misc-mirrors/article_scraper
1
0
Fork 0
mirror of https://gitlab.com/news-flash/article_scraper.git synced 2025-07-07 16:15:32 +02:00
Code Releases Activity

fmt

Browse source
This commit is contained in:
Jan Lukas Gernert 2023-04-26 09:12:55 +02:00
parent fbb6585596
commit 5621a0ea54
1 changed files with 10 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

12
article_scraper/src/full_text_parser/mod.rs
View file

@ -381,7 +381,11 @@ impl FullTextParser {
if let Some(encoding) = Self::get_encoding_from_html(&lossy_string) { if let Some(encoding) = Self::get_encoding_from_html(&lossy_string) {
log::debug!("Encoding extracted from HTML: '{}'", encoding); log::debug!("Encoding extracted from HTML: '{}'", encoding);
if let Some(decoded_html) = Self::decode_html(&bytes, encoding) { if let Some(decoded_html) = Self::decode_html(&bytes, encoding) {
let decoded_html = decoded_html.replacen(&format!("charset=\"{encoding}\""), "charset=\"utf-8\"", 1); let decoded_html = decoded_html.replacen(
&format!("charset=\"{encoding}\""),
"charset=\"utf-8\"",
1,
);
return Ok(decoded_html); return Ok(decoded_html);
} }
} }
@ -389,7 +393,11 @@ impl FullTextParser {
if let Some(encoding) = Self::get_encoding_from_http_header(&headers) { if let Some(encoding) = Self::get_encoding_from_http_header(&headers) {
log::debug!("Encoding extracted from headers: '{}'", encoding); log::debug!("Encoding extracted from headers: '{}'", encoding);
if let Some(decoded_html) = Self::decode_html(&bytes, encoding) { if let Some(decoded_html) = Self::decode_html(&bytes, encoding) {
let decoded_html = decoded_html.replacen(&format!("charset=\"{encoding}\""), "charset=\"utf-8\"", 1); let decoded_html = decoded_html.replacen(
&format!("charset=\"{encoding}\""),
"charset=\"utf-8\"",
1,
);
return Ok(decoded_html); return Ok(decoded_html);
} }
} }