From 1d2c0817103c80c2d9b9622df7515458bf5e492f Mon Sep 17 00:00:00 2001
From: H <43509927+guoyuhao2330@users.noreply.github.com>
Date: Mon, 2 Sep 2024 18:49:09 +0800
Subject: [PATCH] Fix component PubMed  (#2195)

### What problem does this PR solve?


### Type of change

- [x] Bug Fix (non-breaking change which fixes an issue)
---
 agent/component/pubmed.py | 26 +++++++++++---------------
 1 file changed, 11 insertions(+), 15 deletions(-)

diff --git a/agent/component/pubmed.py b/agent/component/pubmed.py
index 19cbdecc0..ff97ec88b 100644
--- a/agent/component/pubmed.py
+++ b/agent/component/pubmed.py
@@ -15,6 +15,7 @@
 #
 from abc import ABC
 from Bio import Entrez
+import re
 import pandas as pd
 import xml.etree.ElementTree as ET
 from agent.settings import DEBUG
@@ -47,21 +48,16 @@ class PubMed(ComponentBase, ABC):
         try:
             Entrez.email = self._param.email
             pubmedids = Entrez.read(Entrez.esearch(db='pubmed', retmax=self._param.top_n, term=ans))['IdList']
-            pubmedcnt = ET.fromstring(
-                Entrez.efetch(db='pubmed', id=",".join(pubmedids), retmode="xml").read().decode("utf-8"))
-            pubmed_res = []
-            for child in pubmedcnt.findall("PubmedArticle"):
-                if child.find("MedlineCitation").find("Article").find("ArticleTitle").text:
-                    title_tmp = 'Title:' + child.find("MedlineCitation").find("Article").find("ArticleTitle").text
-                else:
-                    title_tmp = 'Title:' + "".join(
-                        [childtitle.text for childtitle in
-                         child.find("MedlineCitation").find("Article").find("ArticleTitle")])
-                url_tmp = '\nUrl:<a href=" https://pubmed.ncbi.nlm.nih.gov/' + child.find("MedlineCitation").find(
-                    "PMID").text + '">' + '</a>'
-                abstract_tmp = '\nAbstract:' + child.find("MedlineCitation").find("Article").find("Abstract").find(
-                    "AbstractText").text
-                pubmed_res.append({"content": title_tmp + url_tmp + abstract_tmp})
+            pubmedcnt = ET.fromstring(re.sub(r'<(/?)b>|<(/?)i>', '', Entrez.efetch(db='pubmed', id=",".join(pubmedids),
+                                                                                   retmode="xml").read().decode(
+                "utf-8")))
+            pubmed_res = [{"content": 'Title:' + child.find("MedlineCitation").find("Article").find(
+                "ArticleTitle").text + '\nUrl:<a href=" https://pubmed.ncbi.nlm.nih.gov/' + child.find(
+                "MedlineCitation").find("PMID").text + '">' + '</a>\n' + 'Abstract:' + (
+                                          child.find("MedlineCitation").find("Article").find("Abstract").find(
+                                              "AbstractText").text if child.find("MedlineCitation").find(
+                                              "Article").find("Abstract") else "No abstract available")} for child in
+                          pubmedcnt.findall("PubmedArticle")]
         except Exception as e:
             return PubMed.be_output("**ERROR**: " + str(e))