Remove line breaks

Rejoining text that arrived in pieces

Text copied out of a PDF, an email or a code editor arrives broken at whatever width the original was displayed in. Pasted into a field that wraps by itself, those breaks turn a paragraph into a ragged column.

1. 1Paste the broken text. It is processed in the page, which matters for anything from a contract or a document under embargo.
2. 2Leave paragraph gaps on for prose. Lines inside a paragraph are joined and the blank line between paragraphs stays, which is almost always what you wanted.
3. 3Turn paragraph gaps off when you need everything on one line, such as a value going into a single cell or a single field.
4. 4Pick what replaces the break. A space is right for prose, and nothing is right for a word that was split across a line by hyphenation.
5. 5Leave the collapse option on to clean up the double spaces that appear where a line already ended in a space.

The options and what each one does

Four small decisions cover nearly every version of this problem, and the difference between them is the difference between readable prose and a sentence with its words stuck together.

• Keep paragraphs on: a blank line is treated as a paragraph boundary and preserved, while single breaks inside a block are joined.
• Keep paragraphs off: every break goes and the whole input becomes one continuous line, which is what a spreadsheet cell or a CSV field needs.
• Join with a space: the normal choice for prose, where each line was a continuation of a sentence.
• Join with nothing: for text broken mid-word, where a space would leave the halves separated instead of reunited.
• Join with a comma: for a list that was one item per line and needs to become an inline list.
• Collapse spaces: runs of two or more spaces become one, which cleans up the joins where the original lines already ended in a space.

Where the stray breaks came from

Knowing the source tells you which option to pick, because each source breaks text in a characteristically different way and needs a different join.

• PDFs have no paragraphs. Every visual line is a separate run of text, so a copied page arrives broken at the column width and often with the hyphenation left in.
• Plain text email wraps at 72 to 78 characters by convention, which is why forwarded messages arrive with a hard break every dozen words.
• Code editors and terminals wrap at the window, so a long line copied from a terminal comes back with breaks where the window happened to end.
• Word processors use a paragraph mark that copies as a break, and a manual line break inside a paragraph copies as a different one, so both turn up in the same paste.
• Subtitle and transcript exports break on timing rather than on sense, which is why a transcript reads as fragments until the breaks are removed.
• Windows text uses two characters for a break where other systems use one, and this tool handles both without leaving stray characters behind.

When the line breaks should stay exactly where they are

Removing breaks is destructive in the sense that the original layout cannot be recovered afterwards, so there are places to leave well alone.

• Poetry, lyrics and addresses are line-broken on purpose, and flattening them removes the meaning that the shape was carrying.
• Code is whitespace-sensitive in several languages, and joining lines in Python or YAML does not produce a longer line, it produces something that will not run.
• A social caption written with deliberate paragraph spacing wants the opposite treatment, and there is a generator in this catalogue that keeps those breaks alive through a composer.
• Markdown uses two trailing spaces before a break to force a line break, so removing breaks there changes how the document renders.
• CSV data with quoted fields can contain legitimate line breaks inside a value, and flattening the whole file corrupts the rows rather than tidying them.