@scriptaddicts/gas-error-handler - npm Package Compare versions

Comparing version 1.0.1 to 1.1.0

package.json

 {
 	"name": "@scriptaddicts/gas-error-handler",
-	"version": "1.0.1",
+	"version": "1.1.0",
 	"description": "Methods to handle error in GAS",
@@ -5,0 +5,0 @@ "author": {

src/ErrorHandler.gs.js

@@ -9,2 +9,3 @@ /****************************************************************
  * expBackoff()
+ * urlFetchWithExpBackOff()
  * logError()
@@ -36,51 +37,98 @@ *
 function expBackoff(func) {
+  
+  // execute func() then retry 5 times at most if errors
   for (var n = 0; n < 6; n++) {
-    try {
-      return func();
+    // actual exponential backoff
+    n && Utilities.sleep((Math.pow(2, n-1) * 1000) + (Math.round(Math.random() * 1000)));
+    
+    var response = undefined;
+    var error = undefined;
+    
+    var noError = true;
+    var isUrlFetchResponse = false;
+    
+    // Try / catch func()
+    try { response = func() }
+    catch(err) {
+      error = err;
+      noError = false;
     }
-    catch(e) {
-      if (e.message) {
-        // Check for errors thrown by Google APIs on which there's no need to retry
-        // eg: "Access denied by a security policy established by the administrator of your organization. 
-        //      Please contact your administrator for further assistance."
-        if (e.message.indexOf('Invalid requests') !== -1
-            || e.message.indexOf('Access denied') !== -1
-            || e.message.indexOf('Mail service not enabled') !== -1) {
-          throw e;
-        }
-        
-        // TODO: YAMM specific ?: move to YAMM code
-        else if (e.message.indexOf('response too large') !== -1) {
-          // Thrown after calling Gmail.Users.Threads.get()
-          // maybe because a specific thread contains too many messages
-          // best to skip the thread
-          return null;
-        }
+    
+    
+    // Handle retries on UrlFetch calls with muteHttpExceptions
+    if (noError && typeof response.getResponseCode === "function") {
+      isUrlFetchResponse = true;
+      
+      var responseCode = response.getResponseCode();
+      
+      // Only perform retries on error 500 for now
+      if (responseCode === 500) {
+        error = response;
+        noError = false;
       }
+    }
+    
+    // Return result that is not an error
+    if (noError) return response;
+    
+    
+    // Process error retry
+    if (!isUrlFetchResponse && error.message) {
+      // Check for errors thrown by Google APIs on which there's no need to retry
+      // eg: "Access denied by a security policy established by the administrator of your organization. 
+      //      Please contact your administrator for further assistance."
+      if (error.message.indexOf('Invalid requests') !== -1
+          || error.message.indexOf('Access denied') !== -1
+          || error.message.indexOf('Mail service not enabled') !== -1) {
+        throw error;
+      }
       
-      if (n === 5) {
-        // 'User-rate limit exceeded' is always followed by 'Retry after' + timestamp
-        // Maybe we should parse the timestamp to check how long we need to wait 
-        // and if we should abort or not
-        if (e.message && e.message.indexOf('User-rate limit exceeded') !== -1) {
-          ErrorHandler.logError(e, {
-            shouldInvestigate: true,
-            failedAfter5Retries: true,
-            context: "Exponential Backoff"
-          });
-          return null;
-        }
-        
-        // Investigate on errors that are still happening after 5 retries
-        // Especially error "Not Found" - does it make sense to retry on it?
-        ErrorHandler.logError(e, {
-          failedAfter5Retries: true,
-          context: "Exponential Backoff"
-        });
-        
-        throw e;
+      // TODO: YAMM specific ?: move to YAMM code
+      else if (error.message.indexOf('response too large') !== -1) {
+        // Thrown after calling Gmail.Users.Threads.get()
+        // maybe because a specific thread contains too many messages
+        // best to skip the thread
+        return null;
       }
+    }
+    
+  }
+  
+  
+  // Action after last re-try
+  if (isUrlFetchResponse) {
+    
+    ErrorHandler.logError(new Error(response.getContentText()), {
+      shouldInvestigate: true,
+      failedAfter5Retries: true,
+      urlFetchWithMuteHttpExceptions: true,
+      context: "Exponential Backoff"
+    });
+    
+    return response;
+  }
+  
+  else {
+    // 'User-rate limit exceeded' is always followed by 'Retry after' + timestamp
+    // Maybe we should parse the timestamp to check how long we need to wait 
+    // and if we should abort or not
+    // 'User Rate Limit Exceeded' (without '-') isn't followed by 'Retry after' and it makes sense to retry
+    if (error.message && error.message.indexOf('User-rate limit exceeded') !== -1) {
+      ErrorHandler.logError(error, {
+        shouldInvestigate: true,
+        failedAfter5Retries: true,
+        context: "Exponential Backoff"
+      });
       
-      Utilities.sleep((Math.pow(2, n) * 1000) + (Math.round(Math.random() * 1000)));
+      return null;
     }
+    
+    // Investigate on errors that are still happening after 5 retries
+    // Especially error "Not Found" - does it make sense to retry on it?
+    ErrorHandler.logError(error, {
+      failedAfter5Retries: true,
+      context: "Exponential Backoff"
+    });
+    
+    throw error;
   }
@@ -90,2 +138,20 @@ }
 /**
+ * Helper function to automatically handles exponential backoff on UrlFetch use
+ *
+ * @param {string} url
+ * @param {Object} params
+ * 
+ * @return {UrlFetchApp.HTTPResponse}  - fetch response
+ */
+function urlFetchWithExpBackOff(url, params) {
+  params = params || {};
+  
+  params.muteHttpExceptions = true;
+  
+  return ErrorHandler.expBackoff(function(){
+    return UrlFetchApp.fetch(url, params);
+  });
+}
+
+/**
  * If we simply log the error object, only the error message will be submitted to Stackdriver Logging
@@ -148,2 +214,3 @@ * Best to re-write the error as a new object to get lineNumber & stack trace
   expBackoff: expBackoff,
+  urlFetchWithExpBackOff: urlFetchWithExpBackOff,
   logError: logError
@@ -150,0 +217,0 @@ };

New alerts

Network access

Supply chain risk

This module accesses the network.

Found 1 instance in 1 package

Long strings

Supply chain risk

Contains long string literals, which may be a sign of obfuscated or packed code.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

9042

23.54%

Lines of code

208

30%

Worsened metrics

Number of low supply chain risk alerts

2

Infinity%

Number of medium supply chain risk alerts

2

Infinity%

No dependency changes