diff --git a/.gitignore b/.gitignore index 32b9f67..44ac756 100644 --- a/.gitignore +++ b/.gitignore @@ -12,3 +12,6 @@ docs/_build /build/ /dist/ /pdfminer.six.egg-info/ +tests/*.xml +tests/*.txt +.idea/ diff --git a/pdfminer/__init__.py b/pdfminer/__init__.py index 8644ca8..79844fe 100644 --- a/pdfminer/__init__.py +++ b/pdfminer/__init__.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + # -*- coding: utf-8 -*- __version__ = '20160614' diff --git a/pdfminer/arcfour.py b/pdfminer/arcfour.py index b2ba8a1..5c0e64c 100644 --- a/pdfminer/arcfour.py +++ b/pdfminer/arcfour.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + """ Python implementation of Arcfour encryption algorithm. See https://en.wikipedia.org/wiki/RC4 diff --git a/pdfminer/ascii85.py b/pdfminer/ascii85.py index dfd44e6..7fda67b 100644 --- a/pdfminer/ascii85.py +++ b/pdfminer/ascii85.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + """ Python implementation of ASCII85/ASCIIHex decoder (Adobe version). diff --git a/pdfminer/ccitt.py b/pdfminer/ccitt.py index 5bfda3b..0ed5a18 100644 --- a/pdfminer/ccitt.py +++ b/pdfminer/ccitt.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + # CCITT Fax decoder # # Bugs: uncompressed mode untested. diff --git a/pdfminer/cmapdb.py b/pdfminer/cmapdb.py index 52bba8b..f48fcf8 100644 --- a/pdfminer/cmapdb.py +++ b/pdfminer/cmapdb.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + """ Adobe character mapping (CMap) support. diff --git a/pdfminer/converter.py b/pdfminer/converter.py index 75969b5..c92f57d 100644 --- a/pdfminer/converter.py +++ b/pdfminer/converter.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + # -*- coding: utf-8 -*- import logging import re diff --git a/pdfminer/encodingdb.py b/pdfminer/encodingdb.py index 209ce54..870bd28 100644 --- a/pdfminer/encodingdb.py +++ b/pdfminer/encodingdb.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + import re from .psparser import PSLiteral from .glyphlist import glyphname2unicode diff --git a/pdfminer/fontmetrics.py b/pdfminer/fontmetrics.py index bf0be75..8b3779a 100644 --- a/pdfminer/fontmetrics.py +++ b/pdfminer/fontmetrics.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + """ Font metrics for the Adobe core 14 fonts. diff --git a/pdfminer/glyphlist.py b/pdfminer/glyphlist.py index 10e1008..848b097 100644 --- a/pdfminer/glyphlist.py +++ b/pdfminer/glyphlist.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + """ Mappings from Adobe glyph names to Unicode characters. diff --git a/pdfminer/high_level.py b/pdfminer/high_level.py index bd79158..e7892c5 100644 --- a/pdfminer/high_level.py +++ b/pdfminer/high_level.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python3 +3 # -*- coding: utf-8 -*- """ Functions that encapsulate "usual" use-cases for pdfminer, for use making diff --git a/pdfminer/image.py b/pdfminer/image.py index 1f58019..3bcad3b 100644 --- a/pdfminer/image.py +++ b/pdfminer/image.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + import struct import os import os.path diff --git a/pdfminer/latin_enc.py b/pdfminer/latin_enc.py index bb0c1eb..13886c8 100644 --- a/pdfminer/latin_enc.py +++ b/pdfminer/latin_enc.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + """ Standard encoding tables used in PDF. diff --git a/pdfminer/layout.py b/pdfminer/layout.py index 44cce01..f0fe222 100644 --- a/pdfminer/layout.py +++ b/pdfminer/layout.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + from .utils import INF from .utils import Plane from .utils import get_bound diff --git a/pdfminer/lzw.py b/pdfminer/lzw.py index 30e2205..3ab8147 100644 --- a/pdfminer/lzw.py +++ b/pdfminer/lzw.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + from io import BytesIO import six #Python 2+3 compatibility diff --git a/pdfminer/pdfcolor.py b/pdfminer/pdfcolor.py index 4db483e..6fe6eaa 100644 --- a/pdfminer/pdfcolor.py +++ b/pdfminer/pdfcolor.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + from .psparser import LIT import six #Python 2+3 compatibility diff --git a/pdfminer/pdfdevice.py b/pdfminer/pdfdevice.py index 76725a3..acf7d68 100644 --- a/pdfminer/pdfdevice.py +++ b/pdfminer/pdfdevice.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + from .pdffont import PDFUnicodeNotDefined from . import utils diff --git a/pdfminer/pdfdocument.py b/pdfminer/pdfdocument.py index 43d6b6e..3b2b551 100644 --- a/pdfminer/pdfdocument.py +++ b/pdfminer/pdfdocument.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + import re import struct import logging diff --git a/pdfminer/pdffont.py b/pdfminer/pdffont.py index d83888c..00361fa 100644 --- a/pdfminer/pdffont.py +++ b/pdfminer/pdffont.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + import sys import struct from io import BytesIO diff --git a/pdfminer/pdfinterp.py b/pdfminer/pdfinterp.py index fbc8e37..e7ba104 100644 --- a/pdfminer/pdfinterp.py +++ b/pdfminer/pdfinterp.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + import re import logging from io import BytesIO diff --git a/pdfminer/pdfpage.py b/pdfminer/pdfpage.py index 62c70a6..2cbe58e 100644 --- a/pdfminer/pdfpage.py +++ b/pdfminer/pdfpage.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + import logging from .psparser import LIT from .pdftypes import PDFObjectNotFound diff --git a/pdfminer/pdfparser.py b/pdfminer/pdfparser.py index d19e179..1dc17d7 100644 --- a/pdfminer/pdfparser.py +++ b/pdfminer/pdfparser.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + import logging from io import BytesIO from .psparser import PSStackParser diff --git a/pdfminer/pdftypes.py b/pdfminer/pdftypes.py index 0d74325..2ea6aea 100644 --- a/pdfminer/pdftypes.py +++ b/pdfminer/pdftypes.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + import zlib import logging from .lzw import lzwdecode diff --git a/pdfminer/psparser.py b/pdfminer/psparser.py index be435bd..644a8c0 100644 --- a/pdfminer/psparser.py +++ b/pdfminer/psparser.py @@ -1,4 +1,5 @@ -#!/usr/bin/python +#!/usr/bin/env python + # -*- coding: utf-8 -*- import re diff --git a/pdfminer/rijndael.py b/pdfminer/rijndael.py index d1fd4c7..05cf946 100644 --- a/pdfminer/rijndael.py +++ b/pdfminer/rijndael.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + """ Python implementation of Rijndael encryption algorithm. diff --git a/pdfminer/runlength.py b/pdfminer/runlength.py index 9964e7f..54bc769 100644 --- a/pdfminer/runlength.py +++ b/pdfminer/runlength.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + # # RunLength decoder (Adobe version) implementation based on PDF Reference # version 1.4 section 3.3.4. diff --git a/pdfminer/utils.py b/pdfminer/utils.py index e2638d4..9700abf 100644 --- a/pdfminer/utils.py +++ b/pdfminer/utils.py @@ -1,4 +1,4 @@ -#!/usr/bin/env python + """ Miscellaneous Routines. """ diff --git a/setup.py b/setup.py index 2859607..e5ffecb 100644 --- a/setup.py +++ b/setup.py @@ -1,4 +1,3 @@ -#!/usr/bin/env python #from distutils.core import setup from setuptools import setup from pdfminer import __version__ diff --git a/tests/test_pdfminer_ccitt.py b/tests/test_pdfminer_ccitt.py index f482d88..9f27833 100644 --- a/tests/test_pdfminer_ccitt.py +++ b/tests/test_pdfminer_ccitt.py @@ -1,4 +1,5 @@ -#!/usr/bin/python +#!/usr/bin/env python + # -*- coding: utf-8 -*- from nose.tools import assert_equal, assert_true, assert_false diff --git a/tests/test_pdfminer_crypto.py b/tests/test_pdfminer_crypto.py index eb7fffd..fdabd00 100644 --- a/tests/test_pdfminer_crypto.py +++ b/tests/test_pdfminer_crypto.py @@ -1,4 +1,4 @@ -#!/usr/bin/python +#!/usr/bin/env python # -*- coding: utf-8 -*- from nose.tools import assert_equal diff --git a/tests/test_pdfminer_psparser.py b/tests/test_pdfminer_psparser.py index 0ab7a3d..2fbae8e 100644 --- a/tests/test_pdfminer_psparser.py +++ b/tests/test_pdfminer_psparser.py @@ -1,4 +1,5 @@ -#!/usr/bin/python +#!/usr/bin/env python + # -*- coding: utf-8 -*- from nose.tools import assert_equal, assert_true, assert_false diff --git a/tests/test_tools_dumppdf.py b/tests/test_tools_dumppdf.py index 7b962f7..87d74d2 100644 --- a/tests/test_tools_dumppdf.py +++ b/tests/test_tools_dumppdf.py @@ -1,4 +1,5 @@ -#!/usr/bin/python +#!/usr/bin/env python + # -*- coding: utf-8 -*- import six diff --git a/tests/test_tools_pdf2txt.py b/tests/test_tools_pdf2txt.py index 6d92df6..7ef0ffd 100644 --- a/tests/test_tools_pdf2txt.py +++ b/tests/test_tools_pdf2txt.py @@ -1,4 +1,5 @@ -#!/usr/bin/python +#!/usr/bin/env python + # -*- coding: utf-8 -*- import nose, logging, os diff --git a/tools/conv_afm.py b/tools/conv_afm.py index 846e255..2402a8e 100755 --- a/tools/conv_afm.py +++ b/tools/conv_afm.py @@ -1,4 +1,5 @@ #!/usr/bin/env python + import sys import fileinput diff --git a/tools/conv_cmap.py b/tools/conv_cmap.py index 7e45a89..7ef04d6 100755 --- a/tools/conv_cmap.py +++ b/tools/conv_cmap.py @@ -1,4 +1,5 @@ #!/usr/bin/env python + import sys try: import cPickle as pickle diff --git a/tools/conv_glyphlist.py b/tools/conv_glyphlist.py index 93179ad..2a70438 100755 --- a/tools/conv_glyphlist.py +++ b/tools/conv_glyphlist.py @@ -1,4 +1,5 @@ #!/usr/bin/env python + import sys import fileinput diff --git a/tools/dumppdf.py b/tools/dumppdf.py index 01dc6e4..037c6e0 100755 --- a/tools/dumppdf.py +++ b/tools/dumppdf.py @@ -1,4 +1,5 @@ #!/usr/bin/env python + # # dumppdf.py - dump pdf contents in XML format. # diff --git a/tools/latin2ascii.py b/tools/latin2ascii.py index aa89e1d..4caf1c7 100755 --- a/tools/latin2ascii.py +++ b/tools/latin2ascii.py @@ -1,4 +1,5 @@ #!/usr/bin/env python + # # latin2ascii.py - converts latin1 characters into ascii. # diff --git a/tools/pdf2html.cgi b/tools/pdf2html.cgi index bfd591f..ea04615 100755 --- a/tools/pdf2html.cgi +++ b/tools/pdf2html.cgi @@ -1,4 +1,4 @@ -#!/usr/bin/python -O +#!/usr/bin/env python -O # # pdf2html.cgi - Gateway script for converting PDF into HTML. # diff --git a/tools/pdf2txt.py b/tools/pdf2txt.py index cd40de1..1e8ec0b 100755 --- a/tools/pdf2txt.py +++ b/tools/pdf2txt.py @@ -1,4 +1,5 @@ #!/usr/bin/env python + """ Converts PDF text content (though not images containing text) to plain text, html, xml or "tags". """ diff --git a/tools/prof.py b/tools/prof.py index 5402de4..aaf6712 100644 --- a/tools/prof.py +++ b/tools/prof.py @@ -1,4 +1,5 @@ #!/usr/bin/env python + import sys def prof_main(argv): diff --git a/tools/runapp.py b/tools/runapp.py index b90f962..6b953be 100755 --- a/tools/runapp.py +++ b/tools/runapp.py @@ -1,4 +1,5 @@ #!/usr/bin/env python + ## ## WebApp class runner ##