From 0617663b249f9ca488e5de652108b17d67fbaf45 Mon Sep 17 00:00:00 2001 From: Vika Date: Sat, 29 Jul 2023 21:59:56 +0300 Subject: Moved the entire Kittybox tree into the root --- src/micropub/util.rs | 444 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 444 insertions(+) create mode 100644 src/micropub/util.rs (limited to 'src/micropub/util.rs') diff --git a/src/micropub/util.rs b/src/micropub/util.rs new file mode 100644 index 0000000..940d7c3 --- /dev/null +++ b/src/micropub/util.rs @@ -0,0 +1,444 @@ +use crate::database::Storage; +use kittybox_indieauth::TokenData; +use chrono::prelude::*; +use core::iter::Iterator; +use newbase60::num_to_sxg; +use serde_json::json; +use std::convert::TryInto; + +pub(crate) const DEFAULT_CHANNEL_PATH: &str = "/feeds/main"; +const DEFAULT_CHANNEL_NAME: &str = "Main feed"; +pub(crate) const CONTACTS_CHANNEL_PATH: &str = "/feeds/vcards"; +const CONTACTS_CHANNEL_NAME: &str = "My address book"; +pub(crate) const FOOD_CHANNEL_PATH: &str = "/feeds/food"; +const FOOD_CHANNEL_NAME: &str = "My recipe book"; + +fn get_folder_from_type(post_type: &str) -> String { + (match post_type { + "h-feed" => "feeds/", + "h-card" => "vcards/", + "h-event" => "events/", + "h-food" => "food/", + _ => "posts/", + }) + .to_string() +} + +/// Reset the datetime to a proper datetime. +/// Do not attempt to recover the information. +/// Do not pass GO. Do not collect $200. +fn reset_dt(post: &mut serde_json::Value) -> DateTime { + let curtime: DateTime = Local::now(); + post["properties"]["published"] = json!([curtime.to_rfc3339()]); + chrono::DateTime::from(curtime) +} + +pub fn normalize_mf2(mut body: serde_json::Value, user: &TokenData) -> (String, serde_json::Value) { + // Normalize the MF2 object here. + let me = &user.me; + let folder = get_folder_from_type(body["type"][0].as_str().unwrap()); + let published: DateTime = + if let Some(dt) = body["properties"]["published"][0].as_str() { + // Check if the datetime is parsable. + match DateTime::parse_from_rfc3339(dt) { + Ok(dt) => dt, + Err(_) => reset_dt(&mut body), + } + } else { + // Set the datetime. + // Note: this code block duplicates functionality with the above failsafe. + // Consider refactoring it to a helper function? + reset_dt(&mut body) + }; + match body["properties"]["uid"][0].as_str() { + None => { + let uid = serde_json::Value::String( + me.join( + &(folder.clone() + + &num_to_sxg(published.timestamp_millis().try_into().unwrap())), + ) + .unwrap() + .to_string(), + ); + body["properties"]["uid"] = serde_json::Value::Array(vec![uid.clone()]); + match body["properties"]["url"].as_array_mut() { + Some(array) => array.push(uid), + None => body["properties"]["url"] = body["properties"]["uid"].clone(), + } + } + Some(uid_str) => { + let uid = uid_str.to_string(); + match body["properties"]["url"].as_array_mut() { + Some(array) => { + if !array.iter().any(|i| i.as_str().unwrap_or("") == uid) { + array.push(serde_json::Value::String(uid)) + } + } + None => body["properties"]["url"] = body["properties"]["uid"].clone(), + } + } + } + if let Some(slugs) = body["properties"]["mp-slug"].as_array() { + let new_urls = slugs + .iter() + .map(|i| i.as_str().unwrap_or("")) + .filter(|i| i != &"") + .map(|i| me.join(&((&folder).clone() + i)).unwrap().to_string()) + .collect::>(); + let urls = body["properties"]["url"].as_array_mut().unwrap(); + new_urls.iter().for_each(|i| urls.push(json!(i))); + } + let props = body["properties"].as_object_mut().unwrap(); + props.remove("mp-slug"); + + if body["properties"]["content"][0].is_string() { + // Convert the content to HTML using the `markdown` crate + body["properties"]["content"] = json!([{ + "html": markdown::to_html(body["properties"]["content"][0].as_str().unwrap()), + "value": body["properties"]["content"][0] + }]) + } + // TODO: apply this normalization to editing too + if body["properties"]["mp-channel"].is_array() { + let mut additional_channels = body["properties"]["mp-channel"].as_array().unwrap().clone(); + if let Some(array) = body["properties"]["channel"].as_array_mut() { + array.append(&mut additional_channels); + } else { + body["properties"]["channel"] = json!(additional_channels) + } + body["properties"] + .as_object_mut() + .unwrap() + .remove("mp-channel"); + } else if body["properties"]["mp-channel"].is_string() { + let chan = body["properties"]["mp-channel"] + .as_str() + .unwrap() + .to_owned(); + if let Some(array) = body["properties"]["channel"].as_array_mut() { + array.push(json!(chan)) + } else { + body["properties"]["channel"] = json!([chan]); + } + body["properties"] + .as_object_mut() + .unwrap() + .remove("mp-channel"); + } + if body["properties"]["channel"][0].as_str().is_none() { + match body["type"][0].as_str() { + Some("h-entry") => { + // Set the channel to the main channel... + // TODO find like posts and move them to separate private channel + let default_channel = me.join(DEFAULT_CHANNEL_PATH).unwrap().to_string(); + + body["properties"]["channel"] = json!([default_channel]); + } + Some("h-card") => { + let default_channel = me.join(CONTACTS_CHANNEL_PATH).unwrap().to_string(); + + body["properties"]["channel"] = json!([default_channel]); + } + Some("h-food") => { + let default_channel = me.join(FOOD_CHANNEL_PATH).unwrap().to_string(); + + body["properties"]["channel"] = json!([default_channel]); + } + // TODO h-event + /*"h-event" => { + let default_channel + },*/ + _ => { + body["properties"]["channel"] = json!([]); + } + } + } + body["properties"]["posted-with"] = json!([user.client_id]); + if body["properties"]["author"][0].as_str().is_none() { + body["properties"]["author"] = json!([me.as_str()]) + } + // TODO: maybe highlight #hashtags? + // Find other processing to do and insert it here + return ( + body["properties"]["uid"][0].as_str().unwrap().to_string(), + body, + ); +} + +pub(crate) fn form_to_mf2_json(form: Vec<(String, String)>) -> serde_json::Value { + let mut mf2 = json!({"type": [], "properties": {}}); + for (k, v) in form { + if k == "h" { + mf2["type"] + .as_array_mut() + .unwrap() + .push(json!("h-".to_string() + &v)); + } else if k != "access_token" { + let key = k.strip_suffix("[]").unwrap_or(&k); + match mf2["properties"][key].as_array_mut() { + Some(prop) => prop.push(json!(v)), + None => mf2["properties"][key] = json!([v]), + } + } + } + if mf2["type"].as_array().unwrap().is_empty() { + mf2["type"].as_array_mut().unwrap().push(json!("h-entry")); + } + mf2 +} + +pub(crate) async fn create_feed( + storage: &impl Storage, + uid: &str, + channel: &str, + user: &TokenData, +) -> crate::database::Result<()> { + let path = url::Url::parse(channel).unwrap().path().to_string(); + + let name = match path.as_str() { + DEFAULT_CHANNEL_PATH => DEFAULT_CHANNEL_NAME, + CONTACTS_CHANNEL_PATH => CONTACTS_CHANNEL_NAME, + FOOD_CHANNEL_PATH => FOOD_CHANNEL_NAME, + _ => panic!("Tried to create an unknown default feed!"), + }; + + let (_, feed) = normalize_mf2( + json!({ + "type": ["h-feed"], + "properties": { + "name": [name], + "uid": [channel] + }, + }), + user, + ); + storage.put_post(&feed, user.me.as_str()).await?; + storage.add_to_feed(channel, uid).await +} + +#[cfg(test)] +mod tests { + use super::*; + use serde_json::json; + + fn token_data() -> TokenData { + TokenData { + me: "https://fireburn.ru/".parse().unwrap(), + client_id: "https://quill.p3k.io/".parse().unwrap(), + scope: kittybox_indieauth::Scopes::new(vec![kittybox_indieauth::Scope::Create]), + exp: Some(u64::MAX), + iat: Some(0) + } + } + + #[test] + fn test_form_to_mf2() { + assert_eq!( + super::form_to_mf2_json( + serde_urlencoded::from_str("h=entry&content=something%20interesting").unwrap() + ), + json!({ + "type": ["h-entry"], + "properties": { + "content": ["something interesting"] + } + }) + ) + } + + #[test] + fn test_no_replace_uid() { + let mf2 = json!({ + "type": ["h-card"], + "properties": { + "uid": ["https://fireburn.ru/"], + "name": ["Vika Nezrimaya"], + "note": ["A crazy programmer girl who wants some hugs"] + } + }); + + let (uid, normalized) = normalize_mf2( + mf2.clone(), + &token_data(), + ); + assert_eq!( + normalized["properties"]["uid"][0], mf2["properties"]["uid"][0], + "UID was replaced" + ); + assert_eq!( + normalized["properties"]["uid"][0], uid, + "Returned post location doesn't match UID" + ); + } + + #[test] + fn test_mp_channel() { + let mf2 = json!({ + "type": ["h-entry"], + "properties": { + "uid": ["https://fireburn.ru/posts/test"], + "content": [{"html": "

Hello world!

"}], + "mp-channel": ["https://fireburn.ru/feeds/test"] + } + }); + + let (_, normalized) = normalize_mf2( + mf2.clone(), + &token_data(), + ); + + assert_eq!( + normalized["properties"]["channel"], + mf2["properties"]["mp-channel"] + ); + } + + #[test] + fn test_mp_channel_as_string() { + let mf2 = json!({ + "type": ["h-entry"], + "properties": { + "uid": ["https://fireburn.ru/posts/test"], + "content": [{"html": "

Hello world!

"}], + "mp-channel": "https://fireburn.ru/feeds/test" + } + }); + + let (_, normalized) = normalize_mf2( + mf2.clone(), + &token_data(), + ); + + assert_eq!( + normalized["properties"]["channel"][0], + mf2["properties"]["mp-channel"] + ); + } + + #[test] + fn test_normalize_mf2() { + let mf2 = json!({ + "type": ["h-entry"], + "properties": { + "content": ["This is content!"] + } + }); + + let (uid, post) = normalize_mf2( + mf2, + &token_data(), + ); + assert_eq!( + post["properties"]["published"] + .as_array() + .expect("post['published'] is undefined") + .len(), + 1, + "Post doesn't have a published time" + ); + DateTime::parse_from_rfc3339(post["properties"]["published"][0].as_str().unwrap()) + .expect("Couldn't parse date from rfc3339"); + assert!( + !post["properties"]["url"] + .as_array() + .expect("post['url'] is undefined") + .is_empty(), + "Post doesn't have any URLs" + ); + assert_eq!( + post["properties"]["uid"] + .as_array() + .expect("post['uid'] is undefined") + .len(), + 1, + "Post doesn't have a single UID" + ); + assert_eq!( + post["properties"]["uid"][0], uid, + "UID of a post and its supposed location don't match" + ); + assert!( + uid.starts_with("https://fireburn.ru/posts/"), + "The post namespace is incorrect" + ); + assert_eq!( + post["properties"]["content"][0]["html"] + .as_str() + .expect("Post doesn't have a rich content object") + .trim(), + "

This is content!

", + "Parsed Markdown content doesn't match expected HTML" + ); + assert_eq!( + post["properties"]["channel"][0], "https://fireburn.ru/feeds/main", + "Post isn't posted to the main channel" + ); + assert_eq!( + post["properties"]["author"][0], "https://fireburn.ru/", + "Post author is unknown" + ); + } + + #[test] + fn test_mp_slug() { + let mf2 = json!({ + "type": ["h-entry"], + "properties": { + "content": ["This is content!"], + "mp-slug": ["hello-post"] + }, + }); + + let (_, post) = normalize_mf2( + mf2, + &token_data(), + ); + assert!( + post["properties"]["url"] + .as_array() + .unwrap() + .iter() + .map(|i| i.as_str().unwrap()) + .any(|i| i == "https://fireburn.ru/posts/hello-post"), + "Didn't found an URL pointing to the location expected by the mp-slug semantics" + ); + assert!( + post["properties"]["mp-slug"].as_array().is_none(), + "mp-slug wasn't deleted from the array!" + ) + } + + #[test] + fn test_normalize_feed() { + let mf2 = json!({ + "type": ["h-feed"], + "properties": { + "name": "Main feed", + "mp-slug": ["main"] + } + }); + + let (uid, post) = normalize_mf2( + mf2, + &token_data(), + ); + assert_eq!( + post["properties"]["uid"][0], uid, + "UID of a post and its supposed location don't match" + ); + assert_eq!(post["properties"]["author"][0], "https://fireburn.ru/"); + assert!( + post["properties"]["url"] + .as_array() + .unwrap() + .iter() + .map(|i| i.as_str().unwrap()) + .any(|i| i == "https://fireburn.ru/feeds/main"), + "Didn't found an URL pointing to the location expected by the mp-slug semantics" + ); + assert!( + post["properties"]["mp-slug"].as_array().is_none(), + "mp-slug wasn't deleted from the array!" + ) + } +} -- cgit 1.4.1