diff options
| author |  Alexandre Alapetite <alexandre@alapetite.fr>
| 2025-06-25 16:37:41 +0200 |
|---|---|---|
| committer |  GitHub <noreply@github.com>
| 2025-06-25 16:37:41 +0200 |
| commit | 7c57f38008136202ba7b38e3154ac87be4eefb68 (patch) | |
| tree | 8fb1c8856579e3f1ae880ae5d3e64ccc18339572 /lib/simplepie | |
| parent | 42aeecb5ed6f7e89c795dca0f96ba6497f3e349e (diff) | |
Fix feeds encoded in UTF-16LE (#7691)
* Fix feeds encoded in UTF-16LE
Fix https://github.com/FreshRSS/FreshRSS/issues/7690
https://github.com/FreshRSS/simplepie/pull/40
The final character `>` of a feed is encoded as `3E00` in UTF-16LE, so calling `trim()` was removing the `\x00`, breaking the multibyte encoding and making the feed invalid.
Upstream PR https://github.com/simplepie/simplepie/pull/916
* Trim body for all paths
https://github.com/FreshRSS/simplepie/pull/42
https://github.com/simplepie/simplepie/pull/917
Slight refactor of https://github.com/simplepie/simplepie/pull/916 (https://github.com/FreshRSS/simplepie/pull/40) to cover all paths.
Missing paths included the fsock method without gzip (e.g. deflate or plain).
Diffstat (limited to 'lib/simplepie')
| -rw-r--r-- | lib/simplepie/simplepie/src/File.php | 11 |
1 files changed, 8 insertions, 3 deletions
diff --git a/lib/simplepie/simplepie/src/File.php b/lib/simplepie/simplepie/src/File.php index 6f5f37c3c..a6a444f56 100644 --- a/lib/simplepie/simplepie/src/File.php +++ b/lib/simplepie/simplepie/src/File.php @@ -162,7 +162,7 @@ class File implements Response $parser = new \SimplePie\HTTP\Parser($responseHeaders, true); if ($parser->parse()) { $this->set_headers($parser->headers); - $this->body = trim($parser->body); + $this->body = $parser->body; $this->status_code = $parser->status_code; if ((in_array($this->status_code, [300, 301, 302, 303, 307]) || $this->status_code > 307 && $this->status_code < 400) && ($locationHeader = $this->get_header_line('location')) !== '' && $this->redirects < $redirects) { $this->redirects++; @@ -246,7 +246,7 @@ class File implements Response $this->error = 'Unable to decode HTTP "gzip" stream'; $this->success = false; } else { - $this->body = trim($decompressed); + $this->body = $decompressed; } break; @@ -288,11 +288,16 @@ class File implements Response $this->error = sprintf('file "%s" is not readable', $url); $this->success = false; } else { - $this->body = trim($filebody); + $this->body = $filebody; $this->status_code = 200; } $this->on_http_response(); } + if ($this->success) { + // (Leading) whitespace may cause XML parsing errors so we trim it, + // but we must not trim \x00 to avoid breaking BOM or multibyte characters + $this->body = trim($this->body, " \n\r\t\v"); + } } /** |