Elysia Tools
Navigation
Workflow Playbook
Prepare imported or copied text for reuse by recovering its encoding, choosing a character-width convention, standardizing inline dates or currency, and selecting a final text representation.
Hubs
This workflow is for text that will be reused elsewhere, not for changing a source of record in place. Save the original and a small sample that exposes the problem: mojibake, fullwidth punctuation, mixed date patterns, currency symbols, or MIME escape sequences. Decide the receiving system's charset, width convention, locale, and output format before touching the text.
If the input is email MIME transfer text, run quoted-printable-decoder before interpreting its characters. If the bytes are in the wrong file charset, use file-encoding-converter next and keep the intermediate output. This order prevents later steps from treating escape sequences or mojibake as literal content.
Choose exactly one width direction with fullwidth-to-halfwidth or halfwidth-to-fullwidth. The right choice comes from the receiving field: narrow ASCII-oriented identifiers need the first path, while a CJK presentation rule may need the second. Compare punctuation and digits, not only letters.
Once the representation is stable, use text-date-format-unifier for dates with a declared target pattern and text-currency-normalizer for money expressions with a known code and policy. Record what changed and retain the original value; formatting or conversion of a currency string is not a financial approval.
Finish with the branch that matches the consumer. Use advanced-case-converter for naming conventions, number-to-words for prose that needs spelled-out numbers, and word-separator for line-oriented word lists. If the final destination is MIME, apply quoted-printable-encoder only after all text edits. Validate representative characters, dates, amounts, line breaks, and the downstream format before delivery.
Workflow playbook
Preserve the original first. Decode Quoted-Printable only when the input is MIME transfer text, then use file-encoding-converter when the file charset must be changed. Save the readable intermediate output before editing its characters.
Inspect the consumer's field rules and choose fullwidth-to-halfwidth or halfwidth-to-fullwidth. Apply one direction to the relevant fields, then compare punctuation, digits, and CJK samples with the target convention.
After encoding and width are stable, convert dates and currency expressions that have an explicit target policy. Keep locale, currency code, source value, and review notes so a formatting change is not mistaken for an untracked business decision.
Choose only the needed branch: convert identifiers to the required naming convention, spell out numbers for prose, or split text into one word per line for token or list-oriented consumers. Do not apply these transforms to fields whose meaning depends on the original form.
Use quoted-printable-decoder for MIME transfer text; use file-encoding-converter for a file whose bytes use the wrong charset. If both apply, decode the transfer layer first and then convert the resulting character encoding.
Choose fullwidth-to-halfwidth for ASCII-oriented fields, identifiers, or systems that expect narrow punctuation; choose halfwidth-to-fullwidth for a CJK layout convention. Do not run both directions without a defined target.
Use text-date-format-unifier only when a target date pattern is defined, and text-currency-normalizer only when the currency code, display rule, and conversion policy are known. Preserve values that are outside the intended scope.
Use advanced-case-converter for identifiers, number-to-words when prose requires spelled-out numbers, word-separator for one-token-per-line output, and quoted-printable-encoder only at the end when MIME transport requires it.
Review a before-and-after sample, count changed fields, and confirm the downstream format. If the destination is MIME text, run quoted-printable-encoder last; otherwise deliver the plain or file output without an unnecessary transport layer.