|
Removed some parts of the structure documentation since the assignment |
|
|
Fixed default inputs. |
|
|
Changed inspected context details to more accurately reflect the current policy. |
|
|
Improved context-related documentation and related tests. |
|
|
Speculative adjustment to contexts associated with new instances. |
|
|
Introduced instance templates to the generated raw image. |
|
|
Fixed raw object positioning for functions with static defaults. |
|
|
Introduced a DataObject class to provide a nicer representation of the raw |
|
|
Fixed the output representation of Block instances. |
|
|
Renamed global_namespace attributes to module in data instances. |
|
|
Fix unused objects optimisation for Attr instances with no predictable |
|
|
Moved InspectError usage into the inspect module. |
|
|
Make unused name and object removal more conservative, avoiding issues of |
|
|
Insist on Const instances as keys for the constants dictionary, ensuring that |
|
|
Separated functionality in Importer into distinct Importer and Program classes. |
|
|
Added some support for constant data. |
|
|
Moved raw code production into the Importer class. |
|
|
Moved method name information into the common module. |
|
|
Added status flags to the Importer, ensuring that certain operations (vacuum and |
|
|
Upheld the class/function restrictions on removing objects. |
|
|
Added name accounting. |
|
|
Moved constant accounting to the Importer class. |
|
|
Moved optimisations onto instances of the Importer class, using such instances |
|
|
Add name retrieval method for checking raw output and ensuring that each object |
|
|
Moved block expansion into the micropython.rsvp module's raw processing |
|
|
Added methods to convert from blocks to code sequences. |
|
|
Introduced code blocks in place of labels, potentially leading to increased |
|
|
Added the raw form of Module objects. |
|
|
Fixed unary operator frame preparation. |
|
|
Re-ordered the methods according to corresponding node type. |
|
|
Introduced immediate usage assessment for AssList/AssTuple, along with separate |
|
|
Tidied up the optimiser by removing unnecessary translator/translation |
|
|
Split the attributes2 test into smaller pieces. |
|
|
Tidied the optimisations, moving the work done in the self access optimisation |
|
|
Made the constant attribute accessor optimisation explicit. |
|
|
Moved the Optimiser class into its own module, taking the responsibility for |
|
|
Moved some documentation to separate files in the docs directory, updating some |
|
|
Attempt to remove excessive temporary storage usage for the results of |
|
|
Added some support for augmented assignment. |
|
|
Apply the simple case immediate assignment optimisation to attribute assignment. |
|
|
Moved graph definition production to a separate module. |
|
|
Added notes about related projects. |
|
|
Added a simple graph function producing Graphviz output. |
|
|
Added exception conditions for various RSVP instructions using AttributeError |
|
|
Tidied up the exception code generation to ensure that referenced exceptions are |
|
|
Changed the frame usage calculations. |
|
|
Removed the source storage optimisation, since it can effectively be replaced by |
|
|
Removed the superfluous temporary storage optimisation since it could never work |
|
|
Added a summary of current optimisations. |
|
|
Separated the optimisation methods out into a distinct class. |
|
|
Added the final Return instruction to generated images. |
|
|
Made object construction through MakeObject use details of a supplied class, |
|
|
Added notes about calling initialisers and instantiators, adopting a strategy |
|
|
Added another test of method invocation, explicitly testing instance/class |
|
|
Added descendants (and self) to the collection of attributes for each class, |
|
|
Proposed the replacement of the class table by special attributes in the object |
|
|
Fixed temporary storage and local usage. |
|
|
Changed context loading conditions to depend on the target instead of a context |
|
|
Fixed known_target optimisation. |
|
|
Fixed temp allocation, hopefully. |
|
