feat(blog::components::check+update) update

backend/server/src/constants/config.rs

@@ -1,3 +1,5 @@
+// todo: rebuild with static OnceCell
+
 // Database related
 pub mod database {
     pub mod var_name {

backend/server/src/modules/blog/api/get_post/get_post_by_i64_id.sql

@@ -0,0 +1 @@
+SELECT 

backend/server/src/modules/blog/api/get_post/mod.rs

@@ -1,9 +1,9 @@
 use macros::dev_consume;
 
 use axum::{
+    Json,
     extract::{Path, State},
     response::IntoResponse,
-    Json,
 };
 use serde_json::json;
 use sqlx::PgPool;

backend/server/src/modules/blog/components/check_update/changes.rs

@@ -0,0 +1,30 @@
+use git2::Delta;
+use std::path::PathBuf;
+use uuid::Uuid;
+
+#[derive(Debug)]
+pub enum Change {
+    DELETE {
+        uuid: Uuid,
+    },
+    RENAME {
+        uuid: Uuid,
+        from: PathBuf,
+        to: PathBuf,
+    },
+    UPDATE {
+        uuid: Uuid,
+        content: String,
+    },
+    CREATE {
+        uuid: Uuid,
+        path: PathBuf,
+        content: String,
+    },
+}
+
+#[derive(Debug, Clone)]
+pub struct DeltaIdx {
+    pub delta_type: Delta,
+    pub content: String,
+}

backend/server/src/modules/blog/components/check_update/error.rs

@@ -1,7 +1,7 @@
 use thiserror::Error;
 
 #[derive(Error, Debug)]
-pub enum PostIdxError {
+pub enum Error {
     #[error("missing essential field `{}` for file `{}`", field, filename)]
     MissingField { field: String, filename: String },
 
@@ -46,4 +46,7 @@ pub enum PostIdxError {
         id: String,
         msg: String,
     },
+
+    #[error("git error")]
+    Git2(#[from] git2::Error),
 }

backend/server/src/modules/blog/components/check_update/mod.rs

@@ -6,6 +6,7 @@ use std::{
     path::{Path, PathBuf},
 };
 
+use crate::modules::blog::components::check_update::pull::{analyze_deltas, fetch_deltas};
 use crate::modules::blog::components::check_update::{
     // index::drop_index,
     // utils::{expand_path, index_files, search_md},
@@ -13,14 +14,18 @@ use crate::modules::blog::components::check_update::{
 };
 use crate::scheduler::task_func::AsyncTaskFunc;
 use sqlx::PgPool;
-use tracing::{info, warn, Level};
+use tracing::{info, trace, warn, Level};
 
 // mod error;
 // mod index;
+mod changes;
+mod error;
+mod pull;
 mod utils;
 
 #[derive(Debug)]
 struct Args {
+    db_connection: PgPool,
     git_work_dir: PathBuf,
     git_url: String,
     git_remote_name: String,
@@ -45,7 +50,6 @@ pub fn task(db_connection: PgPool) -> impl AsyncTaskFunc {
     };
 
     // clone & check update
-
     let git_url = match env::var(var_name::BLOG_GIT_URL) {
         Ok(url) => url,
         Err(e) => {
@@ -63,16 +67,33 @@ pub fn task(db_connection: PgPool) -> impl AsyncTaskFunc {
         .unwrap_or(default_value::BLOG_GIT_REMOTE_BRANCH.to_string());
 
     let arc_args = Arc::new(Args {
+        db_connection,
         git_work_dir,
         git_url,
         git_remote_name,
         git_remote_branch,
     });
 
-    move || Box::pin(test(arc_args.clone()))
+    move || Box::pin(workflow(arc_args.clone()))
 }
 
-async fn test(args: Arc<Args>) {
+async fn workflow(args: Arc<Args>) {
-    info!("test");
+    let deltas = match fetch_deltas(
-    dbg!(&args);
+        &args.git_url,
+        &args.git_remote_name,
+        &args.git_remote_branch,
+        &args.git_work_dir,
+    )
+    .await
+    {
+        Ok(updates) => updates,
+        Err(_) => return,
+    };
+
+    if deltas.is_empty() {
+        trace!("blog git already up-to-date");
+        return;
+    }
+
+    let updates = analyze_deltas(deltas, &args.git_work_dir).await;
 }

backend/server/src/modules/blog/components/check_update/pull.rs

@@ -0,0 +1,147 @@
+use crate::modules::blog::components::check_update::changes::{Change, DeltaIdx};
+use crate::modules::blog::components::check_update::error::Error;
+use crate::utils::git;
+use crate::utils::git::FileDelta;
+use git2::{Delta, Repository};
+use markdown::MdFile;
+use std::collections::HashMap;
+use std::path::{Path, PathBuf};
+use tracing::error;
+use uuid::Uuid;
+
+pub async fn fetch_deltas(
+    git_url: &str,
+    git_remote_name: &str,
+    git_branch: &str,
+    git_work_dir: &PathBuf,
+) -> Result<Vec<FileDelta>, Error> {
+    let repo = match Repository::open(git_work_dir) {
+        Ok(repo) => repo,
+        Err(_) => match Repository::clone(git_url, git_work_dir) {
+            Ok(repo) => repo,
+            Err(e) => {
+                error!("failed to clone blog content git({}): {}", &git_url, e);
+                return Err(e.into());
+            }
+        },
+    };
+
+    let updates = match git::sync(git_remote_name, git_branch, &repo) {
+        Ok(updates) => updates,
+        Err(e) => {
+            error!(
+                "failed to sync blog content git({}) updates: {}",
+                &git_url, e
+            );
+            return Err(e.into());
+        }
+    };
+
+    Ok(updates)
+}
+
+pub async fn analyze_deltas(file_deltas: Vec<FileDelta>, git_work_dir: &Path) -> Vec<Change> {
+    dbg!(&file_deltas);
+    let mut changes: Vec<Change> = vec![];
+    let mut merge_deltas: Vec<FileDelta> = vec![];
+
+    for delta in file_deltas.into_iter() {
+        if delta.status == Delta::Modified {
+            let file_path = &git_work_dir.join(
+                delta
+                    .new_path
+                    .expect("cannot parse modified post `new_path`"),
+            );
+            let post = match MdFile::from_file(file_path) {
+                Ok(post) => post,
+                Err(e) => {
+                    error!("md file parsing error: {}", e);
+                    continue;
+                }
+            };
+
+            let id = match extract_post_id(&post) {
+                Some(id) => id,
+                None => {
+                    // todo: add to unresolved error list
+                    error!(
+                        "failed to parse uuid on post {}",
+                        file_path.to_string_lossy()
+                    );
+                    continue;
+                }
+            };
+
+            changes.push(Change::UPDATE {
+                uuid: id,
+                content: post.content,
+            })
+        } else {
+            merge_deltas.push(delta)
+        };
+    }
+
+    // merge some addition and remove deltas to rename/move changes
+    let delta_idx = {
+        let mut delta_idx: HashMap<Uuid, Vec<DeltaIdx>> = HashMap::new();
+        for file_delta in merge_deltas.into_iter() {
+            let file_path = &git_work_dir.join(
+                file_delta
+                    .new_path
+                    .expect("cannot parse modified post `new_path`"),
+            );
+            match file_delta.status {
+                // todo: split impl
+                Delta::Added | Delta::Deleted => {
+                    let file_path_str = file_path.to_string_lossy();
+
+                    let post = match MdFile::from_file(file_path) {
+                        Ok(post) => post,
+                        Err(e) => {
+                            error!("md file parsing error ({}): {}", file_path_str, e);
+                            continue;
+                        }
+                    };
+
+                    let id = match extract_post_id(&post) {
+                        Some(id) => id,
+                        None => {
+                            // todo: add to unresolved error list
+                            error!("failed to parse uuid on post {}", file_path_str);
+                            continue;
+                        }
+                    };
+
+                    let file_delta_idx = DeltaIdx {
+                        delta_type: file_delta.status,
+                        content: post.content,
+                    };
+                    delta_idx
+                        .entry(id)
+                        // todo: change to shallow copy
+                        .and_modify(|v| v.push(file_delta_idx.clone()))
+                        .or_insert(vec![file_delta_idx]);
+                }
+                _ => {
+                    error!("unsupported delta on file {}", file_path.to_string_lossy());
+                    continue;
+                }
+            };
+        }
+        delta_idx
+    };
+
+    dbg!(&delta_idx);
+
+    changes
+}
+
+fn extract_post_id(post: &MdFile) -> Option<Uuid> {
+    match post.meta["uuid"].as_str() {
+        Some(id) => match Uuid::parse_str(id) {
+            Ok(id) => Some(id),
+            Err(_) => None,
+        },
+        None => None,
+    }
+}

backend/server/src/modules/blog/components/check_update/utils.rs

@@ -8,48 +8,6 @@ use chrono::{DateTime, FixedOffset, ParseResult};
 use dirs::home_dir;
 use tracing::warn;
 
-// pub fn search_md(entry: PathBuf) -> Vec<PathBuf> {
-//     if entry.ends_with(".git") {
-//         return vec![];
-//     }
-//     if entry.is_dir() {
-//         if let Ok(files) = entry.read_dir() {
-//             files
-//                 .flat_map(|file| {
-//                     let file_path = file.unwrap().path();
-//                     search_md(file_path)
-//                 })
-//                 .collect()
-//         } else {
-//             vec![]
-//         }
-//     } else if entry.extension().is_some_and(|ext| ext == "md") {
-//         vec![entry]
-//     } else {
-//         vec![]
-//     }
-// }
-
-// pub fn index_files(files: &[PathBuf], db_conn: &Pool<Postgres>) {
-//     let rt = Runtime::new().unwrap();
-//     rt.block_on(async {
-//         let files_clone = files.to_owned();
-//         let handlers = files_clone
-//             .into_iter()
-//             .map(|file: PathBuf| {
-//                 let conn_clone = db_conn.clone();
-//                 spawn(async move {
-//                     if let Err(e) = index_file(&file, &conn_clone).await {
-//                         error!("failed to index markdown file `{}`: {}", file.display(), e);
-//                     }
-//                 })
-//             })
-//             .collect::<Vec<_>>();
-//
-//         block_on(join_all(handlers));
-//     });
-// }
-
 pub fn expand_path(path: String) -> String {
     if path.starts_with('~') {
         home_dir()
@@ -75,7 +33,3 @@ pub fn expand_path(path: String) -> String {
 pub fn parse_date(date_str: &str) -> ParseResult<DateTime<FixedOffset>> {
     DateTime::parse_from_str(format!("{}T120000", date_str).as_str(), DEFAULT_DATE_FORMAT)
 }
-
-// pub fn get_md_file_basename(f_name: &str) -> String {
-//     f_name.replace(".md", "")
-// }

backend/server/src/modules/blog/components/consistency_guard/mod.rs

@@ -0,0 +1 @@
+

backend/server/src/modules/blog/components/mod.rs

@@ -1,2 +1,3 @@
 pub mod check_update;
+mod consistency_guard;
 mod rag;

backend/server/src/modules/blog/components/rag/mod.rs

@@ -0,0 +1 @@
+

backend/server/src/modules/blog/mod.rs

@@ -1,6 +1,8 @@
 mod api;
 mod components;
 
+use crate::modules::blog::components::check_update;
+use crate::scheduler::tasks::async_basic::AsyncBasic;
 use crate::{
     modules::{
         blog::{
@@ -16,9 +18,6 @@ use crate::{
     scheduler::tasks::CronTask,
 };
 
-use crate::modules::blog::components::check_update;
-use crate::scheduler::tasks::async_basic::AsyncBasic;
-use crate::scheduler::tasks::basic::BasicTask;
 use axum::{routing::get, Router};
 use sqlx::PgPool;
 
@@ -46,19 +45,11 @@ impl Module for Blog {
 
     fn get_cron_tasks(&self) -> Vec<(&str, Box<dyn CronTask>)> {
         let db_handler = self.db_handler.clone();
-        vec![
+        vec![(
-            (
+            "Blog Check Update",
-                "Blog Check Update",
+            AsyncBasic::new(check_update::task(db_handler.clone()), "*/5 * * * * *")
-                AsyncBasic::new(check_update::task(db_handler.clone()), "* * * * * *")
+                .unwrap()
-                    .unwrap()
+                .to_task(),
-                    .to_task(),
+        )]
-            ),
-            (
-                "Sync test task",
-                BasicTask::new(|| println!("test lol"), "* * * * * *")
-                    .unwrap()
-                    .to_task(),
-            ),
-        ]
     }
 }

backend/server/src/modules/blog/schema.sql

@@ -2,20 +2,20 @@ CREATE TABLE tags
 (
     id          SERIAL PRIMARY KEY,
     name        VARCHAR(20) UNIQUE NOT NULL,
-    description TEXT
+    description TEXT DEFAULT ''
 );
 
 CREATE TABLE categories
 (
     id          SERIAL PRIMARY KEY,
     name        VARCHAR(20) UNIQUE NOT NULL,
-    description TEXT
+    description TEXT DEFAULT ''
 );
 
 CREATE TABLE posts
 (
-    id             SERIAL PRIMARY KEY,
+    id             UUID UNIQUE                       NOT NULL,
-    identifier     VARCHAR(50) UNIQUE                NOT NULL,
+    identifier     VARCHAR(100) UNIQUE               NOT NULL,
     title          VARCHAR(255)                      NOT NULL,
     subtitle       VARCHAR(255)                      NOT NULL,
     published_date DATE                              NOT NULL,
@@ -27,15 +27,15 @@ CREATE TABLE posts
 
 CREATE TABLE post_tag
 (
-    post SERIAL REFERENCES posts (id),
+    post UUID REFERENCES posts (id),
     tag  SERIAL REFERENCES tags (id)
 );
 
 CREATE TABLE metadata
 (
-    pid        SERIAL REFERENCES posts (id) NOT NULL,
+    pid        UUID REFERENCES posts (id) NOT NULL,
-    meta_key   VARCHAR(50)                  NOT NULL,
+    meta_key   VARCHAR(50)                NOT NULL,
-    meta_value TEXT                         NOT NULL,
+    meta_value TEXT                       NOT NULL,
 
     CONSTRAINT meta_pk PRIMARY KEY (pid, meta_key)
 );

backend/server/src/modules/mod.rs

@@ -2,7 +2,7 @@ use std::{env, num::ParseIntError};
 
 use crate::{
     constants::config::server::{default_value, var_name},
-    scheduler::{tasks::CronTask, Scheduler},
+    scheduler::{Scheduler, tasks::CronTask},
 };
 use macros::error_panic;
 

backend/server/src/scheduler/mod.rs

@@ -20,8 +20,6 @@ pub struct Scheduler {
 impl Scheduler {
     /// Start all tasks in the scheduler
     pub async fn start(&self) {
-        dbg!("entering");
-
         let fut = self
             .task_pool
             .clone()

backend/server/src/scheduler/utils.rs

@@ -3,7 +3,7 @@ use std::{sync::Arc, time::Duration};
 use crate::scheduler::{error::SchedulerError, tasks::CronTask};
 
 use chrono::{DateTime, Utc};
-use tokio::time::{sleep, Sleep};
+use tokio::time::{Sleep, sleep};
 
 pub fn duration_from_now(future: DateTime<Utc>) -> Duration {
     future

backend/server/src/utils/git.rs

@@ -3,12 +3,11 @@
 /// https://github.com/rust-lang/git2-rs/commit/f3b87baed1e33d6c2d94fe1fa6aa6503a071d837
 ///
 use git2::{
-    build, AnnotatedCommit, AutotagOption, Delta, FetchOptions, Reference, Remote, Repository,
+    AnnotatedCommit, AutotagOption, Delta, FetchOptions, Reference, Remote, Repository, build,
 };
 use std::path::PathBuf;
 use tracing::{error, warn};
 
-
 #[derive(Debug, Eq, PartialEq)]
 pub struct FileDelta {
     pub old_path: Option<PathBuf>,
@@ -40,7 +39,6 @@ fn fast_forward(
         None => String::from_utf8_lossy(lb.name_bytes()).to_string(),
     };
     let msg = format!("fast-forward({}) -> {}", name, rc.id());
-    // println!("{}", msg);
     lb.set_target(rc.id(), &msg)?;
     repo.set_head(&name)?;
     repo.checkout_head(Some(build::CheckoutBuilder::default().force()))?;