[Web] Always return -1 as body length in HTTPClientWeb.

Body length cannot be reliably retrieved from the web.
Reading the "content-length" value will return a meaningless value when
the response is compressed, as reading will return uncompressed chunks
in any case, resulting in a mismatch between the detected body size and
the actual size returned by repeatedly calling read_response_body_chunk.

Additionally, while "content-length" is considered a safe CORS header,
"content-encoding" is not, so using the "content-encoding" to decide if
"content-length" is meaningful is not an option either.

We simply must accept the fact that browsers are awful when it comes to
networking APIs.

(cherry picked from commit f4713d235a)
This commit is contained in:
Fabio Alessandrelli 2023-07-24 08:32:44 +02:00 committed by Yuri Sizov
parent f2c8eea60e
commit 5eabd5e04a
4 changed files with 10 additions and 17 deletions

View File

@ -41,6 +41,7 @@
<description>
Returns the response's body length.
[b]Note:[/b] Some Web servers may not send a body length. In this case, the value returned will be [code]-1[/code]. If using chunked transfer encoding, the body length will also be [code]-1[/code].
[b]Note:[/b] This function always returns [code]-1[/code] on the Web platform due to browsers limitations.
</description>
</method>
<method name="get_response_code" qualifiers="const">

View File

@ -149,7 +149,15 @@ Error HTTPClientWeb::get_response_headers(List<String> *r_response) {
}
int64_t HTTPClientWeb::get_response_body_length() const {
return godot_js_fetch_body_length_get(js_id);
// Body length cannot be consistently retrieved from the web.
// Reading the "content-length" value will return a meaningless value when the response is compressed,
// as reading will return uncompressed chunks in any case, resulting in a mismatch between the detected
// body size and the actual size returned by repeatedly calling read_response_body_chunk.
// Additionally, while "content-length" is considered a safe CORS header, "content-encoding" is not,
// so using the "content-encoding" to decide if "content-length" is meaningful is not an option either.
// We simply must accept the fact that browsers are awful when it comes to networking APIs.
// See GH-47597, and GH-79327.
return -1;
}
PackedByteArray HTTPClientWeb::read_response_body_chunk() {

View File

@ -51,7 +51,6 @@ extern int godot_js_fetch_read_headers(int p_id, void (*parse_callback)(int p_si
extern int godot_js_fetch_read_chunk(int p_id, uint8_t *p_buf, int p_buf_size);
extern void godot_js_fetch_free(int p_id);
extern godot_js_fetch_state_t godot_js_fetch_state_get(int p_id);
extern int godot_js_fetch_body_length_get(int p_id);
extern int godot_js_fetch_http_status_get(int p_id);
extern int godot_js_fetch_is_chunked(int p_id);

View File

@ -50,22 +50,17 @@ const GodotFetch = {
return;
}
let chunked = false;
let bodySize = -1;
response.headers.forEach(function (value, header) {
const v = value.toLowerCase().trim();
const h = header.toLowerCase().trim();
if (h === 'transfer-encoding' && v === 'chunked') {
chunked = true;
}
if (h === 'content-length') {
bodySize = parseInt(v, 10);
}
});
obj.status = response.status;
obj.response = response;
obj.reader = response.body.getReader();
obj.chunked = chunked;
obj.bodySize = bodySize;
},
onerror: function (id, err) {
@ -87,7 +82,6 @@ const GodotFetch = {
reading: false,
status: 0,
chunks: [],
bodySize: -1,
};
const id = IDHandler.add(obj);
const init = {
@ -224,15 +218,6 @@ const GodotFetch = {
return p_buf_size - to_read;
},
godot_js_fetch_body_length_get__sig: 'ii',
godot_js_fetch_body_length_get: function (p_id) {
const obj = IDHandler.get(p_id);
if (!obj || !obj.response) {
return -1;
}
return obj.bodySize;
},
godot_js_fetch_is_chunked__sig: 'ii',
godot_js_fetch_is_chunked: function (p_id) {
const obj = IDHandler.get(p_id);