From 21ca9e5f7d12698c152e6eb8e251b85191661724 Mon Sep 17 00:00:00 2001
From: Stony Wang <stony@tianrang-inc.com>
Date: Wed, 31 Jan 2024 23:12:41 +0800
Subject: [PATCH] Change the way to share and clean up temp directory. Add
 _get_temp_path to make sure to access tmp pdf file in the same way. Update
 _save_page parameters to meet the change.

---
 camelot/handlers.py | 58 ++++++++++++++++++++++++++-------------------
 1 file changed, 34 insertions(+), 24 deletions(-)

diff --git a/camelot/handlers.py b/camelot/handlers.py
index 66ee1697..3a2bba1a 100644
--- a/camelot/handlers.py
+++ b/camelot/handlers.py
@@ -1,5 +1,8 @@
 import os
+import shutil
 import sys
+import tempfile
+import weakref
 from pathlib import Path
 from typing import Union
 
@@ -10,7 +13,6 @@
 from .core import TableList
 from .parsers import Lattice
 from .parsers import Stream
-from .utils import TemporaryDirectory
 from .utils import download_url
 from .utils import get_page_layout
 from .utils import get_rotation
@@ -36,6 +38,8 @@ class PDFHandler:
     """
 
     def __init__(self, filepath: Union[StrByteType, Path], pages="1", password=None):
+        self.tempdir = tempfile.mkdtemp()
+        self._finalizer = weakref.finalize(self, shutil.rmtree, self.tempdir)
         if is_url(filepath):
             filepath = download_url(filepath)
         self.filepath: Union[StrByteType, Path] = filepath
@@ -56,15 +60,13 @@ def _get_pages(self, pages):
 
         Parameters
         ----------
-        filepath : str
-            Filepath or URL of the PDF file.
         pages : str, optional (default: '1')
             Comma-separated page numbers.
             Example: '1,3,4' or '1,4-end' or 'all'.
 
         Returns
         -------
-        P : list
+        result : list[int]
             List of int page numbers.
 
         """
@@ -95,24 +97,33 @@ def _get_pages(self, pages):
             result.extend(range(p["start"], p["end"] + 1))
         return sorted(set(result))
 
-    def _save_page(self, filepath: Union[StrByteType, Path], page, temp):
-        """Saves specified page from PDF into a temporary directory.
+    def _get_temp_path(self, page, rotated=False):
+        """Generate page path with temp directory.
 
         Parameters
         ----------
-        filepath : str
-            Filepath or URL of the PDF file.
         page : int
             Page number.
-        temp : str
-            Tmp directory.
+        rotated: bool (default: False)
+            Switch to generate temp file name.
+        """
+        if rotated is False:
+            return os.path.join(self.tempdir, f"page-{page}.pdf")
+        else:
+            return os.path.join(self.tempdir, f"p-{page}_rotated.pdf")
+
+    def _save_page(self, page):
+        """Saves specified page from PDF into a temporary directory.
 
+        Parameters
+        ----------
+        page : int
+            Page number.
         """
-        infile = PdfReader(filepath, strict=False)
+        infile = PdfReader(self.filepath, strict=False)
         if infile.is_encrypted:
             infile.decrypt(self.password)
-        fpath = os.path.join(temp, f"page-{page}.pdf")
-        froot, fext = os.path.splitext(fpath)
+        fpath = self._get_temp_path(page)
         p = infile.pages[page - 1]
         outfile = PdfWriter()
         outfile.add_page(p)
@@ -125,7 +136,7 @@ def _save_page(self, filepath: Union[StrByteType, Path], page, temp):
         vertical_text = get_text_objects(layout, ltype="vertical_text")
         rotation = get_rotation(chars, horizontal_text, vertical_text)
         if rotation != "":
-            fpath_new = "".join([froot.replace("page", "p"), "_rotated", fext])
+            fpath_new = self._get_temp_path(page, rotated=True)
             os.rename(fpath, fpath_new)
             instream = open(fpath_new, "rb")
             infile = PdfReader(instream, strict=False)
@@ -171,14 +182,13 @@ def parse(
             layout_kwargs = {}
 
         tables = []
-        with TemporaryDirectory() as tempdir:
-            for p in self.pages:
-                self._save_page(self.filepath, p, tempdir)
-            pages = [os.path.join(tempdir, f"page-{p}.pdf") for p in self.pages]
-            parser = Lattice(**kwargs) if flavor == "lattice" else Stream(**kwargs)
-            for p in pages:
-                t = parser.extract_tables(
-                    p, suppress_stdout=suppress_stdout, layout_kwargs=layout_kwargs
-                )
-                tables.extend(t)
+        for p in self.pages:
+            self._save_page(p)
+        pages = [self._get_temp_path(p) for p in self.pages]
+        parser = Lattice(**kwargs) if flavor == "lattice" else Stream(**kwargs)
+        for p in pages:
+            t = parser.extract_tables(
+                p, suppress_stdout=suppress_stdout, layout_kwargs=layout_kwargs
+            )
+            tables.extend(t)
         return TableList(sorted(tables))