PDF copy paste gibberish happens when the font inside the file has no reliable map from the shapes you see to real characters. The page looks fine because the shapes are drawn correctly, but the codes behind them mean nothing to your clipboard. Try a second viewer first, and if the symbols persist, run OCR on the page.
What PDF copy paste gibberish looks like
PDF copy paste gibberish comes in a few recognisable patterns, and each points to a likely cause. Text can paste as random symbols, as letters shifted by a fixed step, as empty boxes, or as correct words with a few wrong characters. Reading the pattern before you try a fix saves time.
The table shows example patterns from a line that reads "Opening balance 1,250.00" on screen. The pasted text is made up to illustrate each case.
| What pastes | Pattern | Likely cause | First fix to try |
|---|---|---|---|
ÒÔÅÎÉÎÇ ÂÁÌÁÎÃÅ ±¬²µ°®°° |
Accented or high symbols, one per letter | Font uses its own codes with no map to standard characters | OCR the page |
Pqfojoh cbmbodf 2-361/11 |
Real letters, each one step off | Custom encoding offset from the standard order | Second viewer, then OCR |
□□□□□□□ □□□□□□□ □□□□□□□□ |
Empty boxes | Codes map to characters your paste target cannot display, or to nothing | Paste into a plain text editor to check, then OCR |
Opening balance 1,25O.OO |
Mostly right, a few look-alikes | The text layer came from earlier OCR that misread digits | Redo OCR, or check digits by hand |
Openingbalance1,250.00 |
Correct letters, no spaces | Spaces were never stored as characters | A spacing fix, not a font fix |
| (nothing) | Selection works but pastes nothing | Characters map to no value at all | OCR the page |
The shifted-letters pattern is worth a closer look. If every pasted letter is the next one along the alphabet, it is tempting to shift them back yourself. The last section explains why that is usually a mistake.
The no-spaces pattern looks similar at first glance but has a different cause and a different fix. The words are right; only the gaps are missing. That case is about how characters are placed on the page, not about what they are, and why PDF copy gives no spaces between words covers it.
Why does a PDF font paste as symbols?
A PDF font pastes as symbols when the file gives no way to turn its character codes into standard Unicode characters. A PDF draws each letter by looking up a code in an embedded font. Copying needs a second lookup, from that same code to a real letter, and some files leave that second table out.
The PDF 32000-1:2008 specification lists, in section 9.10.2, the methods a reader can use to map a character code to Unicode. The first is a ToUnicode CMap, a table inside the font entry that pairs each code with its character. Failing that, the reader falls back on a few predefined encodings and standard character sets. If none of these methods produces a value, the specification says "there is no way to determine what the character code represents", and the reader may substitute whatever it chooses. That substitution is the gibberish you see.
Why software produces files like this
A program that writes a PDF often embeds only the letters a document uses, a practice called subsetting, and numbers them in the order they first appear. In such a file, code 1 might be "O", code 2 "p", code 3 "e", and so on. The drawing works because the font knows what shape code 1 is. Whether copying works depends on whether the program also wrote the ToUnicode table that says code 1 means "O". Some reporting systems, older print drivers and conversion steps skip it or write a broken one.
Processing a file again can also lose the table. Re-saving through another tool, optimising for size, or converting to a different PDF version can drop or damage the map even when the original copied cleanly. If an earlier version of the same statement copied fine, ask for that version.
Testing PDF gibberish in a second viewer first
A second viewer is the cheapest test because it tells you whether the file or the app is at fault. Viewers fall back differently when a font map is missing or damaged, so one app may paste symbols while another pastes clean text. If two viewers both paste gibberish, the problem is in the file.
Run the test in this order:
- Copy the same line in a different app. Try a browser's built-in PDF viewer and a desktop reader. Paste each into a plain text editor, not straight into Excel, so formatting does not hide the result.
- Search the PDF for a word you can see. Use the viewer's find command to look for "balance" or a merchant name. If find cannot locate a word that is plainly on the page, the stored text does not match the visible shapes, and every viewer will struggle.
- Try one line, not the whole page. A single garbled font can sit beside clean ones. Headings may use a decorative font with no map while the transaction table uses a plain one that copies fine.
If only some fonts are affected, you may not need to fix the whole file. A statement whose logo text and headings paste as symbols, while dates and amounts paste correctly, can still go into Excel, because the rows you need are clean.
OCR as the fix for PDF gibberish that never copies cleanly
OCR, short for optical character recognition, fixes gibberish by ignoring the broken text and reading the page as a picture instead. It looks at the drawn shapes, which are correct, and produces new characters from them. The cost is that OCR can misread digits, so the result needs checking.
There are two ways to apply OCR to a gibberish file:
- Make a searchable copy. Many PDF tools can run OCR and add a new text layer to the page. For a gibberish file, the tool must be told to replace the existing text rather than skip pages that already have some, or it will leave the broken text in place.
- Turn the pages into images first. Export or print the pages as images, then OCR those. With no old text left, the OCR text is the only text. Some files do this to themselves when they are flattened. How a flattened PDF turns text into an image explains how to tell whether a file has been through that step.
After OCR, check the figures that matter. For a bank statement, the fastest check is arithmetic: the opening balance plus credits minus debits should equal the closing balance. A gap points to a misread digit.
If you split the file to treat only the gibberish pages, you may end up with several pieces. Combining PDF files into one before you convert covers keeping the months in order and avoiding duplicate pages when you put them back together.
A different shortcut applies when the statement also exists as a web page in online banking. Printing that page to PDF from the browser builds a new file from the page's own text, so it does not inherit the original file's broken fonts. Saving a statement page with print to PDF on iPhone shows the hidden gesture for doing that on a phone.
Is PDF copy paste gibberish ever partly right?
PDF copy paste gibberish is often partly right, and the right part is usually the digits. Some files map the numbers and common punctuation correctly but leave letters unmapped, or the reverse. When the amounts and dates paste cleanly, you may be able to convert the file and fix only the descriptions.
Three partial cases come up often:
- Digits fine, letters garbled. Amounts, dates and balances paste correctly. Descriptions are symbols. The sheet is usable for totals and reconciliation, and you can type the few descriptions you need or OCR only that column.
- One font garbled. Headers, footers or a bank's branded font paste as symbols, and the transaction table is clean. This is common and mostly harmless, because the garbled lines are not rows you want.
- Some characters wrong. Most text is right, but a few letters such as "fi" or "fl" paste as one odd symbol. These are ligatures, two letters drawn as one shape, and a missing map for that one shape gives a single wrong character.
Partial gibberish needs a careful check rather than a full redo. Look at the amount column first. If every amount is a real number, Excel will sum it and the balance check will tell you whether any digit is off. Then decide whether the descriptions matter enough to fix.
FastBankConvert is getting ready to open as a bank statement converter that tells you before download if the balances do not add up, which is the same check you would run by hand on a partly garbled file. Get started is the place to join.
Gibberish you should not fix by find and replace
Find and replace looks like a quick fix when every letter is shifted by the same step, but it rarely holds up. The shift often changes between fonts, sizes or pages, a single code can stand for different shapes in different fonts, and a replacement that fixes one line can quietly corrupt the next.
Here is how it goes wrong. Suppose "Pqfojoh" should read "Opening", so you replace each character with the one before it. That works on the heading. But the amounts use a different font in the same file, where code values were assigned in a different order. Your replacement now turns "1,250.00", which was correct, into something else. If you replace on the whole sheet, the balance column breaks without any warning.
Other reasons to avoid hand-fixing:
- Subset fonts number letters by first use. Two statements from the same bank can map the same code to different letters, so a fix list built for one file fails on the next.
- Digits and letters may share codes across fonts. A rule meant for descriptions can change amounts.
- You cannot see what was lost. If a code maps to nothing, there is no character to replace, and the gap stays invisible until a total is wrong.
Use find and replace only for a small, known problem, such as one ligature symbol that always stands for "fi" in descriptions, and only in the description column. For anything wider, OCR the page and check the numbers, or ask the sender for a copy produced without the broken fonts.