(zeppelin) branch master updated: [ZEPPELIN-6495] Use URLSearchParams in terminal getParams helper

[gyeongtae]

This is an automated email from the ASF dual-hosted git repository.

ParkGyeongTae pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/zeppelin.git


The following commit(s) were added to refs/heads/master by this push:
     new 0b8559a237 [ZEPPELIN-6495] Use URLSearchParams in terminal getParams 
helper
0b8559a237 is described below

commit 0b8559a23799be0bd77d239f8be0c4c3f8b4dfee
Author: 김예나 <101786858+kimye...@users.noreply.github.com>
AuthorDate: Tue Jul 14 09:32:08 2026 +0900

    [ZEPPELIN-6495] Use URLSearchParams in terminal getParams helper
    
    ### What is this PR for?
    The `%sh.terminal` frontend helper `getParams(key)` in 
`shell/src/main/resources/html/js/index.js` parsed the query string by hand: it 
built a `RegExp` over `location.search`, sliced it with `substr(1)`, and 
decoded matches with the deprecated global `unescape(...)`. This is harder to 
read than the platform API and can diverge from standard URL parsing for 
encoded/repeated/special characters.
    
    This PR replaces that logic with the standard `URLSearchParams` API. 
`URLSearchParams.get(key)` returns `null` for missing keys, so the existing 
behavior for the `noteId`/`paragraphId` parameters consumed by the terminal 
page is preserved, while relying on standard URL decoding instead of the 
deprecated `unescape`.
    
    ```js
    function getParams(key) {
        var params = new URLSearchParams(location.search);
        return params.get(key);
    }
    ```
    
    The `noteId`/`paragraphId`/`t` values the server generates (see 
`TerminalInterpreter#createTerminalDashboard`) are plain alphanumeric IDs and a 
numeric timestamp, so decoded values are identical to the previous 
implementation for all real inputs.
    
    ### What type of PR is it?
    Refactoring
    
    ### Todos
    * [x] - Replace `substr(...)`/`unescape(...)` in `getParams` with 
`URLSearchParams`
    
    ### What is the Jira issue?
    * https://issues.apache.org/jira/browse/ZEPPELIN-6495
    
    ### How should this be tested?
    * The changed file is a static, vendored frontend resource (`index.js`); 
the `shell` module has no JavaScript test harness (no `package.json`, and the 
existing `TerminalInterpreterTest` covers the Java interpreter, not this 
script), so no automated JS test is added.
    * Behavior was verified equivalent to the previous implementation:
      * encoded value (e.g. `%2F`) → decoded (`/`)
      * missing key → `null`
      * empty value (`key=`) → `""`
    * Optional manual check: open a `%sh.terminal` paragraph so the terminal 
dashboard loads `...?noteId=<id>&paragraphId=<id>&t=<ts>`, and confirm the 
terminal connects and `TERMINAL_READY` carries the correct 
`noteId`/`paragraphId`.
    
    ### Screenshots (if appropriate)
    N/A
    
    ### Questions:
    * Does the license files need to update? No — the file already carries the 
Apache License 2.0 header and no new file is added.
    * Is there breaking changes for older versions? No.
    * Does this needs documentation? No.
    
    
    Closes #5294 from kimyenac/ZEPPELIN-6495.
    
    Signed-off-by: ParkGyeongTae <gyeong...@apache.org>
---
 shell/src/main/resources/html/js/index.js | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/shell/src/main/resources/html/js/index.js 
b/shell/src/main/resources/html/js/index.js
index f9cc7b7c78..a1a96b8ad7 100644
--- a/shell/src/main/resources/html/js/index.js
+++ b/shell/src/main/resources/html/js/index.js
@@ -113,12 +113,8 @@ function setupHterm() {
 }
 
 function getParams(key) {
-    var reg = new RegExp("(^|&)" + key + "=([^&]*)(&|$)");
-    var r = location.search.substr(1).match(reg);
-    if (r != null) {
-        return unescape(r[2]);
-    }
-    return null;
+    var params = new URLSearchParams(location.search);
+    return params.get(key);
 };
 
 // This will be whatever normal entry/initialization point your project uses.