heldout_pool/codereverser__casparser-220edcfd7a16/solution/patch.diff

diff --git a/casparser/parsers/__init__.py b/casparser/parsers/__init__.py
index a56ad4c..f2d50d6 100644
--- a/casparser/parsers/__init__.py
+++ b/casparser/parsers/__init__.py
@@ -90,61 +90,71 @@ def read_cas_pdf(
     # parser / investor extractor calls — every pypdfium2 open re-runs
     # the password decrypt + content-stream parse, so the savings on
     # multi-page detailed statements are significant.
+    # Open the document once and ALWAYS close it before returning.
+    # pypdfium2 tracks pages / text-pages as children of the document;
+    # leaving the document open leaks those handles and makes pdfium emit
+    # "objects still open" at interpreter / library teardown. `close()`
+    # cascades to all child handles created during parsing. The parsed
+    # `data` is plain pydantic models holding no pdfium references, so it
+    # is safe to return after the document is closed.
     doc = _open_document(filename, password)
-
-    file_type = detect_file_type(filename, password, _doc=doc)
-    if file_type == FileType.UNKNOWN:
-        raise CASParseError(
-            "Could not identify the CAS issuer. Supported issuers are "
-            "CAMS, KFintech, NSDL, and CDSL."
-        )
-
-    if file_type in (FileType.CAMS, FileType.KFINTECH):
-        cas_type = detect_cas_type(filename, password, _doc=doc)
-        if cas_type == CASFileType.DETAILED:
-            from . import cams_detailed
-
-            data: Union[CASData, NSDLCASData] = cams_detailed.parse(
-                filename,
-                password,
-                file_type=file_type,
-                _doc=doc,
-            )
-        elif cas_type == CASFileType.SUMMARY:
-            from . import cams_summary
-
-            data = cams_summary.parse(
-                filename,
-                password,
-                file_type=file_type,
-                _doc=doc,
-            )
-        else:
+    try:
+        file_type = detect_file_type(filename, password, _doc=doc)
+        if file_type == FileType.UNKNOWN:
             raise CASParseError(
-                "Could not identify whether this is a DETAILED or " "SUMMARY CAMS / KFin statement."
+                "Could not identify the CAS issuer. Supported issuers are "
+                "CAMS, KFintech, NSDL, and CDSL."
             )
-        if sort_transactions and isinstance(data, CASData):
-            data = _sort_transactions(data)
-    elif file_type == FileType.NSDL:
-        from . import nsdl
 
-        data = nsdl.parse_nsdl(
-            filename,
-            password,
-            file_type=FileType.NSDL,
-            _doc=doc,
-        )
-    elif file_type == FileType.CDSL:
-        from . import cdsl
+        if file_type in (FileType.CAMS, FileType.KFINTECH):
+            cas_type = detect_cas_type(filename, password, _doc=doc)
+            if cas_type == CASFileType.DETAILED:
+                from . import cams_detailed
 
-        data = cdsl.parse_cdsl(
-            filename,
-            password,
-            file_type=FileType.CDSL,
-            _doc=doc,
-        )
-    else:  # pragma: no cover — handled above
-        raise CASParseError(f"Unsupported file type: {file_type}")
+                data: Union[CASData, NSDLCASData] = cams_detailed.parse(
+                    filename,
+                    password,
+                    file_type=file_type,
+                    _doc=doc,
+                )
+            elif cas_type == CASFileType.SUMMARY:
+                from . import cams_summary
+
+                data = cams_summary.parse(
+                    filename,
+                    password,
+                    file_type=file_type,
+                    _doc=doc,
+                )
+            else:
+                raise CASParseError(
+                    "Could not identify whether this is a DETAILED or "
+                    "SUMMARY CAMS / KFin statement."
+                )
+            if sort_transactions and isinstance(data, CASData):
+                data = _sort_transactions(data)
+        elif file_type == FileType.NSDL:
+            from . import nsdl
+
+            data = nsdl.parse_nsdl(
+                filename,
+                password,
+                file_type=FileType.NSDL,
+                _doc=doc,
+            )
+        elif file_type == FileType.CDSL:
+            from . import cdsl
+
+            data = cdsl.parse_cdsl(
+                filename,
+                password,
+                file_type=FileType.CDSL,
+                _doc=doc,
+            )
+        else:  # pragma: no cover — handled above
+            raise CASParseError(f"Unsupported file type: {file_type}")
+    finally:
+        doc.close()
 
     if output == "dict":
         return data
diff --git a/casparser/parsers/cdsl.py b/casparser/parsers/cdsl.py
index 0b82d30..cf3774b 100644
--- a/casparser/parsers/cdsl.py
+++ b/casparser/parsers/cdsl.py
@@ -59,6 +59,13 @@ INE_ISIN_RE = re.compile(r"^IN[E9][0-9A-Z]{8}\d$")
 # `0.196` unit balance prints as `.196`), so the integer part is
 # optional when a decimal part is present.
 NUMERIC_RE = re.compile(r"^-?(?:[\d,]+(?:\.\d+)?|\.\d+)$")
+# A long CDSL folio wraps its tail onto the next display line, which the
+# extractor emits as a separate cell — e.g. folio "91012112582/0" splits
+# into "910121125" | "82/0". The tail is a `<digits>/<digits>` token; we
+# splice it back onto the folio head (the on-page hyphen at the wrap is
+# only a line-break indicator, absent from the authoritative
+# "Folio No :" block).
+FOLIO_TAIL_RE = re.compile(r"\d+/\d+")
 
 PERIOD_RE = re.compile(
     r"(?:for\s+the\s+period\s+from|statement\s+for\s+the\s+period\s+from)\s+"
@@ -543,14 +550,29 @@ def _parse_mf_holdings_row(
 
     # Folio = cell right after ISIN (`<digits>/<digits>` or just digits).
     folio = None
+    folio_end = isin_idx + 1
     if isin_idx + 1 < len(block.cells):
         folio = block.cells[isin_idx + 1].text.strip() or None
 
+    # A long folio wraps its tail into the NEXT cell:
+    #   "910121125" | "82/0" | DIRECT | units | ...
+    # The real folio "91012112582/0" is split across two cells. Splice
+    # the `<digits>/<digits>` tail back on (no separator — the on-page
+    # hyphen is just the wrap indicator); otherwise the folio is
+    # truncated to its head and the tail is mistaken for the
+    # distribution-mode column below.
+    if folio and isin_idx + 2 < len(block.cells):
+        tail = block.cells[isin_idx + 2].text.strip()
+        if FOLIO_TAIL_RE.fullmatch(tail):
+            folio = folio + tail
+            folio_end = isin_idx + 2
+
     # Discriminate 13-cell (has ARN/DIRECT) vs 7-cell (no such column).
-    has_distrib_col = isin_idx + 2 < len(block.cells) and not _looks_numeric(
-        block.cells[isin_idx + 2].text
-    )
-    data_start = isin_idx + (3 if has_distrib_col else 2)
+    # The distribution-mode discriminator is the cell right after the
+    # folio (which may span two cells when it wrapped).
+    disc_idx = folio_end + 1
+    has_distrib_col = disc_idx < len(block.cells) and not _looks_numeric(block.cells[disc_idx].text)
+    data_start = disc_idx + (1 if has_distrib_col else 0)
     numerics = [c.text.strip() for c in block.cells[data_start:] if _looks_numeric(c.text)]
     if len(numerics) < 3:
         return None