search for multi-byte encoded strings in binary data.

Author: Jens Getreu
Date: 2017-01-28
Version: 1.4.3
Copyright: Apache License, Version 2.0 (for details see COPYING section)
Manual section:1
Manual group:Forensic Tools


stringsext [options] [-e ENC...] [--] [FILE...]
stringsext [options] [-e ENC...] [--] [-]


stringsext is a Unicode enhancement of the GNU strings tool with additional functionalities: stringsext recognizes Cyrillic, CJKV characters and other scripts in all supported multi-byte-encodings, while GNU strings fails in finding any of these scripts in UTF-16 and many other encodings.

stringsext is mainly useful for determining the Unicode content of non-text files: It prints all graphic character sequences in FILE or stdin that are at least MIN bytes long.

Unlike GNU strings stringsext can be configured to search for valid characters not only in ASCII but also in many other input encodings, e.g.: utf-8, utf-16be, utf-16le, big5-2003, euc-jp, koi8-r and many others. --list-encodings shows a list of valid encoding names based on the WHATWG Encoding Standard. When more than one encoding is specified, the scan is performed in different threads simultaneously.

When searching for UTF-16 encoded strings, 96% of all possible two byte sequences, interpreted as UTF-16 code unit, relate directly to a Unicode code point. As a result, the probability of encountering valid Unicode characters in a random byte stream, interpreted as UTF-16, is also 96%. In order to reduce this big number of false positives, stringsext provides a parameterizable Unicode-block-filter. See --encodings option for more details.

stringsext reads its input data from FILE. With no FILE, or when FILE is -, it reads standard input stdin.

When invoked with stringsext -e ascii -c i stringsext can be used as GNU strings replacement.

Under Windows a Unicode editor is required. For first tests wordpad should do. Choose the Courier new font or Segoe UI symbol font to see the flag symbols ⚑ (U+2691).


-c MODE, --control-chars=MODE

Determine if and how control characters are printed.

The search algorithm first scans for valid character sequences which are then re-encoded into UTF-8 strings containing graphic (printable) and control (non-printable) characters.

When MODE is set to p all valid (control and graphic) characters are printed. Warning: Control characters may contain a harmful payload. An attacker may exploit a vulnerability of your terminal or post processing software. Use with caution.

MODE r will never print any control character but instead indicate their position: Control characters in valid strings are first grouped and then replaced with the Unicode replacement character '�' (U+FFFD). This mode is most useful together with --radix because it keeps the whole valid character sequence in one line allowing post-processing the output with line oriented tools like grep. To ease post-processing, the output in MODE r is formatted slightly different from other modes: instead of indenting the byte-counter, the encoding name and the found string with spaces as separator, only one tab is inserted.

When MODE is i all control characters are silently ignored. They are first grouped and then replaced with a newline character.

See the output of --help for the default value of MODE.

-e ENC, --encoding=ENC

Set (multiple) input search encodings.


Search for strings in encoded in ENCNAME. Encoding names ENCNAME are denoted following the WATHWG standard. --list-encodings prints a list of available encodings.
Print only strings at least min bytes long. The length is measured in UTF-8 encoded bytes. MIN overwrites the general --bytes MIN option for this ENC only.

Restrict the search to characters within UNICODEBLOCK. This can be used to search for a certain script or to reduce false positives, especially when searching for UTF-16 encoded strings. See for a list of scripts and their corresponding Unicode-block-ranges. UNICODEBLOCK has the following syntax:


XXXXXX and YYYYYY are the lower and upper bounds of the Unicode block in hexadecimal. The prefix U+ can be omitted. The default value for this optional range is U+0..U+10FFFF which means "no filter" or "print all characters whatever their Unicode code-point is". For performance reasons the filter is implemented with a logical bit-mask. If necessary, the given UNICODEBLOCK is enlarged to be representable as a bit-mask. In this case a warning specifying the enlarged UNICODEBLOCK is emitted.

When a second optional UNICODEBLOCK is given, the total Unicode-point search range is the union of the first and the second.

See the output of --help for the default value of ENC.

-f, --print-file-name
Print the name of the file before each string.
-h, --help
Print a synopsis of available options and default values.
-l, --list-encodings
List available encodings as WHATWG Encoding Standard names and exit.
-n MIN, --bytes=MIN
Print only strings at least MIN bytes long. The length is measured in UTF-8 encoded bytes. --help shows the default value.
-p FILE, --output=FILE
Print to FILE instead of stdout.
-s SPLIT-MIN, --split_bytes=SPLIT-MIN

Print only split pieces at least SPLIT-MIN bytes long. The length is measured in UTF-8 encoded bytes and applies to all scanners. SPLIT-MIN=1 (default) ensures that no byte can get lost (never any true negatives, but false positives possible). With a value SPLIT-MIN>1 the first or the second piece can get lost, but the probability of false positives is reduced.

You only need this option when your output contains too many flag symbols ⚑ next to very short strings.

Explanation: In some rare circumstances a graphic string is split into two smaller pieces (see LIMITATIONS). Their cutting edges are labelled with a flag symbol ⚑ (U+2691). This option controls the minimum length of a split piece to be printed.

-t RADIX, --radix=RADIX

Print the offset within the file before each valid string. The single character argument specifies the radix of the offset: o for octal, x for hexadecimal, or d for decimal. When a valid string is split into several graphic character sequences, the cut-off point is labelled according to the --control-chars option and no additional offset is printed at the cut-off point.

The exception to the above is --encoding=ascii --control-chars=i for which the offset is always printed before each graphic character sequence.

When the output of stringsext is piped to another filter you may consider --control-chars=r to keep multi-line strings in one line.

-v, --version
Print version info and exit.


other values


List available encodings:

stringsext -l

Search for UTF-8 strings and strings in UTF-16 Big-Endian encoding:

stringsext -e utf-8  -e utf-16be  someimage.raw

Same, but read from stream:

cat someimage.raw | stringsext -e utf-8  -e utf-16be  -

The above is also useful when reading a non-file device:

cat /dev/sda1  | stringsext -e utf-8  -e utf-16be  -

When used with pipes -c r is required:

stringsext -e iso-8859-7  -c r  -t x  someimage.raw | grep "Ιστορία"

Reduce the number of false positives, when scanning an image file for UTF-16. In the following example we search for Cyrillic, Arabic and Siriac strings, which may contain these additional these symbols: \t !"#$%&'()*+,-./0123456789:;<=>?

stringsext -e UTF-16le,30,U+20..U+3f,U+400..U+07ff someimage.raw

The same but shorter:

stringsext -e UTF-16le,30,20..3f,400..07ff someimage.raw

Combine Little-Endian and Big-Endian scanning:

stringsext -e UTF-16be,20,U+0..U+3FF -e UTF-16le,20,U+0..U+3FF someimage.raw

The following settings are designed to produce bit-identical output with GNU strings:

stringsext -e ascii -c i         # equals `strings`
stringsext -e ascii -c i -t d    # equals `strings -t d`
stringsext -e ascii -c i -t x    # equals `strings -t x`
stringsext -e ascii -c i -t o    # equals `strings -t o`

The following examples perform the same search, but the output format is slightly different:

stringsext -e UTF-16LE,10,0..7f  # equals `strings -n 10 -e l`
stringsext -e UTF-16BE,10,0..7f  # equals `strings -n 10 -e b`


A valid string is a sequence a valid characters according to the encoding chosen with --encoding. A valid string may contain control characters and graphic (visible and human readable) characters. stringsext is a tool to extract sequences of graphic characters out of a binary data stream.

A scanner is defined with the --encoding ENC option. Multiple scanners operate in parallel. The search field is divided into input chunks of WIN_LEN bytes (see source code for exact size) in size. A scanner is a module that extracts valid character sequences, valid strings, of an input chunk.

A valid string is then fed into a filter that extracts multiple graphic strings out of a valid string. A filter may apply additional criteria such as MIN or UNICODEBLOCK.


  1. Valid strings smaller than FINISH_STR_BUF are never cut. When a valid string exceeds WIN_LEN bytes it is always cut. It may happen that at the cutting edge locates a short graphic string that is then split into two pieces which are printed on separate lines. stringsext labels such a cutting edge with two flag symbols ⚑ (U+2691). Furthermore, one or both of those pieces may then become to short to meet the --bytes condition. In order not to loose any bytes of a piece the --bytes option is not observed for split strings. The downside of this is the appearance of some undesirable false positives. Therefore the --split-bytes option allows to set an additional condition to control the appearance of these false positives: The SPLIT-MIN value determines the minimum number of bytes a split piece must have to be printed. Note that with a value SPLIT-MIN > 1 some bytes of the split graphic string may not appear in the output. Therefore the default is SPLIT-MIN = 1.

    In practice, the above limitation occurs only when the search field contains large vectors of Null (0x00) terminated strings. For most multi-byte encodings, as well as for the Unicode-scanner, the Null (0x00) character is regarded as a valid control character. Thus the Unicode scanner will detect such a string vector as one big string which might exceed the WIN_LEN buffer size.

    For searching in large Null (0x00) terminated string vectors, the ASCII scanner is recommended. The ASCII scanner regards Null (0x00) as an invalid character, so the string vector will be detected as a sequence of short distinguishable valid strings. These short strings will most likely never exceed the WIN_LEN buffer and therefore will never be split. In such a scenario it is a good practise to run Unicode and ASCII scanners in parallel.

    Summary: It is guaranteed that valid strings not longer than FINISH_STR_BUF are never split. However, when the size of a valid string exceeds FINISH_STR_BUF bytes it may be split into two or more valid strings and then filtered separately. Note that this limitation refers to the valid string length. A valid string may consist of several graphic strings. If a valid string is longer than WIN_LEN bytes, it is always split. To know the values of the constants please refer to the definition in the source code of your stringsext build. Original values are: FINISH_STR_BUF = 6144 bytes, WIN_LEN = 14342 bytes.

  2. It is guaranteed that all string sequences are detected and printed according to the search criteria. However due to potential false positives when interpreting binary data as multi-byte-strings, it may happen that the first characters of a valid string may not be recognised immediately. In practice, this effect occurs very rarely and the scanner synchronises with the correct character boundaries quickly.


Project website:


Copyright (C) 2016 Jens Getreu

Licenced under the Apache Licence, Version 2.0 (the "Licence"); you may not use this file except in compliance with the Licence. You may obtain a copy of the Licence at

Unless required by applicable law or agreed to in writing, software distributed under the Licence is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the Licence for the specific language governing permissions and limitations under the Licence.