minor update

Cargo.toml

@@ -39,11 +39,11 @@ zstd-sys = { version = "=2.0.9", default-features = false }
 serde = { version = "1.0" }
 
 # Use our fork until https://github.com/apache/opendal/pull/5530
-opendal = { git = "https://github.com/XiangpengHao/opendal.git", default-features = false, features = [
+opendal = { git = "https://github.com/XiangpengHao/opendal.git", default-features = false, rev = "8b6ea6", features = [
 	"services-s3",
 	"services-http",
 ] }
-object_store_opendal = { git = "https://github.com/XiangpengHao/opendal.git", features = [
+object_store_opendal = { git = "https://github.com/XiangpengHao/opendal.git", rev = "8b6ea6", features = [
 	"send_wrapper",
 ] }
 

README.md

@@ -21,6 +21,11 @@ Online at: https://parquet-viewer.xiangpeng.systems
 For example, [`parquet-viewer.xiangpeng.systems/?url=https://raw.githubusercontent.com/tobilg/public-cloud-provider-ip-ranges/main/data/providers/all.parquet`](https://parquet-viewer.xiangpeng.systems/?url=https://raw.githubusercontent.com/tobilg/public-cloud-provider-ip-ranges/main/data/providers/all.parquet) will load the file from github.
 `parquet-viewer` is smart enough to only download the data that is relevant to your query, usually a few KBs, even if the file is large.
 
+- You can use `parquet-viewer.py` in `utils` to open a local file. Only works on Chrome or Firefox (not Safari).
+```bash
+./parquet-viewer.py /path/to/your/file.parquet
+```
+
 
 
 ## Development

src/file_reader.rs

@@ -1,9 +1,10 @@
-use std::sync::Arc;
+use std::sync::{Arc, LazyLock};
 
 use datafusion::execution::object_store::ObjectStoreUrl;
 use leptos::prelude::*;
 use leptos::wasm_bindgen::{prelude::Closure, JsCast};
 use leptos_router::hooks::query_signal;
+use object_store::memory::InMemory;
 use object_store::path::Path;
 use object_store::{ObjectStore, PutPayload};
 use object_store_opendal::OpendalStore;
@@ -12,7 +13,10 @@ use parquet::arrow::async_reader::{AsyncFileReader, ParquetObjectReader};
 use url::Url;
 use web_sys::js_sys;
 
-use crate::{ParquetTable, INMEMORY_STORE, SESSION_CTX};
+use crate::{ParquetTable, SESSION_CTX};
+
+pub(crate) static INMEMORY_STORE: LazyLock<Arc<InMemory>> =
+    LazyLock::new(|| Arc::new(InMemory::new()));
 
 const S3_ENDPOINT_KEY: &str = "s3_endpoint";
 const S3_ACCESS_KEY_ID_KEY: &str = "s3_access_key_id";
@@ -45,7 +49,7 @@ pub fn FileReader(
     set_error_message: WriteSignal<Option<String>>,
     set_parquet_table: WriteSignal<Option<ParquetTable>>,
 ) -> impl IntoView {
-    let (active_tab, set_active_tab) = signal("url".to_string());
+    let (active_tab, set_active_tab) = signal("file".to_string());
 
     let (url_query, set_url_query) = query_signal::<String>("url");
     let default_url = {
@@ -258,6 +262,7 @@ pub fn FileReader(
         Some(url) => {
             // user provided an url, set it and run it.
             set_url.set(url);
+            set_active_tab.set("url".to_string());
             on_url_submit();
         }
         None => set_url.set(DEFAULT_URL.to_string()),

src/main.rs

@@ -5,13 +5,12 @@ use datafusion::{
     physical_plan::ExecutionPlan,
     prelude::{SessionConfig, SessionContext},
 };
-use file_reader::FileReader;
+use file_reader::{FileReader, INMEMORY_STORE};
 use leptos_router::{
     components::Router,
     hooks::{query_signal, use_query_map},
 };
 
-use object_store::memory::InMemory;
 use query_results::{export_to_csv_inner, export_to_parquet_inner, QueryResult, QueryResultView};
 use schema::SchemaSection;
 
@@ -38,9 +37,6 @@ use query_input::{execute_query_inner, QueryInput};
 mod settings;
 use settings::Settings;
 
-pub(crate) static INMEMORY_STORE: LazyLock<Arc<InMemory>> =
-    LazyLock::new(|| Arc::new(InMemory::new()));
-
 pub(crate) static SESSION_CTX: LazyLock<Arc<SessionContext>> = LazyLock::new(|| {
     let mut config = SessionConfig::new();
     config.options_mut().sql_parser.dialect = "PostgreSQL".to_string();

utils/parquet-viewer.py

@@ -1,25 +1,32 @@
 #!/usr/bin/env python3
 # encoding: utf-8
 
-from http.server import HTTPServer, SimpleHTTPRequestHandler
+from http.server import HTTPServer, BaseHTTPRequestHandler
+from socketserver import ThreadingMixIn
 import argparse
 import os
+import webbrowser
 
 
-class CORSRequestHandler(SimpleHTTPRequestHandler):
+class ThreadedHTTPServer(ThreadingMixIn, HTTPServer):
+    daemon_threads = True
+
+
+class CORSRequestHandler(BaseHTTPRequestHandler):
     def __init__(self, *args, file_path=None, **kwargs):
         self.file_path = file_path
         super().__init__(*args, **kwargs)
 
+    def log_message(self, format, *args):
+        pass
+
     def end_headers(self):
-        # Add CORS headers to allow all origins
         self.send_header('Access-Control-Allow-Origin', '*')
         self.send_header('Access-Control-Allow-Methods', 'GET, HEAD, OPTIONS')
         self.send_header('Access-Control-Allow-Headers', '*')
         super().end_headers()
 
     def do_OPTIONS(self):
-        # Handle preflight requests
         self.send_response(200)
         self.end_headers()
 
@@ -58,6 +65,7 @@ def do_GET(self):
             self.send_header('Content-Length', str(content_length))
             self.send_header('Content-type', 'application/vnd.apache.parquet')
             self.send_header('Content-Disposition', f'attachment; filename="{os.path.basename(self.file_path)}"')
+            self.send_header("Connection", "keep-alive")
             self.end_headers()
 
             with open(self.file_path, 'rb') as f:
@@ -73,11 +81,6 @@ def do_GET(self):
             self.send_error(500, f"Internal server error: {str(e)}")
 
     def do_HEAD(self):
-        expected_path = '/' + os.path.basename(self.file_path)
-        if self.path != expected_path:
-            self.send_error(404, f"File not found. Please use {expected_path}")
-            return
-
         try:
             file_size = os.path.getsize(self.file_path)
             self.send_response(200)
@@ -92,17 +95,24 @@ def do_HEAD(self):
             self.send_error(500, f"Internal server error: {str(e)}")
 
 def main():
-    parser = argparse.ArgumentParser(description='Start a CORS-enabled HTTP server')
+    parser = argparse.ArgumentParser(description='Open a local parquet file in parquet-viewer')
+    parser.add_argument('file', type=str,
+                       help='Path to the parquet file')
     parser.add_argument('-p', '--port', type=int, default=8003,
                        help='Port to run the server on (default: 8003)')
-    parser.add_argument('-f', '--file', type=str, required=True,
-                       help='Path to the file to serve')
+    parser.add_argument('--no-open', action='store_true',
+                       help='Do not open the browser automatically')
     args = parser.parse_args()
 
     file_name = os.path.basename(args.file)
     handler = lambda *handler_args: CORSRequestHandler(*handler_args, file_path=args.file)
-    httpd = HTTPServer(('127.0.0.1', args.port), handler)
-    print(f'Open in your browser: https://parquet-viewer.xiangpeng.systems/?url=http://127.0.0.1:{args.port}/{file_name}')
+    httpd = ThreadedHTTPServer(('127.0.0.1', args.port), handler)
+    url = f'https://parquet-viewer.xiangpeng.systems/?url=http://127.0.0.1:{args.port}/{file_name}'
+    print(f'Opening in your browser:\n{url}')
+
+    if not args.no_open:
+        webbrowser.open(url)
+
     httpd.serve_forever()