diffpdf 0.3.0__tar.gz → 0.3.1__tar.gz

{diffpdf-0.3.0 → diffpdf-0.3.1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: diffpdf
-Version: 0.3.0
+Version: 0.3.1
 Summary: A tool for comparing PDF files
 Project-URL: Homepage, https://github.com/JustusRijke/DiffPDF
 Project-URL: Issues, https://github.com/JustusRijke/DiffPDF/issues

{diffpdf-0.3.0 → diffpdf-0.3.1}/src/diffpdf/__init__.py

@@ -15,13 +15,13 @@ def diffpdf(
     output_dir: str | Path | None = None,
     verbosity: int = 0,
     save_log: bool = False,
-) -> None:
+) -> bool:
     ref_path = Path(reference) if isinstance(reference, str) else reference
     actual_path = Path(actual) if isinstance(actual, str) else actual
     out_path = Path(output_dir) if isinstance(output_dir, str) else output_dir
 
     logger = setup_logging(verbosity, save_log)
-    compare_pdfs(ref_path, actual_path, threshold, dpi, out_path, logger)
+    return compare_pdfs(ref_path, actual_path, threshold, dpi, out_path, logger)
 
 
 __all__ = ["diffpdf", "__version__"]

{diffpdf-0.3.0 → diffpdf-0.3.1}/src/diffpdf/cli.py

@@ -37,7 +37,10 @@ def cli(reference, actual, threshold, dpi, output_dir, verbosity, save_log):
     logger.debug("Debug logging enabled")
 
     try:
-        compare_pdfs(reference, actual, threshold, dpi, output_dir, logger)
+        if compare_pdfs(reference, actual, threshold, dpi, output_dir, logger):
+            sys.exit(0)
+        else:
+            sys.exit(1)
     except Exception as e:  # pragma: no cover
         logger.critical(f"Error: {e}", exc_info=True)
         sys.exit(2)

diffpdf-0.3.1/src/diffpdf/comparators.py

@@ -0,0 +1,31 @@
+from pathlib import Path
+
+from .hash_check import check_hash
+from .page_check import check_page_counts
+from .text_check import check_text_content
+from .visual_check import check_visual_content
+
+
+def compare_pdfs(
+    ref: Path, actual: Path, threshold: float, dpi: int, output_dir: Path | None, logger
+) -> bool:
+    logger.info("[1/4] Checking file hashes...")
+    if check_hash(ref, actual):
+        logger.info("Files are identical (hash match)")
+        return True
+    logger.info("Hashes differ, continuing checks")
+
+    logger.info("[2/4] Checking page counts...")
+    if not check_page_counts(ref, actual, logger):
+        return False
+
+    logger.info("[3/4] Checking text content...")
+    if not check_text_content(ref, actual, logger):
+        return False
+
+    logger.info("[4/4] Checking visual content...")
+    if not check_visual_content(ref, actual, threshold, dpi, output_dir, logger):
+        return False
+
+    logger.info("PDFs are equivalent")
+    return True

{diffpdf-0.3.0 → diffpdf-0.3.1}/src/diffpdf/hash_check.py

@@ -1,5 +1,4 @@
 import hashlib
-import sys
 from pathlib import Path
 
 
@@ -11,14 +10,8 @@ def compute_file_hash(filepath: Path) -> str:
     return sha256.hexdigest()
 
 
-def check_hash(ref: Path, actual: Path, logger) -> None:
-    logger.info("[1/4] Checking file hashes...")
-
+def check_hash(ref: Path, actual: Path) -> bool:
     ref_hash = compute_file_hash(ref)
     actual_hash = compute_file_hash(actual)
 
-    if ref_hash == actual_hash:
-        logger.info("Files are identical (hash match)")
-        sys.exit(0)
-
-    logger.info("Hashes differ, continuing checks")
+    return ref_hash == actual_hash

{diffpdf-0.3.0 → diffpdf-0.3.1}/src/diffpdf/page_check.py

@@ -1,4 +1,3 @@
-import sys
 from pathlib import Path
 
 import fitz
@@ -11,14 +10,13 @@ def get_page_count(pdf_path: Path) -> int:
     return count
 
 
-def check_page_counts(ref: Path, actual: Path, logger) -> None:
-    logger.info("[2/4] Checking page counts...")
-
+def check_page_counts(ref: Path, actual: Path, logger) -> bool:
     ref_count = get_page_count(ref)
     actual_count = get_page_count(actual)
 
     if ref_count != actual_count:
         logger.error(f"Page count mismatch: expected {ref_count}, got {actual_count}")
-        sys.exit(1)
+        return False
 
     logger.info(f"Page counts match ({ref_count} pages)")
+    return True

{diffpdf-0.3.0 → diffpdf-0.3.1}/src/diffpdf/text_check.py

@@ -1,6 +1,5 @@
 import difflib
 import re
-import sys
 from pathlib import Path
 from typing import Iterable
 
@@ -33,9 +32,7 @@ def generate_diff(
     return diff
 
 
-def check_text_content(ref: Path, actual: Path, logger) -> None:
-    logger.info("[3/4] Checking text content...")
-
+def check_text_content(ref: Path, actual: Path, logger) -> bool:
     # Extract text and remove whitespace
     ref_text = re.sub(r"\s+", " ", extract_text(ref)).strip()
     actual_text = re.sub(r"\s+", " ", extract_text(actual)).strip()
@@ -44,6 +41,7 @@ def check_text_content(ref: Path, actual: Path, logger) -> None:
         diff = generate_diff(ref_text, ref, actual_text, actual)
         diff_text = "\n".join(diff)
         logger.error(f"Text content mismatch:\n {diff_text}")
-        sys.exit(1)
+        return False
 
-    logger.info("Text content matches")
+    logger.info("Text content identical")
+    return True

{diffpdf-0.3.0 → diffpdf-0.3.1}/src/diffpdf/visual_check.py

@@ -1,4 +1,3 @@
-import sys
 from pathlib import Path
 
 import fitz
@@ -10,7 +9,7 @@ def render_page_to_image(pdf_path: Path, page_num: int, dpi: int) -> Image.Image
     doc = fitz.open(pdf_path)
     page = doc[page_num]
     pix = page.get_pixmap(dpi=dpi)
-    img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
+    img = Image.frombytes("RGB", (pix.width, pix.height), pix.samples)
     doc.close()
     return img
 
@@ -33,9 +32,7 @@ def compare_images(
 
 def check_visual_content(
     ref: Path, actual: Path, threshold: float, dpi: int, output_dir: Path | None, logger
-) -> None:
-    logger.info("[4/4] Checking visual content...")
-
+) -> bool:
     if output_dir is not None:
         output_dir.mkdir(parents=True, exist_ok=True)
 
@@ -64,6 +61,7 @@ def check_visual_content(
 
     if failing_pages:
         logger.error(f"Visual mismatch on pages: {', '.join(map(str, failing_pages))}")
-        sys.exit(1)
+        return False
 
     logger.info("Visual content matches")
+    return True

diffpdf-0.3.1/tests/test_api.py

@@ -0,0 +1,12 @@
+from pathlib import Path
+
+from diffpdf import diffpdf
+
+TEST_ASSETS_DIR = Path(__file__).parent / "assets"
+
+
+def test_diffpdf():
+    assert diffpdf(
+        TEST_ASSETS_DIR / "pass/identical-A.pdf",
+        TEST_ASSETS_DIR / "pass/identical-B.pdf",
+    )

diffpdf-0.3.0/src/diffpdf/comparators.py

@@ -1,22 +0,0 @@
-import sys
-from pathlib import Path
-
-from .hash_check import check_hash
-from .page_check import check_page_counts
-from .text_check import check_text_content
-from .visual_check import check_visual_content
-
-
-def compare_pdfs(
-    ref: Path, actual: Path, threshold: float, dpi: int, output_dir: Path | None, logger
-) -> None:
-    check_hash(ref, actual, logger)
-
-    check_page_counts(ref, actual, logger)
-
-    check_text_content(ref, actual, logger)
-
-    check_visual_content(ref, actual, threshold, dpi, output_dir, logger)
-
-    logger.info("PDFs are equivalent")
-    sys.exit(0)

diffpdf-0.3.0/tests/test_api.py

@@ -1,16 +0,0 @@
-from pathlib import Path
-
-import pytest
-
-from diffpdf import diffpdf
-
-TEST_ASSETS_DIR = Path(__file__).parent / "assets"
-
-
-def test_diffpdf():
-    with pytest.raises(SystemExit) as exc_info:
-        diffpdf(
-            TEST_ASSETS_DIR / "pass/identical-A.pdf",
-            TEST_ASSETS_DIR / "pass/identical-B.pdf",
-        )
-    assert exc_info.value.code == 0