https://github.com/trananhtung/textstat
Readability metrics for English text (Flesch, Flesch-Kincaid, Gunning Fog, SMOG, ARI, Coleman-Liau). Zero-dep, no_std.
https://github.com/trananhtung/textstat
flesch nlp no-std readability rust text-processing textstat
Last synced: 1 day ago
JSON representation
Readability metrics for English text (Flesch, Flesch-Kincaid, Gunning Fog, SMOG, ARI, Coleman-Liau). Zero-dep, no_std.
- Host: GitHub
- URL: https://github.com/trananhtung/textstat
- Owner: trananhtung
- License: apache-2.0
- Created: 2026-06-21T10:12:07.000Z (8 days ago)
- Default Branch: main
- Last Pushed: 2026-06-22T14:47:57.000Z (7 days ago)
- Last Synced: 2026-06-22T16:17:00.400Z (6 days ago)
- Topics: flesch, nlp, no-std, readability, rust, text-processing, textstat
- Language: Rust
- Size: 14.6 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- Changelog: CHANGELOG.md
- License: LICENSE-APACHE
Awesome Lists containing this project
README
# textstat
[](#contributors-)
[](https://crates.io/crates/textstat)
[](https://docs.rs/textstat)
[](https://github.com/trananhtung/textstat/actions/workflows/ci.yml)
[](#license)
[](#no_std)
**Readability metrics for English text.** Flesch Reading Ease, Flesch-Kincaid
Grade, Gunning Fog, SMOG, Automated Readability Index, Coleman-Liau — plus the
word / sentence / syllable counts they're built on. Inspired by Python's
[`textstat`](https://pypi.org/project/textstat/). Zero dependencies, `#![no_std]`.
```rust
let text = "The cat sat on the mat. The dog ran fast.";
assert_eq!(textstat::lexicon_count(text), 10); // words
assert_eq!(textstat::sentence_count(text), 2);
assert_eq!(textstat::syllable_count(text), 10);
let ease = textstat::flesch_reading_ease(text); // ~117 (very easy)
let grade = textstat::flesch_kincaid_grade(text); // ~ -1.8 (well below grade 1)
```
## Why textstat?
Rust's `readability` crates are *article extractors* (arc90/Mozilla Readability) —
they pull the main content out of a web page. **None of them compute readability
*scores*.** `textstat` fills that gap: drop-in functions for the standard formulas,
useful for content tooling, SEO, writing assistants, education, and accessibility
(WCAG) checks.
## Install
```toml
[dependencies]
textstat = "0.1"
```
## Metrics
| Function | Returns |
| --- | --- |
| `flesch_reading_ease` | 0–100+ score (higher = easier) |
| `flesch_kincaid_grade` | U.S. school grade level |
| `gunning_fog` | U.S. school grade level |
| `smog_index` | U.S. school grade level |
| `automated_readability_index` | U.S. school grade level |
| `coleman_liau_index` | U.S. school grade level |
| `reading_time(text, wpm)` | estimated seconds to read |
### Counts
`lexicon_count` (words), `sentence_count`, `syllable_count`, `syllables(word)`,
`polysyllabic_count`, `char_count` (non-whitespace), `letter_count`.
## Accuracy
Syllables are estimated with a fast English heuristic (vowel groups + a silent-`e`
rule, including accented Latin vowels), not a pronunciation dictionary, so scores
are *close to* — not bit-identical with — dictionary-based tools. The metric
**formulas** are the standard published ones.
- `sentence_count` skips decimal points (`3.14`) and initialism dots (`U.S.A.`);
trailing abbreviation dots (`Dr.`) may still add one.
- Best results are on English text. Non-Latin scripts (no Latin vowels) fall back
to one syllable per word, so scores stay bounded rather than correct.
- Empty input yields `0` everywhere — no panics, no division by zero.
## no_std
`textstat` is `#![no_std]` (needs only `alloc`) with a dependency-free Newton's
-method `sqrt`, so it builds for bare-metal targets such as `thumbv7em-none-eabi`.
## Contributors ✨
This project follows the [all-contributors](https://github.com/all-contributors/all-contributors) specification. Contributions of any kind are welcome — code, docs, bug reports, ideas, reviews! See the [emoji key](https://allcontributors.org/docs/en/emoji-key) for how each contribution is recognized, and open a PR or issue to get involved.
Thanks goes to these wonderful people:
## License
Licensed under either of [Apache-2.0](LICENSE-APACHE) or [MIT](LICENSE-MIT) at
your option.
