Added option to strip indirect citations

paperqa/docs.py

@@ -35,6 +35,7 @@
     maybe_is_text,
     md5sum,
     name_in_text,
+    strip_citations,
 )
 
 
@@ -59,6 +60,8 @@ class Docs(BaseModel, arbitrary_types_allowed=True, smart_union=True):
     memory: bool = False
     memory_model: Optional[BaseChatMemory] = None
     jit_texts_index: bool = False
+    # This is used to strip indirect citations that come up from the summary llm
+    strip_citations: bool = True
 
     # TODO: Not sure how to get this to work
     # while also passing mypy checks
@@ -505,6 +508,9 @@ async def process(match):
                 raise e
             if "not applicable" in context.lower() or "not relevant" in context.lower():
                 return None
+            if self.strip_citations:
+                # remove citations that collide with our grounded citations (for the answer LLM)
+                context = strip_citations(context)
             c = Context(
                 context=context,
                 text=Text(

paperqa/utils.py

@@ -97,3 +97,11 @@ def get_llm_name(llm: BaseLanguageModel) -> str:
         return llm.model_name  # type: ignore
     except AttributeError:
         return llm.model  # type: ignore
+
+
+def strip_citations(text: str) -> str:
+    # Combined regex for identifying citations (see unit tests for examples)
+    citation_regex = r"\b[\w\-]+\set\sal\.\s\([0-9]{4}\)|\((?:[^\)]*?[a-zA-Z][^\)]*?[0-9]{4}[^\)]*?)\)"
+    # Remove the citations from the text
+    text = re.sub(citation_regex, "", text, flags=re.MULTILINE)
+    return text

paperqa/version.py

@@ -1 +1 @@
-__version__ = "3.11.2"
+__version__ = "3.12.0"

setup.py

@@ -20,7 +20,7 @@
         "pypdf",
         "pydantic<2",
         "langchain>=0.0.303",
-        "openai >= 0.27.8",
+        "openai <1",
         "faiss-cpu",
         "PyCryptodome",
         "html2text",

tests/test_paperqa.py

@@ -15,14 +15,88 @@
 from paperqa.chains import get_score
 from paperqa.readers import read_doc
 from paperqa.types import Doc
-from paperqa.utils import maybe_is_html, maybe_is_text, name_in_text, strings_similarity
+from paperqa.utils import (
+    maybe_is_html,
+    maybe_is_text,
+    name_in_text,
+    strings_similarity,
+    strip_citations,
+)
 
 
 class TestHandler(AsyncCallbackHandler):
     async def on_llm_new_token(self, token: str, **kwargs: Any) -> None:
         print(token)
 
 
+# Assume strip_citations is imported or defined in this file.
+
+
+def test_single_author():
+    text = "This was first proposed by (Smith 1999)."
+    assert strip_citations(text) == "This was first proposed by ."
+
+
+def test_multiple_authors():
+    text = "Recent studies (Smith et al. 1999) show that this is true."
+    assert strip_citations(text) == "Recent studies  show that this is true."
+
+
+def test_multiple_citations():
+    text = "As discussed by several authors (Smith et al. 1999; Johnson 2001; Lee et al. 2003)."
+    assert strip_citations(text) == "As discussed by several authors ."
+
+
+def test_citations_with_pages():
+    text = "This is shown in (Smith et al. 1999, p. 150)."
+    assert strip_citations(text) == "This is shown in ."
+
+
+def test_citations_without_space():
+    text = "Findings by(Smith et al. 1999)were significant."
+    assert strip_citations(text) == "Findings bywere significant."
+
+
+def test_citations_with_commas():
+    text = "The method was adopted by (Smith, 1999, 2001; Johnson, 2002)."
+    assert strip_citations(text) == "The method was adopted by ."
+
+
+def test_citations_with_text():
+    text = "This was noted (see Smith, 1999, for a review)."
+    assert strip_citations(text) == "This was noted ."
+
+
+def test_no_citations():
+    text = "There are no references in this text."
+    assert strip_citations(text) == "There are no references in this text."
+
+
+def test_malformed_citations():
+    text = "This is a malformed citation (Smith 199)."
+    assert strip_citations(text) == "This is a malformed citation (Smith 199)."
+
+
+def test_edge_case_citations():
+    text = "Edge cases like (Smith et al.1999) should be handled."
+    assert strip_citations(text) == "Edge cases like  should be handled."
+
+
+def test_citations_with_special_characters():
+    text = "Some names have dashes (O'Neil et al. 2000; Smith-Jones 1998)."
+    assert strip_citations(text) == "Some names have dashes ."
+
+
+def test_citations_with_nonstandard_chars():
+    text = (
+        "In non-English languages, citations might look different (Müller et al. 1999)."
+    )
+    assert (
+        strip_citations(text)
+        == "In non-English languages, citations might look different ."
+    )
+
+
 def test_ablations():
     tests_dir = os.path.dirname(os.path.abspath(__file__))
     doc_path = os.path.join(tests_dir, "paper.pdf")