typo in setup.py

Replace extension checking with python-magic

.github/workflows/python-app.yml

@@ -16,19 +16,20 @@ jobs:
 
     steps:
     - uses: actions/checkout@v2
-    - name: Set up Python 3.8
+    - name: Set up Python 3.7
       uses: actions/setup-python@v2
       with:
-        python-version: 3.8
+        python-version: 3.7
     - name: Install dependencies
       run: |
         python -m pip install --upgrade pip
-        pip install pycodestyle pytest coverage
-        if [ -f requirements.txt ]; then pip install -r requirements.txt; fi
+        pip install flake8
+        python setup.py develop
+        if [ -f requirements/base.txt ]; then pip install -r requirements/base.txt; fi
         sudo apt install poppler-utils pdftk
-    - name: Lint with pycodestyle
+    - name: Lint with flake8
       run: |
-        pycodestyle mglib --count --max-line-length=127 --statistics
-    - name: Test with pytest
+        flake8 mglib
+    - name: Run tests
       run: |
-        coverage run -m pytest
+        python test/run.py

README.md

@@ -11,3 +11,8 @@ Python Package containing modules shared across all [Papermerge Project](https:/
 ## Run tests
 
     python test/run.py
+
+
+## Requirements
+
+    python >= 3.7

mglib/pdfinfo.py

@@ -2,6 +2,7 @@ import os
 import re
 import subprocess
 import logging
+from magic import from_file
 
 from .conf import settings
 from .exceptions import FileTypeNotSupported
@@ -63,18 +64,18 @@ def get_pagecount(filepath):
     if os.path.isdir(filepath):
         raise ValueError("Filepath %s is a directory!" % filepath)
 
-    base, ext = os.path.splitext(filepath)
+    mime_type = from_file(filepath, mime=True)
 
     # pure images (png, jpeg) have only one page :)
-    if ext and ext.lower() in ('.jpeg', '.png', '.jpg'):
+    if mime_type in ['image/png', 'image/jpeg', 'image/jpg']:
         # whatever png/jpg image is there - it is
         # considered by default one page document.
         return 1
 
-    if ext and ext.lower() in ('.tiff', ):
+    if mime_type == 'image/tiff':
         return get_tiff_pagecount(filepath)
 
-    if ext and ext.lower() not in ('.pdf', '.tiff'):
+    if mime_type != 'application/pdf':
         raise FileTypeNotSupported(
             "Only jpeg, png, pdf and tiff are handled by this"
             " method"

mglib/stapler.py

@@ -285,8 +285,6 @@ def reorder_pages(
 
 
 def delete_pages(src, dst, page_numbers):
-    page_count = get_pagecount(src)
-
     cmd = [
         settings.BINARY_STAPLER,
         "del",

mglib/utils.py

@@ -87,7 +87,7 @@ def get_assigns_after_delete(total_pages, deleted_pages):
     ]
     """
     if total_pages < len(deleted_pages):
-        err_msg = f"total_pages < deleted_pages"
+        err_msg = "total_pages < deleted_pages"
         raise ValueError(err_msg)
 
     # only numbers of pages which were not deleted
@@ -136,7 +136,7 @@ def try_load_config(config_locations, config_env_var_name):
             try:
                 cfg_papermerge, cfg_file_found = load_config(config_file)
             except FileNotFoundError:
-                err_msg = f"Failed attempted to read" +\
+                err_msg = "Failed attempted to read" +\
                     f" configuration file '{config_file}'" +\
                     f" pointed by environment variable '{config_env_var_name}'"
                 raise FileNotFoundError(err_msg)

requirements/base.txt

@@ -0,0 +1 @@
+python-magic

setup.py

@@ -6,7 +6,7 @@ with open("README.md", "r") as fh:
 
 setup(
     name="mglib",
-    version="1.2.8",
+    version="1.3.0",
     author="Eugen Ciur",
     author_email="eugen@papermerge.com",
     url="https://github.com/papermerge/mglib",
@@ -14,7 +14,7 @@ setup(
     long_description=long_description,
     long_description_content_type="text/markdown",
     license="Apache 2.0 License",
-    keywords="common, pacakge, shared, papermerge",
+    keywords="common, package, shared, papermerge, pdf, ocr, dms",
     packages=find_packages(),
     classifiers=[
         "Programming Language :: Python :: 3",

test/data/berlin.jpeg

@@ -1,2 +1,2 @@
-I am not even binary!
+ÿØÿØI am not even binary!
 The idea is to test pdfinfo.get_pagecount

test/data/berlin.jpg

@@ -1 +1 @@
-well... I am text! But who cares? The idea is to test pdfinfo.get_pagecount
+ÿØÿîwell... I am text! But who cares? The idea is to test pdfinfo.get_pagecount