From 850688a35096bfd78e863e41c62b0aae46de6c6f Mon Sep 17 00:00:00 2001
From: Harsh <115716485+Kaos599@users.noreply.github.com>
Date: Wed, 30 Oct 2024 19:58:02 +0530
Subject: [PATCH] Create uncompress.py

---
 pdfly/uncompress.py | 50 +++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 50 insertions(+)
 create mode 100644 pdfly/uncompress.py

diff --git a/pdfly/uncompress.py b/pdfly/uncompress.py
new file mode 100644
index 0000000..d0636ce
--- /dev/null
+++ b/pdfly/uncompress.py
@@ -0,0 +1,50 @@
+# uncompress.py
+"""Provides functionality to uncompress PDF files."""
+from pathlib import Path
+import zlib
+from typing import cast
+
+from pypdf import PdfReader, PdfWriter
+from pypdf.generic import IndirectObject, StreamObject
+
+
+def main(pdf: Path, output: Path) -> None:
+    reader = PdfReader(pdf)
+    writer = PdfWriter()
+
+    for page in reader.pages:
+        if "/Contents" in page:
+            contents = page["/Contents"]
+            if isinstance(contents, IndirectObject):
+                contents = contents.get_object()
+                
+            if isinstance(contents, list):
+                for content_obj in contents:
+                    if isinstance(content_obj, IndirectObject):
+                        content_stream = content_obj.get_object()
+                        decompress_content_stream(content_stream)
+            elif isinstance(contents, StreamObject): # type: ignore[unreachable]
+                decompress_content_stream(contents)
+
+
+        writer.add_page(page)
+
+    with open(output, "wb") as fp:
+        writer.write(fp)
+
+    orig_size = pdf.stat().st_size
+    uncomp_size = output.stat().st_size
+    ratio = uncomp_size / orig_size
+    print(f"Original Size  : {orig_size:,}")
+    print(f"Uncompressed Size: {uncomp_size:,} ({ratio * 100:.1f}% of original)")
+
+
+def decompress_content_stream(content: StreamObject) -> None:  # type: ignore[type-arg]
+    if content.get("/Filter") == "/FlateDecode":
+        try:
+            compressed_data = content.get_data()
+            uncompressed_data = zlib.decompress(compressed_data)
+            content.set_data(uncompressed_data)
+            content.update({"/Filter": None}) # type: ignore[arg-type]
+        except zlib.error as e:
+            print(f"Decompression error: {e}")