kemitix/paperoni
1
0
Fork 0
Code Issues 4 Pull requests 3 Projects Releases Packages Wiki Activity

Merge pull request #12 from kxt/11-image-status-codes

Browse source 
Check response status for fetched images
This commit is contained in:
Kenneth Gitere 2021-05-10 10:58:23 +03:00 committed by GitHub
commit 312dff95e2
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 70 additions and 61 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

107
src/http.rs
View file

@ -72,6 +72,60 @@ pub async fn fetch_html(url: &str) -> Result<HTMLResource, PaperoniError> {
})
}
type ImgItem<'a> = (&'a str, String, Option<String>);
async fn process_img_response<'a>(
img_response: &mut surf::Response,
url: &'a str,
) -> Result<ImgItem<'a>, ImgError> {
if !img_response.status().is_success() {
let kind = ErrorKind::HTTPError(format!(
"Non-success HTTP status code ({})",
img_response.status()
));
return Err(ImgError::with_kind(kind));
}
let img_content: Vec<u8> = match img_response.body_bytes().await {
Ok(bytes) => bytes,
Err(e) => return Err(e.into()),
};
let img_mime = img_response
.content_type()
.map(|mime| mime.essence().to_string());
let img_ext = match img_response
.content_type()
.map(|mime| map_mime_subtype_to_ext(mime.subtype()).to_string())
{
Some(mime_str) => mime_str,
None => return Err(ErrorKind::HTTPError("Image has no Content-Type".to_owned()).into()),
};
let mut img_path = std::env::temp_dir();
img_path.push(format!("{}.{}", hash_url(url), &img_ext));
let mut img_file = match File::create(&img_path).await {
Ok(file) => file,
Err(e) => return Err(e.into()),
};
match img_file.write_all(&img_content).await {
Ok(_) => (),
Err(e) => return Err(e.into()),
}
Ok((
url,
img_path
.file_name()
.map(|os_str_name| {
os_str_name
.to_str()
.expect("Unable to get image file name")
.to_string()
})
.unwrap(),
img_mime,
))
}
pub async fn download_images(
extractor: &mut Extractor,
article_origin: &Url,
@ -102,53 +156,9 @@ pub async fn download_images(
bar.set_message(format!("Downloading images [{}/{}]", img_idx + 1, img_count).as_str());
match req.await {
Ok(mut img_response) => {
let process_response = async {
let img_content: Vec<u8> = match img_response.body_bytes().await {
Ok(bytes) => bytes,
Err(e) => return Err(e.into()),
};
let img_mime = img_response
.content_type()
.map(|mime| mime.essence().to_string());
let img_ext = match img_response
.content_type()
.map(|mime| map_mime_subtype_to_ext(mime.subtype()).to_string())
{
Some(mime_str) => mime_str,
None => {
return Err(ErrorKind::HTTPError(
"Image has no Content-Type".to_owned(),
)
.into())
}
};
let mut img_path = std::env::temp_dir();
img_path.push(format!("{}.{}", hash_url(&url), &img_ext));
let mut img_file = match File::create(&img_path).await {
Ok(file) => file,
Err(e) => return Err(e.into()),
};
match img_file.write_all(&img_content).await {
Ok(_) => (),
Err(e) => return Err(e.into()),
}
Ok((
url,
img_path
.file_name()
.map(|os_str_name| {
os_str_name
.to_str()
.expect("Unable to get image file name")
.to_string()
})
.unwrap(),
img_mime,
))
};
process_response.await.map_err(|mut e: ImgError| {
let process_response =
process_img_response(&mut img_response, url.as_ref()).await;
process_response.map_err(|mut e: ImgError| {
e.set_url(url);
e
})
@ -162,8 +172,7 @@ pub async fn download_images(
});
// A utility closure used when update the value of an image source after downloading is successful
let replace_existing_img_src =
|img_item: (&String, String, Option<String>)| -> (String, Option<String>) {
let replace_existing_img_src = |img_item: ImgItem| -> (String, Option<String>) {
let (img_url, img_path, img_mime) = img_item;
let img_ref = extractor
.article()