Statistics

import { analyzeUrdu, countWords, countSentences, splitWords, splitSentences } from "urdu-text-utils";

analyzeUrdu("پاکستان ایک خوبصورت ملک ہے۔ اس کی آبادی زیادہ ہے۔");

{
  "characters": 49,
  "charactersNoSpaces": 40,
  "words": 10,
  "sentences": 2,
  "paragraphs": 1,
  "urduPercentage": 100,
  "diacritics": 0,
  "digits": 0,
  "averageWordsPerSentence": 5,
  "readingTimeMinutes": 0.1
}

Counting words

countWords("پاکستان ایک خوبصورت ملک ہے");  // 5
countWords("پاکستان ایک خوبصورت ملک ہے۔"); // 5 — the ۔ is not a word
countWords("آپ کیسے ہیں؟");                // 3

Splitting on /\s+/ alone leaves ہے۔ attached and counts Urdu punctuation as part of the word. The split here covers whitespace plus Urdu punctuation (، ؛ ؟ ۔ ٫ ٬) and the ASCII set.

splitWords(text) returns the tokens themselves — the starting point for an index, a tag extractor or a highlighter.

Counting sentences

countSentences("یہ پہلا جملہ ہے۔ یہ دوسرا ہے۔"); // 2
countSentences("آپ کیسے ہیں؟ میں ٹھیک ہوں۔");    // 2
countSentences("ایک جملہ۔");                      // 1 — no phantom empty sentence

Terminators are ۔ (U+06D4), ؟ (U+061F), !, . and …. Note the Urdu full stop is a distinct character from the ASCII period — splitting on . alone finds nothing in real Urdu text.

Fields

Field	Notes
characters	Every codepoint, including spaces and diacritics
charactersNoSpaces	Whitespace removed
words	Same as countWords
sentences	Same as countSentences
paragraphs	Blocks separated by a blank line
urduPercentage	Arabic-script letters as a share of all letters, 0–100
diacritics	Combining marks present
digits	Digits in any of the three systems
averageWordsPerSentence	Rounded to one decimal
readingTimeMinutes	At 180 words per minute

Reading speed is set at 180 wpm rather than the ~230 wpm usually quoted for English: the script is denser and short vowels are inferred rather than read. Treat it as an editorial estimate.

urduPercentage is a useful moderation signal — a submission at 20% in an Urdu-only form is probably English with a few pasted words.

Try it

normalizeUrdu(text) stripDiacritics

Input

كيا حال ہے

Output

کیا حال ہے