diff options
Diffstat (limited to 'src')
-rw-r--r-- | src/character.rs | 3 | ||||
-rw-r--r-- | src/main.rs | 26 | ||||
-rw-r--r-- | src/pre_process.rs | 13 |
3 files changed, 42 insertions, 0 deletions
diff --git a/src/character.rs b/src/character.rs new file mode 100644 index 0000000..9c548c1 --- /dev/null +++ b/src/character.rs | |||
@@ -0,0 +1,3 @@ | |||
1 | pub struct Character { | ||
2 | |||
3 | } \ No newline at end of file | ||
diff --git a/src/main.rs b/src/main.rs new file mode 100644 index 0000000..ca9b1c5 --- /dev/null +++ b/src/main.rs | |||
@@ -0,0 +1,26 @@ | |||
1 | extern crate regex; | ||
2 | use regex::Regex; | ||
3 | |||
4 | extern crate walkdir; | ||
5 | use walkdir::WalkDir; | ||
6 | |||
7 | use std::io::prelude::*; | ||
8 | use std::fs::File; | ||
9 | use std::env; | ||
10 | use std::path::{Path, PathBuf}; | ||
11 | |||
12 | mod pre_process; | ||
13 | |||
14 | fn main() { | ||
15 | let raw_files = env::var("RAW_FILES").unwrap_or("S:\\grilist\\acd\\acd_character_parser\\characters\\".into()); | ||
16 | let base_path = Path::new(&raw_files); | ||
17 | |||
18 | for entry in WalkDir::new(base_path).min_depth(1).into_iter().filter_map(|e| e.ok()) { | ||
19 | println!("{:?}", entry.path()); | ||
20 | let mut f = File::open(entry.path()).expect("could not open file"); | ||
21 | let mut buf = String::new(); | ||
22 | f.read_to_string(&mut buf).unwrap(); | ||
23 | |||
24 | let buf = pre_process::strip_irrelevant_content(&buf); | ||
25 | } | ||
26 | } | ||
diff --git a/src/pre_process.rs b/src/pre_process.rs new file mode 100644 index 0000000..aa55bd8 --- /dev/null +++ b/src/pre_process.rs | |||
@@ -0,0 +1,13 @@ | |||
1 | pub fn strip_irrelevant_content(s: &str) -> String { | ||
2 | let mut retn = ""; | ||
3 | match s.find(r#"<div class=profile id=profile>"#) { | ||
4 | Some(pos) => retn = &s[pos..], | ||
5 | None => (), | ||
6 | }; | ||
7 | |||
8 | match s.find(r#"<INPUT style="font-size: 2em;" TYPE=SUBMIT NAME="votes" VALUE="Cast Votes">"#) { | ||
9 | Some(pos) => retn = &s[..pos], | ||
10 | None => (), | ||
11 | }; | ||
12 | return retn.into(); | ||
13 | } \ No newline at end of file | ||