In Django 3.2 before 3.2.25, 4.2 before 4.2.11, and 5.0 before 5.0.3, the django.utils.text.Truncator.words() method (with html=True) and the truncatewords_html template filter are subject to a potential regular expression denial-of-service attack via a crafted string. NOTE: this issue exists because of an incomplete fix for CVE-2019-14232 and CVE-2023-43665.

An issue was discovered in Django 3.2 before 3.2.24, 4.2 before 4.2.10, and Django 5.0 before 5.0.2. The intcomma template filter was subject to a potential denial-of-service attack when used with very long strings.

In Django 3.2 before 3.2.22, 4.1 before 4.1.12, and 4.2 before 4.2.6, the django.utils.text.Truncator chars() and words() methods (when used with html=True) are subject to a potential DoS (denial of service) attack via certain inputs with very long, potentially malformed HTML text. The chars() and words() methods are used to implement the truncatechars_html and truncatewords_html template filters, which are thus also vulnerable. NOTE: this issue exists because of an incomplete fix for CVE-2019-14232.

In Django 3.2 before 3.2.21, 4.1 before 4.1.11, and 4.2 before 4.2.5, django.utils.encoding.uri_to_iri() is subject to a potential DoS (denial of service) attack via certain inputs with a very large number of Unicode characters.

An issue was discovered in Django 3.2 before 3.2.23, 4.1 before 4.1.13, and 4.2 before 4.2.7. The NFKC normalization is slow on Windows. As a consequence, django.contrib.auth.forms.UsernameField is subject to a potential DoS (denial of service) attack via certain inputs with a very large number of Unicode characters.

In Django 3.2 before 3.2.20, 4 before 4.1.10, and 4.2 before 4.2.3, EmailValidator and URLValidator are subject to a potential ReDoS (regular expression denial of service) attack via a very large number of domain name labels of emails and URLs.

In Django 3.2 before 3.2.19, 4.x before 4.1.9, and 4.2 before 4.2.1, it was possible to bypass validation when using one form field to upload multiple files. This multiple upload has never been supported by forms.FileField or forms.ImageField (only the last uploaded file was validated). However, Django's "Uploading multiple files" documentation suggested otherwise.

An issue was discovered in the Multipart Request Parser in Django 3.2 before 3.2.18, 4.0 before 4.0.10, and 4.1 before 4.1.7. Passing certain inputs (e.g., an excessive number of parts) to multipart forms could result in too many open files or memory exhaustion, and provided a potential vector for a denial-of-service attack.

In Django 3.2 before 3.2.17, 4.0 before 4.0.9, and 4.1 before 4.1.6, the parsed values of Accept-Language headers are cached in order to avoid repetitive parsing. This leads to a potential denial-of-service vector via excessive memory usage if the raw value of Accept-Language headers is very large.

In Django 3.2 before 3.2.16, 4.0 before 4.0.8, and 4.1 before 4.1.2, internationalized URLs were subject to a potential denial of service attack via the locale parameter, which is treated as a regular expression.

An issue was discovered in the HTTP FileResponse class in Django 3.2 before 3.2.15 and 4.0 before 4.0.7. An application is vulnerable to a reflected file download (RFD) attack that sets the Content-Disposition header of a FileResponse when the filename is derived from user-supplied input.

An issue was discovered in Django 3.2 before 3.2.14 and 4.0 before 4.0.6. The Trunc() and Extract() database functions are subject to SQL injection if untrusted data is used as a kind/lookup_name value. Applications that constrain the lookup name and kind choice to a known safe list are unaffected.

An issue was discovered in Django 2.2 before 2.2.28, 3.2 before 3.2.13, and 4.0 before 4.0.4. QuerySet.annotate(), aggregate(), and extra() methods are subject to SQL injection in column aliases via a crafted dictionary (with dictionary expansion) as the passed **kwargs.

A SQL injection issue was discovered in QuerySet.explain() in Django 2.2 before 2.2.28, 3.2 before 3.2.13, and 4.0 before 4.0.4. This occurs by passing a crafted dictionary (with dictionary expansion) as the **options argument, and placing the injection payload in an option name.

An issue was discovered in MultiPartParser in Django 2.2 before 2.2.27, 3.2 before 3.2.12, and 4.0 before 4.0.2. Passing certain inputs to multipart forms could result in an infinite loop when parsing files.

The {% debug %} template tag in Django 2.2 before 2.2.27, 3.2 before 3.2.12, and 4.0 before 4.0.2 does not properly encode the current context. This may lead to XSS.

An issue was discovered in Django 2.2 before 2.2.26, 3.2 before 3.2.11, and 4.0 before 4.0.1. UserAttributeSimilarityValidator incurred significant overhead in evaluating a submitted password that was artificially large in relation to the comparison values. In a situation where access to user registration was unrestricted, this provided a potential vector for a denial-of-service attack.

An issue was discovered in Django 2.2 before 2.2.26, 3.2 before 3.2.11, and 4.0 before 4.0.1. Due to leveraging the Django Template Language's variable resolution logic, the dictsort template filter was potentially vulnerable to information disclosure, or an unintended method call, if passed a suitably crafted key.

Storage.save in Django 2.2 before 2.2.26, 3.2 before 3.2.11, and 4.0 before 4.0.1 allows directory traversal if crafted filenames are directly passed to it.

In Django 2.2 before 2.2.25, 3.1 before 3.1.14, and 3.2 before 3.2.10, HTTP requests for URLs with trailing newlines could bypass upstream access control based on URL paths.

Django 3.1.x before 3.1.13 and 3.2.x before 3.2.5 allows QuerySet.order_by SQL injection if order_by is untrusted input from a client of a web application.

Django before 2.2.24, 3.x before 3.1.12, and 3.2.x before 3.2.4 has a potential directory traversal via django.contrib.admindocs. Staff members could use the TemplateDetailView view to check the existence of arbitrary files. Additionally, if (and only if) the default admindocs templates have been customized by application developers to also show file contents, then not only the existence but also the file contents would have been exposed. In other words, there is directory traversal outside of the template root directories.

In Django 2.2 before 2.2.24, 3.x before 3.1.12, and 3.2 before 3.2.4, URLValidator, validate_ipv4_address, and validate_ipv46_address do not prohibit leading zero characters in octal literals. This may allow a bypass of access control that is based on IP addresses. (validate_ipv4_address and validate_ipv46_address are unaffected with Python 3.9.5+..) .

In Django 2.2 before 2.2.22, 3.1 before 3.1.10, and 3.2 before 3.2.2 (with Python 3.9.5+), URLValidator does not prohibit newlines and tabs (unless the URLField form field is used). If an application uses values with newlines in an HTTP response, header injection can occur. Django itself is unaffected because HttpResponse prohibits newlines in HTTP headers.

In Django 2.2 before 2.2.21, 3.1 before 3.1.9, and 3.2 before 3.2.1, MultiPartParser, UploadedFile, and FieldFile allowed directory traversal via uploaded files with suitably crafted file names.

In Django 2.2 before 2.2.20, 3.0 before 3.0.14, and 3.1 before 3.1.8, MultiPartParser allowed directory traversal via uploaded files with suitably crafted file names. Built-in upload handlers were not affected by this vulnerability.

Django Channels 3.x before 3.0.3 allows remote attackers to obtain sensitive information from a different request scope. The legacy channels.http.AsgiHandler class, used for handling HTTP type requests in an ASGI environment prior to Django 3.0, did not correctly separate request scopes in Channels 3.0. In many cases this would result in a crash but, with correct timing, responses could be sent to the wrong client, resulting in potential leakage of session identifiers and other sensitive data. Note that this affects only the legacy Channels provided class, and not Django's similar ASGIHandler, available from Django 3.0.

The package python/cpython from 0 and before 3.6.13, from 3.7.0 and before 3.7.10, from 3.8.0 and before 3.8.8, from 3.9.0 and before 3.9.2 are vulnerable to Web Cache Poisoning via urllib.parse.parse_qsl and urllib.parse.parse_qs by using a vector called parameter cloaking. When the attacker can separate query parameters using a semicolon (;), they can cause a difference in the interpretation of the request between the proxy (running with default configuration) and the server. This can result in malicious requests being cached as completely safe ones, as the proxy would usually not see the semicolon as a separator, and therefore would not include it in a cache key of an unkeyed parameter.

In Django 2.2 before 2.2.18, 3.0 before 3.0.12, and 3.1 before 3.1.6, the django.utils.archive.extract method (used by "startapp --template" and "startproject --template") allows directory traversal via an archive with absolute paths or relative paths with dot segments.

An issue was discovered in Django 2.2 before 2.2.16, 3.0 before 3.0.10, and 3.1 before 3.1.1 (when Python 3.7+ is used). The intermediate-level directories of the filesystem cache had the system's standard umask rather than 0o077.

An issue was discovered in Django 2.2 before 2.2.16, 3.0 before 3.0.10, and 3.1 before 3.1.1 (when Python 3.7+ is used). FILE_UPLOAD_DIRECTORY_PERMISSIONS mode was not applied to intermediate-level directories created in the process of uploading files. It was also not applied to intermediate-level collected static directories when using the collectstatic management command.

An issue was discovered in Django 2.2 before 2.2.13 and 3.0 before 3.0.7. Query parameters generated by the Django admin ForeignKeyRawIdWidget were not properly URL encoded, leading to a possibility of an XSS attack.

An issue was discovered in Django 2.2 before 2.2.13 and 3.0 before 3.0.7. In cases where a memcached backend does not perform key validation, passing malformed cache keys could result in a key collision, and potential data leakage.

Django 1.11 before 1.11.29, 2.2 before 2.2.11, and 3.0 before 3.0.4 allows SQL Injection if untrusted data is used as a tolerance parameter in GIS functions and aggregates on Oracle. By passing a suitably crafted tolerance to GIS functions and aggregates on Oracle, it was possible to break escaping and inject malicious SQL.

Django 1.11 before 1.11.28, 2.2 before 2.2.10, and 3.0 before 3.0.3 allows SQL Injection if untrusted data is used as a StringAgg delimiter (e.g., in Django applications that offer downloads of data as a series of rows with a user-specified column delimiter). By passing a suitably crafted delimiter to a contrib.postgres.aggregates.StringAgg instance, it was possible to break escaping and inject malicious SQL.

Django before 1.11.27, 2.x before 2.2.9, and 3.x before 3.0.1 allows account takeover. A suitably crafted email address (that is equal to an existing user's email address after case transformation of Unicode characters) would allow an attacker to be sent a password reset token for the matched user account. (One mitigation in the new releases is to send password reset tokens only to the registered user email address.)

Django 2.1 before 2.1.15 and 2.2 before 2.2.8 allows unintended model editing. A Django model admin displaying inline related models, where the user has view-only permissions to a parent model but edit permissions to the inline model, would be presented with an editing UI, allowing POST requests, for updating the inline model. Directly editing the view-only parent model was not possible, but the parent model's save() method was called, triggering potential side effects, and causing pre and post-save signal handlers to be invoked. (To resolve this, the Django admin is adjusted to require edit permissions on the parent model in order for inline models to be editable.)

An issue was discovered in Django 1.11.x before 1.11.23, 2.1.x before 2.1.11, and 2.2.x before 2.2.4. Due to an error in shallow key transformation, key and index lookups for django.contrib.postgres.fields.JSONField, and key lookups for django.contrib.postgres.fields.HStoreField, were subject to SQL injection. This could, for example, be exploited via crafted use of "OR 1=1" in a key or index name to return all records, using a suitably crafted dictionary, with dictionary expansion, as the **kwargs passed to the QuerySet.filter() function.

An issue was discovered in Django 1.11.x before 1.11.23, 2.1.x before 2.1.11, and 2.2.x before 2.2.4. If passed certain inputs, django.utils.encoding.uri_to_iri could lead to significant memory usage due to a recursion when repercent-encoding invalid UTF-8 octet sequences.

An issue was discovered in Django 1.11.x before 1.11.23, 2.1.x before 2.1.11, and 2.2.x before 2.2.4. Due to the behaviour of the underlying HTMLParser, django.utils.html.strip_tags would be extremely slow to evaluate certain inputs containing large sequences of nested incomplete HTML entities.

An issue was discovered in Django 1.11.x before 1.11.23, 2.1.x before 2.1.11, and 2.2.x before 2.2.4. If django.utils.text.Truncator's chars() and words() methods were passed the html=True argument, they were extremely slow to evaluate certain inputs due to a catastrophic backtracking vulnerability in a regular expression. The chars() and words() methods are used to implement the truncatechars_html and truncatewords_html template filters, which were thus vulnerable.

An issue was discovered in Django 1.11 before 1.11.22, 2.1 before 2.1.10, and 2.2 before 2.2.3. An HTTP request is not redirected to HTTPS when the SECURE_PROXY_SSL_HEADER and SECURE_SSL_REDIRECT settings are used, and the proxy connects to Django via HTTPS. In other words, django.http.HttpRequest.scheme has incorrect behavior when a client uses HTTP.

An issue was discovered in Django 1.11 before 1.11.21, 2.1 before 2.1.9, and 2.2 before 2.2.2. The clickable Current URL value displayed by the AdminURLFieldWidget displays the provided value without validating it as a safe URL. Thus, an unvalidated value stored in the database, or a value provided as a URL query parameter payload, could result in an clickable JavaScript link.

Django 1.11.x before 1.11.19, 2.0.x before 2.0.11, and 2.1.x before 2.1.6 allows Uncontrolled Memory Consumption via a malicious attacker-supplied value to the django.utils.numberformat.format() function.

In Django 1.11.x before 1.11.18, 2.0.x before 2.0.10, and 2.1.x before 2.1.5, an Improper Neutralization of Special Elements in Output Used by a Downstream Component issue exists in django.views.defaults.page_not_found(), leading to content spoofing (in a 404 error page) if a user fails to recognize that a crafted URL has malicious content.

An issue was discovered in Django 2.1 before 2.1.2, in which unprivileged users can read the password hashes of arbitrary accounts. The read-only password widget used by the Django Admin to display an obfuscated password hash was bypassed if a user has only the "view" permission (new in Django 2.1), resulting in display of the entire password hash to those users. This may result in a vulnerability for sites with legacy user accounts using insecure hashes.

django.middleware.common.CommonMiddleware in Django 1.11.x before 1.11.15 and 2.0.x before 2.0.8 has an Open Redirect.

An issue was discovered in Django 2.0 before 2.0.3, 1.11 before 1.11.11, and 1.8 before 1.8.19. If django.utils.text.Truncator's chars() and words() methods were passed the html=True argument, they were extremely slow to evaluate certain inputs due to a catastrophic backtracking vulnerability in a regular expression. The chars() and words() methods are used to implement the truncatechars_html and truncatewords_html template filters, which were thus vulnerable.

An issue was discovered in Django 2.0 before 2.0.3, 1.11 before 1.11.11, and 1.8 before 1.8.19. The django.utils.html.urlize() function was extremely slow to evaluate certain inputs due to catastrophic backtracking vulnerabilities in two regular expressions (only one regular expression for Django 1.8.x). The urlize() function is used to implement the urlize and urlizetrunc template filters, which were thus vulnerable.

django.contrib.auth.forms.AuthenticationForm in Django 2.0 before 2.0.2, and 1.11.8 and 1.11.9, allows remote attackers to obtain potentially sensitive information by leveraging data exposure from the confirm_login_allowed() method, as demonstrated by discovering whether a user account is inactive.

In Django 1.10.x before 1.10.8 and 1.11.x before 1.11.5, HTML autoescaping was disabled in a portion of the template for the technical 500 debug page. Given the right circumstances, this allowed a cross-site scripting attack. This vulnerability shouldn't affect most production sites since you shouldn't run with "DEBUG = True" (which makes this page accessible) in your production settings.