored as their original text rather than converted to spoken form.

Edge Cases¶

Two-digit years: 15/08/25 is matched, but four-digit years produce better spoken output.
Single-digit day/month: 5/8/2025 is recognized the same as 05/08/2025.
Month name variations: Both Aug and August are recognized. The full name is used in output.
Trailing dot on month: Aug. 15, 2025 is recognized.
Date vs fraction conflict: Dates are extracted before fractions run, so 15/08/2025 is never misinterpreted as a fraction.
Dates in URLs: URL extraction runs before date extraction, so dates inside URLs are handled as part of the URL, not as standalone dates.