Skip to content

slang

Indonesian slang normalization engine.

The slang object powers BASA's informal text preprocessing, including slang replacement and repeated-character reduction.

Import

from basa import slang

Basic Usage

slang.normalize(
    "gw gkkkk ngerti bngtttt"
)

Output:

'saya tidak mengerti banget'

Character Reduction

Repeated characters are automatically simplified.

slang.normalize(
    "sihhhhhhh"
)

Output:

'sih'

Dictionary-Based Replacement

slang.normalize(
    "udh makan blm?"
)

Output:

'sudah makan belum?'

Inspecting the Dictionary

You can inspect the supported slang words and their standard replacements, or count the total number of entries:

# Get all supported words (alphabetically sorted)
words = slang.supported_words()

print(words["gw"])
# 'saya'
print(words["gk"])
# 'tidak'

# Count total entries
print(len(slang))
# 1308

Adding Custom Entries

Users can extend the dictionary dynamically.

slang.add({
    "wfo": "work from office",
    "wfh": "work from home"
})

Example:

slang.normalize(
    "hari ini wfh"
)

Output:

'hari ini work from home'

Dictionary Coverage

BASA currently provides:

  • General conversation slang
  • Social media abbreviations
  • Internet expressions
  • Common Indonesian informal forms

Future releases may include:

  • Javanese slang
  • Sundanese slang
  • Regional dictionaries
  • Domain-specific vocabularies

Design Philosophy

The slang engine prioritizes:

  • Conservative replacements
  • Explicit mappings
  • Fast regex-based processing
  • Developer extensibility
  • Indonesian-first normalization