Syllables

The main building-blocks of predicate-expressions are the syllables, which may massively shorten the resulting string by allowing to leave out parts of the target identifier, and thereby make the process of inputting desired filters less tedious. Another advantage would be that one only needs to remember parts of said identifiers, and can thereby explore a massive realm of available matches in a playful and stepwise manner.

Definition

Since I do not plan on constraining the supported character-set to merely ASCII, I'll have to define syllables in terms of what they aren't, rather than the set of characters which do make them up. They do not begin with a digit (0-9), and they cannot contain hyphens (-), since the latter is what separates individual syllables; every character other than the first may be a digit.

Examples:

• ✅ abc2de3
• ❌ 2abcde
• ❌ two-syllables

Identifiers

As mentioned, individual syllables now string together using hyphens (-) to make up an identifier matching unit. Before diving into the process of matching syllables against identifiers as retrieved from mojang's language-files, let's first look at a few examples, as well as normalization, like it's automatically performed internally.

Raw Identifier	Normalized Result
Minecart with Chest	Minecart-with-Chest
Music Disc C418 - mellohi	Music-Disc-C418-mellohi
Example "one"	Example-'one'
Example_(two)	Example-[two]

Spaces as well as underscores are interpreted as delimiters to syllables within identifiers and thereby become replaced by hyphens. Double-quotes are substituted by single-quotes, just as parentheses are substituted by brackets - both to avoid ambiguity with syntax soon to be discussed.

Matching

When testing an identifier matching unit, made up of one or more syllables joined by hyphens (-), against an identifier, the following simple yet effective algorithm takes place; let's call the former query-syllables, and the latter target-syllables. An identifier matches if no more pending query-syllables remain - namely if each and every one of them could be matched against a part of said identifier. Once a query-syllable matched, it's exact occurrence within the target is removed, so that no other remaining query-syllable can match against it. Let's look at a few examples.

Target: Minecart-with-Chest
Query: car-ches
Remaining Target: Mine-t-with-t
Remaining Query:
Matched: ✅

Target: Diamantbrustplatte
Query: dia-bru
Remaining Target: mant-stplatte
Remaining Query:
Matched: ✅

If the same syllable is contained multiple times, all occurrences need to match one-to-one.

Target: A-Only-Once
Query: a-a-onl
Remaining Target: y-once
Remaining Query: a
Matched: ❌

Target: A-Only-Twice-A
Query: a-a-onl
Remaining Target: y-twice
Remaining Query:
Matched: ✅

Negation

In the case that a syllable is not to be matched within the target, but should rather not exist for the match to be a success, simply prepend it by a bang (!). The algorithm is similar to positive testing, just that the query is removed if it doesn't exist, and kept if it exists, as to leave a remainder, and thereby cause a mismatch. Let's look at an example with multiple target candidates.

Target 1: Acacia-Sign
Target 2: Bamboo-Hanging-Sign
Query: sign-!hang
Remaining Target 1: Acacia
Remaining Target 2: Bamboo-ing
Remaining Query Target 1:
Remaining Query Target 2: !hang
Matched Target 1: ✅
Matched Target 2: ❌

Negation becomes highly useful when constructing custom material-groups, as discussed on the corresponding predicate page.