Create LICENSE and README for project
First commit for godiffx, just boilerplate based off of
pydiffx.
Testing done: NA
4552441df26f5bc1f71e0957b915eb5d294e2c88
Jordan
Create basic support files for diffx.
Created errors.go, sections.go, options.go, and text.go. These will
all be based on their python equivalents from pydiffx. At the moment,
there is no code within them besides stating which package they are
apart of.
No functionality introduced so no testing done.
53615b11759d77a06b2a0d6aa8f387beeb1eced7
Jordan
Write basic constants for options.go
Added the basic functions to options.go based on what is in
options.py for pydiffx. This was accomplished via enums using iota
in go. The implementation may be tweaked overtime once I actually need
to start using them.
The functionality of these was tested in another go file. This was
also the first time I tried testing the godiffx module locally and
it worked!
ed55e923e78ead35565ee3f344a234516cfc3921
Jordan
Wrote code for sections.go
sections.go is based off of sections.py, and the functionality was
replicated as close as possible to that file. One hiccup was that
Go does not support sets natively, but an easy work around is to use
sets and just set all of the keys to true.
Code was tested by creating the struct in another file that imported
this module. The default parameter method was tested and properly adds
in the default values and the getters were all functioning as expected.
3659b78ffe8b8cc27d23a5d1f497f2bb0ab00a21
Jordan
Update documentation for sections.go
Basic documentation via comments was added to the code in sections.go.
No new functionality was introduced so there was no additional testing.
398e1cd37ede5380265ee716bb93f7217bfa9996
Jordan
Write all custom errors for godiffx.
These are based on the errors that were created for pydiffx. Go
does not use errors in the same way, but an error struct was created
and an error can be returned whenever needed by other functions.
One note is that at the moment, there is no unicode support since I am not
very familiar with it or implementing it, but I assume it will be as
simple as changing the type and possibly importing the library in the
future.
This was tested by importing the module locally and testing each of
the different types of errors. The init function was validated to
ensure it would not overwrite an existing error and the Err function
was tested to ensure it printed properly. I also tested to ensure
the baseDiffXError was not accessible.
37da0122f047f79a5cb92e0aeba4d0b5711196ec
Jordan
Add documentation and fix line value for DiffXParseError
Basic documentation was added to all of the functions and errors.
They are all pretty simple definitions based off of what was used in
the pydiffx version.
There was also a modification made to the way line and column values are
treated in DiffXParseError. Originally this was set to just use the values
as they are passed in. However, pydiffx assumes these are 0 based so the
message increments it by 1. This change was tested by importing the
module locally in another file.
b9ade0ac8f01e150291ac24e2bebc38b01276cbc
Jordan
Create text values and write SplitLines
Following the example from text.py, the maps have been created for
godiffx as well. One thing to note is the utils folder is now gone.
This is because when creating a module in Go, you actually don't
use nested directories since those will then belong to a new package.
Another change is that instead of using a map for the objects like in
python, structs were created along with a way to initialize their
default values. The reasoning behind this is that BOMS uses tuples to
store multiple values in pydiffx, but tuples don't exist in Go.
The SplitLines method was also written. This took a fair bit of time
while I experimented and learned about the best way to handle bytes.
The bytes.Buffer seems to be a very efficient data structure that comes
with some functionality. Time will tell if this is the right way to go!
All of the objects have been tested in the separate project using this
module and I have run SplitLines with what I believe to be the proper
behaviour.
ddb24d486e20a4f3365cb1186c647575d2ff3458
Jordan
Fix Errors and Write StripBom Method.
I made a silly mistake when creating the errors by naming the
error functions 'Err()' instead of 'Error()' so they were not
quite working like errors yet, so that has been resolved.
I also wrote the StripBom method which will remove a given BOM from
a byte buffer if it exists. In order to do this, I made a method that
can retrieve the BOM given a string of the encoding type.
I tested out the methods again in main. To test StripBom, I created a
buffer that held "\uFEFF is a string that starts with a Byte Order Mark"
and compared its length to before and after running the function. The buffer
that the function returns will be shorter but it will not mutate the buffer
passed in.
87a9119601141817aa81cef3e8b163a7d20b4f7b
Jordan
Created test files and wrote init tests for Errors.go
All of the test files were created and can be run with the command
'go test'. I have written all of the Init tests for errors.go except
for the base class since I am not 100% sure how to test private methods.
However, I don't think it is too important since all of the other test cases
work.
0576404dacb4b9f6ac9f5eceb9c29ee97c022b2a
Jordan
Finish tests for errors.go
Finished writing the rest of the tests for errors.go, specifically
all of the tests for the Error() methods.
The only testing that was done was running the test suite and seeing
all of the new tests pass.
40257068d62ca72958a447221529a04876ca908a
Jordan
Add testing for sections.go
The testing itself is pretty trivial, since it just ensure the
init methods works properly and that the set functions return
the proper sets. The ValidSectionStates function has a very long
test since there are so many different pieces that need to be checked.
Not sure if this can be simplified some way in the future.
cb86e91b7b51ac9be3ed52c35fd9aab20baefa90
Jordan
Wrote unit tests for finished functions in text.go
While there is still a bit of work to be done for text.go, the unit
tests for the finished functions have now been added in. These tests
try to test for every edge case I can think of, but suggestions are
welcome!
I also do not know if there is a better way to run these tests.
Currently, each test is rather large with a number of small pieces
it tests for a function. It seems like Go tests usually have one test
per function instead of multiple tests for different edge cases of a
function, but if I find something new then I will make adjustments.
The ony strategy I saw is to make a struct with test data, but I feel
it would bloat the code a bit due to the number of byte buffers needed.
6f73cf4159218d416aa97234e19481982e6c2c94
Jordan
Create new error for invalid encoding.
Since it seemed relatively straightforward to add my own encoding
and there weren't any great go modules I found with a quick search,
I have started creating a function to encode in some given utf scheme.
It makes sense to begin by creating an error that can be thrown in case
someone provides an invalid string.
A go test has been created as well in the same manner as all of the
other error tests to ensure InvalidEncodingError works properly.
13bebc89fe5764f1da87e795230cab2eaddd1cb4
Jordan
Create encode method for text.go
I created an encode function based off of what Python does. There
weren't any libraries I found quick that seemed to cover the different
utf styles. I based their implementation off of the returns for unix
and dos newlines.
I have not created tests for the function yet, but I manually tested it
by using the function in another project and comparing the output to Python
b5ef19b54466ccd7e3b3a961188974203b0c15f4
Jordan
Add GetNewlineForType function in text.go
This function is the same as the version in pydiffx. One note is that
it returns a byte slice instead of a buffer, but I thought this would
make more sense since it will usually only be a few bytes at most.
There is no testing created for this function yet, it will be written
at a later time. But it was manually tested with utf8 and utf16.
0951ce3c23a99b2d1b5e53128ebdd85feb28e413
Jordan
Finish manually testing all functions for text.go
All functions have now been written and manually tested in text.go.
I thought I had things working on Saturday, but there was still an
issue related to an EOF error that I had to sort out.
Manual testing was done by creating a buffer that had both unix and
dos newlines as well as an empty one where it assumed we used unix.
No tests have been written for the newline guessing or getting.
5678c06d947b2c40ff5e73814a2ae1af36812da4
Jordan
Write tests for line ending methods.
This commit just adds a couple tests to ensure the GuessLineEnding
and GetNewlineForType functions work properly.
3225e18cd1e33cf1f473cedfff7eb37fdb9df982
Jordan
Wrote unit test for encoding function
Just a quick unit test for encoding function. It's very inclusive
and covers every possible return.
I noticed that Go only thinks I have ~90% coverage for my tests, so I
will likely spend some time trying to identify the edge cases I missed.
1a455560529844e0eb504cb130b8136534fbf337
Jordan
