Matching logic

Before going into the automatic guild definition and automatic channel definition, we first need to explain how their patterns are matched. Automatic guild definition, automatic channel definition and also Automatic responder determine the match by considering a logic pattern. The logic pattern can either be an actual logical operations (and, or, not) or a text-matching operation.

Text matching operations

These operations are used for matching actual text. Two types currently exist. They are the RegEx type, for matching based on a regex pattern, and a contains type, for matching when the text input contains the (single) configured keyword.

Caution

Text matching is lower-case. This means that all the message content will be interpreted as lower-case characters.

When writing patterns, make sure the pattern words are lower-case.

regex

This can be used for matching a RegEx (regular expression pattern). It is probably the most desirable as it allows flexible pattern matching. For example, if we want the message to contain the words “buy” and “nft”, we can write our pattern like this: buy.*nft. The former pattern will match when the message contains both words - “buy” and “nft”, and “nft” appears after “buy”. The .* means match any character (.) zero or infinite times (*).

Here are some example messages the buy.*nft pattern will match:

  • “Where can I buy the NFTs?”

  • “May I buy the dragon NFT from you?”

  • “I would really want to buy the bunny NFT.”

Additionally, RegEx supports the logical OR operation. This can be done by separating multiple RegEx patterns with the | character. For example, the buy.*nft|sell.*nft pattern will match the text message if any of the 2 patterns (separated by |) matches.

Caution

RegEx is sensitive to spaces around the | (the logical OR) character. When a space is inserted into a RegEx pattern, it is considered as something that needs to appear for the text to be matched.

For example, pattern hello | world would match if the text contains the “hello “ sequence or the “ world” sequence (notice the spaces).

Thus, unless you want spaces to be matched as well, write buy.*nft|sell.*nft instead of buy.*nft | sell.*nft! This is especially important for server (guild) names.

For testing RegEx patterns, the following site is recommended: https://regex101.com/.

contains

This can be used for matching text messages containing a certain a word. As a parameter it accepts the word (keyword) used for checking. Usually, contains would be used alongside logical operations, such as or_, to match any of the multiple words.

Logical operations

Logical operations are used to combine multiple text matching operations, as well as other nested logical operations. Themselves, they do not match the text inside a message.

and_

Represents a logical AND operation. and_ evaluates to true when all of the operands inside evaluate to true.

For example, if we write:

1and_(contains('buy'), contains('nft'), contains('dragon'))

then the text message will be matched only if it contains all of the words “buy”, “nft” and “dragon” (in any order). The above example would in a human-readable form look like contains('buy') and contains('nft') and contains('dragon'), where the contains('word') evaluates to a human-readable form of if 'word' is in message.

or_

Represents a logical OR operation. or_ evaluates to true when any of the operands inside evaluate to true.

For example, if we write:

1or_(contains('buy'), contains('nft'), contains('dragon'))

then the text message will be matched only if it contains any of the words “buy”, “nft” and “dragon” (in any order). The above example would in a human-readable form look like contains('buy') or contains('nft') or contains('dragon').

not_

Represents a logical NOT operation. not_ accepts a single operand and evaluates to true when that operand is false. Basically, it negates the operand.

For example, if we write:

1and_(contains('buy'), not_(contains('dragon')))

then the text message will be matched only if it contains the word “buy” but doesn’t contain the word “dragon”. The above example would in a human-readable form look like contains('buy') and not contains('dragon').