inflect

URL slugs, English pluralisation, ordinals and identifier case conversion. Pure - no capabilities.

ecko get github.com/ecko-lang/inflect

import inflect

Pure computation: it declares no capabilities, so it cannot touch the network, the filesystem or the environment.

Version 0.10.1 - source - MIT.

fail(message)

The error this package raises.

is_alpha(c)

Is this character an ASCII letter?

is_digit(c)

Is this character a digit?

is_upper(c)

Is this an uppercase ASCII letter?

words(text)

words("helloWorldAgain") -> ["hello", "world", "again"].

Splits on anything that is not a letter or digit, and additionally at a lowercase-to-uppercase boundary. A run of capitals is kept together until the last one, so XMLHttpRequest gives ["xml", "http", "request"] rather than eleven single letters.

starts_word(cs, idx)

Does a new word begin at idx? True at a lowercase-to-uppercase step (helloWorld), and at the last capital of a run followed by a lowercase letter (XMLHttp breaks before the H, not before the T).

capitalize(word)

Capitalise the first character, leaving the rest alone.

translit_table()

Accented Latin characters and what they become. Transliterating rather than dropping matters: "Café Münster" stripped is "caf-mnster", which is unreadable and collides with unrelated titles. Multi-character expansions (ß to ss, æ to ae) are the reason this is a table and not arithmetic.

transliterate(text)

Replace accented characters with their ASCII equivalents. Anything unknown is left for slug to drop.

slug(text, opts = {})

slug("Hello, World!") -> "hello-world".

Options: separator (default "-") and max_length (default 0, no limit). A length limit cuts at a word boundary rather than mid-word, and never leaves a dangling separator.

clip(parts, separator, max_length)

Keep whole words while the joined result fits. Cutting mid-word would produce a slug that reads as a different, wrong word.

opt(m, k, fallback)

get with a default.

uncountable()

Words whose plural is the same as their singular. Applying a rule to any of these produces "sheeps" or "informations".

irregular()

Irregular singular-to-plural pairs, in one direction; singular reverses this table rather than keeping a second one that could drift out of step.

keeps_f()

Words ending in -f or -fe that do NOT take -ves. English has both, and the rule alone would give "rooves" and "chiefs" the same treatment.

o_takes_es()

Words ending in -o that take -es. Most take a plain -s (photos, pianos), so the shorter list is the exceptional one.

is_vowel(c)

Is this a vowel?

tail(word, n)

The last n characters of a word.

head(word, n)

Everything but the last n characters.

match_case(original, result)

Re-apply the casing of original to result: a leading capital stays a leading capital, and an all-caps word stays all-caps. A CMS pluralises headings and route names, so "Post" must not come back as "posts".

plural(word)

The plural of a word. Case is preserved.

plural_rule(w)

The regular rules, applied to an already-lowercased word.

singular(word)

The singular of a word. Case is preserved.

singular_rule(w)

The regular rules in reverse.

pluralize(count, word)

"1 post" / "2 posts". The number is included; word_for gives just the word.

word_for(count, word)

The right form of word for count, without the number.

ordinal(n)

ordinal(21) -> "21st".

The rule is not "look at the last digit": 11, 12 and 13 take "th" despite ending in 1, 2 and 3. That trap is the whole reason this function exists.

snake(text)

snake_case.

kebab(text)

kebab-case.

camel(text)

camelCase - first word lowercase, the rest capitalised.

pascal(text)

PascalCase - every word capitalised.

title(text)

Title Case - every word capitalised, spaces between.

case_names()

The shape names to_case accepts.

to_case(text, name)

to_case(text, "snake"). An unknown name raises rather than passing the input through, because a silent passthrough turns a typo into a subtle data bug.