Elysia Tools
Navigation
Workflow Playbook
A plain-text presentation workflow for centering lines and marking selected words, phrases, or patterns while keeping code and data semantics outside the transformation.
Hubs
Start with plain text that you can inspect before and after the presentation pass. Keep the original unchanged, confirm the source encoding before pasting, and record the display medium. If the input contains indentation, delimiters, identifiers, serialized values, or code, identify those boundaries before using either tool.
Use text-centerer when lines need a centered appearance. Choose a fixed target length when the destination has a known width; choose auto-detection when the longest processed line should establish the width. Set fillChar deliberately. A single-character fill is the safest choice when an exact width matters, because the tool repeats the supplied fill string.
Review the space policy before running the tool. With trimming disabled, leading whitespace is still removed; with trimming enabled, trailing whitespace is removed as well. The tool does not truncate lines that exceed the target. Those behaviors make it suitable for presentation lines, not for indentation-sensitive source or data.
Use text-highlighter after centering if the reviewer needs selected terms to stand out. Choose words, phrase, or regex according to the input, then set case sensitivity and whole-word behavior. Select a marker style and output mode that the destination accepts. Wrapping markers such as bold, brackets, or Markdown code are easier to audit than uppercase when the spelling of an identifier matters. Test regex input on a small sample and inspect for unintended matches.
Compare short lines, long lines, blank lines, indented lines, and highlighted matches with the canonical copy. Confirm the intended width, fill, marker style, encoding, and line breaks. This emphasis is visual-only: highlighting is not a semantic parser and is not an accessibility proof. Use the marked result for human review, while code, structured data, identifiers, and machine-readable values continue to come from the untouched source.
Workflow playbook
Duplicate the plain-text input, confirm its encoding, and record the audience and display medium. Decide whether indentation, delimiters, identifiers, or serialized values must remain byte-for-byte represented in the canonical copy.
Run text-centerer with a fixed targetLength or with auto-detection when the longest processed line should set the width. Select a deliberate fillChar, use a single-character fill when exact width matters, and enable trimSpaces only when removing boundary spaces is acceptable.
Pass the centered review copy to text-highlighter, then choose words, phrase, or regex, case sensitivity, and whole-word behavior. Select a marker such as bold, brackets, or Markdown code for review; avoid uppercase when the matched spelling or identifier must stay unchanged, and use HTML mode only for a destination that accepts it.
Use a fixed target length when the display has a known width. Use auto-detection when the longest processed line should define the width. A line longer than the target is not truncated, so it must be reviewed separately.
text-centerer removes leading whitespace even when trimming is off; enabling trim also removes trailing whitespace. Do not center lines when indentation or boundary spaces carry meaning.
Choose words for separate terms, phrase for one literal phrase, or regex only when a pattern is necessary. Prefer wrapping markers such as bold, brackets, or Markdown code for semantic-sensitive review; uppercase changes the matched text.
Use the transformed copy only as a presentation aid for human review. Keep the original as the source of truth because highlighting is not a semantic parser and is not proof of accessibility.
Check representative short, long, blank, indented, and marked lines against the canonical input. Confirm the marker count and visible terms, investigate an invalid regex or unexpected spacing, and deliver the marked copy only to a human-review destination, not as parsed code or data.