Add an option to fetch and parse href="data:image"

Some sites are using base64 encoded inline images for favicons.
This will try to match those with some sane checks and return that.
These icons will have lower prio then the icons with a normal URL.
This commit is contained in:
BlackDex 2019-11-22 13:16:12 +01:00
parent cbadf00941
commit b209c1bc4d
3 changed files with 54 additions and 17 deletions

32
Cargo.lock generated
View File

@ -89,6 +89,7 @@ dependencies = [
"chashmap 2.2.2 (registry+https://github.com/rust-lang/crates.io-index)",
"chrono 0.4.9 (registry+https://github.com/rust-lang/crates.io-index)",
"data-encoding 2.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
"data-url 0.1.0 (git+https://github.com/servo/rust-url?rev=7f1bd6ce1c2fde599a757302a843a60e714c5f72)",
"derive_more 0.99.2 (registry+https://github.com/rust-lang/crates.io-index)",
"diesel 1.4.3 (registry+https://github.com/rust-lang/crates.io-index)",
"diesel_migrations 1.4.0 (registry+https://github.com/rust-lang/crates.io-index)",
@ -355,6 +356,14 @@ name = "data-encoding"
version = "2.1.2"
source = "registry+https://github.com/rust-lang/crates.io-index"
[[package]]
name = "data-url"
version = "0.1.0"
source = "git+https://github.com/servo/rust-url?rev=7f1bd6ce1c2fde599a757302a843a60e714c5f72#7f1bd6ce1c2fde599a757302a843a60e714c5f72"
dependencies = [
"matches 0.1.8 (registry+https://github.com/rust-lang/crates.io-index)",
]
[[package]]
name = "derive_more"
version = "0.99.2"
@ -925,7 +934,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
dependencies = [
"matches 0.1.8 (registry+https://github.com/rust-lang/crates.io-index)",
"unicode-bidi 0.3.4 (registry+https://github.com/rust-lang/crates.io-index)",
"unicode-normalization 0.1.9 (registry+https://github.com/rust-lang/crates.io-index)",
"unicode-normalization 0.1.10 (registry+https://github.com/rust-lang/crates.io-index)",
]
[[package]]
@ -935,7 +944,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
dependencies = [
"matches 0.1.8 (registry+https://github.com/rust-lang/crates.io-index)",
"unicode-bidi 0.3.4 (registry+https://github.com/rust-lang/crates.io-index)",
"unicode-normalization 0.1.9 (registry+https://github.com/rust-lang/crates.io-index)",
"unicode-normalization 0.1.10 (registry+https://github.com/rust-lang/crates.io-index)",
]
[[package]]
@ -1281,7 +1290,7 @@ dependencies = [
"openssl-probe 0.1.2 (registry+https://github.com/rust-lang/crates.io-index)",
"openssl-sys 0.9.52 (registry+https://github.com/rust-lang/crates.io-index)",
"schannel 0.1.16 (registry+https://github.com/rust-lang/crates.io-index)",
"security-framework 0.3.3 (registry+https://github.com/rust-lang/crates.io-index)",
"security-framework 0.3.4 (registry+https://github.com/rust-lang/crates.io-index)",
"security-framework-sys 0.3.3 (registry+https://github.com/rust-lang/crates.io-index)",
"tempfile 3.1.0 (registry+https://github.com/rust-lang/crates.io-index)",
]
@ -2067,7 +2076,7 @@ dependencies = [
[[package]]
name = "security-framework"
version = "0.3.3"
version = "0.3.4"
source = "registry+https://github.com/rust-lang/crates.io-index"
dependencies = [
"core-foundation 0.6.4 (registry+https://github.com/rust-lang/crates.io-index)",
@ -2189,6 +2198,11 @@ dependencies = [
"maybe-uninit 2.0.0 (registry+https://github.com/rust-lang/crates.io-index)",
]
[[package]]
name = "smallvec"
version = "1.0.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
[[package]]
name = "soup"
version = "0.4.1"
@ -2584,10 +2598,10 @@ dependencies = [
[[package]]
name = "unicode-normalization"
version = "0.1.9"
version = "0.1.10"
source = "registry+https://github.com/rust-lang/crates.io-index"
dependencies = [
"smallvec 0.6.13 (registry+https://github.com/rust-lang/crates.io-index)",
"smallvec 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)",
]
[[package]]
@ -2850,6 +2864,7 @@ dependencies = [
"checksum crypto-mac 0.3.0 (registry+https://github.com/rust-lang/crates.io-index)" = "dba62c86c26dcba13c278afcaac0c7452486fe604a2668a0dfa4e0edc98d8a9e"
"checksum crypto-mac 0.7.0 (registry+https://github.com/rust-lang/crates.io-index)" = "4434400df11d95d556bac068ddfedd482915eb18fe8bea89bc80b6e4b1c179e5"
"checksum data-encoding 2.1.2 (registry+https://github.com/rust-lang/crates.io-index)" = "f4f47ca1860a761136924ddd2422ba77b2ea54fe8cc75b9040804a0d9d32ad97"
"checksum data-url 0.1.0 (git+https://github.com/servo/rust-url?rev=7f1bd6ce1c2fde599a757302a843a60e714c5f72)" = "<none>"
"checksum derive_more 0.99.2 (registry+https://github.com/rust-lang/crates.io-index)" = "2159be042979966de68315bce7034bb000c775f22e3e834e1c52ff78f041cae8"
"checksum devise 0.3.0 (git+https://github.com/SergioBenitez/Devise.git?rev=e58b3ac9a)" = "<none>"
"checksum devise_codegen 0.3.0 (git+https://github.com/SergioBenitez/Devise.git?rev=e58b3ac9a)" = "<none>"
@ -3034,7 +3049,7 @@ dependencies = [
"checksum scheduled-thread-pool 0.2.2 (registry+https://github.com/rust-lang/crates.io-index)" = "bd07742e081ff6c077f5f6b283f12f32b9e7cc765b316160d66289b74546fbb3"
"checksum scopeguard 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)" = "b42e15e59b18a828bbf5c58ea01debb36b9b096346de35d941dcb89009f24a0d"
"checksum sct 0.5.0 (registry+https://github.com/rust-lang/crates.io-index)" = "2f5adf8fbd58e1b1b52699dc8bed2630faecb6d8c7bee77d009d6bbe4af569b9"
"checksum security-framework 0.3.3 (registry+https://github.com/rust-lang/crates.io-index)" = "301c862a6d0ee78f124c5e1710205965fc5c553100dcda6d98f13ef87a763f04"
"checksum security-framework 0.3.4 (registry+https://github.com/rust-lang/crates.io-index)" = "8ef2429d7cefe5fd28bd1d2ed41c944547d4ff84776f5935b456da44593a16df"
"checksum security-framework-sys 0.3.3 (registry+https://github.com/rust-lang/crates.io-index)" = "e31493fc37615debb8c5090a7aeb4a9730bc61e77ab10b9af59f1a202284f895"
"checksum semver 0.9.0 (registry+https://github.com/rust-lang/crates.io-index)" = "1d7eb9ef2c18661902cc47e535f9bc51b78acd254da71d375c2f6720d9a40403"
"checksum semver-parser 0.7.0 (registry+https://github.com/rust-lang/crates.io-index)" = "388a1df253eca08550bef6c72392cfe7c30914bf41df5269b68cbd6ff8f570a3"
@ -3048,6 +3063,7 @@ dependencies = [
"checksum siphasher 0.2.3 (registry+https://github.com/rust-lang/crates.io-index)" = "0b8de496cf83d4ed58b6be86c3a275b8602f6ffe98d3024a869e124147a9a3ac"
"checksum slab 0.4.2 (registry+https://github.com/rust-lang/crates.io-index)" = "c111b5bd5695e56cffe5129854aa230b39c93a305372fdbb2668ca2394eea9f8"
"checksum smallvec 0.6.13 (registry+https://github.com/rust-lang/crates.io-index)" = "f7b0758c52e15a8b5e3691eae6cc559f08eee9406e548a4477ba4e67770a82b6"
"checksum smallvec 1.0.0 (registry+https://github.com/rust-lang/crates.io-index)" = "4ecf3b85f68e8abaa7555aa5abdb1153079387e60b718283d732f03897fcfc86"
"checksum soup 0.4.1 (registry+https://github.com/rust-lang/crates.io-index)" = "16eb6b0678654a57009598ed84610f2afa5fadb22f3815e9f23dc5eab1056031"
"checksum spin 0.5.2 (registry+https://github.com/rust-lang/crates.io-index)" = "6e63cff320ae2c57904679ba7cb63280a3dc4613885beafb148ee7bf9aa9042d"
"checksum stable_deref_trait 1.1.1 (registry+https://github.com/rust-lang/crates.io-index)" = "dba1a27d3efae4351c8051072d619e3ade2820635c3958d826bfea39d59b54c8"
@ -3091,7 +3107,7 @@ dependencies = [
"checksum unicase 1.4.2 (registry+https://github.com/rust-lang/crates.io-index)" = "7f4765f83163b74f957c797ad9253caf97f103fb064d3999aea9568d09fc8a33"
"checksum unicase 2.6.0 (registry+https://github.com/rust-lang/crates.io-index)" = "50f37be617794602aabbeee0be4f259dc1778fabe05e2d67ee8f79326d5cb4f6"
"checksum unicode-bidi 0.3.4 (registry+https://github.com/rust-lang/crates.io-index)" = "49f2bd0c6468a8230e1db229cff8029217cf623c767ea5d60bfbd42729ea54d5"
"checksum unicode-normalization 0.1.9 (registry+https://github.com/rust-lang/crates.io-index)" = "09c8070a9942f5e7cfccd93f490fdebd230ee3c3c9f107cb25bad5351ef671cf"
"checksum unicode-normalization 0.1.10 (registry+https://github.com/rust-lang/crates.io-index)" = "f0d98d53dfd9509a7c7f36fa8857b8f1fb699edbddd7dc2fb688db2ae5d0b2c1"
"checksum unicode-xid 0.1.0 (registry+https://github.com/rust-lang/crates.io-index)" = "fc72304796d0818e357ead4e000d19c9c174ab23dc11093ac919054d20a6a7fc"
"checksum unicode-xid 0.2.0 (registry+https://github.com/rust-lang/crates.io-index)" = "826e7639553986605ec5979c7dd957c7895e93eabed50ab2ffa7f6128a75097c"
"checksum untrusted 0.6.2 (registry+https://github.com/rust-lang/crates.io-index)" = "55cd1f4b4e96b46aeb8d4855db4a7a9bd96eeeb5c6a1ab54593328761642ce2f"

View File

@ -105,6 +105,7 @@ handlebars = "2.0.2"
# For favicon extraction from main website
soup = "0.4.1"
regex = "1.3.1"
data-url = "0.1.0"
# Required for SSL support for PostgreSQL
openssl = { version = "0.10.25", optional = true }
@ -123,3 +124,6 @@ rocket_contrib = { git = 'https://github.com/SergioBenitez/Rocket', rev = 'b95b6
# Use git version for timeout fix #706
lettre = { git = 'https://github.com/lettre/lettre', rev = '24d694db3be017d82b1cdc8bf9da601420b31bb0' }
lettre_email = { git = 'https://github.com/lettre/lettre', rev = '24d694db3be017d82b1cdc8bf9da601420b31bb0' }
# For favicon extraction from main website
data-url = { git = 'https://github.com/servo/rust-url', package="data-url", rev = '7f1bd6ce1c2fde599a757302a843a60e714c5f72' }

View File

@ -238,7 +238,7 @@ fn get_icon_url(domain: &str) -> Result<(Vec<Icon>, String), Error> {
let favicons = soup
.tag("link")
.attr("rel", Regex::new(r"icon$|apple.*icon")?) // Only use icon rels
.attr("href", Regex::new(r"(?i)\w+\.(jpg|jpeg|png|ico)(\?.*)?$")?) // Only allow specific extensions
.attr("href", Regex::new(r"(?i)\w+\.(jpg|jpeg|png|ico)(\?.*)?$|^data:image.*base64")?) // Only allow specific extensions
.find_all();
// Loop through all the found icons and determine it's priority
@ -373,15 +373,32 @@ fn download_icon(domain: &str) -> Result<Vec<u8>, Error> {
let mut buffer = Vec::new();
use data_url::DataUrl;
for icon in iconlist.iter().take(5) {
match get_page_with_cookies(&icon.href, &cookie_str) {
Ok(mut res) => {
info!("Downloaded icon from {}", icon.href);
res.copy_to(&mut buffer)?;
break;
}
Err(_) => info!("Download failed for {}", icon.href),
};
if icon.href.starts_with("data:image") {
let datauri = DataUrl::process(&icon.href).unwrap();
// Check if we are able to decode the data uri
match datauri.decode_to_vec() {
Ok((body, _fragment)) => {
// Also check if the size is atleast 67 bytes, which seems to be the smallest png i could create
if body.len() >= 67 {
buffer = body;
break;
}
}
_ => warn!("data uri is invalid")
};
} else {
match get_page_with_cookies(&icon.href, &cookie_str) {
Ok(mut res) => {
info!("Downloaded icon from {}", icon.href);
res.copy_to(&mut buffer)?;
break;
}
Err(_) => info!("Download failed for {}", icon.href),
};
}
}
if buffer.is_empty() {