error handling (coderabbitai#56)

<!-- This is an auto-generated comment: release notes by openai -->
### Summary by OpenAI

**Release Notes**

This pull request includes improvements to error handling in the
`review.ts` file. The changes made aim to improve the reliability and
robustness of the code by handling potential errors more gracefully.
This is a "Bug fix" that increases the `limit` variable and adds
try-catch blocks around various sections of code to handle errors from
OpenAI and the GitHub API.
<!-- end of auto-generated comment: release notes by openai -->

src/review.ts

@@ -13,7 +13,7 @@ const token = core.getInput('token')
 const octokit = new Octokit({auth: `token ${token}`})
 const context = github.context
 const repo = context.repo
-const limit = pLimit(2)
+const limit = pLimit(4)
 
 const MAX_TOKENS_FOR_EXTRA_CONTENT = 2500
 
@@ -170,15 +170,20 @@ export const codeReview = async (
         const file_diff_tokens = tokenizer.get_token_count(file_diff)
         if (file_diff_tokens < MAX_TOKENS_FOR_EXTRA_CONTENT) {
           // summarize diff
-          const [summarize_resp] = await bot.chat(
-            prompts.render_summarize_file_diff(ins),
-            summarize_begin_ids
-          )
-          if (!summarize_resp) {
-            core.info('summarize: nothing obtained from openai')
+          try {
+            const [summarize_resp] = await bot.chat(
+              prompts.render_summarize_file_diff(ins),
+              summarize_begin_ids
+            )
+            if (!summarize_resp) {
+              core.info('summarize: nothing obtained from openai')
+              return null
+            } else {
+              return [filename, summarize_resp]
+            }
+          } catch (error) {
+            core.warning(`summarize: error from openai: ${error}`)
             return null
-          } else {
-            return [filename, summarize_resp]
           }
         }
       }
@@ -262,15 +267,19 @@ Tips:
           if (file_content.length > 0) {
             const file_content_tokens = tokenizer.get_token_count(file_content)
             if (file_content_tokens < MAX_TOKENS_FOR_EXTRA_CONTENT) {
-              // review file
-              const [resp, review_file_ids] = await bot.chat(
-                prompts.render_review_file(ins),
-                next_review_ids
-              )
-              if (!resp) {
-                core.info('review: nothing obtained from openai')
-              } else {
-                next_review_ids = review_file_ids
+              try {
+                // review file
+                const [resp, review_file_ids] = await bot.chat(
+                  prompts.render_review_file(ins),
+                  next_review_ids
+                )
+                if (!resp) {
+                  core.info('review: nothing obtained from openai')
+                } else {
+                  next_review_ids = review_file_ids
+                }
+              } catch (error) {
+                core.warning(`review: error from openai: ${error}`)
               }
             } else {
               core.info(
@@ -282,15 +291,19 @@ Tips:
           if (file_diff.length > 0) {
             const file_diff_tokens = tokenizer.get_token_count(file_diff)
             if (file_diff_tokens < MAX_TOKENS_FOR_EXTRA_CONTENT) {
-              // review diff
-              const [resp, review_diff_ids] = await bot.chat(
-                prompts.render_review_file_diff(ins),
-                next_review_ids
-              )
-              if (!resp) {
-                core.info('review: nothing obtained from openai')
-              } else {
-                next_review_ids = review_diff_ids
+              try {
+                // review diff
+                const [resp, review_diff_ids] = await bot.chat(
+                  prompts.render_review_file_diff(ins),
+                  next_review_ids
+                )
+                if (!resp) {
+                  core.info('review: nothing obtained from openai')
+                } else {
+                  next_review_ids = review_diff_ids
+                }
+              } catch (error) {
+                core.warning(`review: error from openai: ${error}`)
               }
             } else {
               core.info(
@@ -313,33 +326,41 @@ Tips:
               core.warning('No pull request found, skipping.')
               continue
             }
-            // get existing comments on the line
-            const all_chains = await commenter.get_conversation_chains_at_line(
-              context.payload.pull_request.number,
-              filename,
-              line,
-              COMMENT_REPLY_TAG
-            )
 
-            if (all_chains.length > 0) {
-              ins.comment_chain = all_chains
-            } else {
-              ins.comment_chain = 'no previous comments'
+            try {
+              // get existing comments on the line
+              const all_chains =
+                await commenter.get_conversation_chains_at_line(
+                  context.payload.pull_request.number,
+                  filename,
+                  line,
+                  COMMENT_REPLY_TAG
+                )
+
+              if (all_chains.length > 0) {
+                ins.comment_chain = all_chains
+              } else {
+                ins.comment_chain = 'no previous comments'
+              }
+            } catch (e: any) {
+              core.warning(
+                `Failed to get comments: ${e}, skipping. backtrace: ${e.stack}`
+              )
             }
 
-            const [response, patch_ids] = await bot.chat(
-              prompts.render_review_patch(ins),
-              next_review_ids
-            )
-            if (!response) {
-              core.info('review: nothing obtained from openai')
-              continue
-            }
-            next_review_ids = patch_ids
-            if (!options.review_comment_lgtm && response.includes('LGTM')) {
-              continue
-            }
             try {
+              const [response, patch_ids] = await bot.chat(
+                prompts.render_review_patch(ins),
+                next_review_ids
+              )
+              if (!response) {
+                core.info('review: nothing obtained from openai')
+                continue
+              }
+              next_review_ids = patch_ids
+              if (!options.review_comment_lgtm && response.includes('LGTM')) {
+                continue
+              }
               await commenter.review_comment(
                 context.payload.pull_request.number,
                 commits[commits.length - 1].sha,