Skip to content
Commits on Dec 14, 2015
  1. @koriroys

    Update rspec to 3.4 with transpec syntax conversion and manual edits

    koriroys committed
    Convert specs to RSpec 2.99.2 syntax with Transpec
    This conversion is done by Transpec 3.1.1 with the following command:
    * 950 conversions
        from: obj.should
          to: expect(obj).to
    * 270 conversions
        from: == expected
          to: eq(expected)
    * 27 conversions
        from: obj.should_receive(:message)
          to: expect(obj).to receive(:message)
    * 18 conversions
        from: lambda { }.should
          to: expect { }.to
    * 17 conversions
        from: be_false
          to: be_falsey
    * 12 conversions
        from: obj.should_not
          to: expect(obj).not_to
    * 10 conversions
        from: be_true
          to: be_truthy
    * 7 conversions
        from: =~ /pattern/
          to: match(/pattern/)
    * 6 conversions
        from: mock('something')
          to: double('something')
    * 5 conversions
        from: stub('something')
          to: double('something')
    * 2 conversions
        from: lambda { }.should_not
          to: expect { }.not_to
    * 2 conversions
        from: pending 'is an example' { }
          to: skip 'is an example' { }
    * 1 conversion
        from: it { should ... }
          to: it { ... }
    * 1 conversion
        from: obj.should_not_receive(:message)
          to: expect(obj).not_to receive(:message)
    * 1 conversion
        from: pending
          to: skip
    For more details:
    Additional fix:
     -> Assertion config changes
    "For RSpec 3, we've removed expect_with :stdlib and instead
    opted for explicit :test_unit and :minitest options:"
    :stdlib is no longer a valid option, so changing it to :minitest
Commits on Dec 11, 2015
  1. remove minitest dependency

    * convert the few specs that uised minitest to use vanilla rspec
    * They were originally added in commit aa2c270, with the justification
      that they dropped spec times from ~192s to ~34 seconds. Somewhere
      between then and now, that became less of an issue - possibly due to
      upgrading to rspec 3 or improvements to MRI.
    * On my current laptop with MRI 2.2.3, the spec/callback_spec.rb complete
      in < 8 seconds with both minitest and rspec
  2. add minitest as a development dependency

    * required in some specs to keep them fast. Previously we used test/unit
      from the stdlib, but it's been removed in MRi >= 2.2
Commits on Jun 26, 2013
  1. Merge branch 'lzw-fix'

  2. add a spec for the fix in c5866da

Commits on Jun 23, 2013
  1. add an integration spec for times control char glyphs

    * we shouldn't raise an exception when a PDF uses the built in
      times-roman font and includes a control character glyph
Commits on May 12, 2013
  1. ensure we can decrypt PDFs with no ID

    * the PDF spec says encrypted files MUST have a document ID
    * Yet Acrobat reader works just fine when they don't so let's copy their
  2. Merge branch 'page-orientation'

  3. use a context block to make these spec lines < 100 chars

    * so that the quality specs continue to pass on this branch
  4. remove some unused spec PDFs

  5. move page orientation detection into a dedicated class

    * significantly easier to unit test
Commits on May 7, 2013
  1. Adding sample files used to test page orientation

    Lyle Mullican committed
Commits on Apr 27, 2013
  1. @nmadura

    Activated TrueType width calculator

    nmadura committed with
    The TrueType width calculator was essentially unused, and was being set
    to TypeOneOrThree width calculator, upon activating it, found an error,
    I fixed that as well.
Commits on Apr 26, 2013
  1. Add spec for new method orientation()

    Lyle Mullican committed
Commits on Apr 7, 2013
  1. GlyphHash should recognise names with AXXXXX

    James Healy committed with
    * where X is a digit
  2. Zapf Dingbat and Symbol fonts with no explicit encoding should defaul…

    James Healy committed
    …t to the correct encodings
    * fixes text extraction from these fonts
Commits on Mar 7, 2013
  1. fix a failing spec for RegisterReceiver#series

    James Healy committed
    * thanks to Peter Suschlik for reporting
Commits on Mar 6, 2013
  1. @splattael
  2. @splattael
  3. @pat

    Handle files with extra characters after the EOF marker.

    pat committed
    In this provided file, the suffix is mb, but this patch will ignore anything after the EOF marker, provided it's part of the same line.
Commits on Mar 2, 2013
  1. oops, this should have been included in ac28f59

    James Healy committed
  2. PDFs in the wild can have a cmap bfrange containing > 255 characters

    James Healy committed
    * I'm not sure why I added this check in the first place. Maybe because
      the cmap spec says only 255 chars are allowed. If that's the case I
      guess we follow the usual policy of "do what adobe acrobat does" and
      support ranges that aren't spec compliant.
    * thanks to AJarreck for reporting it
    * closes #87
Commits on Feb 25, 2013
  1. ensure we can extract text from pages with built in fonts

    James Healy committed
    * and no difference table
  2. update GlyphHash to support two-way translations

    James Healy committed
    * name to unicode codepoint
    * unicode codepoint to name
Commits on Feb 12, 2013
  1. assume 8.5" x 11" pages if the MediaBox is missing

    James Healy committed
    * the PDF spec requires a MediaBox, so any PDF that leaves it out isn't
    * Acrobat seems to just assume the page is 8.5" by 11", so users expect
      these non-compliant files to just work. We'd better copy the behaviour
Commits on Jan 28, 2013
  1. @chip
Commits on Jan 18, 2013
  1. add a forgotten spec file

  2. re-jig PageState#font_size to work with rotated text

    * The final font size is a combination of the text_font_size and any scaling.
    * Closes #74
Something went wrong with that request. Please try again.