feat: 增加get_pdf_from_elsevier

apps/resm/models.py

@@ -1,5 +1,7 @@
 from django.db import models
 from apps.utils.models import BaseModel
+from django.conf import settings
+import os
 # Create your models here.
 
 class Paper(BaseModel):
@@ -43,6 +45,32 @@ class Paper(BaseModel):
     )
     search_word_first = models.TextField(default="cement")
 
+    def init_save_dir(self):
+        publication_date = self.publication_date
+        paper_dir = os.path.join(
+            settings.BASE_DIR,
+            "media/papers",
+            str(publication_date.year),
+            str(publication_date.month),
+            str(publication_date.day)
+        )
+        os.makedirs(paper_dir, exist_ok=True)
+        return paper_dir
+    
+    def save_file_xml(self, content):
+        safe_doi = self.doi.replace("/", "_")
+        paper_file = os.path.join(self.init_save_dir(), f"{safe_doi}.xml")
+        with open(paper_file, "wb") as f:
+            f.write(content.encode("utf-8"))
+
+    def save_file_pdf(self, content):
+        safe_doi = self.doi.replace("/", "_")
+        paper_file = os.path.join(self.init_save_dir(), f"{safe_doi}.pdf")
+        with open(paper_file, "wb") as f:
+            f.write(content)
+    
+
+    
 class PaperAbstract(BaseModel):
     paper = models.OneToOneField(
         Paper,

apps/resm/tasks.py

@@ -120,7 +120,8 @@ def get_abstract_from_elsevier(publication_year: int = None, number_of_task:int
                     continue
 
                 ns = {"dc": "http://purl.org/dc/elements/1.1/",
-                "ce": "http://www.elsevier.com/xml/common/dtd"}
+                "ce": "http://www.elsevier.com/xml/common/dtd",
+                "xocs": "http://www.elsevier.com/xml/xocs/dtd",}
                 abstract = root.xpath("//dc:description/text()", namespaces=ns)
                 if abstract:
                     PaperAbstract.objects.update_or_create(
@@ -140,25 +141,16 @@ def get_abstract_from_elsevier(publication_year: int = None, number_of_task:int
 
                 paras = root.xpath("//ce:para", namespaces=ns)
                 has_fulltext = len(paras) > 0
+                if has_fulltext is False:
+                    rawtexts = root.xpath("//xocs:rawtext/text()",namespaces=ns)
+                    if rawtexts and len(rawtexts[0].strip()) > 2000:
+                        has_fulltext = True
                 if has_fulltext:
                     paper.has_fulltext = True
                     paper.has_fulltext_xml = True
                     paper.fetch_status = "fulltext_ready"
                     
-                publication_date = paper.publication_date
-                paper_dir = os.path.join(
-                    settings.BASE_DIR,
-                    "media/papers",
-                    str(publication_date.year),
-                    str(publication_date.month),
-                    str(publication_date.day)
-                )
-                os.makedirs(paper_dir, exist_ok=True)
-
-                safe_doi = paper.doi.replace("/", "_")
-                paper_file = os.path.join(paper_dir, f"{safe_doi}.xml")
-                with open(paper_file, "wb") as f:
-                    f.write(xml_str.encode("utf-8"))
+                paper.save_file_xml(xml_str)
                 paper.save(update_fields=["has_abstract", 
                 "has_abstract_xml", "has_fulltext", 
                 "has_fulltext_xml", "update_time", "fetch_status", "fail_reason"])
@@ -175,4 +167,50 @@ def get_abstract_from_elsevier(publication_year: int = None, number_of_task:int
             },
             countdown=5,
         )
-    return f'{err_msg}, remaining {qs.count()} papers'
+    return f'{err_msg}, remaining {qs.count()} papers'
+
+
+def is_elsevier_pdf_task_enabled():
+    return cache.get("elsevier_pdf_task_enabled", True)
+
+
+@shared_task(base=CustomTask)
+def get_pdf_from_elsevier(number_of_task=100):
+    """
+    获取elsevier全文
+    """
+    if not is_elsevier_pdf_task_enabled():
+        return "stoped"
+    qs = Paper.objects.filter(has_fulltext=True, has_fulltext_pdf=False)
+    err_msg = ""
+    with requests.Session() as req:
+        for paper in qs[:number_of_task]:
+            if not is_elsevier_pdf_task_enabled():
+                break
+            params = {
+                "apiKey": ELSEVIER_APIKEY,
+                "httpAccept": "application/pdf"
+            }
+            try:
+                res = req.get(
+                    f"https://api.elsevier.com/content/article/doi/{paper.doi}",
+                    params=params,
+                    timeout=(3, 15)
+                )
+            except requests.RequestException:
+                err_msg = "elsevier_request_error"
+                break
+            if res.status_code == 200 and res.headers["content-type"] == "application/pdf":
+                paper.save_file_pdf(res.content)
+                paper.has_fulltext_pdf = True
+                paper.save(update_fields=["has_fulltext_pdf", "update_time"])
+    if is_elsevier_pdf_task_enabled():
+        current_app.send_task(
+            "apps.resm.tasks.get_pdf_from_elsevier",
+            kwargs={
+                "number_of_task": number_of_task,
+            },
+            countdown=5,
+        )
+    return f'{err_msg}, remaining {qs.count()} papers'
+