repo string | pull_number int64 | instance_id string | issue_numbers list | base_commit string | patch string | test_patch string | problem_statement string | hints_text string | created_at string | version string | updated_at string | environment_setup_commit string | FAIL_TO_PASS list | PASS_TO_PASS list | FAIL_TO_FAIL list | PASS_TO_FAIL list | source_dir string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
rinja-rs/askama | 673 | rinja-rs__askama-673 | [
"671"
] | 358f7cd07dc42ba4189d2661461ff0b43a27c304 | diff --git a/askama_shared/src/generator.rs b/askama_shared/src/generator.rs
--- a/askama_shared/src/generator.rs
+++ b/askama_shared/src/generator.rs
@@ -246,7 +246,7 @@ struct Generator<'a, S: std::hash::BuildHasher> {
next_ws: Option<&'a str>,
// Whitespace suppression from the previous non-literal. Will b... | diff --git a/askama_shared/src/lib.rs b/askama_shared/src/lib.rs
--- a/askama_shared/src/lib.rs
+++ b/askama_shared/src/lib.rs
@@ -701,5 +705,14 @@ mod tests {
)
.unwrap();
assert_eq!(config.whitespace, WhitespaceHandling::Preserve);
+
+ let config = Config::new(
+ r#"
+ ... | Trim all but 1 whitespace jinja condition
@vallentin suggested [here](https://github.com/djc/askama/pull/664#pullrequestreview-945253861) that we could extend the "trim_all_whitespace" feature to be able to trim all whitespace characters but one.
So currently, we have:
```jinja
<div> {%- if foo -%} blabla {%-... | I'm on board with all that, including the proposed operator.
Ok, then I'll make a PR soon. :)
I like that! For readability I'd like it best if a run of whitespaces becomes a single '\n' if it contains an '\n', or a space otherwise.
I'm fine with implementing it this way. Anyone else has an opinion maybe?
Yeah, I think ... | 2022-04-21T14:51:49Z | 0.11 | 2022-04-26T08:31:21Z | 7b6f1df433a7f11612608644342b898cd6be8ff5 | [
"filters::json::tests::test_json",
"filters::tests::test_abs",
"filters::tests::test_center",
"filters::tests::test_capitalize",
"filters::tests::test_filesizeformat",
"filters::tests::test_indent",
"filters::tests::test_into_f64",
"filters::tests::test_into_isize",
"filters::tests::test_join",
"f... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 652 | rinja-rs__askama-652 | [
"651"
] | b14982f97ffd20039286171d56e6fcfab21f56bc | diff --git a/askama_shared/src/filters/mod.rs b/askama_shared/src/filters/mod.rs
--- a/askama_shared/src/filters/mod.rs
+++ b/askama_shared/src/filters/mod.rs
@@ -330,20 +330,14 @@ where
/// Capitalize a value. The first character will be uppercase, all others lowercase.
pub fn capitalize<T: fmt::Display>(s: T) -> ... | diff --git a/askama_shared/src/filters/mod.rs b/askama_shared/src/filters/mod.rs
--- a/askama_shared/src/filters/mod.rs
+++ b/askama_shared/src/filters/mod.rs
@@ -655,6 +649,9 @@ mod tests {
assert_eq!(capitalize(&"").unwrap(), "".to_string());
assert_eq!(capitalize(&"FoO").unwrap(), "Foo".to_string()... | Capitalize does not work with non ascii chars
Capitalize filter only works with ascii chars not chars like å, ä and ö.
| This custom filter does what i expected capitalize to do.
```rust
mod filters {
pub fn cap(s: &str) -> ::askama::Result<String> {
match s.chars().next() {
Some(c) => {
if c.is_lowercase() {
let mut replacement: String = c.to_uppercase().collect();
... | 2022-03-26T15:32:24Z | 0.11 | 2022-03-26T17:50:42Z | 7b6f1df433a7f11612608644342b898cd6be8ff5 | [
"filters::tests::test_capitalize"
] | [
"filters::tests::test_abs",
"filters::tests::test_center",
"filters::tests::test_filesizeformat",
"filters::json::tests::test_json",
"filters::tests::test_indent",
"filters::tests::test_into_f64",
"filters::tests::test_into_isize",
"filters::tests::test_join",
"filters::tests::test_linebreaks",
"f... | [
"filters::yaml::tests::test_yaml"
] | [] | auto_2025-06-03 |
rinja-rs/askama | 359 | rinja-rs__askama-359 | [
"357"
] | 17b9d06814cee84bfd57b73e1941b63187ec5f65 | diff --git a/askama_shared/src/filters/mod.rs b/askama_shared/src/filters/mod.rs
--- a/askama_shared/src/filters/mod.rs
+++ b/askama_shared/src/filters/mod.rs
@@ -249,7 +249,7 @@ pub fn indent(s: &dyn fmt::Display, width: &usize) -> Result<String> {
#[cfg(feature = "num-traits")]
/// Casts number to f64
-pub fn int... | diff --git a/askama_shared/src/filters/mod.rs b/askama_shared/src/filters/mod.rs
--- a/askama_shared/src/filters/mod.rs
+++ b/askama_shared/src/filters/mod.rs
@@ -460,22 +460,22 @@ mod tests {
#[test]
#[allow(clippy::float_cmp)]
fn test_into_f64() {
- assert_eq!(into_f64(1).unwrap(), 1.0 as f64);
... | into_f64, into_isize unusable
`{{ number | into_f64 }}` expands to `into_f64(&number)`, which doesn't compile, because borrowed number doesn't `impl NumCast`.
| Thanks for the report. Would you mind submitting a patch that adds the appropriate `*` operators?
Maybe it's better to have info_f64 take `&T`?
That's fine, too! Whatever works in the context of Askama filters. | 2020-09-16T01:05:50Z | 0.10 | 2020-09-16T09:49:22Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"filters::tests::test_abs",
"filters::tests::test_filesizeformat",
"filters::tests::test_center",
"filters::tests::test_capitalize",
"filters::json::tests::test_json",
"filters::tests::test_into_f64",
"filters::tests::test_indent",
"filters::tests::test_into_isize",
"filters::tests::test_lower",
"... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 304 | rinja-rs__askama-304 | [
"285",
"196"
] | cff49453a851029f87b35512f5234a999fea3e6c | diff --git a/askama_shared/src/generator.rs b/askama_shared/src/generator.rs
--- a/askama_shared/src/generator.rs
+++ b/askama_shared/src/generator.rs
@@ -909,7 +909,9 @@ impl<'a, S: std::hash::BuildHasher> Generator<'a, S> {
Expr::StrLit(s) => self.visit_str_lit(buf, s),
Expr::CharLit(s) => s... | diff --git a/askama_shared/src/parser.rs b/askama_shared/src/parser.rs
--- a/askama_shared/src/parser.rs
+++ b/askama_shared/src/parser.rs
@@ -1078,6 +1098,34 @@ mod tests {
super::parse("{{ strvar|e }}", &Syntax::default());
}
+ #[test]
+ fn test_parse_var_call() {
+ assert_eq!(
+ ... | Call methods on root context
Is there a syntax for calling methods implemented directly on the 'context'? A bare `function()` doesn't appear to do the trick
Static method call
Can I use static method in the template?
| I think you can call `self.function()`? Documentation patches welcome. 😊
Can you give a concrete example? Presumably you have tried it?
I tried like so:
`{% let a = crate::a() %}
`
Did you get an error? What was the error? Did you look at the generated code?
The print attribute is "all", but it shows only `message: ... | 2020-03-15T11:07:43Z | 0.9 | 2020-03-18T21:33:57Z | cff49453a851029f87b35512f5234a999fea3e6c | [
"filters::tests::test_capitalize",
"filters::tests::test_center",
"filters::tests::test_filesizeformat",
"filters::json::tests::test_json",
"filters::tests::test_indent",
"filters::tests::test_linebreaks",
"filters::tests::test_linebreaksbr",
"filters::tests::test_join",
"filters::tests::test_lower"... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 861 | rinja-rs__askama-861 | [
"860"
] | 43e92aa3b6b9cd967a70bd0fd54d1f087d6ed76b | diff --git a/askama_parser/src/expr.rs b/askama_parser/src/expr.rs
--- a/askama_parser/src/expr.rs
+++ b/askama_parser/src/expr.rs
@@ -248,15 +248,15 @@ impl<'a> Suffix<'a> {
}
fn r#macro(i: &'a str) -> IResult<&'a str, Self> {
- fn nested_parenthesis(i: &str) -> IResult<&str, ()> {
+ fn neste... | diff --git a/askama_parser/src/tests.rs b/askama_parser/src/tests.rs
--- a/askama_parser/src/tests.rs
+++ b/askama_parser/src/tests.rs
@@ -788,3 +788,10 @@ fn test_parse_array() {
)],
);
}
+
+#[test]
+fn fuzzed_unicode_slice() {
+ let d = "{eeuuu{b&{!!&{!!11{{
+ 0!(!1q҄א!)!!!!!!n!";
+ as... | Fuzzing askama_parser results in panic
Hi, fuzzing `askama_parser` resulted in panic at following line.
https://github.com/djc/askama/blob/43e92aa3b6b9cd967a70bd0fd54d1f087d6ed76b/askama_parser/src/expr.rs#L290
I suppose it happens because crash input contains Cyrillic letters which are multi-byte and we need exact b... | 2023-09-11T09:42:00Z | 0.12 | 2023-09-11T10:19:21Z | 4dab5f91ba15e7c238ddf6223ec7b45eef32cab4 | [
"tests::fuzzed_unicode_slice"
] | [
"tests::change_delimiters_parse_filter",
"tests::test_missing_space_after_kw",
"tests::test_invalid_block - should panic",
"tests::test_parse_comments",
"tests::test_associativity",
"tests::test_parse_const",
"tests::test_parse_numbers",
"tests::test_odd_calls",
"tests::test_parse_root_path",
"tes... | [] | [] | auto_2025-06-03 | |
rinja-rs/askama | 837 | rinja-rs__askama-837 | [
"836"
] | 9de9af4a006021a63f705e420be4cdef3eb6af82 | diff --git a/askama_derive/src/generator.rs b/askama_derive/src/generator.rs
--- a/askama_derive/src/generator.rs
+++ b/askama_derive/src/generator.rs
@@ -1373,7 +1373,7 @@ impl<'a> Generator<'a> {
}
Expr::Group(ref inner) => self.visit_group(buf, inner)?,
Expr::Call(ref obj, ref ... | diff --git a/askama_derive/src/parser/tests.rs b/askama_derive/src/parser/tests.rs
--- a/askama_derive/src/parser/tests.rs
+++ b/askama_derive/src/parser/tests.rs
@@ -221,6 +221,51 @@ fn test_parse_root_path() {
);
}
+#[test]
+fn test_rust_macro() {
+ let syntax = Syntax::default();
+ assert_eq!(
+ ... | Parse error when using macro with expanded module path
I am making a crate that allows people to embed icons from Iconify at compile time, but I ran into a little snag.
Tried on main and 0.12. Only happens when using a macro, normal functions work fine.
The intended usage is:
```jsx
<body>
<h1>Hello</h1>
... | Huh, I guess there's some subtlety in the parser relating to the combination of a path, the macro call and the filter. Care to submit a PR?
Sure, I'll take a jab at it.
Macro invocations are parsed specially, but they should actually be simply suffix expressions, and be handled like normal function calls:
https://gi... | 2023-07-05T21:49:42Z | 0.12 | 2023-07-24T09:39:14Z | 4dab5f91ba15e7c238ddf6223ec7b45eef32cab4 | [
"config::tests::find_absolute",
"config::tests::find_relative",
"config::tests::add_syntax",
"config::tests::escape_modes",
"config::tests::add_syntax_two",
"config::tests::find_relative_sub",
"config::tests::get_source",
"config::tests::find_relative_nonexistent - should panic",
"config::tests::ill... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 965 | rinja-rs__askama-965 | [
"962",
"961"
] | 4dab5f91ba15e7c238ddf6223ec7b45eef32cab4 | diff --git a/askama/Cargo.toml b/askama/Cargo.toml
--- a/askama/Cargo.toml
+++ b/askama/Cargo.toml
@@ -1,6 +1,6 @@
[package]
name = "askama"
-version = "0.12.1"
+version = "0.13.0"
description = "Type-safe, compiled Jinja-like templates for Rust"
documentation = "https://docs.rs/askama"
keywords = ["markup", "temp... | diff --git a/askama_derive/src/config.rs b/askama_derive/src/config.rs
--- a/askama_derive/src/config.rs
+++ b/askama_derive/src/config.rs
@@ -299,9 +299,9 @@ pub(crate) fn get_template_source(tpl_path: &Path) -> std::result::Result<String
static CONFIG_FILE_NAME: &str = "askama.toml";
static DEFAULT_SYNTAX_NAME: &st... | Remove the need to import askama itself when using integration crates
While wording on https://github.com/djc/askama/issues/961, I wandered off to find the cause for the need as the title goes. A rough scan reveals that askama_derive generates some uses of `::askama::`, for example, this:
https://github.com/djc/aska... | Well, I wouldn't want to do this in a way that requires adding another intermediate allocation, but perhaps a newtype with a `Display` implementation could do the job. I'd be open to reviewing such a PR.
How about checking for existence of `::askama::`, and if so, splitting the `&str` into head, prefix and tail, then 1... | 2024-02-13T14:31:39Z | 0.12 | 2024-03-05T10:34:46Z | 4dab5f91ba15e7c238ddf6223ec7b45eef32cab4 | [
"config::tests::find_absolute",
"config::tests::find_relative",
"config::tests::find_relative_sub",
"config::tests::add_syntax",
"config::tests::add_syntax_two",
"config::tests::get_source",
"config::tests::find_relative_nonexistent - should panic",
"config::tests::test_config_whitespace_error",
"co... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 951 | rinja-rs__askama-951 | [
"924"
] | 29b25505b496510217a39606a5f72884867ef492 | diff --git a/askama_parser/src/lib.rs b/askama_parser/src/lib.rs
--- a/askama_parser/src/lib.rs
+++ b/askama_parser/src/lib.rs
@@ -349,9 +349,9 @@ fn path_or_identifier(i: &str) -> ParseResult<'_, PathOrIdentifier<'_>> {
let rest = rest.as_deref().unwrap_or_default();
// The returned identifier can be assum... | diff --git a/testing/tests/simple.rs b/testing/tests/simple.rs
--- a/testing/tests/simple.rs
+++ b/testing/tests/simple.rs
@@ -484,3 +484,23 @@ fn test_num_literals() {
"[90, -90, 90, 2, 56, 240, 10.5, 10.5, 100000000000, 105000000000]",
);
}
+
+#[allow(non_snake_case)]
+#[derive(askama::Template)]
+#[te... | 0.12.1 only allow lower-case variable name in template
I don't know if I asked a stupid question, but I spend one day to find it ! my rust version is 1.74.0, I run it on my win11 and WSL ubuntu 22.04, see the same error below
```rust
use askama::Template;
#[derive(Template)]
#[template(path = "hello.html")]
str... | I'm new to axum, I have learn the docs https://djc.github.io/askama/askama.html, but it seems not have the request of lower case
I suggest you get the macro expansion via Rust Analyzer or via the [debugging](https://djc.github.io/askama/debugging.html) features. I guess the code generator applies some heuristics that t... | 2024-01-18T09:50:05Z | 0.12 | 2024-01-18T10:23:22Z | 4dab5f91ba15e7c238ddf6223ec7b45eef32cab4 | [
"test_attr",
"test_comment",
"test_composition",
"test_constants",
"test_define_string_var",
"test_else",
"test_else_if",
"test_empty",
"test_escape",
"test_func_ref_call",
"test_generics",
"test_if",
"test_literals",
"test_index",
"test_literals_escape",
"test_minus",
"test_negation... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 943 | rinja-rs__askama-943 | [
"330"
] | 84c2094e871b7ef16ab51c77b4a3568794c75638 | diff --git a/askama_derive/src/generator.rs b/askama_derive/src/generator.rs
--- a/askama_derive/src/generator.rs
+++ b/askama_derive/src/generator.rs
@@ -1210,6 +1210,20 @@ impl<'a> Generator<'a> {
Ok(DisplayWrap::Wrapped)
}
+ fn _visit_as_ref_filter(
+ &mut self,
+ buf: &mut Buffer,
+... | diff --git a/testing/tests/filters.rs b/testing/tests/filters.rs
--- a/testing/tests/filters.rs
+++ b/testing/tests/filters.rs
@@ -309,3 +309,17 @@ fn test_json_script() {
r#"<script>var user = "\u003c/script\u003e\u003cbutton\u003eHacked!\u003c/button\u003e"</script>"#
);
}
+
+#[derive(askama::Template)... | Add support for reference operator (i.e. `{% let x = &y %}`)
Sometimes while working with deeply nested structures, it would be useful to create an alias to a field, like say `{% let h = &parent.child.hash %}`
At the moment the `&` unary operator is not supported.
| Not supporting this has been a design goal from the start. If I remember correctly, your example should already work without `&`.
> If I remember correctly, your example should already work without `&`.
Unfortunately it doesn't (`hash` here is an owned `HashMap`, and all other structures are owned by `self`):
~~~~
... | 2024-01-10T14:11:35Z | 0.12 | 2024-01-12T10:10:23Z | 4dab5f91ba15e7c238ddf6223ec7b45eef32cab4 | [
"filter_escape",
"filter_format",
"filter_fmt",
"filter_opt_escaper_html",
"filter_opt_escaper_none",
"into_numbers_fmt",
"test_filter_let_filter",
"test_filter_truncate",
"test_join",
"test_json",
"test_json_attribute",
"test_json_attribute2",
"test_json_script",
"test_my_filter",
"test... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 911 | rinja-rs__askama-911 | [
"904"
] | 6cbfde04514a90d4e24350c21ef490c40666d820 | diff --git a/askama_derive/src/generator.rs b/askama_derive/src/generator.rs
--- a/askama_derive/src/generator.rs
+++ b/askama_derive/src/generator.rs
@@ -1575,6 +1575,16 @@ impl<'a> Generator<'a> {
}
buf.write(name);
}
+ Target::OrChain(targets) => match target... | diff --git /dev/null b/testing/templates/match-enum-or.html
new file mode 100644
--- /dev/null
+++ b/testing/templates/match-enum-or.html
@@ -0,0 +1,8 @@
+The card is
+{%- match suit %}
+ {%- when Suit::Clubs or Suit::Spades -%}
+ {{ " black" }}
+ {%- when Suit::Diamonds or Suit::Hearts -%}
+ {{ " red" }}
+... | Feature request: enhance match to include multiple targets
This has previously been discussed at https://github.com/djc/askama/issues/711#issuecomment-1794340598 but I would like to move it to its own issue as it can be implemented separately.
The request is to enhance Askana's syntax for the `match` block (see http... | I'm happy to review a PR for this as proposed. | 2023-11-19T18:50:18Z | 0.12 | 2023-11-22T13:56:14Z | 4dab5f91ba15e7c238ddf6223ec7b45eef32cab4 | [
"test_match_custom_enum",
"test_match_literal_char",
"test_match_literal",
"test_match_literal_num",
"test_match_no_whitespace",
"test_match_option",
"test_match_option_bool",
"test_match_option_result_option",
"test_match_ref_deref",
"test_match_without_with_keyword",
"test_match_with_comment"
... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 910 | rinja-rs__askama-910 | [
"885"
] | 80238d7f48fd86ef939e74df9fdc9678ee78a208 | diff --git a/askama_derive/src/generator.rs b/askama_derive/src/generator.rs
--- a/askama_derive/src/generator.rs
+++ b/askama_derive/src/generator.rs
@@ -727,7 +727,50 @@ impl<'a> Generator<'a> {
args.len()
)));
}
- for (expr, arg) in std::iter::zip(args, &def.args) {
+ ... | diff --git a/testing/tests/macro.rs b/testing/tests/macro.rs
--- a/testing/tests/macro.rs
+++ b/testing/tests/macro.rs
@@ -95,3 +95,31 @@ fn test_macro_self_arg() {
let t = MacroSelfArgTemplate { s: "foo" };
assert_eq!(t.render().unwrap(), "foo");
}
+
+#[derive(Template)]
+#[template(
+ source = "{%- macr... | Allow to have named arguments in macro
It'd pretty nice to be able to have named arguments. For example:
```
{% macro menu_link(href, text, target) %}
...
{% endmacro menu_link %}
```
Then to be able to call it like this:
```
{% menu_link(href="/something", text="link", target="_blank") %}
```
It make... | I worry that this would be a little foreign to users -- that is, I'm not sure it passes the design test of "it should be obvious how this gets compiled to Rust code".
Does it really matter? If this is a big enough concern, we can add a new entry in the askama book explaining how it is generated.
It matters to me -- I t... | 2023-11-17T13:50:00Z | 0.12 | 2023-11-28T10:54:25Z | 4dab5f91ba15e7c238ddf6223ec7b45eef32cab4 | [
"test_one_func",
"test_one_func_binop",
"test_one_func_self",
"test_one_func_index",
"test_coerce",
"test_path_ext_html",
"test_path_ext_html_and_ext_txt",
"test_path_ext_html_jinja",
"test_path_ext_html_jinja_and_ext_txt",
"test_path_ext_jinja",
"test_path_ext_jinja_and_ext_txt",
"filter_esca... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 401 | rinja-rs__askama-401 | [
"333",
"221",
"107"
] | 5057b75e7752f2186157fb3e1f5e5d2d0c5ff880 | diff --git a/askama_shared/src/generator.rs b/askama_shared/src/generator.rs
--- a/askama_shared/src/generator.rs
+++ b/askama_shared/src/generator.rs
@@ -667,8 +667,26 @@ impl<'a, S: std::hash::BuildHasher> Generator<'a, S> {
", _loop_item) in ::askama::helpers::TemplateLoop::new({}) {{",
... | diff --git a/testing/tests/loops.rs b/testing/tests/loops.rs
--- a/testing/tests/loops.rs
+++ b/testing/tests/loops.rs
@@ -67,3 +67,39 @@ fn test_for_range() {
"foo (first)\nfoo (last)\nbar\nbar\nfoo\nbar\nbar\n"
);
}
+
+#[derive(Template)]
+#[template(source = "{% for i in [1, 2, 3] %}{{ i }}{% endfor %... | for-loop on values of a BTreeMap inside a template does not compile
Hi Askama team,
I'm hitting the an issue with for-loops inside template on `(some BTreeMap).values()`.
## Preliminary: what DOES work
Consider first the following piece of code, which works, demonstrating that the `values()` function of `BTreeMa... | Allocating a `Vec` just to appease `djc/askama` seems unfortunate. I wonder whether adding something like the `&`-syntax could improve things:
```
{% for &foo in map.values() %}
```
I encountered the same issue trying to iterate over a `BTreeMap` using `.rev()`:
```jinja
{% for (date, day_videos) in videos.video... | 2020-12-15T06:21:33Z | 0.10 | 2021-02-18T12:18:18Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"test_for",
"test_for_range",
"test_precedence_for",
"test_nested_for"
] | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 399 | rinja-rs__askama-399 | [
"397",
"397"
] | 5b01e605914a49f0b9e71e7dbe7c17ef1de2c522 | diff --git a/askama_shared/src/generator.rs b/askama_shared/src/generator.rs
--- a/askama_shared/src/generator.rs
+++ b/askama_shared/src/generator.rs
@@ -444,8 +444,8 @@ impl<'a, S: std::hash::BuildHasher> Generator<'a, S> {
Node::Cond(ref conds, ws) => {
self.write_cond(ctx, buf,... | diff --git a/testing/tests/gen_ws_tests.py b/testing/tests/gen_ws_tests.py
--- a/testing/tests/gen_ws_tests.py
+++ b/testing/tests/gen_ws_tests.py
@@ -4,12 +4,23 @@
from itertools import product, chain, zip_longest, tee
+# The amount of branches to generate
+BRANCHES = 2 # branches
+
IF, ELSE_IF, ELSE, END_IF = 0... | Unwanted whitespace in match arm
For a Template struct containing a field `ver: Option<&'a str>` and the following template:
```jinja2
{% let suffix -%}
{% match ver -%}
{% when Some with (ver) -%}
{% let suffix = format!("-{}", ver) -%}
{% when None -%}
{% let suffix = String::new() -%}
{% endmat... | I'm wondering where the whitespace between the `match` and the first `when` should go. Whether it should go into the first match arm or before the match expression (thus "into" all match arms). The whitespace before `endmatch` is self-explanatory.
In short:
```jinja
{% match ... %}[this whitespace]{% when ... %}... | 2020-12-12T16:51:50Z | 0.10 | 2020-12-13T22:08:20Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"test_match_custom_enum",
"test_match_literal_char",
"test_match_literal",
"test_match_literal_num",
"test_match_option",
"test_match_ref_deref",
"test_extra_whitespace",
"test_match_ws"
] | [
"test_match_no_whitespace",
"test_cond_ws"
] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 394 | rinja-rs__askama-394 | [
"377",
"378"
] | 810d5ad5067085cc5e4be19903b5804b1ff13430 | diff --git a/askama_shared/src/generator.rs b/askama_shared/src/generator.rs
--- a/askama_shared/src/generator.rs
+++ b/askama_shared/src/generator.rs
@@ -505,8 +505,9 @@ impl<'a, S: std::hash::BuildHasher> Generator<'a, S> {
let mut has_else = false;
for (i, &(cws, ref cond, ref nodes)) in conds.iter... | diff --git /dev/null b/testing/tests/gen_ws_tests.py
new file mode 100644
--- /dev/null
+++ b/testing/tests/gen_ws_tests.py
@@ -0,0 +1,112 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+
+from itertools import product, chain, zip_longest, tee
+
+
+IF, ELSE_IF, ELSE, END_IF = 0, 1, 2, 3
+
+NL = "\\n"
+dash = lambda... | Whitespace issue with if conditions
Askama seems to be including whitespace from inside an if condition even when the condition is not fulfilled. Here is my template file:
```jinja
{%- if some_condition -%}
# WARNING
{% endif -%}
# FILE START
```
What I expected is:
- When `some_condition` is `true`:
... | It sure sounds like a bug! Let me look at some code.
So the code generation for if statements is here:
https://github.com/djc/askama/blob/main/askama_shared/src/generator.rs#L502
I don't immediately see anything obviously wrong. Maybe you can contribute a test case and/or use the debugging facility (add `print = ... | 2020-12-03T18:13:39Z | 0.10 | 2020-12-03T21:14:33Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"test_cond_ws"
] | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 393 | rinja-rs__askama-393 | [
"331",
"331"
] | a199defeca2dfc6aa3e972acca82c96db07f99e9 | diff --git a/askama_shared/src/parser.rs b/askama_shared/src/parser.rs
--- a/askama_shared/src/parser.rs
+++ b/askama_shared/src/parser.rs
@@ -1,7 +1,7 @@
use nom::branch::alt;
use nom::bytes::complete::{escaped, is_not, tag, take_until};
use nom::character::complete::{anychar, char, digit1};
-use nom::combinator::{... | diff --git a/askama_shared/src/parser.rs b/askama_shared/src/parser.rs
--- a/askama_shared/src/parser.rs
+++ b/askama_shared/src/parser.rs
@@ -1136,6 +1138,25 @@ mod tests {
);
}
+ #[test]
+ fn test_parse_root_path() {
+ let syntax = Syntax::default();
+ assert_eq!(
+ supe... | Allow paths to start with `::`
[Discovered while working on #315 and #327.]
At the moment `{{ std::string::String::new()}}` works but `{{ ::std::string::String::new() }}` doesn't.
Allow paths to start with `::`
[Discovered while working on #315 and #327.]
At the moment `{{ std::string::String::new()}}` works bu... | Alright, that should be fixed.
Alright, that should be fixed. | 2020-12-02T18:03:44Z | 0.10 | 2020-12-02T19:48:44Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"parser::tests::test_parse_root_path",
"test_attr",
"test_composition",
"test_comment",
"test_define_string_var",
"test_else",
"test_else_if",
"test_empty",
"test_escape",
"test_func_ref_call",
"test_generics",
"test_if",
"test_literals_escape",
"test_index",
"test_literals",
"test_nes... | [
"filters::tests::test_abs",
"filters::json::tests::test_json",
"filters::tests::test_capitalize",
"filters::tests::test_center",
"filters::tests::test_filesizeformat",
"filters::tests::test_into_f64",
"filters::tests::test_into_isize",
"filters::tests::test_indent",
"filters::tests::test_join",
"f... | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 391 | rinja-rs__askama-391 | [
"381"
] | f4065b09b91f5d00efa5644915cdd83bfb7d393a | diff --git a/askama_shared/src/parser.rs b/askama_shared/src/parser.rs
--- a/askama_shared/src/parser.rs
+++ b/askama_shared/src/parser.rs
@@ -573,36 +573,32 @@ fn expr_rust_macro(i: &[u8]) -> IResult<&[u8], Expr> {
macro_rules! expr_prec_layer {
( $name:ident, $inner:ident, $op:expr ) => {
fn $name(i: &... | diff --git a/askama_shared/src/parser.rs b/askama_shared/src/parser.rs
--- a/askama_shared/src/parser.rs
+++ b/askama_shared/src/parser.rs
@@ -1078,12 +1074,13 @@ pub fn parse<'a>(src: &'a str, syntax: &'a Syntax<'a>) -> Result<Vec<Node<'a>>,
#[cfg(test)]
mod tests {
+ use super::{Expr, Node, WS};
use crate... | The operator parser gets precedences wrong
Consider the expression `a + b == c`. If we changed the code generated from `{lhs} {op} {rhs}` to include parentheses around the entire thing, we get `(a + (b == c))`, which is incorrect, because _(i)_ the `+` and `==` operators combined are not associative (and I think that's... | So precedence is handled here:
https://github.com/djc/askama/blob/main/askama_shared/src/parser.rs#L610
I forgot how I wrote this code in the first place, but I'm guessing the associativity aspect is not handled correctly. | 2020-12-01T08:40:04Z | 0.10 | 2020-12-01T09:04:25Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"parser::tests::test_associativity",
"parser::tests::test_precedence"
] | [
"filters::tests::test_abs",
"filters::tests::test_capitalize",
"filters::tests::test_center",
"filters::json::tests::test_json",
"filters::tests::test_filesizeformat",
"filters::tests::test_into_f64",
"filters::tests::test_into_isize",
"filters::tests::test_indent",
"filters::tests::test_linebreaks"... | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 374 | rinja-rs__askama-374 | [
"368"
] | ac7d9e8031b4df489a0d58dc3459bc931f1ff870 | diff --git a/askama_derive/Cargo.toml b/askama_derive/Cargo.toml
--- a/askama_derive/Cargo.toml
+++ b/askama_derive/Cargo.toml
@@ -25,5 +25,4 @@ warp = []
[dependencies]
askama_shared = { version = "0.10.4", path = "../askama_shared", default-features = false }
proc-macro2 = "1"
-quote = "1"
syn = "1"
diff --git a/... | diff --git a/testing/Cargo.toml b/testing/Cargo.toml
--- a/testing/Cargo.toml
+++ b/testing/Cargo.toml
@@ -15,6 +15,7 @@ serde_json = { version = "1.0", optional = true }
[dev-dependencies]
criterion = "0.3"
+trybuild = "1.0"
[[bench]]
name = "all"
diff --git /dev/null b/testing/tests/ui.rs
new file mode 100644... | Consider using `compile_error!` instead of panicking
I just tried out askama for the first time and my IDE (using rust-analyzer) reported that my code failed to compile because `proc-macro derive panicked`. This error message is not very helpful as I had to go to the command line and ask cargo what the real problem was... | The simple approach using what syn provides is not enough because many of the potential panics originate from the template parser, which doesn't process Rust code as such. Still, it would surely be nice to use something better than `panic!()`, it just hasn't been much of a priority for me. In the same space, I've also ... | 2020-11-02T15:50:35Z | 0.10 | 2020-11-04T00:28:09Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"ui"
] | [
"filter_escape",
"filter_fmt",
"filter_format",
"into_numbers_fmt",
"test_filter_let_filter",
"test_filter_truncate",
"test_join",
"test_nested_filter_ref",
"test_my_filter",
"test_vec_join",
"main",
"test_include",
"test_deep",
"test_different_module",
"test_empty_child",
"test_let_bl... | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 500 | rinja-rs__askama-500 | [
"494"
] | 49252d2457f280026c020d0df46733578eb959a5 | diff --git a/askama_shared/src/generator.rs b/askama_shared/src/generator.rs
--- a/askama_shared/src/generator.rs
+++ b/askama_shared/src/generator.rs
@@ -658,6 +658,12 @@ impl<'a, S: std::hash::BuildHasher> Generator<'a, S> {
", _loop_item) in ::askama::helpers::TemplateLoop::new(((&{}).into_iter())) ... | diff --git a/testing/tests/loops.rs b/testing/tests/loops.rs
--- a/testing/tests/loops.rs
+++ b/testing/tests/loops.rs
@@ -1,3 +1,5 @@
+use std::ops::Range;
+
use askama::Template;
#[derive(Template)]
diff --git a/testing/tests/loops.rs b/testing/tests/loops.rs
--- a/testing/tests/loops.rs
+++ b/testing/tests/loops... | move occurs because `item1.list2` has type `Vec<Box<dyn MyTrait + Send>>`, which does not implement the `Copy` trait
I got this error with `main` but not with `0.10.5`.
```
error[E0507]: cannot move out of `item1.list2` which is behind a shared reference
--> src/main.rs:3:10
|
3 | #[derive(Template)]
| ... | As a workaround, it should work if you change `item1.list2` to `item1.list2.iter()`. However, I'll look into why that fails, as right now I don't recall the reason why it _would_ fail.
@vallentin would be great if you can look into this, thanks!
@djc I will. I've just been a bit busy the past week and this week, but ... | 2021-06-23T16:29:30Z | 0.10 | 2021-06-23T19:41:31Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"test_for",
"test_for_array",
"test_for_method_call",
"test_for_index",
"test_for_range",
"test_for_path_call",
"test_for_zip_ranges",
"test_nested_for",
"test_precedence_for"
] | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 499 | rinja-rs__askama-499 | [
"498"
] | b318d7cbcded2c6dfc66bbe19687f1246a9a9eab | diff --git a/askama_shared/src/generator.rs b/askama_shared/src/generator.rs
--- a/askama_shared/src/generator.rs
+++ b/askama_shared/src/generator.rs
@@ -717,6 +717,7 @@ impl<'a, S: std::hash::BuildHasher> Generator<'a, S> {
let mut names = Buffer::new(0);
let mut values = Buffer::new(0);
+ ... | diff --git /dev/null b/testing/templates/nested-macro-args.html
new file mode 100644
--- /dev/null
+++ b/testing/templates/nested-macro-args.html
@@ -0,0 +1,9 @@
+{%- macro outer(first) -%}
+{%- call inner(first, "second") -%}
+{%- endmacro -%}
+
+{%- macro inner(first, second) -%}
+{{ first }} {{ second }}
+{%- endmac... | Error "proc-macro derive produced unparseable tokens" with nested macro call
I was working on updating [UniFFI](https://github.com/mozilla/uniffi-rs/) to the latest `main` branch of Askama, and ran into a case where the `#[derive(Template)]` seems to generate invalid code. Below is a reduced example:
```
use askama... | 2021-06-21T06:39:19Z | 0.10 | 2021-06-22T09:41:51Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"test_deep_import",
"test_nested",
"test_macro",
"test_import",
"test_short_circuit"
] | [] | [] | [] | auto_2025-06-03 | |
rinja-rs/askama | 486 | rinja-rs__askama-486 | [
"245"
] | 92df4d1fe49e8fde5ca13f13b8236102bc16b969 | diff --git a/askama_escape/src/lib.rs b/askama_escape/src/lib.rs
--- a/askama_escape/src/lib.rs
+++ b/askama_escape/src/lib.rs
@@ -129,7 +129,6 @@ impl Escaper for Html {
b'&' => escaping_body!(start, i, fmt, bytes, "&"),
b'"' => escaping_body!(start, i, fmt, bytes, ""... | diff --git a/testing/tests/filters.rs b/testing/tests/filters.rs
--- a/testing/tests/filters.rs
+++ b/testing/tests/filters.rs
@@ -21,7 +21,7 @@ fn filter_escape() {
};
assert_eq!(
s.render().unwrap(),
- "// my <html> is "unsafe" & \
+ "// my <html> i... | askama_escape Why escape slashes?
Sorry, I just don't seem to understand what harm slashes can do in html.
[Why does it escape slashes?](https://github.com/djc/askama/blob/467f4ade19fa34983de7e6f6d81c6b4d5ff140fe/askama_escape/src/lib.rs#L134)
| Because the OWASP recommends doing so.
https://security.stackexchange.com/questions/49852/why-should-xss-filters-escape-forward-slash
[OWASP Cross Site Scripting Prevention Cheat Sheet](https://github.com/OWASP/CheatSheetSeries/blob/master/cheatsheets/Cross_Site_Scripting_Prevention_Cheat_Sheet.md#rule-1---html-esc... | 2021-05-16T20:39:00Z | 0.10 | 2021-05-17T19:33:59Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"filter_escape"
] | [
"filter_format",
"filter_fmt",
"into_numbers_fmt",
"test_filter_let_filter",
"test_filter_truncate",
"test_join",
"test_my_filter",
"test_nested_filter_ref",
"test_vec_join",
"test_attr",
"test_comment",
"test_composition",
"test_constants",
"test_define_string_var",
"test_else",
"test... | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 458 | rinja-rs__askama-458 | [
"457",
"457"
] | 96a4328d642191e4925a8f822c9d2cf311f2f9ec | diff --git a/askama_shared/src/generator.rs b/askama_shared/src/generator.rs
--- a/askama_shared/src/generator.rs
+++ b/askama_shared/src/generator.rs
@@ -156,10 +156,7 @@ impl<'a, S: std::hash::BuildHasher> Generator<'a, S> {
buf.writeln("}")?;
buf.writeln("fn extension(&self) -> Option<&'static st... | diff --git a/askama_shared/src/input.rs b/askama_shared/src/input.rs
--- a/askama_shared/src/input.rs
+++ b/askama_shared/src/input.rs
@@ -227,3 +242,51 @@ impl FromStr for Print {
})
}
}
+
+#[cfg(test)]
+mod tests {
+ use super::*;
+
+ #[test]
+ fn test_ext() {
+ assert_eq!(extension(Pat... | Proper file extension for templates for plugin integration?
First of all, let me congratulate you for your work. I'm developing an application with Askama and Rocket, and I'm very happy with the results. However, I recently hit an interesting bump. This was using VS Code as my development environment:
- As the templ... | 2021-03-09T05:45:50Z | 0.10 | 2021-03-10T12:09:59Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"filters::tests::test_abs",
"filters::json::tests::test_json",
"filters::tests::test_capitalize",
"filters::tests::test_filesizeformat",
"filters::tests::test_center",
"filters::tests::test_indent",
"filters::tests::test_into_f64",
"filters::tests::test_into_isize",
"filters::tests::test_linebreaks"... | [] | [] | [] | auto_2025-06-03 | |
rinja-rs/askama | 433 | rinja-rs__askama-433 | [
"432",
"291"
] | 560d219c269bbf291a4f78e8ef3ffeb0d02ffdef | diff --git a/askama_shared/src/generator.rs b/askama_shared/src/generator.rs
--- a/askama_shared/src/generator.rs
+++ b/askama_shared/src/generator.rs
@@ -1721,8 +1721,11 @@ impl MapChain<'_, &str, LocalMeta> {
}
fn resolve_or_self(&self, name: &str) -> String {
- self.resolve(name)
- .unw... | diff --git a/testing/tests/simple.rs b/testing/tests/simple.rs
--- a/testing/tests/simple.rs
+++ b/testing/tests/simple.rs
@@ -72,6 +72,36 @@ fn test_variables_no_escape() {
);
}
+#[derive(Template)]
+#[template(
+ source = "{{ foo }} {{ foo_bar }} {{ FOO }} {{ FOO_BAR }} {{ self::FOO }} {{ self::FOO_BAR }} ... | Updated book to include urlencode and urlencode_strict
Forgot to include these in #429.
Question: passing a const, not as a field.
I'm trying to reference a constant in my template, since I reuse it all over the place (including across templates) and it'd be a pain to have to update it everywhere it needs to change. ... |
As a workaround, try referencing it through a path, for example, from the crate root?
Askama tries not to use its knowledge of the context fields, and instead keeps track of local variables, such that everything non-local is treated as a context field. Maybe this can be improved, although I think there were some bene... | 2021-01-11T18:51:24Z | 0.10 | 2021-01-13T07:58:54Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"test_attr",
"test_comment",
"test_composition",
"test_define_string_var",
"test_else",
"test_else_if",
"test_empty",
"test_escape",
"test_func_ref_call",
"test_generics",
"test_if",
"test_index",
"test_literals",
"test_minus",
"test_literals_escape",
"test_nested_attr",
"test_negati... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 426 | rinja-rs__askama-426 | [
"424",
"424"
] | c29ecd68714bddf5e27a9e347c902faa23b2a545 | diff --git a/askama_shared/src/parser.rs b/askama_shared/src/parser.rs
--- a/askama_shared/src/parser.rs
+++ b/askama_shared/src/parser.rs
@@ -539,7 +539,7 @@ fn filter(i: &[u8]) -> IResult<&[u8], (&str, Option<Vec<Expr>>)> {
}
fn expr_filtered(i: &[u8]) -> IResult<&[u8], Expr> {
- let (i, (obj, filters)) = tupl... | diff --git a/askama_shared/src/parser.rs b/askama_shared/src/parser.rs
--- a/askama_shared/src/parser.rs
+++ b/askama_shared/src/parser.rs
@@ -1126,7 +1126,41 @@ mod tests {
#[test]
fn test_parse_filter() {
- super::parse("{{ strvar|e }}", &Syntax::default()).unwrap();
+ use Expr::*;
+ ... | Filters have precedence over unary operators
Performing e.g. `{{ -2|abs }}` is parsed as `{{ -(2|abs) }}` resulting in `-2` being rendered.
Filters have precedence over unary operators
Performing e.g. `{{ -2|abs }}` is parsed as `{{ -(2|abs) }}` resulting in `-2` being rendered.
| Oops -- we should fix that!
Oops -- we should fix that! | 2021-01-05T14:44:35Z | 0.10 | 2021-01-05T15:15:20Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"parser::tests::test_parse_filter"
] | [
"filters::tests::test_abs",
"filters::tests::test_center",
"filters::json::tests::test_json",
"filters::tests::test_capitalize",
"filters::tests::test_filesizeformat",
"filters::tests::test_indent",
"filters::tests::test_into_isize",
"filters::tests::test_into_f64",
"filters::tests::test_lower",
"... | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 423 | rinja-rs__askama-423 | [
"404",
"404"
] | c29ecd68714bddf5e27a9e347c902faa23b2a545 | diff --git a/askama_shared/src/filters/json.rs b/askama_shared/src/filters/json.rs
--- a/askama_shared/src/filters/json.rs
+++ b/askama_shared/src/filters/json.rs
@@ -8,8 +8,8 @@ use serde::Serialize;
///
/// This will panic if `S`'s implementation of `Serialize` decides to fail,
/// or if `T` contains a map with no... | diff --git a/askama_shared/src/filters/json.rs b/askama_shared/src/filters/json.rs
--- a/askama_shared/src/filters/json.rs
+++ b/askama_shared/src/filters/json.rs
@@ -22,6 +22,8 @@ mod tests {
#[test]
fn test_json() {
+ assert_eq!(json(Html, true).unwrap().to_string(), "true");
+ assert_eq!(js... | Version 0.10 places Copy value literals behind a reference where 0.9 did not
I have a large collection of templates that compiles perfectly fine with version 0.9 of askama, but when I try to upgrade to 0.10, I get a bunch of type errors because a bunch of `Copy` types that were previously passed as values now get place... | Sorry for breaking your templates! Are these types being passed to macros, or to custom filters? Can you give an example?
We have some things in the works that will hopefully fix this.
Just off the top of my head, boolean literals, i.e. `true` and `false` are both passed to functions as `&true` and `&false`, and sim... | 2021-01-02T22:39:30Z | 0.10 | 2021-01-05T15:17:19Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"filters::tests::test_abs",
"filters::tests::test_center",
"filters::tests::test_capitalize",
"filters::json::tests::test_json",
"filters::tests::test_into_isize",
"filters::tests::test_into_f64",
"filters::tests::test_indent",
"filters::tests::test_filesizeformat",
"filters::tests::test_linebreaks"... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 411 | rinja-rs__askama-411 | [
"329",
"329"
] | c7697cbd406dce0962618e99a6b78116b14fdb1c | diff --git a/askama_shared/src/generator.rs b/askama_shared/src/generator.rs
--- a/askama_shared/src/generator.rs
+++ b/askama_shared/src/generator.rs
@@ -876,16 +876,34 @@ impl<'a, S: std::hash::BuildHasher> Generator<'a, S> {
match *var {
Target::Name(name) => {
- if !self.local... | diff --git /dev/null b/testing/templates/let-shadow.html
new file mode 100644
--- /dev/null
+++ b/testing/templates/let-shadow.html
@@ -0,0 +1,22 @@
+{%- let a = 1 -%}
+{%- let b -%}
+
+{%- if cond -%}
+ {%- let b = 22 -%}
+ {{ b }}-
+
+ {%- let b = 33 -%}
+ {{ a }}-{{ b }}-
+{%- else -%}
+ {%- let b = 2... | Add support for redefining the same variable with `let` (just as Rust permits)
[Discovered while working on #315 and #327.]
The following statements, although allowed by Rust, are disallowed by `askama`:
~~~~
{% let x = 1 %}
{% let x = 2 %}
~~~~
The failure is:
~~~~
cannot assign twice to immutable variable... | I'm guessing this should be easy to fix.
Any news on this? It's very important when splitting template into small parts with ability to `include` them and setting some variables before include.
I haven't had a chance to work on it, and it's not much of a priority for me right now. I might be able to mentor you if you w... | 2020-12-23T14:44:36Z | 0.10 | 2020-12-25T21:42:49Z | 49252d2457f280026c020d0df46733578eb959a5 | [
"test_if_let",
"test_let",
"test_let_decl",
"test_self_iter",
"test_let_tuple"
] | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 1,097 | rinja-rs__askama-1097 | [
"1086"
] | 53b4b518f9a230665029560df038c318b2e55458 | diff --git a/askama_parser/src/expr.rs b/askama_parser/src/expr.rs
--- a/askama_parser/src/expr.rs
+++ b/askama_parser/src/expr.rs
@@ -270,7 +270,7 @@ impl<'a> Expr<'a> {
separated_list0(char(','), ws(move |i| Self::parse(i, level))),
Self::Array,
),
- ... | diff --git a/testing/tests/loops.rs b/testing/tests/loops.rs
--- a/testing/tests/loops.rs
+++ b/testing/tests/loops.rs
@@ -83,6 +83,16 @@ fn test_for_array() {
assert_eq!(t.render().unwrap(), "123");
}
+#[derive(Template)]
+#[template(source = "{% for i in [1, 2, 3, ] %}{{ i }}{% endfor %}", ext = "txt")]
+stru... | Allow Trailing Commas in Arrays Inside Template Expressions
Basically what the title says.
To give an example of what I mean, this here compiles just fine:
```html
<nav class="flex justify-between items-center max-w-7xl gap-x-12">
{% for (route, name) in [
("/about", "About"),
("/contact",... | Should be easy enough to implement, do you want to try submitting a PR? You'll want to start here:
https://github.com/djc/askama/blob/main/askama_parser/src/expr.rs#L263
Oh yeah sure! I'll see to it once I have the time. Doesn't look too complicated at first glance. | 2024-09-27T14:22:20Z | 0.13 | 2024-09-28T09:55:54Z | 53b4b518f9a230665029560df038c318b2e55458 | [
"test_for",
"test_for_array",
"test_for_cycle",
"test_for_cycle_dynamic",
"test_for_cycle_empty",
"test_for_destructoring_ref_tuple",
"test_for_enumerate",
"test_for_destructoring_tuple",
"test_for_in_if",
"test_for_index",
"test_for_method_call",
"test_for_path_call",
"test_for_range",
"t... | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 1,093 | rinja-rs__askama-1093 | [
"1063"
] | 668bd6f2c1f60dc25143360976a5c775901af889 | diff --git a/askama_derive/src/generator.rs b/askama_derive/src/generator.rs
--- a/askama_derive/src/generator.rs
+++ b/askama_derive/src/generator.rs
@@ -822,6 +822,7 @@ impl<'a> Generator<'a> {
};
let locals = MapChain::with_parent(&self.locals);
let mut child = Self::new(self.input, self.c... | diff --git /dev/null b/testing/templates/fragment-include.html
new file mode 100644
--- /dev/null
+++ b/testing/templates/fragment-include.html
@@ -0,0 +1,9 @@
+{% extends "fragment-base.html" %}
+
+{% block body %}
+{% include "included.html" %}
+{% endblock %}
+
+{% block other_body %}
+<p>Don't render me.</p>
+{% en... | Include statement has no effect when used from a block fragment
When using a block fragment include statements don't seem to work.
See https://github.com/OscartGiles/askama/commit/74162b9453486a53a4de40bba905b4be31dcf3f0 for an example.
Is this intented behaviour or a bug?
| 2024-09-11T03:39:27Z | 0.13 | 2024-09-11T17:40:47Z | 53b4b518f9a230665029560df038c318b2e55458 | [
"test_fragment_include"
] | [
"test_fragment_nested_block",
"test_fragment_simple",
"test_fragment_nested_super",
"test_fragment_super",
"test_fragment_unused_expression",
"test_specific_block"
] | [] | [] | auto_2025-06-03 | |
rinja-rs/askama | 1,057 | rinja-rs__askama-1057 | [
"1056"
] | 627d58bc3b93c484f190d0711a4f9a9e26bc062c | diff --git a/askama_parser/src/node.rs b/askama_parser/src/node.rs
--- a/askama_parser/src/node.rs
+++ b/askama_parser/src/node.rs
@@ -369,24 +369,20 @@ pub struct Cond<'a> {
impl<'a> Cond<'a> {
fn parse(i: &'a str, s: &State<'_>) -> ParseResult<'a, Self> {
- let mut p = tuple((
+ let (i, (_, pws,... | diff --git /dev/null b/testing/tests/if.rs
new file mode 100644
--- /dev/null
+++ b/testing/tests/if.rs
@@ -0,0 +1,26 @@
+use askama::Template;
+
+#[derive(Template)]
+#[template(
+ source = r#"{%- if s == "" -%}
+empty
+{%- else if s == "b" -%}
+b
+{%- elif s == "c" -%}
+c
+{%- else -%}
+else
+{%- endif -%}"#,
+ ... | Add `elif` keyword
Hi,
we recently migrated from Tera to Askama and overall it was quite straightforward. One thing that did not work out of the box was the `elif` keyword, so it would be nice to include that. In addition, `elif` seems to be [supported by jinja](https://jinja.palletsprojects.com/en/3.1.x/templates/... | I guess I'm open to having that added as an alias for `else if`. Would you be able to submit a PR?
The code is here: https://github.com/djc/askama/blob/main/askama_parser/src/node.rs#L375.
Ah cool. :D | 2024-05-23T10:19:41Z | 0.13 | 2024-05-23T12:28:00Z | 53b4b518f9a230665029560df038c318b2e55458 | [
"test_if_let",
"test_if_let_else",
"test_if_let_shadowing",
"test_if_let_struct",
"test_if_let_struct_ref"
] | [] | [] | [] | auto_2025-06-03 |
rinja-rs/askama | 1,029 | rinja-rs__askama-1029 | [
"1022"
] | 03690ab46bbb9cc10f688ebfe003986f9cf4999c | diff --git a/askama_derive/src/generator.rs b/askama_derive/src/generator.rs
--- a/askama_derive/src/generator.rs
+++ b/askama_derive/src/generator.rs
@@ -705,7 +705,7 @@ impl<'a> Generator<'a> {
break;
}
}
- let current_buf = mem::take(&mut self.buf_writable);
+ let... | diff --git /dev/null b/testing/templates/blocks.txt
new file mode 100644
--- /dev/null
+++ b/testing/templates/blocks.txt
@@ -0,0 +1,11 @@
+{% block index %}
+Section: {{ s1 }}
+{% endblock %}
+
+{% block section -%}
+ [
+ {%- for value in values -%}
+ {{ value }}
+ {%- endfor -%}
+ ]
+{%- endblock %}
diff --gi... | Compilation error when using nested template blocks with in a single template file
Hello! I Encountered a compilation error when using[ block fragments](https://github.com/djc/askama/pull/824) with Vec<&str> in a single template file. The error does not occur when the template is split into separate files.
This is w... | I had the same issue, investigating what's going on. | 2024-05-02T15:40:27Z | 0.13 | 2024-05-16T12:07:05Z | 53b4b518f9a230665029560df038c318b2e55458 | [
"test_fragment_nested_block",
"test_fragment_simple",
"test_fragment_nested_super",
"test_fragment_super",
"test_fragment_unused_expression"
] | [
"test_deep",
"test_different_module",
"test_empty_child",
"test_flat_deep",
"test_let_block",
"test_named_end",
"test_nested_blocks",
"test_use_base_directly",
"test_simple_extends",
"test_attr",
"test_comment",
"test_constants",
"test_composition",
"test_else",
"test_define_string_var",... | [] | [] | auto_2025-06-03 |
tracel-ai/burn | 2,224 | tracel-ai__burn-2224 | [
"2080"
] | a9abd8f7462bb2c4976f8338ef3e25c5aa7d57eb | diff --git a/crates/burn-jit/src/kernel/interpolate/bicubic.rs b/crates/burn-jit/src/kernel/interpolate/bicubic.rs
--- a/crates/burn-jit/src/kernel/interpolate/bicubic.rs
+++ b/crates/burn-jit/src/kernel/interpolate/bicubic.rs
@@ -100,6 +100,7 @@ impl<E: JitElement> InterpolateBicubicShader<E> {
cpa!(scope, ... | diff --git a/crates/burn-import/onnx-tests/tests/test_onnx.rs b/crates/burn-import/onnx-tests/tests/test_onnx.rs
--- a/crates/burn-import/onnx-tests/tests/test_onnx.rs
+++ b/crates/burn-import/onnx-tests/tests/test_onnx.rs
@@ -962,7 +962,6 @@ mod tests {
}
#[test]
- #[ignore = "https://github.com/tracel-... | NaN for `[N, C, 1, W]` interpolate inputs with bilinear and bicubic modes for burn-wgpu and burn-ndarray backends
**Describe the bug**
NaN values when the input tensor is `[N, C, 1, W]` shape.
**To Reproduce**
1. Enable `test_1d_bilinear` test in `crates/burn-tensor/src/tests/module/bilinear_interpolate.rs` and... | Ignored tests just in case the code gets deleted:
```rust
#[test]
#[ignore = "https://github.com/tracel-ai/burn/issues/2080"]
fn test_1d_bicubic() {
// Initialize the model without weights (because the exported file does not contain them)
let device = Default::default();
/... | 2024-08-30T15:12:45Z | 0.15 | 2024-09-01T20:37:48Z | 3b51c26958128502d60fb35029c43d9b686b816c | [
"tests::resize_with_scales_1d_linear"
] | [
"tests::greater_scalar",
"tests::greater_or_equal_scalar",
"tests::log",
"tests::expand_shape",
"tests::expand",
"tests::flatten",
"tests::less_or_equal_scalar",
"tests::resize_with_sizes",
"tests::shape",
"tests::clip_opset7",
"tests::greater",
"tests::globalavrpool_1d_2d",
"tests::greater_... | [] | [] | auto_2025-06-10 |
tracel-ai/burn | 1,902 | tracel-ai__burn-1902 | [
"1893"
] | a04da9a2850c3c8b32d4552eab1c4240e27e7983 | diff --git a/crates/burn-core/src/module/param/primitive.rs b/crates/burn-core/src/module/param/primitive.rs
--- a/crates/burn-core/src/module/param/primitive.rs
+++ b/crates/burn-core/src/module/param/primitive.rs
@@ -21,6 +21,12 @@ where
}
fn load_record(self, record: Self::Record) -> Self {
+ let ... | diff --git a/crates/burn-core/src/module/param/primitive.rs b/crates/burn-core/src/module/param/primitive.rs
--- a/crates/burn-core/src/module/param/primitive.rs
+++ b/crates/burn-core/src/module/param/primitive.rs
@@ -267,3 +281,28 @@ impl_module_tuple!([L0, L1, L2, L3, L4, L5, L6][0, 1, 2, 3, 4, 5, 6]);
impl_module_... | The `model.load_record()` method turns off the activation function during the forward pass
**Describe the bug**
The `model.load_record()` method turns off the activation function during the forward pass.
**To Reproduce**
1. Initialize the model with the ReLU activation function.
2. Perform a forward pass with inp... | Hmm well ReLU has no parameters so it's not saved with the weights, which explains why it's initialized to the default value `None` when loading the state from the saved record. For an optional layer with parameters it will work. In the meantime, you can manually set the activation yourself after loading the weights.
... | 2024-06-17T13:20:13Z | 0.14 | 2024-06-18T20:45:22Z | 978ac6c4ecbb4e223c6f0f6361fae982f89e5fde | [
"module::param::primitive::tests::dont_override_constant_module_when_loading_none_record"
] | [
"lr_scheduler::cosine::test::config_initial_lr_too_high - should panic",
"lr_scheduler::cosine::test::config_initial_lr_too_low - should panic",
"lr_scheduler::cosine::test::config_min_lr_too_high - should panic",
"lr_scheduler::cosine::test::config_min_lr_too_low - should panic",
"lr_scheduler::cosine::tes... | [] | [] | auto_2025-06-10 |
tracel-ai/burn | 923 | tracel-ai__burn-923 | [
"912"
] | 64e58b44639e3265ba002a0f71d06a7b584ece9e | diff --git a/burn-train/src/renderer/tui/progress.rs b/burn-train/src/renderer/tui/progress.rs
--- a/burn-train/src/renderer/tui/progress.rs
+++ b/burn-train/src/renderer/tui/progress.rs
@@ -1,22 +1,20 @@
-use crate::renderer::TrainingProgress;
-
use super::TerminalFrame;
+use crate::renderer::TrainingProgress;
use r... | diff --git a/burn-train/src/renderer/tui/progress.rs b/burn-train/src/renderer/tui/progress.rs
--- a/burn-train/src/renderer/tui/progress.rs
+++ b/burn-train/src/renderer/tui/progress.rs
@@ -167,9 +232,29 @@ mod tests {
};
let starting_epoch = 8;
- let progress = calculate_progress(&progress,... | Autotune causes inaccurate estimated training time
The autotune mechanism runs many benchmarks at the beginning of a training execution, then the best operations are stored and used directly throughout the rest of the training.
However, the estimated training time shown in the UI is computed with the time spent on ... | @nathanielsimard 's simple solution: wait about 30 secs before computing the estimated time | 2023-11-01T20:33:01Z | 0.2 | 2023-11-03T12:57:12Z | 306619629797ad4bd631a60560c198e3113f25bf | [
"checkpoint::strategy::lastn::tests::should_always_delete_lastn_epoch_if_higher_than_one",
"checkpoint::strategy::composed::tests::should_delete_when_both_deletes",
"metric::acc::tests::test_accuracy_without_padding",
"metric::acc::tests::test_accuracy_with_padding",
"renderer::tui::progress::tests::calcula... | [] | [] | [] | auto_2025-06-10 |
tracel-ai/burn | 888 | tracel-ai__burn-888 | [
"834"
] | d021c7d7e8d82ddf38d396e1a05d8fdf8244a87e | diff --git a/burn-core/src/optim/adamw.rs b/burn-core/src/optim/adamw.rs
--- a/burn-core/src/optim/adamw.rs
+++ b/burn-core/src/optim/adamw.rs
@@ -150,7 +150,7 @@ impl AdaptiveMomentumW {
let factor = 1.0 - self.beta_2;
let moment_2 = grad.powf(2.0).mul_scalar(factor);
- AdaptiveM... | diff --git a/burn-core/src/optim/adamw.rs b/burn-core/src/optim/adamw.rs
--- a/burn-core/src/optim/adamw.rs
+++ b/burn-core/src/optim/adamw.rs
@@ -228,7 +228,7 @@ mod tests {
assert_eq!(state_optim_before.len(), state_optim_after.len());
}
- const ASSERT_PRECISION: usize = 6;
+ const ASSERT_PRECIS... | AdamW NaN
When learning with AdamW, we get NaNs.
I did not investigate in details, but we had a similar problem with Adam.
I think it's related to a division by zero occurring because we start at time 0. In Adam we start at time 1 to fix that problem, but changing this value might have an incidence on the algorit... | 2023-10-23T15:28:20Z | 0.2 | 2023-10-24T18:48:43Z | 306619629797ad4bd631a60560c198e3113f25bf | [
"optim::adamw::tests::test_adam_optimizer_no_nan"
] | [
"module::param::constant::tests::empty_module_with_phantom",
"lr_scheduler::noam::tests::test_function_increase_and_decrease",
"grad_clipping::base::tests::test_clip_by_norm",
"nn::attention::mask::tests::test_generate_autoregressive_mask",
"nn::attention::mask::tests::test_generate_padding_mask",
"data::... | [] | [] | auto_2025-06-10 | |
tracel-ai/burn | 815 | tracel-ai__burn-815 | [
"814"
] | d7e9e750992229ed6a47101341f4630705fd564c | diff --git a/burn-train/src/learner/builder.rs b/burn-train/src/learner/builder.rs
--- a/burn-train/src/learner/builder.rs
+++ b/burn-train/src/learner/builder.rs
@@ -258,9 +258,9 @@ where
if self.log_to_file {
self.init_logger();
}
- let renderer = self
- .renderer
- ... | diff --git a/burn-train/src/metric/dashboard/tui/progress.rs b/burn-train/src/metric/dashboard/tui/progress.rs
--- a/burn-train/src/metric/dashboard/tui/progress.rs
+++ b/burn-train/src/metric/dashboard/tui/progress.rs
@@ -130,6 +139,7 @@ fn format_eta(eta_secs: u64) -> String {
#[cfg(test)]
mod tests {
use supe... | Dashboard shows wrong remaining time estimate when resuming from checkpoint
**Describe the bug**
<!-- A clear and concise description of what the bug is. -->
Resuming a training from a checkpoint will show the progress bar let's say at 25% and remaining time is not calculated based on the time per batch and number of... | We should compute the time remaining ignoring the previous epochs. The estimate will probably be fine in that case. | 2023-09-18T22:03:22Z | 8.0 | 2023-09-21T12:52:08Z | bdcd5873c526ab27699692c00f9025ae80619948 | [
"metric::dashboard::tui::progress::tests::test_format_eta",
"metric::acc::tests::test_accuracy_without_padding",
"metric::acc::tests::test_accuracy_with_padding",
"metric::dashboard::tui::full_history::tests::test_points",
"metric::dashboard::tui::recent_history::tests::test_push_update_bounds_max_y",
"me... | [] | [] | [] | auto_2025-06-10 |
tracel-ai/burn | 799 | tracel-ai__burn-799 | [
"795"
] | dbc5f0a340136251170138f8825062b42ed2d096 | diff --git a/burn-import/src/burn/node/clip.rs b/burn-import/src/burn/node/clip.rs
--- a/burn-import/src/burn/node/clip.rs
+++ b/burn-import/src/burn/node/clip.rs
@@ -1,5 +1,5 @@
use super::{Node, NodeCodegen};
-use crate::burn::{BurnImports, Scope, TensorType, Type};
+use crate::burn::{Scope, TensorType, Type};
use ... | diff --git a/burn-import/src/burn/node/clip.rs b/burn-import/src/burn/node/clip.rs
--- a/burn-import/src/burn/node/clip.rs
+++ b/burn-import/src/burn/node/clip.rs
@@ -73,7 +70,6 @@ mod tests {
graph.register_input_output(vec!["tensor1".to_string()], vec!["tensor2".to_string()]);
let expected = quote... | How to use .clamp() with a generic backend?
I fear I've probably missed something here, so a pointer in the right direction would be appreciated. As I understand it, we're supposed to define our model with just B: Backend, and only use a concrete definition in our entrypoint. We ran into trouble using .clamp() with thi... | Does this happen for other ops or just with .clamp()? I wander if I correctly implemented this op. Here is the PR responsible for this introduction: https://github.com/burn-rs/burn/pull/550
Yes, you have to use element conversion trait, but the API should be generic over the element type, so I'll make a fix.
```rust... | 2023-09-12T15:32:02Z | 8.0 | 2023-09-12T18:20:15Z | bdcd5873c526ab27699692c00f9025ae80619948 | [
"burn::node::clip::tests::codegen_nodes_max",
"burn::node::clip::tests::codegen_nodes_min_max",
"burn::node::clip::tests::codegen_nodes_min"
] | [
"burn::node::binary::tests::test_binary_codegen_sub_scalars",
"burn::node::dropout::tests::test_codegen",
"burn::node::global_avg_pool::tests::test_codegen_2d",
"burn::node::batch_norm::tests::test_codegen",
"burn::node::conv1d::tests::test_codegen",
"burn::node::avg_pool2d::tests::test_codegen",
"burn:... | [] | [] | auto_2025-06-10 |
tracel-ai/burn | 572 | tracel-ai__burn-572 | [
"541"
] | 597eab524d7232a9798487f094c51ffc2c81f545 | diff --git a/burn-autodiff/src/ops/module.rs b/burn-autodiff/src/ops/module.rs
--- a/burn-autodiff/src/ops/module.rs
+++ b/burn-autodiff/src/ops/module.rs
@@ -38,12 +38,11 @@ impl<B: Backend> ModuleOps<ADBackendDecorator<B>> for ADBackendDecorator<B> {
}
fn embedding_backward(
- weights: ADTensor<B, ... | diff --git /dev/null b/burn-autodiff/src/tests/adaptive_avgpool2d.rs
new file mode 100644
--- /dev/null
+++ b/burn-autodiff/src/tests/adaptive_avgpool2d.rs
@@ -0,0 +1,64 @@
+#[burn_tensor_testgen::testgen(ad_adaptive_avg_pool2d)]
+mod tests {
+ use super::*;
+ use burn_tensor::module::adaptive_avg_pool2d;
+ us... | Adaptive average pooling
Implement adaptive average pooling in all backends.
- [ ] ndarray
- [ ] tch
- [ ] wgpu
- [ ] autodiff
| 2023-08-01T15:30:16Z | 7.0 | 2023-08-04T14:24:02Z | 8808ee265ed366b0e263ec7221854df0c956993f | [
"burn::node::unary::tests::test_unary_codegen_sigmoid",
"burn::node::concat::tests::test_codegen_concat",
"burn::node::reshape::tests::test_codegen_nodes",
"burn::node::batch_norm::tests::test_codegen",
"burn::node::unary::tests::test_unary_codegen_transpose",
"burn::node::unary::tests::test_unary_codegen... | [
"tensor::api::check::tests::binary_ops_devices - should panic",
"tensor::api::check::tests::binary_ops_shapes_no_broadcast - should panic",
"tensor::api::check::tests::binary_ops_shapes_with_broadcast",
"tensor::api::check::tests::index_range_exceed_number_of_dimensions - should panic",
"tensor::api::check:... | [] | [] | auto_2025-06-10 | |
tracel-ai/burn | 390 | tracel-ai__burn-390 | [
"388"
] | bff752b1a86208be6515d6393b752c748f2206cd | diff --git a/burn-tensor/src/tensor/data.rs b/burn-tensor/src/tensor/data.rs
--- a/burn-tensor/src/tensor/data.rs
+++ b/burn-tensor/src/tensor/data.rs
@@ -1,9 +1,9 @@
use alloc::format;
+use alloc::string::String;
use alloc::vec::Vec;
use crate::{tensor::Shape, Element, ElementConversion};
-use libm::{pow, round... | diff --git a/burn-tensor/src/tensor/data.rs b/burn-tensor/src/tensor/data.rs
--- a/burn-tensor/src/tensor/data.rs
+++ b/burn-tensor/src/tensor/data.rs
@@ -387,4 +407,30 @@ mod tests {
let data = Data::from([3.0, 5.0, 6.0]);
assert_eq!(data.shape, Shape::new([3]));
}
+
+ #[test]
+ fn should_... | Intermittent test failure
CI has failed due to this error but it worked when restarted:
```
error: test failed, to rerun pass `--lib`
failures:
---- nn::transformer::decoder::tests::test_autoregressive_norm_last stdout ----
thread 'nn::transformer::decoder::tests::test_autoregressive_norm_last' panicked at ... | 2023-06-06T18:10:48Z | 5.0 | 2023-06-06T18:34:25Z | 8ed0775dfdd76ef950fce94fb3fbf7a0aaad957c | [
"tensor::data::tests::should_assert_appox_eq_limit"
] | [
"tensor::api::check::tests::binary_ops_shapes_no_broadcast - should panic",
"tensor::api::check::tests::binary_ops_shapes_with_broadcast",
"tensor::api::check::tests::binary_ops_devices - should panic",
"tensor::api::check::tests::index_range_exceed_dimension - should panic",
"tensor::api::check::tests::ind... | [] | [] | auto_2025-06-10 | |
tracel-ai/burn | 1,682 | tracel-ai__burn-1682 | [
"1672"
] | 6c708527b933d67a49bca29e119ff6b4f1ebcb1e | diff --git a/crates/burn-core/src/nn/linear.rs b/crates/burn-core/src/nn/linear.rs
--- a/crates/burn-core/src/nn/linear.rs
+++ b/crates/burn-core/src/nn/linear.rs
@@ -64,9 +64,14 @@ impl<B: Backend> Linear<B> {
///
/// # Shapes
///
- /// - input: `[..., any, d_input]`
- /// - output: `[..., any, d_... | diff --git a/crates/burn-core/src/nn/linear.rs b/crates/burn-core/src/nn/linear.rs
--- a/crates/burn-core/src/nn/linear.rs
+++ b/crates/burn-core/src/nn/linear.rs
@@ -150,4 +155,23 @@ mod tests {
assert_eq!(result.into_data(), expected_result.into_data());
}
+
+ #[test]
+ fn test_linear_1d() {
+ ... | Support forward without a batch dimension
### Feature description
Support forward without a batch dimension
### Feature motivation
Linear model import from pytorch cannot be used as it for example.
Imported model work with tensor of rank 1 but Burn Linear doesn't not support tensor of rank 1 at this time.
As d... | 2024-04-22T16:41:55Z | 0.14 | 2024-04-22T22:39:09Z | 978ac6c4ecbb4e223c6f0f6361fae982f89e5fde | [
"nn::linear::tests::test_linear_1d"
] | [
"lr_scheduler::cosine::test::config_initial_lr_too_high - should panic",
"lr_scheduler::cosine::test::config_initial_lr_too_low - should panic",
"lr_scheduler::cosine::test::config_min_lr_too_high - should panic",
"lr_scheduler::cosine::test::config_min_lr_too_low - should panic",
"lr_scheduler::cosine::tes... | [] | [] | auto_2025-06-10 | |
tracel-ai/burn | 1,657 | tracel-ai__burn-1657 | [
"1642"
] | 7705fd9c25cf6ca201da5652605cf2f46c363086 | diff --git a/crates/burn-import/src/burn/node/unary.rs b/crates/burn-import/src/burn/node/unary.rs
--- a/crates/burn-import/src/burn/node/unary.rs
+++ b/crates/burn-import/src/burn/node/unary.rs
@@ -198,8 +198,9 @@ impl UnaryNode {
Self::new(input, output, UnaryNodeKind::Tanh, Rc::new(function))
}
- ... | diff --git a/crates/burn-import/onnx-tests/tests/onnx_tests.rs b/crates/burn-import/onnx-tests/tests/onnx_tests.rs
--- a/crates/burn-import/onnx-tests/tests/onnx_tests.rs
+++ b/crates/burn-import/onnx-tests/tests/onnx_tests.rs
@@ -685,18 +685,23 @@ mod tests {
let model: transpose::Model<Backend> = transpose::... | Incorrect implementation of Transpose ONNX operation
**Describe the bug**
This PR https://github.com/tracel-ai/burn/pull/546/files incorrectly implemented transpose operation. Instead of using permute operator, this uses transpose Burn operator which only flips last two dimensions.
| CCing @laggui , @nathanielsimard | 2024-04-18T14:41:01Z | 0.14 | 2024-04-19T13:34:04Z | 978ac6c4ecbb4e223c6f0f6361fae982f89e5fde | [
"burn::node::binary::tests::test_binary_codegen_div_scalars",
"burn::node::unary::tests::test_unary_codegen_tanh",
"burn::node::unary::tests::test_unary_neg_tensor",
"burn::node::unsqueeze::tests::test_codegen_nodes",
"burn::node::conv_transpose_2d::tests::test_codegen",
"burn::node::binary::tests::test_b... | [] | [] | [] | auto_2025-06-10 |
tracel-ai/burn | 1,188 | tracel-ai__burn-1188 | [
"1112"
] | f5ac5d8e9f7743bb82b36bc3ab824ec621b95c69 | diff --git a/backend-comparison/Cargo.toml b/backend-comparison/Cargo.toml
--- a/backend-comparison/Cargo.toml
+++ b/backend-comparison/Cargo.toml
@@ -23,7 +23,7 @@ ndarray-blas-openblas = ["burn/ndarray", "burn/openblas"]
tch-cpu = ["burn/tch"]
tch-gpu = ["burn/tch"]
wgpu = ["burn/wgpu"]
-wgpu-fusion = ["burn/wgpu"... | diff --git a/burn-compute/src/client.rs b/burn-compute/src/client.rs
--- a/burn-compute/src/client.rs
+++ b/burn-compute/src/client.rs
@@ -72,12 +72,18 @@ where
}
/// Executes the fastest kernel in the autotune operation, using (cached) runtime benchmarks
- pub fn execute_autotune(
+ pub fn autotune_e... | Support autotune of fused compute shaders
When creating new shaders, we should be able to execute them with different launch options based on benchmarks performed by the autotune system.
| 2024-01-29T23:02:56Z | 0.12 | 2024-02-01T17:58:26Z | f5ac5d8e9f7743bb82b36bc3ab824ec621b95c69 | [
"memory_management::simple::tests::never_dealloc_strategy_never_deallocs",
"memory_management::simple::tests::period_tick_dealloc_strategy_should_dealloc_after_period",
"memory_management::simple::tests::slice_strategy_maximum_bytes",
"memory_management::simple::tests::can_mut_with_single_tensor_reference",
... | [] | [] | [] | auto_2025-06-10 | |
tracel-ai/burn | 1,087 | tracel-ai__burn-1087 | [
"911"
] | b17ac2dfba451fe24e22bbe2a63de84277b0c09e | diff --git a/.gitignore b/.gitignore
--- a/.gitignore
+++ b/.gitignore
@@ -5,7 +5,8 @@ Cargo.lock
.DS_Store
.cargo/config.toml
+.dir-locals.el
.idea
.vscode
-.fleet
.vs
+.fleet
diff --git a/burn-common/Cargo.toml b/burn-common/Cargo.toml
--- a/burn-common/Cargo.toml
+++ b/burn-common/Cargo.toml
@@ -29,6 +29,7 @@... | diff --git a/burn-compute/Cargo.toml b/burn-compute/Cargo.toml
--- a/burn-compute/Cargo.toml
+++ b/burn-compute/Cargo.toml
@@ -30,9 +39,14 @@ derive-new = { workspace = true }
spin = { workspace = true }
log = { workspace = true }
hashbrown = { workspace = true }
+dirs = { workspace = true , optional = true}
+serde ... | Autotune: benchmarks cached on disk
The autotune mechanism uses a cache to check if, for a specific operation on an input size within a specific range, the benchmarks have already been computed. If so, the fastest operation in that setting on your machine is stored in the cache and we just use this one. That way, it is... | From tuner in burn-compute:
```rust
fn autotuning(
&mut self,
autotune_operation_set: Box<dyn AutotuneOperationSet<S::AutotuneKey>>,
client: &ComputeClient<S, C>,
) -> Box<dyn AutotuneOperation> {
let key = autotune_operation_set.key();
let autotunables = autotune_... | 2023-12-20T04:21:34Z | 0.12 | 2024-01-05T19:19:33Z | f5ac5d8e9f7743bb82b36bc3ab824ec621b95c69 | [
"memory_management::simple::tests::never_dealloc_strategy_never_deallocs",
"memory_management::simple::tests::period_tick_dealloc_strategy_should_dealloc_after_period",
"memory_management::simple::tests::slice_strategy_maximum_bytes",
"memory_management::simple::tests::can_mut_with_single_tensor_reference",
... | [] | [] | [] | auto_2025-06-10 |
tracel-ai/burn | 978 | tracel-ai__burn-978 | [
"636"
] | cdf54d0b40d513ea1ae2cf1a4335b60e6248e039 | diff --git a/burn-core/src/nn/pool/avg_pool2d.rs b/burn-core/src/nn/pool/avg_pool2d.rs
--- a/burn-core/src/nn/pool/avg_pool2d.rs
+++ b/burn-core/src/nn/pool/avg_pool2d.rs
@@ -20,7 +20,7 @@ pub struct AvgPool2dConfig {
pub padding: PaddingConfig2d,
/// If the padding is counted in the denominator when computin... | diff --git a/burn-import/onnx-tests/tests/avg_pool2d/avg_pool2d.py b/burn-import/onnx-tests/tests/avg_pool2d/avg_pool2d.py
--- a/burn-import/onnx-tests/tests/avg_pool2d/avg_pool2d.py
+++ b/burn-import/onnx-tests/tests/avg_pool2d/avg_pool2d.py
@@ -10,14 +10,20 @@ class Model(nn.Module):
def __init__(self):
... | Support for `count_include_pad=False` for AvgPool1d and AvgPool2d
#### Description
The current implementation of the `AvgPool2d` layer in the Burn framework assumes that `count_include_pad=True`, meaning that zero-padding is included in the average pooling calculation. While this approach is valid and useful in so... | I am reopening it to work on onnx part. I will take care of it.
@antimora Yup I forgot it sorry !
@nathanielsimard I started working on ONNX changes but I may not be able to complete before a planned release. It appears a little more involved because PyTorch adds Pads node instead of accounting in AvgPool logic itself.... | 2023-11-21T00:53:50Z | 0.2 | 2023-11-22T14:28:21Z | 306619629797ad4bd631a60560c198e3113f25bf | [
"burn::node::avg_pool2d::tests::test_codegen"
] | [
"burn::node::base::tests::test_codegen_clone_tensor",
"burn::node::binary::tests::test_binary_codegen_add_scalars",
"burn::node::binary::tests::test_binary_codegen_div_scalars",
"burn::node::binary::tests::test_binary_codegen_div",
"burn::node::binary::tests::test_binary_codegen_sub",
"burn::node::conv2d:... | [] | [] | auto_2025-06-10 |
tracel-ai/burn | 2,186 | tracel-ai__burn-2186 | [
"2185"
] | c29ed43441fc5b3286749816c9069ed90f178a77 | diff --git a/crates/burn-core/src/nn/norm/layer.rs b/crates/burn-core/src/nn/norm/layer.rs
--- a/crates/burn-core/src/nn/norm/layer.rs
+++ b/crates/burn-core/src/nn/norm/layer.rs
@@ -67,7 +67,7 @@ impl<B: Backend> LayerNorm<B> {
pub fn forward<const D: usize>(&self, input: Tensor<B, D>) -> Tensor<B, D> {
... | diff --git a/crates/burn-core/src/nn/norm/layer.rs b/crates/burn-core/src/nn/norm/layer.rs
--- a/crates/burn-core/src/nn/norm/layer.rs
+++ b/crates/burn-core/src/nn/norm/layer.rs
@@ -122,6 +122,27 @@ mod tests {
output.to_data().assert_approx_eq(&expected, 3);
}
+ #[test]
+ fn layer_norm_forward_l... | LayerNorm does not normalize its input correctly
**Describe the bug**
Both from pytorch documentation...:
https://pytorch.org/docs/stable/generated/torch.nn.LayerNorm.html#torch.nn.LayerNorm
```
y = (x − E[x]) / sqrt(Var[x] + ϵ)∗γ + β
```
... and from pytorch ATen implementation...:
https://github.com/pytorch/... | 2024-08-20T07:32:27Z | 0.14 | 2024-08-20T11:47:34Z | 978ac6c4ecbb4e223c6f0f6361fae982f89e5fde | [
"nn::norm::layer::tests::layer_norm_forward_large_epsilon"
] | [
"lr_scheduler::cosine::test::config_initial_lr_too_high - should panic",
"grad_clipping::base::tests::test_clip_by_value",
"grad_clipping::base::tests::test_clip_by_norm",
"lr_scheduler::cosine::test::config_initial_lr_too_low - should panic",
"lr_scheduler::cosine::test::config_min_lr_too_low - should pani... | [] | [] | auto_2025-06-10 | |
tokio-rs/axum | 1,934 | tokio-rs__axum-1934 | [
"1931"
] | 377d73a55ba7bdbb1a47084c684fb1334b200f1b | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -7,7 +7,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
# Unreleased
-- None.
+- **fixed:** Don't allow extracting `MatchedPath` in fallbacks ([#1934])
+- **fixed:** Fix pani... | diff --git a/axum/src/extract/matched_path.rs b/axum/src/extract/matched_path.rs
--- a/axum/src/extract/matched_path.rs
+++ b/axum/src/extract/matched_path.rs
@@ -176,6 +176,7 @@ mod tests {
Router,
};
use http::{Request, StatusCode};
+ use hyper::Body;
#[crate::test]
async fn extracti... | Axum 0.6.15 seems to break ServeDir completely
- [x] I have looked for existing issues (including closed) about this
## Bug Report
### Version
`cargo tree | grep axum`
```
axum v0.6.15
│ ├── axum-core v0.3.4
│ ├── axum-core v0.3.4
```
`cargo tree | grep tower-http`
```
tower-http
├── tower-http v0... | Not sure if your actual routing setup is more complex than that but you don't need to crate a `Router` just to put `ServeDir` into a `fallback_service`. You simply do:
```rust
let routes_all = Router::new().fallback_service(ServeDir::new("./"));
```
This works because `ServeDir` implements `Service` which is th... | 2023-04-14T09:58:06Z | 0.6 | 2023-04-17T11:20:56Z | c100650464e96cc9b48a37290760f288d6311ed8 | [
"error_handling::traits",
"body::stream_body::stream_body_traits",
"extract::connect_info::traits",
"extract::host::tests::forwarded_parsing",
"extract::path::de::tests::test_parse_error_at_index_error",
"extract::path::de::tests::test_parse_error_at_key_error",
"extract::path::de::tests::test_parse_err... | [] | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 1,889 | tokio-rs__axum-1889 | [
"1884"
] | 24f8dc53f49d8946418f103314ab6a31ff68be5c | diff --git a/axum-extra/CHANGELOG.md b/axum-extra/CHANGELOG.md
--- a/axum-extra/CHANGELOG.md
+++ b/axum-extra/CHANGELOG.md
@@ -7,7 +7,9 @@ and this project adheres to [Semantic Versioning].
# Unreleased
-- None.
+- **added:** Add `OptionalPath` extractor ([#1889])
+
+[#1889]: https://github.com/tokio-rs/axum/pull/... | diff --git /dev/null b/axum-extra/src/extract/optional_path.rs
new file mode 100644
--- /dev/null
+++ b/axum-extra/src/extract/optional_path.rs
@@ -0,0 +1,102 @@
+use axum::{
+ async_trait,
+ extract::{path::ErrorKind, rejection::PathRejection, FromRequestParts, Path},
+ RequestPartsExt,
+};
+use serde::de::De... | OptionalPath extractor
- [x] I have looked for existing issues (including closed) about this
## Feature Request
### Motivation
We've been getting many people asking about an easy way of extracting a path segment value only if a placeholder exists at the route the handler was called from.
### Proposal
Add... | ```
.route("/", get(handler)
.route("/:path", get(handler))
```
```rs
pub(crate) async fn handler(
path: Option<Path<String>>,
) -> impl IntoResponse
{
..
}
```
Yes, that works most of the time, but it doesn't reject the request for `/:path` requests with invalid UTF-8. Or m... | 2023-03-27T17:03:41Z | 0.6 | 2023-04-09T12:23:15Z | c100650464e96cc9b48a37290760f288d6311ed8 | [
"extract::cached::tests::works",
"extract::cookie::tests::signed_cannot_access_invalid_cookies",
"extract::cookie::tests::plaintext_cookies",
"extract::with_rejection::tests::extractor_rejection_is_transformed",
"extract::cookie::tests::private_cookies",
"extract::cookie::tests::signed_cookies_with_custom... | [] | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 1,826 | tokio-rs__axum-1826 | [
"1738"
] | 1327a598ce07c4642a2cca2c20f1093d7ff123f7 | diff --git a/axum-macros/CHANGELOG.md b/axum-macros/CHANGELOG.md
--- a/axum-macros/CHANGELOG.md
+++ b/axum-macros/CHANGELOG.md
@@ -7,7 +7,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
# Unreleased
-- None.
+- **fixed:** Improve `#[debug_handler]` message for known generic
+... | diff --git /dev/null b/axum-macros/tests/debug_handler/fail/json_not_deserialize.rs
new file mode 100644
--- /dev/null
+++ b/axum-macros/tests/debug_handler/fail/json_not_deserialize.rs
@@ -0,0 +1,9 @@
+use axum::Json;
+use axum_macros::debug_handler;
+
+struct Struct {}
+
+#[debug_handler]
+async fn handler(foo: Json<... | `#[debug_handler]` could suggest `Deserialize` for `Json`
- [x] I have looked for existing issues (including closed) about this
## Feature Request
### Motivation
While trying to teach someone axum I suggested the use of the `#[debug_handler]` macro in order to find out why their handler wasn't working. The err... | Thats a regression I'm surprised we missed actually. Thought we had a test for that but oh well.
On 0.5 the error is
```
error[E0277]: the trait bound `for<'de> NotDeserialize: serde::de::Deserialize<'de>` is not satisfied
--> src/main.rs:19:21
|
19 | async fn handler(_: Json<NotDeserialize>) {}
| ... | 2023-03-07T13:14:46Z | 0.6 | 2023-03-10T08:51:05Z | c100650464e96cc9b48a37290760f288d6311ed8 | [
"debug_handler::ui"
] | [
"tests/typed_path/fail/missing_capture.rs [should fail to compile]",
"tests/typed_path/fail/missing_field.rs [should fail to compile]",
"tests/typed_path/fail/not_deserialize.rs [should fail to compile]",
"tests/typed_path/fail/route_not_starting_with_slash.rs [should fail to compile]",
"tests/typed_path/fa... | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 1,810 | tokio-rs__axum-1810 | [
"1781"
] | 5606ea3f9ecb738c1cd4321fa0bedc4fc2adc29f | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -9,11 +9,13 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- **fixed:** Add `#[must_use]` attributes to types that do nothing unless used ([#1809])
- **fixed:** Add `#[must_use]... | diff --git a/axum/src/typed_header.rs b/axum/src/typed_header.rs
--- a/axum/src/typed_header.rs
+++ b/axum/src/typed_header.rs
@@ -175,19 +177,35 @@ mod tests {
async fn typed_header() {
async fn handle(
TypedHeader(user_agent): TypedHeader<headers::UserAgent>,
+ TypedHeader(cookie... | TypedHeader shouldn't reject missing headers by default
- [x] I have looked for existing issues (including closed) about this
## Feature Request
TypedHeader should allow specific Header type to handle missing headers instead of immediately rejecting the request.
### Motivation
`Header` trait already allows ... | Seems reasonable!
axum just calls [`typed_try_get`](https://docs.rs/headers/0.3.8/src/headers/map_ext.rs.html#41). Could the issue be there?
Seems like it, that optimization (?) of returning `Ok(None)` when the size hint is 0 seems wrong according to the original issue description.
Traced it down through git blame to h... | 2023-03-03T12:24:36Z | 0.6 | 2023-03-03T15:13:25Z | c100650464e96cc9b48a37290760f288d6311ed8 | [
"typed_header::tests::typed_header"
] | [
"error_handling::traits",
"body::stream_body::stream_body_traits",
"extract::connect_info::traits",
"extract::host::tests::forwarded_parsing",
"extract::path::de::tests::test_parse_error_at_key_error",
"extract::path::de::tests::test_parse_error_at_index_error",
"extract::path::de::tests::test_parse_err... | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 1,469 | tokio-rs__axum-1469 | [
"1467"
] | 70833b9f4f8c528a88e7d97fe5193a37f4bf51e3 | diff --git a/axum/src/middleware/from_extractor.rs b/axum/src/middleware/from_extractor.rs
--- a/axum/src/middleware/from_extractor.rs
+++ b/axum/src/middleware/from_extractor.rs
@@ -201,7 +201,7 @@ where
impl<T, E, B, S> Service<Request<B>> for FromExtractor<T, E, S>
where
E: FromRequestParts<S> + 'static,
- ... | diff --git a/axum/src/middleware/from_extractor.rs b/axum/src/middleware/from_extractor.rs
--- a/axum/src/middleware/from_extractor.rs
+++ b/axum/src/middleware/from_extractor.rs
@@ -305,9 +304,10 @@ where
#[cfg(test)]
mod tests {
use super::*;
- use crate::{handler::Handler, routing::get, test_helpers::*, Ro... | `from_extractor` not compatible with `RequestBodyLimitLayer`
I tried adding the `RequestBodyLimitLayer` to my application. This works for everything, except the routes that contain a layer built with `from_extractor`.
My `axum` version is `0.5.16`, `tower-http` is `0.3.4` and `rustc` is `1.63.0`.
## A minimized e... | 2022-10-10T16:24:03Z | 0.14 | 2022-10-11T11:11:51Z | beb5ebbe8abb5f7e9afec08ed86e878d6272375e | [
"body::stream_body::stream_body_traits",
"error_handling::traits",
"extract::connect_info::traits",
"extract::host::tests::forwarded_parsing",
"extract::path::de::tests::test_parse_error_error",
"extract::path::de::tests::test_parse_error_at_index_error",
"extract::path::de::tests::test_parse_error_at_k... | [] | [] | [] | auto_2025-06-06 | |
tokio-rs/axum | 1,305 | tokio-rs__axum-1305 | [
"1290"
] | fa51cf52664d2e33fe37ca4252879bb1c73e87e9 | diff --git a/axum-macros/CHANGELOG.md b/axum-macros/CHANGELOG.md
--- a/axum-macros/CHANGELOG.md
+++ b/axum-macros/CHANGELOG.md
@@ -14,10 +14,13 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
enum but instead generate `type Rejection = axum::response::Response`. Use the
new `#[... | diff --git a/axum-macros/src/from_request.rs b/axum-macros/src/from_request.rs
--- a/axum-macros/src/from_request.rs
+++ b/axum-macros/src/from_request.rs
@@ -537,29 +734,57 @@ fn impl_enum_by_extracting_all_at_once(
let path_span = path.span();
- Ok(quote_spanned! {path_span=>
- #[::axum::async_trai... | Add `#[derive(FromRequestParts)]` macro
## Feature Request
### Motivation
I'm pretty sure many current users of the `FromRequest` macro are using the resulting type as a non-terminal extractor. After #1272, this will no longer be supported.
### Proposal
Add `FromRequestParts` macro after #1272, but before t... | 2022-08-23T12:12:37Z | 0.5 | 2022-08-23T19:14:03Z | fa51cf52664d2e33fe37ca4252879bb1c73e87e9 | [
"tests/typed_path/fail/missing_capture.rs [should fail to compile]",
"tests/typed_path/fail/missing_field.rs [should fail to compile]",
"tests/typed_path/fail/not_deserialize.rs [should fail to compile]",
"tests/typed_path/fail/route_not_starting_with_slash.rs [should fail to compile]",
"tests/typed_path/fa... | [] | [] | [] | auto_2025-06-06 | |
tokio-rs/axum | 3,059 | tokio-rs__axum-3059 | [
"3056"
] | 6d30c579e04cce85a161dd72f8129ad96aed482d | diff --git a/axum/src/extract/connect_info.rs b/axum/src/extract/connect_info.rs
--- a/axum/src/extract/connect_info.rs
+++ b/axum/src/extract/connect_info.rs
@@ -92,6 +92,17 @@ const _: () = {
*stream.remote_addr()
}
}
+
+ impl<'a, L, F> Connected<serve::IncomingStream<'a, serve::TapIo<L,... | diff --git a/axum/src/serve.rs b/axum/src/serve.rs
--- a/axum/src/serve.rs
+++ b/axum/src/serve.rs
@@ -434,6 +434,7 @@ mod tests {
extract::connect_info::Connected,
handler::{Handler, HandlerWithoutStateExt},
routing::get,
+ serve::ListenerExt,
Router,
};
diff --git a/a... | into_make_service_with_connect_info not compile anymore
<!--
Thank you for reporting an issue.
Please fill in as much of the template below as you're able.
-->
after this PR https://github.com/tokio-rs/axum/pull/2941
with `tap_io` called on `listener`,
`axum::serve(listener, app.into_make_service_with_co... | Thanks for the report! I will look into it soon. | 2024-12-01T11:50:12Z | 1.0 | 2024-12-02T07:36:52Z | 6d30c579e04cce85a161dd72f8129ad96aed482d | [
"extract::connect_info::traits",
"extract::connect_info::tests::custom",
"error_handling::traits",
"extract::matched_path::tests::can_extract_nested_matched_path_in_middleware_on_nested_router_with_nest_service",
"extract::connect_info::tests::socket_addr",
"extract::matched_path::tests::can_extract_neste... | [] | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 3,051 | tokio-rs__axum-3051 | [
"2651"
] | 9ddb25f0897f7334e4e85ca26a81db272df1f02f | diff --git a/axum/src/routing/mod.rs b/axum/src/routing/mod.rs
--- a/axum/src/routing/mod.rs
+++ b/axum/src/routing/mod.rs
@@ -200,6 +200,10 @@ where
#[doc(alias = "scope")] // Some web frameworks like actix-web use this term
#[track_caller]
pub fn nest(self, path: &str, router: Router<S>) -> Self {
+ ... | diff --git a/axum/src/extract/matched_path.rs b/axum/src/extract/matched_path.rs
--- a/axum/src/extract/matched_path.rs
+++ b/axum/src/extract/matched_path.rs
@@ -324,14 +324,14 @@ mod tests {
"/{*path}",
any(|req: Request| {
Router::new()
- .nest("/", Route... | Consider disallowing root nesting
Before the 0.8.0 release, I would like to come back to `nest("/", router)` and `nest_service("/", svc)`. I believe I suggested disallowing them (in favor of `merge` and `fallback_service`) before, but if so there must have been some pushback by David since we still allow both.
Any poi... | The only thing I found is this: https://github.com/tokio-rs/axum/discussions/2012
… which suggests I'm wrong about previous push-back and we should just disallow root nesting :)
@jplatte are you planning to deprecate the `nest` function or just improve it? This is very handy and makes the code much readable.
@dancixx ... | 2024-11-26T18:29:20Z | 1.0 | 2024-11-26T19:59:42Z | 6d30c579e04cce85a161dd72f8129ad96aed482d | [
"routing::tests::nest::nest_router_at_empty_path - should panic",
"routing::tests::nest::nest_service_at_empty_path - should panic",
"routing::tests::nest::nest_service_at_root - should panic",
"routing::tests::nest::nest_router_at_root - should panic"
] | [
"error_handling::traits",
"extract::connect_info::traits",
"extract::path::de::tests::test_deserialize_key_value",
"extract::path::de::tests::test_parse_error_at_index_error",
"extract::path::de::tests::test_parse_error_at_key_error",
"extract::path::de::tests::test_parse_error_at_key_error_multiple",
"... | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 3,039 | tokio-rs__axum-3039 | [
"3038"
] | 7e59625778698410ff40ab968120cfea825a6d9a | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -8,6 +8,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
# Unreleased
- **fixed:** Skip SSE incompatible chars of `serde_json::RawValue` in `Event::json_data` ([#2992])
+- **fix... | diff --git a/axum/src/extract/path/mod.rs b/axum/src/extract/path/mod.rs
--- a/axum/src/extract/path/mod.rs
+++ b/axum/src/extract/path/mod.rs
@@ -973,4 +973,25 @@ mod tests {
r#"Invalid URL: Cannot parse `res` with value `456456-123-456456`: UUID parsing failed: invalid group count: expected 5, found 3"#
... | Hit unreachable code when deserializing a [char; 2] inside Path
### Description
Hi, I used this struct inside a `Path`
```
#[derive(Debug, Deserialize)]
struct MoreChars {
first_two: [char; 2],
second_two: [char; 2],
crate_name: String,
}
```
with the route `.route("/:first_two/:second_two/:cr... | Can reproduce on `main`. | 2024-11-19T20:28:12Z | 1.0 | 2024-11-19T20:57:31Z | 6d30c579e04cce85a161dd72f8129ad96aed482d | [
"extract::path::tests::regression_3038"
] | [
"error_handling::traits",
"extract::connect_info::tests::custom",
"extract::connect_info::tests::both_mock_and_real_connect_info",
"extract::connect_info::traits",
"extract::connect_info::tests::socket_addr",
"extract::matched_path::tests::can_extract_nested_matched_path_in_middleware_on_nested_router",
... | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 2,992 | tokio-rs__axum-2992 | [
"2981"
] | 65ad603701c64db42bb759d35722cf7ee6864842 | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -7,6 +7,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
# Unreleased
+- **fixed:** Skip SSE incompatible chars of `serde_json::RawValue` in `Event::json_data` ([#2992])
- **b... | diff --git a/axum/Cargo.toml b/axum/Cargo.toml
--- a/axum/Cargo.toml
+++ b/axum/Cargo.toml
@@ -117,7 +117,7 @@ quickcheck = "1.0"
quickcheck_macros = "1.0"
reqwest = { version = "0.12", default-features = false, features = ["json", "stream", "multipart"] }
serde = { version = "1.0", features = ["derive"] }
-serde_js... | `Event::json_data` can construct an invalid SSE event by passing a `serde_json::value::RawValue`
<!--
Thank you for reporting an issue.
Please fill in as much of the template below as you're able.
-->
- [x] I have looked for existing issues (including closed) about this
## Bug Report
When using `Event::js... | I don't really understand the second solution.
The third one sounds reasonable to me.
@SabrinaJewson do you have an opinion on this?
Yeah, third solution sounds reasonable. | 2024-10-16T17:35:52Z | 1.0 | 2024-10-17T14:49:39Z | 6d30c579e04cce85a161dd72f8129ad96aed482d | [
"response::sse::tests::valid_json_raw_value_chars_stripped"
] | [
"error_handling::traits",
"extract::connect_info::traits",
"extract::path::de::tests::test_parse_error_at_index_error",
"extract::path::de::tests::test_parse_error_at_key_error",
"extract::path::de::tests::test_parse_error_at_key_error_multiple",
"extract::path::de::tests::test_parse_error_error",
"extr... | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 2,903 | tokio-rs__axum-2903 | [
"2251"
] | 0712a46cd9f2bece74160e83085411aa0c174be1 | diff --git /dev/null b/axum/src/docs/routing/method_not_allowed_fallback.md
new file mode 100644
--- /dev/null
+++ b/axum/src/docs/routing/method_not_allowed_fallback.md
@@ -0,0 +1,38 @@
+Add a fallback [`Handler`] for the case where a route exists, but the method of the request is not supported.
+
+Sets a fallback on ... | diff --git a/axum/src/routing/tests/fallback.rs b/axum/src/routing/tests/fallback.rs
--- a/axum/src/routing/tests/fallback.rs
+++ b/axum/src/routing/tests/fallback.rs
@@ -329,6 +329,51 @@ async fn merge_router_with_fallback_into_empty() {
assert_eq!(res.text().await, "outer");
}
+#[crate::test]
+async fn mna_fa... | Add a method-not-allowed fallback on the Router level
- [x] I have looked for existing issues (including closed) about this
## Feature Request
### Motivation
It's a common mistake to expect `Router::fallback` to be triggered on method-not-allowed.
Furthermore, as shown by [#1005](https://github.com/tokio-rs/a... | I wonder if we should just apply the normal `Router::fallback` to both 404 and 405 (keeping the `Allow` header of course). Do you think that would work for most cases? Is it often that you want different fallbacks for 404 and 405?
Wdym by keeping the allow header? You mean setting it if the user-provided fallback handl... | 2024-09-08T18:55:33Z | 1.0 | 2024-10-12T09:36:57Z | 6d30c579e04cce85a161dd72f8129ad96aed482d | [
"error_handling::traits",
"extract::connect_info::traits",
"extract::host::tests::forwarded_parsing",
"extract::path::de::tests::test_parse_error_at_key_error",
"extract::path::de::tests::test_parse_error_at_index_error",
"extract::path::de::tests::test_parse_error_error",
"extract::path::de::tests::tes... | [] | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 868 | tokio-rs__axum-868 | [
"865"
] | 422a883cb2a81fa6fbd2f2a1affa089304b7e47b | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -71,6 +71,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
be accepted but most likely result in bugs ([#823])
- **breaking:** `Headers` has been removed. Arrays of tuples dire... | diff --git a/axum/src/json.rs b/axum/src/json.rs
--- a/axum/src/json.rs
+++ b/axum/src/json.rs
@@ -244,4 +264,19 @@ mod tests {
assert!(valid_json_content_type("application/cloudevents+json").await);
assert!(!valid_json_content_type("text/json").await);
}
+
+ #[tokio::test]
+ async fn inval... | HTTP Status on Invalid JSON Request
When I send a request where the `content-type` is not `application/json` I get the answer `415 Unsupported Media Type` and if the JSON parsing fails I get `422 Unprocessable Entity`
Two issues:
1. This is not documented correctly: https://docs.rs/axum/latest/axum/struct.Json.ht... | 👍🏼 for distinguishing syntactic errors and content errors. `serde_json::Error` has a `classify()` method that allows to check what failed exactly.
> This is not documented correctly: https://docs.rs/axum/latest/axum/struct.Json.html In this documentation both of them would result in 400 Bad Request
Good catch. Tha... | 2022-03-17T17:15:56Z | 0.4 | 2022-03-18T15:53:41Z | 2e5d56a9b125142817378883b098470d05ab0597 | [
"json::tests::invalid_json_syntax"
] | [
"error_handling::traits",
"body::stream_body::stream_body_traits",
"extract::path::de::tests::test_unsupported_type_error_tuple",
"extract::path::de::tests::test_parse_tuple_ignoring_additional_fields",
"extract::multipart::tests::content_type_with_encoding",
"extract::path::de::tests::test_wrong_number_o... | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 755 | tokio-rs__axum-755 | [
"747"
] | 409a6651f59c4443e7cd6aa05c6f9485348ceed9 | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -54,6 +54,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- **fixed:** Set `Allow` header when responding with `405 Method Not Allowed` ([#733])
- **fixed:** Correctly set the `C... | diff --git a/axum/src/routing/tests/mod.rs b/axum/src/routing/tests/mod.rs
--- a/axum/src/routing/tests/mod.rs
+++ b/axum/src/routing/tests/mod.rs
@@ -645,3 +645,36 @@ async fn head_content_length_through_hyper_server_that_hits_fallback() {
let res = client.head("/").send().await;
assert_eq!(res.headers()["co... | Incorrect `content-length` header in reply to HEAD request when chunked transfer enconding is used
## Bug Report
This is a followup to #730. Unfortunately, I found a case in which it still does not work correctly.
### Version
axum main branch, commit 16ee83a
### Platform
Debian Linux 12 (“bookworm”)
Lin... | 2022-02-12T16:28:48Z | 0.4 | 2022-02-14T09:56:09Z | 2e5d56a9b125142817378883b098470d05ab0597 | [
"routing::tests::head_with_middleware_applied"
] | [
"body::stream_body::stream_body_traits",
"error_handling::traits",
"extract::connect_info::traits",
"extract::extractor_middleware::traits",
"extract::path::de::tests::test_parse_error_at_key_error",
"extract::path::de::tests::test_parse_error_at_index_error",
"extract::form::tests::test_form_query",
... | [] | [] | auto_2025-06-06 | |
tokio-rs/axum | 682 | tokio-rs__axum-682 | [
"681"
] | 5512ebcd23d261a2be5ee07676721b7637ffc054 | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -9,8 +9,12 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- **fixed:** Fix using incorrect path prefix when nesting `Router`s at `/` ([#691])
- **fixed:** Make `nest("", service... | diff --git a/axum/src/routing/tests/mod.rs b/axum/src/routing/tests/mod.rs
--- a/axum/src/routing/tests/mod.rs
+++ b/axum/src/routing/tests/mod.rs
@@ -354,6 +354,30 @@ async fn with_and_without_trailing_slash() {
assert_eq!(res.text().await, "without tsr");
}
+// for https://github.com/tokio-rs/axum/issues/681
... | Use HTTP 308 instead of 301 for trailing slash redirects
## Feature Request
### Motivation
Performing a POST to `/some/url/` (with the trailing slash) does an HTTP 301 (introduced in #410) to `/some/url`. However, HTTP 301 transforms a POST to a GET, which is undesired.
### Proposal
Use [HTTP 308](https://d... | 2022-01-01T02:57:12Z | 0.4 | 2022-01-12T14:58:13Z | 2e5d56a9b125142817378883b098470d05ab0597 | [
"routing::tests::with_trailing_slash_post",
"routing::tests::without_trailing_slash_post",
"routing::tests::wildcard_with_trailing_slash"
] | [
"error_handling::traits",
"body::stream_body::stream_body_traits",
"extract::connect_info::traits",
"extract::extractor_middleware::traits",
"extract::path::de::tests::test_parse_error_at_key_error",
"extract::path::de::tests::test_unsupported_type_error_nested_data_structure",
"extract::path::de::tests... | [] | [] | auto_2025-06-06 | |
tokio-rs/axum | 644 | tokio-rs__axum-644 | [
"637"
] | 4df4e07e10b13d05413c1da1a56d8f56ddbfb6ab | diff --git a/axum-core/CHANGELOG.md b/axum-core/CHANGELOG.md
--- a/axum-core/CHANGELOG.md
+++ b/axum-core/CHANGELOG.md
@@ -7,7 +7,36 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
# Unreleased
-- None.
+- **breaking:** Using `HeaderMap` as an extractor will no longer remove the... | diff --git a/axum-core/src/extract/request_parts.rs b/axum-core/src/extract/request_parts.rs
--- a/axum-core/src/extract/request_parts.rs
+++ b/axum-core/src/extract/request_parts.rs
@@ -76,27 +65,20 @@ where
}
}
+/// Clone the headers from the request.
+///
+/// Prefer using [`TypedHeader`] to extract only the... | Change return type of `Field::content_type`
It should return some kind of string, or a type axum owns, to avoid the public dependency on mime.
`Mime` implements `AsRef<str>` meaning we can probably return `&str` but should verify that it returns the whole mime type and not just a part.
| Why do you prefer a string over `mime`'s enum types? Personally I like being able to `match` on valid mime types.
It looks like [`mime` has been unmaintained for a while](https://github.com/hyperium/mime/issues/135) (and will likely continue to be). I'm not sure what a good alternative is though. [`http-types` has m... | 2021-12-21T09:47:37Z | 0.4 | 2022-06-11T11:06:23Z | 2e5d56a9b125142817378883b098470d05ab0597 | [
"response::headers::tests::with_body",
"body::test_try_downcast",
"response::headers::tests::with_status_and_body",
"response::headers::tests::vec_of_strings",
"response::headers::tests::vec_of_header_name_and_value",
"response::headers::tests::invalid_header_name",
"response::headers::tests::invalid_he... | [] | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 588 | tokio-rs__axum-588 | [
"587"
] | 628921bcad878ab4396705e0798494fbc9ebd3c5 | diff --git a/axum-debug/CHANGELOG.md b/axum-debug/CHANGELOG.md
--- a/axum-debug/CHANGELOG.md
+++ b/axum-debug/CHANGELOG.md
@@ -7,7 +7,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
# Unreleased
-- None.
+- Fix `Result<impl IntoResponse, Error>` generating invalid code ([#588]... | diff --git /dev/null b/axum-debug/tests/pass/result_impl_into_response.rs
new file mode 100644
--- /dev/null
+++ b/axum-debug/tests/pass/result_impl_into_response.rs
@@ -0,0 +1,132 @@
+use axum::{
+ async_trait,
+ extract::{FromRequest, RequestParts},
+ response::IntoResponse,
+};
+use axum_debug::debug_handle... | `#[debug_handler]` doesn't work with `Result<impl IntoResponse, SomeError>`
This produces code that doesn't compile
```rust
#[debug_handler]
async fn handler() -> Result<impl IntoResponse, Infallible> {
Ok(())
}
```
Error is:
```
error[E0720]: cannot resolve opaque type
--> examples/hello-world/sr... | 2021-12-04T14:46:46Z | 0.4 | 2021-12-04T17:54:11Z | 2e5d56a9b125142817378883b098470d05ab0597 | [
"tests/pass/result_impl_into_response.rs [should pass]",
"tests/pass/self_receiver.rs [should pass]",
"ui"
] | [
"tests/fail/argument_not_extractor.rs [should fail to compile]",
"tests/fail/attrs.rs [should fail to compile]",
"tests/fail/extract_self_mut.rs [should fail to compile]",
"tests/fail/extract_self_ref.rs [should fail to compile]",
"tests/fail/generics.rs [should fail to compile]",
"tests/fail/not_a_functi... | [] | [] | auto_2025-06-06 | |
tokio-rs/axum | 533 | tokio-rs__axum-533 | [
"528"
] | 939995e80edaf4e882a6e2decb07db5a7c0a2c06 | diff --git a/axum-handle-error-extract/src/lib.rs b/axum-handle-error-extract/src/lib.rs
--- a/axum-handle-error-extract/src/lib.rs
+++ b/axum-handle-error-extract/src/lib.rs
@@ -281,8 +281,8 @@ where
let future = Box::pin(async move {
match inner.oneshot(req).await {
- Ok(res) =>... | diff --git a/axum-handle-error-extract/src/lib.rs b/axum-handle-error-extract/src/lib.rs
--- a/axum-handle-error-extract/src/lib.rs
+++ b/axum-handle-error-extract/src/lib.rs
@@ -138,7 +138,7 @@
#![cfg_attr(test, allow(clippy::float_cmp))]
use axum::{
- body::{box_body, BoxBody, Bytes, Full, HttpBody},
+ body... | Rename `box_body` to `boxed`
As its consistent with `http_body::Body::boxed`.
Will be deprecated in 0.3.4 and removed in 0.4
| 2021-11-17T12:44:21Z | 0.3 | 2021-11-17T12:59:30Z | 939995e80edaf4e882a6e2decb07db5a7c0a2c06 | [
"src/lib.rs - (line 7)",
"src/lib.rs - (line 54)",
"body::stream_body::stream_body_traits",
"error_handling::traits",
"extract::connect_info::traits",
"extract::extractor_middleware::traits",
"extract::path::de::tests::test_parse_struct",
"extract::form::tests::test_form_query",
"extract::path::de::... | [] | [] | [] | auto_2025-06-06 | |
tokio-rs/axum | 530 | tokio-rs__axum-530 | [
"528"
] | b9dfea636071c2c194c8251612beba9d0515e9fe | diff --git a/axum-handle-error-extract/src/lib.rs b/axum-handle-error-extract/src/lib.rs
--- a/axum-handle-error-extract/src/lib.rs
+++ b/axum-handle-error-extract/src/lib.rs
@@ -281,8 +281,8 @@ where
let future = Box::pin(async move {
match inner.oneshot(req).await {
- Ok(res) =>... | diff --git a/axum-handle-error-extract/src/lib.rs b/axum-handle-error-extract/src/lib.rs
--- a/axum-handle-error-extract/src/lib.rs
+++ b/axum-handle-error-extract/src/lib.rs
@@ -138,7 +138,7 @@
#![cfg_attr(test, allow(clippy::float_cmp))]
use axum::{
- body::{box_body, BoxBody, Bytes, Full, HttpBody},
+ body... | Rename `box_body` to `boxed`
As its consistent with `http_body::Body::boxed`.
Will be deprecated in 0.3.4 and removed in 0.4
| 2021-11-17T08:54:59Z | 0.3 | 2021-11-17T12:31:11Z | 939995e80edaf4e882a6e2decb07db5a7c0a2c06 | [
"src/lib.rs - (line 54)",
"src/lib.rs - (line 7)",
"body::stream_body::stream_body_traits",
"error_handling::traits",
"extract::connect_info::traits",
"extract::extractor_middleware::traits",
"extract::path::de::tests::test_parse_map",
"extract::path::de::tests::test_parse_single_value",
"extract::f... | [] | [] | [] | auto_2025-06-06 | |
tokio-rs/axum | 529 | tokio-rs__axum-529 | [
"488"
] | f9a437d0813c0e2f784ca90eccccb4fbad9126cb | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -28,10 +28,15 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- **breaking:** The `Handler<B, T>` trait is now defined as `Handler<T, B =
Body>`. That is the type parameters have... | diff --git a/axum/src/routing/tests/fallback.rs b/axum/src/routing/tests/fallback.rs
--- a/axum/src/routing/tests/fallback.rs
+++ b/axum/src/routing/tests/fallback.rs
@@ -49,25 +49,3 @@ async fn or() {
assert_eq!(res.status(), StatusCode::OK);
assert_eq!(res.text().await, "fallback");
}
-
-#[tokio::test]
-as... | Panic if merging two routers that each have a fallback
Currently when you `Router::merge` two routes that each have a fallback it'll pick the fallback of the router on the right hand side, and if only one has a fallback it'll pick that one. This might lead users to think that multiple fallbacks are supported (see https... | 👍 for panicking. Silently discarding a fallback service will almost definitely end up causing issue for someone. | 2021-11-16T22:44:32Z | 0.3 | 2021-11-18T21:19:07Z | 939995e80edaf4e882a6e2decb07db5a7c0a2c06 | [
"routing::tests::merging_routers_with_fallbacks_panics - should panic",
"routing::tests::nesting_router_with_fallbacks_panics - should panic"
] | [
"error_handling::traits",
"body::stream_body::stream_body_traits",
"extract::connect_info::traits",
"extract::extractor_middleware::traits",
"extract::path::de::tests::test_parse_struct",
"extract::form::tests::test_form_query",
"extract::path::de::tests::test_parse_map",
"extract::form::tests::test_i... | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 423 | tokio-rs__axum-423 | [
"419"
] | 8fe4eaf1d5a17fc1919afc64f7c2e8c890069653 | diff --git a/src/routing/mod.rs b/src/routing/mod.rs
--- a/src/routing/mod.rs
+++ b/src/routing/mod.rs
@@ -440,8 +440,7 @@ where
LayeredResBody::Error: Into<BoxError>,
{
let layer = ServiceBuilder::new()
- .layer_fn(Route)
- .layer_fn(CloneBoxService::new)
+ .laye... | diff --git a/src/tests/mod.rs b/src/tests/mod.rs
--- a/src/tests/mod.rs
+++ b/src/tests/mod.rs
@@ -684,8 +684,6 @@ async fn middleware_still_run_for_unmatched_requests() {
assert_eq!(COUNT.load(Ordering::SeqCst), 2);
}
-// TODO(david): middleware still run for empty routers
-
pub(crate) fn assert_send<T: Send>... | First parameterized path component stripped from logs
## Bug Report
<!--
Thank you for reporting an issue.
Please fill in as much of the template below as you're able.
-->
### Version
<!--
List the versions of all `axum` crates you are using. The easiest way to get
this information is using `cargo tre... | 2021-10-26T17:06:24Z | 0.3 | 2021-10-26T17:30:00Z | 939995e80edaf4e882a6e2decb07db5a7c0a2c06 | [
"tests::nest::outer_middleware_still_see_whole_url"
] | [
"error_handling::traits",
"body::stream_body::stream_body_traits",
"extract::extractor_middleware::traits",
"extract::connect_info::traits",
"extract::path::de::tests::test_parse_struct",
"extract::path::de::tests::test_parse_map",
"extract::path::de::tests::test_parse_single_value",
"extract::form::t... | [] | [] | auto_2025-06-06 | |
tokio-rs/axum | 412 | tokio-rs__axum-412 | [
"386"
] | e43bdf0ecf3ad8a8de7694658f73143e1ca066fc | diff --git a/CHANGELOG.md b/CHANGELOG.md
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -136,6 +136,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
without trailing a slash.
- **breaking:** `EmptyRouter` has been renamed to `MethodNotAllowed` as its only
used in method routers... | diff --git a/src/tests/mod.rs b/src/tests/mod.rs
--- a/src/tests/mod.rs
+++ b/src/tests/mod.rs
@@ -1,6 +1,7 @@
#![allow(clippy::blacklisted_name)]
use crate::error_handling::HandleErrorLayer;
+use crate::extract::MatchedPath;
use crate::BoxError;
use crate::{
extract::{self, Path},
diff --git a/src/tests/mod... | Allow access to matched route pattern(s) from middlewares
## Feature Request
As far as I can tell it's not possible to access information about if a route matched and if so, which route pattern (or patterns in case of nested routes) was matched. This is something a lot of services in environments I've worked needs.
... | Actually I'm not sure its possible to fix. Any middleware you apply run before the router sees the request. So if the router added the matched pattern to a request extension middleware wouldn't see the extension because they run _before_ the router. You'd only be able to see the extension in handlers but that seems les... | 2021-10-25T21:34:02Z | 0.3 | 2021-10-25T21:38:30Z | 939995e80edaf4e882a6e2decb07db5a7c0a2c06 | [
"body::stream_body::stream_body_traits",
"error_handling::traits",
"extract::connect_info::traits",
"extract::extractor_middleware::traits",
"extract::path::de::tests::test_parse_struct",
"extract::form::tests::test_form_query",
"extract::path::de::tests::test_parse_map",
"extract::form::tests::test_i... | [] | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 1,189 | tokio-rs__axum-1189 | [
"1175"
] | 7e7a2f2058690e8286187c1bd7ac999dcfa506b6 | diff --git /dev/null b/examples/handle-head-request/Cargo.toml
new file mode 100644
--- /dev/null
+++ b/examples/handle-head-request/Cargo.toml
@@ -0,0 +1,13 @@
+[package]
+name = "example-handle-head-request"
+version = "0.1.0"
+edition = "2021"
+publish = false
+
+[dependencies]
+axum = { path = "../../axum" }
+tokio... | diff --git /dev/null b/examples/handle-head-request/src/main.rs
new file mode 100644
--- /dev/null
+++ b/examples/handle-head-request/src/main.rs
@@ -0,0 +1,82 @@
+//! Run with
+//!
+//! ```not_rust
+//! cd examples && cargo run -p example-handle-head-request
+//! ```
+
+use axum::response::{IntoResponse, Response};
+u... | Add an example for a GET-or-HEAD handler that special-cases HEAD
Registering a `GET` route also registers a `HEAD` route that calls the same handler. axum will strip the response body if this handler is called with `HEAD`, but that is potentially much slower than never generating a response body in the first place. We ... | I'd be happy to tackle this one as well, though I understand if you want to reserve an issue for the upcoming TWiR and new contributors.
Maybe we wait a few days and if nobody else shows interest you do it?
Hi, @jplatte
I found this from TWIR.
I am interested in contributing, let me grab this.
Thanks. | 2022-07-23T11:10:28Z | 0.5 | 2022-07-27T15:41:40Z | fa51cf52664d2e33fe37ca4252879bb1c73e87e9 | [
"tests::test_something",
"tests::not_found",
"tests::hello_world",
"tests::json",
"tests::multiple_request",
"tests::the_real_deal"
] | [] | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 957 | tokio-rs__axum-957 | [
"954"
] | 8084b242d5b78e53267e63c693dadf71e313320c | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -9,8 +9,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- **added:** Add `response::ErrorResponse` and `response::Result` for
`IntoResponse`-based error handling ([#921])
+-... | diff --git a/axum/src/extract/extractor_middleware.rs b/axum/src/extract/extractor_middleware.rs
--- a/axum/src/extract/extractor_middleware.rs
+++ b/axum/src/extract/extractor_middleware.rs
@@ -2,324 +2,15 @@
//!
//! See [`extractor_middleware`] for more details.
-use super::{FromRequest, RequestParts};
-use crate... | Move `axum::extract::extractor_middleware` to `axum::middleware::from_extractor`
## Feature Request
### Motivation
<!--
Please describe the use case(s) or other motivation for the new feature.
-->
### Proposal
<!--
How should the new feature be implemented, and why? Add any considered
drawbacks.
-->
... | Sounds good to me! We didn't use to have the middleware module but that is probably a better home for it.
You up for making a PR? We just have to make its not a breaking change by re-exporting it from the old location.
Also, next time please fill in the issue template. Thats what its there for.
If we want to depr... | 2022-04-22T14:54:04Z | 0.5 | 2022-04-24T15:31:17Z | fa51cf52664d2e33fe37ca4252879bb1c73e87e9 | [
"body::stream_body::stream_body_traits",
"error_handling::traits",
"extract::connect_info::traits",
"extract::path::de::tests::test_parse_error_at_index_error",
"extract::form::tests::test_form_query",
"extract::form::tests::test_incorrect_content_type",
"extract::path::de::tests::test_parse_error_at_ke... | [] | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 892 | tokio-rs__axum-892 | [
"859"
] | 2e5d56a9b125142817378883b098470d05ab0597 | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -88,6 +88,9 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- **changed:** Update to tokio-tungstenite 0.17 ([#791])
- **breaking:** `Redirect::{to, temporary, permanent}` now acce... | diff --git a/axum/src/extract/connect_info.rs b/axum/src/extract/connect_info.rs
--- a/axum/src/extract/connect_info.rs
+++ b/axum/src/extract/connect_info.rs
@@ -161,7 +161,7 @@ mod tests {
let app = Router::new().route("/", get(handler));
let server = Server::from_tcp(listener)
... | `Router::into_make_service_with_connect_info()` has unnecessary `Target` parameter that can fail to be inferred
## Bug Report
### Version
axum 0.4.8
### Crates
axum
### Description
[`Router::into_make_service_with_connect_info()`](https://docs.rs/axum/0.4.8/axum/struct.Router.html#method.into_make_ser... | > As soon as you have one C that implements Connected for multiple targets, that parameter needs to be specified explicitly. This is very annoying.
What is your use case for this?
> It probably does produce a better error message when the wrong type is used, but having to write out the connection type is frustrat... | 2022-03-31T16:37:01Z | 0.4 | 2022-03-31T16:49:50Z | 2e5d56a9b125142817378883b098470d05ab0597 | [
"error_handling::traits",
"body::stream_body::stream_body_traits",
"extract::connect_info::traits",
"extract::extractor_middleware::traits",
"extract::form::tests::test_form_query",
"extract::form::tests::test_form_body",
"extract::form::tests::test_incorrect_content_type",
"extract::path::de::tests::... | [] | [] | [] | auto_2025-06-06 |
tokio-rs/axum | 2,881 | tokio-rs__axum-2881 | [
"2879"
] | bd3eaa78866127bb9106972f55190cf9fcf3305c | diff --git a/axum/CHANGELOG.md b/axum/CHANGELOG.md
--- a/axum/CHANGELOG.md
+++ b/axum/CHANGELOG.md
@@ -11,10 +11,12 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
- **added:** `axum::serve::Serve::tcp_nodelay` and `axum::serve::WithGracefulShutdown::tcp_nodelay` ([#2653])
- **adde... | diff --git a/axum/src/serve.rs b/axum/src/serve.rs
--- a/axum/src/serve.rs
+++ b/axum/src/serve.rs
@@ -544,6 +554,10 @@ mod tests {
routing::get,
Router,
};
+ use std::{
+ future::pending,
+ net::{IpAddr, Ipv4Addr},
+ };
#[allow(dead_code, unused_must_use)]
async f... | add means to find the socket address for Serve and WithGracefulShutdown
- [*] I have looked for existing issues (including closed) about this
## Feature Request
I would like to get the socket address of `Serve` and `WithGracefulShutdown`.
### Motivation
I am the maintainer of Axum Test, and the initial moti... | 2024-08-18T13:32:04Z | 0.7 | 2024-08-25T09:16:13Z | 4179d11797bd8880c74a306b7e05e9d0f081ca1d | [
"error_handling::traits",
"extract::connect_info::traits",
"extract::host::tests::forwarded_parsing",
"extract::path::de::tests::test_parse_error_at_index_error",
"extract::path::de::tests::test_parse_error_at_key_error",
"extract::path::de::tests::test_parse_error_error",
"extract::path::de::tests::tes... | [] | [] | [] | auto_2025-06-06 | |
tokio-rs/axum | 2,645 | tokio-rs__axum-2645 | [
"2641"
] | 5db62e8452f3c34bab83f3fdfacffa1946b91af1 | diff --git a/axum-extra/src/extract/optional_path.rs b/axum-extra/src/extract/optional_path.rs
--- a/axum-extra/src/extract/optional_path.rs
+++ b/axum-extra/src/extract/optional_path.rs
@@ -28,7 +28,7 @@ use serde::de::DeserializeOwned;
///
/// let app = Router::new()
/// .route("/blog", get(render_blog))
-/// ... | diff --git a/axum-extra/src/extract/optional_path.rs b/axum-extra/src/extract/optional_path.rs
--- a/axum-extra/src/extract/optional_path.rs
+++ b/axum-extra/src/extract/optional_path.rs
@@ -75,7 +75,7 @@ mod tests {
let app = Router::new()
.route("/", get(handle))
- .route("/:num", g... | Update `matchit` to 0.8
- [x] I have looked for existing issues (including closed) about this
### Proposal
`matchit` version 0.8 was just [released](https://github.com/ibraheemdev/matchit/releases/tag/v0.8.0).
It includes changes to the syntax so now instead of `/path/:arg/*wildcard` it uses `/path/{arg}/{*wil... | 2024-03-12T23:28:55Z | 0.7 | 2024-10-10T20:46:34Z | 4179d11797bd8880c74a306b7e05e9d0f081ca1d | [
"tests/typed_path/pass/customize_rejection.rs [should pass]",
"tests/typed_path/pass/into_uri.rs [should pass]",
"tests/typed_path/pass/named_fields_struct.rs [should pass]",
"tests/typed_path/pass/option_result.rs [should pass]",
"tests/typed_path/pass/tuple_struct.rs [should pass]",
"tests/typed_path/pa... | [
"tests/debug_middleware/fail/doesnt_take_next.rs [should fail to compile]",
"tests/debug_middleware/fail/next_not_last.rs [should fail to compile]",
"tests/debug_middleware/fail/takes_next_twice.rs [should fail to compile]",
"tests/debug_middleware/pass/basic.rs [should pass]",
"debug_handler::ui_debug_midd... | [] | [] | auto_2025-06-06 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.