PDF Text Cleaner

How to clean copied PDF text

1. Paste the original plain text to start cleaning automatically. Choose paragraph repair or spacing only, and enable optional changes only when needed.
2. Review the updated excerpts and decide whether each uncertain hyphen should be removed, retained or left unchanged. Each decision updates the result automatically.
3. Copy the cleaned result when processing finishes. Changes to the source or settings update the result and reset hyphen choices; typing waits for a short pause. Clear removes both texts and restores default settings.

What this cleaner changes

PDF copy-paste often inserts a newline at each visual line ending. This tool joins eligible lines, but a newline alone cannot reveal the author’s intent. It preserves blank paragraphs, recognized lists, fenced or indented code, tabular lines and standalone URLs, emails and paths. Sentence-ending punctuation also prevents a join. Headings, poetry and citations are not detected perfectly; review changes or use spacing only.

A line-final hyphen can split a word or belong to a compound. Each occurrence has its own decision and stays unchanged by default. Repairs always start from your current original, so resetting choices never compounds earlier edits. Cleaning an already cleaned result can expose new joins; it is not guaranteed to be idempotent.

Automatic joining omits spaces only between adjacent Han characters. It is not language detection and cannot infer every Japanese, Korean, Thai or Khmer boundary. Optional changes cover only U+00AD, ff/fi/fl/ffi/ffl, U+00A0/U+202F and repeated ASCII spaces. Protected structures remain exact. Tabs, trailing hard-break spaces, combining accents, emoji, shaping controls and other Unicode characters are preserved.

Examples: before → after

A wrapped ␊ sentence. → A wrapped sentence. (␊ denotes a line break, without the surrounding display spaces.)

inter-␊national → unchanged by default; choose removal for international, or retention for inter-national. state-␊of-the-art → state-of-the-art with retention.

中文␊文本 → 中文文本 in Automatic mode. file → file only with ligature expansion. First paragraph.␊␊Second paragraph. keeps both line breaks.

Local processing and limits

Processing is local to your browser; this tool does not upload or persist drafts. Shared site analytics and advertising can still make network requests. Editors and review excerpts use the site’s masking attributes. The 500,000 UTF-16-unit limit is not a visible-character count; emoji may use more than one unit. Oversized input is retained without truncation.