Regularized Text

A Regularized Text is a character string that starts with a lowercase character ([a-z]), and has subsequent characters that match regex [a-z]_[0-9]-, and doesn't end with a - character. In other words, every text that matches the regex ^[a-z]([a-z]_[0-9]-)*(?<=[^-])$ is a regularized text.

Purpose

Regularized texts are used primarily to ensure consistency and predictability in naming and referencing elements within TEv2. It is used for several reasons:

• Ease of Processing: Regularized texts are easier to process and handle by various software systems, as they follow a predictable pattern. This uniformity is particularly important for tasks like sorting, searching, and indexing.
• Avoiding Conflicts: By following a strict pattern, regularized texts help prevent issues like naming conflicts or syntax errors that might arise from irregular naming conventions.
• Interoperability: Standardized text formats ensure better interoperability between different systems and software components. When different systems follow the same naming conventions, it's easier to integrate and share data across them.
• Readability and Maintenance: For IT personnel, especially those who are not programmers, having a consistent naming convention (like regularized text) makes it easier to understand and maintain systems. It reduces the learning curve and aids in quicker identification of elements.

Use of Regularized Texts within TEv2

Within TEv2, regularized texts are used:

• to construct more complex identifiers such as term refs, term ids, terminology identifiers, URLs, etc.
• as a class of texts, with members such as termType, term, grouptags, and scopetag.
• to represent form phrases in MRG entries (note that the formPhrases field of curated texts do not typically contain regularized texts)
• etc.

Regularizing Texts (#regularization-process)

The process that TEv2 tools use if they need to convert a normal, human readable text into a regularized texts, consists of the following steps:

1. convert the text to lowercase;
2. remove all characters at the beginning of the text that do not match regex [a-z].
3. replace all sequences of characters that may not appear in a regularized text with the --character;
4. replace all sequences of -characters with a single - character, such that in the resulting text there are no two consecutive - characters;
5. remove any leading and/or trailing - characters.

Here's a table that shows how a variety of input texts would be converted into regularized texts:

#	Input Text	Regularized Text
1	example	example
2	ex@mple	ex-mple
3	Ex4mPLe 4	ex4mple-4
4	(example):	example
5	EX(ample)	ex-ample
6	1#-_23ex3mple	ex3mple
7	ex--am@#ple123	ex-am-ple
8	**e!x@a#m$p%l^e**	e-x-a-m-p-l-e

---

Notes