summaryrefslogtreecommitdiff
path: root/tests
Commit message (Collapse)AuthorAge
...
* Fix Array() to create empty arrayJames R. Barlow2018-08-16
|
* Update copyright fileJames R. Barlow2018-08-14
|
* Fix error in parsing inline images with masksJames R. Barlow2018-08-10
| | | | | | | | | | via ocrmypdf, error is: ERROR - parsed object (offset 7): unknown token while reading object; treating as string Occurs because PdfInlineImage reparses the dictionary that defines it, but this had the effect of converting the PDF token 'true' to a Python bool with value True, and to a string 'True' which is not a legal PDF token for reparsing.
* pybind11 >2.2.3 seems to fix trampolines on Python 3.7James R. Barlow2018-08-04
|
* Document and test show_xref_tableJames R. Barlow2018-07-18
| | | | Add missing ostream redirect
* Add Pdf.remove_unreferenced_resourcesJames R. Barlow2018-07-18
|
* Convert missing file to proper FileNotFoundErrorJames R. Barlow2018-07-17
|
* Clarify comments about behavior of custom type casterJames R. Barlow2018-07-17
|
* Use qpdf 8.1.0's newUnicodeString for utf-8 stringsJames R. Barlow2018-07-17
|
* Support Name.Thing -> Name('/Thing') like img2pdfJames R. Barlow2018-07-17
| | | | | Seems like it is a lot cleaner, and a little metaclass never hurt anyone.
* Add ability to delete multiple pages at once in a sliceJames R. Barlow2018-07-15
| | | | Also elided a naked PyObject*.
* Reorganize IPython helpers add them to test suiteJames R. Barlow2018-07-14
|
* More array and dict test coverageJames R. Barlow2018-07-13
|
* Improve dictionary construction interface and more not-convertiblesJames R. Barlow2018-07-13
|
* Add test for attempt to encode not convertible objectsJames R. Barlow2018-07-13
|
* Refactor dictionary['key']=valueJames R. Barlow2018-07-13
| | | | | | | Removal the ability to set a key to None; invite the user to use del instead. The reason for this is that in PDFs and QPDF setting a key to null is equivalent to deletion. Seems best to remove potential source of confusion.
* Revise Object.write() to use kwargsJames R. Barlow2018-07-13
| | | | | It's more natural in Python to omit a parameter than use an explicit None, and this helps with normalizing the treatment of None's.
* Add several missing open and save features to pikepdf.PdfJames R. Barlow2018-07-13
|
* Remove objecthandle_decodeJames R. Barlow2018-07-13
| | | | | | It was only ever implemented for scalars, and we know that do in a custom type converter. Updated test code to deal with several tests that were not longer meaningful.
* Do reference counting without a special functionJames R. Barlow2018-07-13
|
* Fix refcount test was always returning TrueJames R. Barlow2018-07-12
| | | | Fortunately it worked anyway
* Disable Object._parse_stream in Py3.7James R. Barlow2018-07-04
| | | | | Trampolines seem to fail... these aren't public API so wait for pybind11 to catch up.
* Improve handling of trivial palettesJames R. Barlow2018-06-26
|
* Better testing for image accessJames R. Barlow2018-06-25
|
* Test case for indirect reprJames R. Barlow2018-06-25
|
* Add more tests for pages.p(N)James R. Barlow2018-06-25
|
* Add slice test for different length replacementJames R. Barlow2018-06-25
|
* Add name equality testJames R. Barlow2018-06-25
|
* Fix same_owner_as -> is_owned_byJames R. Barlow2018-06-25
|
* Rename Object.check_owner to Object.same_owner_asJames R. Barlow2018-06-25
|
* pylint fixes of test and support filesJames R. Barlow2018-06-25
|
* Internal reorganization to objects, modelsJames R. Barlow2018-06-24
|
* Improve PdfMatrixJames R. Barlow2018-06-24
| | | | Remove nonfunctioning setters, add operators.
* Remove a lot of numerical code made unnecessary by 'unboxing'James R. Barlow2018-06-24
| | | | | Native representation of Python objects means we don't need to implement duck typing on our non-native objects.
* Add objgen/get_object testJames R. Barlow2018-06-24
|
* pylintingJames R. Barlow2018-06-23
|
* Convert Real to decimal.Decimal and fix testsJames R. Barlow2018-06-08
|
* Unbox int, bool, real, all passingJames R. Barlow2018-06-08
| | | | | Some tests weakened to the point that they're not properly checking edge cases where we can't convert Python types to QPDF types.
* Gently improve Null unboxingJames R. Barlow2018-06-07
|
* Fix repr(Operator(...)) not pure expressionJames R. Barlow2018-06-07
|
* Add test cases for repr()James R. Barlow2018-06-07
|
* PdfImage: better testsJames R. Barlow2018-06-06
|
* Fix PermissionErrors due to Windows locking semanticsJames R. Barlow2018-06-04
|
* Fix test suite failure from $HOME not being defined on WindowsJames R. Barlow2018-06-04
|
* Allow str(Integer) - fixesJames R. Barlow2018-05-28
| | | | RuntimeError: operation for string attempted on object of type integer
* Fix unit'ed vars in OperandGrouperJames R. Barlow2018-05-25
|
* Add support for accessing inline imagesJames R. Barlow2018-05-24
|
* Forbid constructing pikepdf.Object() directlyJames R. Barlow2018-05-22
| | | | libqpdf defines no public constructor, so it crashes.
* Use qpdf 7.0.0 API for parsing page contentsJames R. Barlow2018-05-21
|
* Do repr() recursion limit with Python CAPI tooJames R. Barlow2018-05-20
|
generated by cgit v1.2.3 (git 2.25.1) at 2024-06-29 17:04:24 +0000