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

Refactor inline_css_str_to_map to use a better tokenizer

Browse source
This commit is contained in:
Kenneth Gitere 2020-11-24 08:29:00 +03:00
parent aff4054ca9
commit cdfbc2b3f6
1 changed files with 69 additions and 16 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

85
src/moz_readability/mod.rs
View file

@ -763,14 +763,66 @@ impl Readability {
} }
/// Converts an inline CSS string to a [HashMap] of property and value(s) /// Converts an inline CSS string to a [HashMap] of property and value(s)
fn inline_css_str_to_map(css_str: &str) -> HashMap<&str, &str> { fn inline_css_str_to_map(css_str: &str) -> HashMap<String, String> {
css_str enum State {
.split(";") ReadProp,
.filter(|split_str| !split_str.trim().is_empty()) ReadVal,
.map(|str_pair| { ReadQuot,
let mut vals = str_pair.split(":"); ReadDquot,
(vals.next().unwrap().trim(), vals.next().unwrap().trim()) }
}) let mut decl: (Option<String>, Option<String>) = (None, None);
let mut chars = css_str.chars();
let mut state = State::ReadProp;
let mut token = String::new();
let mut tokens = vec![];
while let Some(c) = chars.next() {
match state {
State::ReadProp => {
if c != ':' {
token.push(c);
} else {
state = State::ReadVal;
decl.0 = Some(token.trim().to_string());
token.clear();
}
}
State::ReadVal => {
if c == '\'' {
state = State::ReadQuot;
token.push(c);
} else if c == '"' {
state = State::ReadDquot;
token.push(c);
} else if c == ';' {
state = State::ReadProp;
decl.1 = Some(token.trim().to_string());
tokens.push(decl.clone());
token.clear();
} else {
token.push(c);
}
}
State::ReadQuot => {
token.push(c);
if c == '\'' {
state = State::ReadVal;
}
}
State::ReadDquot => {
token.push(c);
if c == '"' {
state = State::ReadVal;
}
}
}
}
if !token.is_empty() {
decl.1 = Some(token.trim().to_string());
tokens.push(decl);
}
tokens
.into_iter()
.map(|tok_pair| (tok_pair.0.unwrap(), tok_pair.1.unwrap()))
.collect() .collect()
} }
@ -2399,18 +2451,19 @@ mod test {
use std::collections::HashMap; use std::collections::HashMap;
let css_str = "display: flex; height: 200px; width: 250px; justify-content: center; align-items: center; border: 2px solid black"; let css_str = "display: flex; height: 200px; width: 250px; justify-content: center; align-items: center; border: 2px solid black";
let mut css_map = HashMap::new(); let mut css_map = HashMap::new();
css_map.insert("display", "flex"); css_map.insert("display".to_string(), "flex".to_string());
css_map.insert("height", "200px"); css_map.insert("height".to_string(), "200px".to_string());
css_map.insert("width", "250px"); css_map.insert("width".to_string(), "250px".to_string());
css_map.insert("justify-content", "center"); css_map.insert("justify-content".to_string(), "center".to_string());
css_map.insert("align-items", "center"); css_map.insert("align-items".to_string(), "center".to_string());
css_map.insert("border", "2px solid black"); css_map.insert("border".to_string(), "2px solid black".to_string());
let css_str_to_vec = Readability::inline_css_str_to_map(css_str); let css_str_to_vec = Readability::inline_css_str_to_map(css_str);
assert_eq!(css_map, css_str_to_vec); assert_eq!(css_map, css_str_to_vec);
let mut css_map = HashMap::new(); let mut css_map = HashMap::new();
css_map.insert("color", "red"); css_map.insert("color".to_string(), "red".to_string());
assert_eq!(css_map, Readability::inline_css_str_to_map("color: red;")); css_map.insert("background-image".to_string(), "url('data:image/jpeg;base64,/wgARCAALABQDASIAAhEBAxEB/8QAFwABAQEBAAAAAAAAAAAAAAAAAgADBP/')".to_string());
assert_eq!(css_map, Readability::inline_css_str_to_map("color: red;background-image: url('data:image/jpeg;base64,/wgARCAALABQDASIAAhEBAxEB/8QAFwABAQEBAAAAAAAAAAAAAAAAAgADBP/')"));
} }
#[test] #[test]