#  `replace`

Regex replacement on page content streams
## Usage
> pdftl `<input>` `replace` `[<spec>...]` `output` `<output>`

## Details
The `replace` operation performs replacement of parts of
page content streams, based on regular expressions. in the
PDF file.  Page ranges can be specified. The default page
range is all pages. The `<spec>` specification is:

```
  [optional page range]/<from>/<to>/[count]
```

where `<from>` and `<to>` are strings describing regular
expressions, as described at
<https://docs.python.org/3/library/re.html>.

The delimiter `/` can be replaced with any other non-alphnumeric
character. It must break the `<spec>` into exactly 4 parts (where the
first may be empty). The delimiter is defined as the final character
of `<spec>`, ignoring digits.

Any trailing digits are interpreted as `count`, which is the
maximum number of times the expression will be matched for
each page content stream.

Before and after the replacement is applied, the page
content stream is normalized (see the [`normalize`](<normalize.md>) operation), which
results in it appearing with one operator per line.

By default, the replacement also recurses into Form XObjects referenced
by the page. Pass ``recurse=false`` to restrict replacement to the
top-level page content stream only.

*See also* the [`dump_streams`](<dump_streams.md>) operation, which lets you examine normalized content
streams so you can formulate a regular expression to use with `replace`.
## Examples

> Replace red with blue on pages 1-3
```
pdftl in.pdf replace '1-3/1 0 0 (RG|rg)/0 0 1 \1/' output out.pdf
```

> Replace foo with bar on all pages, skipping Form XObjects
```
pdftl in.pdf replace recurse=false '/foo/bar/' output out.pdf
```


**Tags**: in_place, content_stream, dangerous

*Source: pdftl.operations.replace*

*Read online: [https://pdftl.readthedocs.io/en/latest/operations/replace.html](https://pdftl.readthedocs.io/en/latest/operations/replace.html)*

*Type: Operation*