use cache by default, add clear-cache argument

Cargo.lock

@@ -89,6 +89,12 @@ version = "1.3.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "bef38d45163c2f1dde094a7dfd33ccf595c92905c8f8f4fdc18d06fb1037718a"
 
+[[package]]
+name = "bitflags"
+version = "2.4.2"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "ed570934406eb16438a4e976b1b4500774099c13b8cb96eec99f620f05090ddf"
+
 [[package]]
 name = "cc"
 version = "1.0.83"
@@ -165,7 +171,7 @@ version = "0.25.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "e64e6c0fbe2c17357405f7c758c1ef960fce08bdfb2c03d88d2a18d7e09c4b67"
 dependencies = [
- "bitflags",
+ "bitflags 1.3.2",
  "crossterm_winapi",
  "libc",
  "mio",
@@ -199,6 +205,27 @@ version = "0.1.13"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "56254986775e3233ffa9c4d7d3faaf6d36a2c09d30b20687e9f88bc8bafc16c8"
 
+[[package]]
+name = "directories"
+version = "5.0.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "9a49173b84e034382284f27f1af4dcbbd231ffa358c0fe316541a7337f376a35"
+dependencies = [
+ "dirs-sys",
+]
+
+[[package]]
+name = "dirs-sys"
+version = "0.4.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "520f05a5cbd335fae5a99ff7a6ab8627577660ee5cfd6a94a6a929b52ff0321c"
+dependencies = [
+ "libc",
+ "option-ext",
+ "redox_users",
+ "windows-sys 0.48.0",
+]
+
 [[package]]
 name = "env_filter"
 version = "0.1.0"
@@ -349,6 +376,17 @@ version = "0.2.153"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "9c198f91728a82281a64e1f4f9eeb25d82cb32a5de251c6bd1b5154d63a8e7bd"
 
+[[package]]
+name = "libredox"
+version = "0.0.1"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "85c833ca1e66078851dba29046874e38f08b2c883700aa29a03ddd3b23814ee8"
+dependencies = [
+ "bitflags 2.4.2",
+ "libc",
+ "redox_syscall",
+]
+
 [[package]]
 name = "lock_api"
 version = "0.4.11"
@@ -404,6 +442,12 @@ version = "1.19.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "3fdb12b2476b595f9358c5161aa467c2438859caa136dec86c26fdd2efe17b92"
 
+[[package]]
+name = "option-ext"
+version = "0.2.0"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "04744f49eae99ab78e0d5c0b603ab218f515ea8cfe5a456d7629ad883a3b6e7d"
+
 [[package]]
 name = "parking_lot"
 version = "0.12.1"
@@ -473,7 +517,18 @@ version = "0.4.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "4722d768eff46b75989dd134e5c353f0d6296e5aaa3132e776cbdb56be7731aa"
 dependencies = [
- "bitflags",
+ "bitflags 1.3.2",
+]
+
+[[package]]
+name = "redox_users"
+version = "0.4.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a18479200779601e498ada4e8c1e1f50e3ee19deb0259c25825a98b5603b2cb4"
+dependencies = [
+ "getrandom",
+ "libredox",
+ "thiserror",
 ]
 
 [[package]]
@@ -524,6 +579,7 @@ name = "risp"
 version = "0.1.0"
 dependencies = [
  "clap",
+ "directories",
  "env_logger",
  "log",
  "pretty_assertions",

Cargo.toml

@@ -16,6 +16,7 @@ log = "0.4"
 tqdm = "0.6"
 toml = "0.8"
 clap = { version = "4.5.0", features = ["derive"] }
+directories = "5.0"
 
 [dev-dependencies]
 pretty_assertions = "1.4"

src/main.rs

@@ -1,5 +1,5 @@
 use clap::{command, Parser};
-use risp::law::Law;
+use risp::{law::Law, misc::clear_cache};
 
 #[derive(Parser, Debug)]
 #[command(version, about, long_about = None)]
@@ -7,12 +7,22 @@ struct Args {
     /// Path to the config of a law text
     #[arg(short, long)]
     config: String,
+
+    /// Clears the cache (downloaded laws + paragraphs)
+    clear_cache: bool,
 }
 
 fn main() {
     env_logger::init();
 
     let args = Args::parse();
+
+    if args.clear_cache {
+        if let Err(e) = clear_cache() {
+            println!("Failed to clear cache: {e:?}");
+        }
+    }
+
     let law = Law::from_config(&args.config).unwrap();
 
     law.to_md();

src/misc.rs

@@ -1,11 +1,11 @@
-use std::io;
+use std::{fs, io, path::Path};
 
 use time::{format_description, OffsetDateTime};
 
 #[derive(Debug)]
 #[allow(dead_code)]
 pub struct Error {
-    msg: String,
+    pub(crate) msg: String,
 }
 
 impl Error {
@@ -57,3 +57,40 @@ pub(crate) fn current_date() -> String {
     let format = format_description::parse("[year]-[month]-[day]").unwrap(); //unwrap okay, supplied format is fine
     local_date.format(&format).expect("Failed to format date")
 }
+
+#[cfg(not(test))]
+pub(crate) fn get_cache_dir() -> Result<String, Error> {
+    let cache_dir = directories::BaseDirs::new().ok_or(Error {
+        msg: "directories crate could not find basedirs.".into(),
+    })?;
+    let cache_dir = cache_dir.cache_dir();
+    Ok(format!("{}/risp/", cache_dir.to_str().unwrap()))
+}
+
+#[cfg(test)]
+pub(crate) fn get_cache_dir() -> Result<String, Error> {
+    Ok("./data/cache/".into())
+}
+
+pub fn clear_cache() -> Result<(), Error> {
+    Ok(delete_all_in_dir(Path::new(&get_cache_dir()?))?)
+}
+
+fn delete_all_in_dir<P: AsRef<Path>>(dir_path: P) -> std::io::Result<()> {
+    let entries = fs::read_dir(dir_path)?;
+
+    for entry in entries {
+        let entry = entry?;
+        let path = entry.path();
+
+        if path.is_dir() {
+            delete_all_in_dir(&path)?;
+            fs::remove_dir(&path)?;
+        } else {
+            // If it's a file, delete it
+            fs::remove_file(&path)?;
+        }
+    }
+
+    Ok(())
+}

src/overview/mod.rs

@@ -2,10 +2,12 @@
 
 mod ris_structure;
 
+use std::path::Path;
+
 use log::info;
 use serde::Deserialize;
 
-use crate::misc::{current_date, Error};
+use crate::misc::{current_date, get_cache_dir, Error};
 
 use ris_structure::OgdSearchResult;
 
@@ -82,30 +84,10 @@ fn parse_from_str(content: &str, skip_first: bool) -> Result<(bool, Vec<String>)
     Ok((true, ret))
 }
 
-#[cfg(not(test))]
-/// Fetches the json content of the given overview (`law_id`) from the RIS API.
-///
-/// # Errors
-/// Fails if `ureq` can't create a connection, probably because there's no internet connection? (Or RIS is not online.)
-fn fetch_page(overview_id: usize, page: usize) -> Result<String, Error> {
-    Ok(
-        ureq::post("https://data.bka.gv.at/ris/api/v2.6/Bundesrecht")
-            .send_form(&[
-                ("Applikation", "BrKons"),
-                ("Gesetzesnummer", &format!("{overview_id}")),
-                ("DokumenteProSeite", "OneHundred"),
-                ("Seitennummer", &format!("{page}")),
-                ("Fassung.FassungVom", &current_date()),
-            ])?
-            .into_string()?,
-    )
-}
-
-#[cfg(test)]
 fn fetch_page(overview_id: usize, page: usize) -> Result<String, Error> {
     use std::fs;
 
-    let expected_filename = format!("./data/cache/law-{overview_id}-{page}");
+    let expected_filename = format!("{}law-{overview_id}-{page}", get_cache_dir()?);
 
     match fs::read_to_string(&expected_filename) {
         Ok(data) => Ok(data),
@@ -120,6 +102,11 @@ fn fetch_page(overview_id: usize, page: usize) -> Result<String, Error> {
                     ("Fassung.FassungVom", &current_date()),
                 ])?
                 .into_string()?;
+            let path = Path::new(&expected_filename);
+            if let Some(parent) = path.parent() {
+                // Try to create the directory (and any necessary parent directories)
+                fs::create_dir_all(parent).expect("Unable to create directory");
+            }
             fs::write(expected_filename, &data).expect("Unable to write file");
             Ok(data)
         }

src/paragraph/mod.rs

@@ -1,9 +1,18 @@
 //! Deals with getting all paragraphs for a given law text
 mod parser;
 
+use std::{
+    fs,
+    hash::{DefaultHasher, Hash, Hasher},
+    path::Path,
+};
+
 use log::info;
 
-use crate::{law::LawBuilder, misc::Error};
+use crate::{
+    law::LawBuilder,
+    misc::{get_cache_dir, Error},
+};
 
 use self::parser::Risdok;
 
@@ -57,30 +66,23 @@ impl Parser {
     }
 }
 
-#[cfg(not(test))]
 fn fetch(url: &str) -> Result<String, Error> {
-    Ok(ureq::get(url).call()?.into_string()?)
-}
-
-#[cfg(test)]
-fn fetch(url: &str) -> Result<String, Error> {
-    use std::{
-        collections::hash_map::DefaultHasher,
-        fs,
-        hash::{Hash, Hasher},
-    };
-
     let mut hasher = DefaultHasher::new();
     url.hash(&mut hasher);
     let hash = format!("{:x}", hasher.finish());
 
-    let expected_filename = format!("./data/cache/par-{hash}");
+    let expected_filename = format!("{}par-{hash}", get_cache_dir()?);
 
     match fs::read_to_string(&expected_filename) {
         Ok(data) => Ok(data),
         Err(_) => {
             info!("Not finding url {url} in the cache, downloading...");
             let data = ureq::get(url).call()?.into_string()?;
+            let path = Path::new(&expected_filename);
+            if let Some(parent) = path.parent() {
+                // Try to create the directory (and any necessary parent directories)
+                fs::create_dir_all(parent).expect("Unable to create directory");
+            }
             fs::write(expected_filename, &data).expect("Unable to write file");
             Ok(data)
         }